diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24590 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8211143695014663,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002346041055718475,
+      "grad_norm": 3.6450053402509406,
+      "learning_rate": 0.0,
+      "loss": 1.2572,
+      "step": 1
+    },
+    {
+      "epoch": 0.000469208211143695,
+      "grad_norm": 4.3485356326437845,
+      "learning_rate": 2.3419203747072602e-08,
+      "loss": 1.302,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007038123167155425,
+      "grad_norm": 6.519353527947624,
+      "learning_rate": 4.6838407494145204e-08,
+      "loss": 1.3035,
+      "step": 3
+    },
+    {
+      "epoch": 0.00093841642228739,
+      "grad_norm": 8.604850613314342,
+      "learning_rate": 7.02576112412178e-08,
+      "loss": 1.3578,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011730205278592375,
+      "grad_norm": 5.384869815415941,
+      "learning_rate": 9.367681498829041e-08,
+      "loss": 1.3102,
+      "step": 5
+    },
+    {
+      "epoch": 0.001407624633431085,
+      "grad_norm": 3.0808542461257487,
+      "learning_rate": 1.1709601873536301e-07,
+      "loss": 1.36,
+      "step": 6
+    },
+    {
+      "epoch": 0.0016422287390029325,
+      "grad_norm": 4.818428908064201,
+      "learning_rate": 1.405152224824356e-07,
+      "loss": 1.3193,
+      "step": 7
+    },
+    {
+      "epoch": 0.00187683284457478,
+      "grad_norm": 8.881983711459453,
+      "learning_rate": 1.639344262295082e-07,
+      "loss": 1.3436,
+      "step": 8
+    },
+    {
+      "epoch": 0.0021114369501466276,
+      "grad_norm": 6.209731653543138,
+      "learning_rate": 1.8735362997658082e-07,
+      "loss": 1.3472,
+      "step": 9
+    },
+    {
+      "epoch": 0.002346041055718475,
+      "grad_norm": 3.8408158741152936,
+      "learning_rate": 2.107728337236534e-07,
+      "loss": 1.3155,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025806451612903226,
+      "grad_norm": 6.645338741206632,
+      "learning_rate": 2.3419203747072603e-07,
+      "loss": 1.2661,
+      "step": 11
+    },
+    {
+      "epoch": 0.00281524926686217,
+      "grad_norm": 16.87412496284319,
+      "learning_rate": 2.5761124121779863e-07,
+      "loss": 1.3359,
+      "step": 12
+    },
+    {
+      "epoch": 0.0030498533724340176,
+      "grad_norm": 3.774480076443614,
+      "learning_rate": 2.810304449648712e-07,
+      "loss": 1.2923,
+      "step": 13
+    },
+    {
+      "epoch": 0.003284457478005865,
+      "grad_norm": 3.8836693486151312,
+      "learning_rate": 3.044496487119438e-07,
+      "loss": 1.3094,
+      "step": 14
+    },
+    {
+      "epoch": 0.0035190615835777126,
+      "grad_norm": 5.2500074581781675,
+      "learning_rate": 3.278688524590164e-07,
+      "loss": 1.2843,
+      "step": 15
+    },
+    {
+      "epoch": 0.00375366568914956,
+      "grad_norm": 4.0449453982333266,
+      "learning_rate": 3.51288056206089e-07,
+      "loss": 1.305,
+      "step": 16
+    },
+    {
+      "epoch": 0.003988269794721408,
+      "grad_norm": 5.250683179693039,
+      "learning_rate": 3.7470725995316163e-07,
+      "loss": 1.3284,
+      "step": 17
+    },
+    {
+      "epoch": 0.004222873900293255,
+      "grad_norm": 7.063096507845875,
+      "learning_rate": 3.9812646370023426e-07,
+      "loss": 1.3711,
+      "step": 18
+    },
+    {
+      "epoch": 0.004457478005865103,
+      "grad_norm": 5.389525286489735,
+      "learning_rate": 4.215456674473068e-07,
+      "loss": 1.2755,
+      "step": 19
+    },
+    {
+      "epoch": 0.00469208211143695,
+      "grad_norm": 13.47512068224741,
+      "learning_rate": 4.449648711943794e-07,
+      "loss": 1.3309,
+      "step": 20
+    },
+    {
+      "epoch": 0.004926686217008798,
+      "grad_norm": 8.690321624905424,
+      "learning_rate": 4.6838407494145205e-07,
+      "loss": 1.3033,
+      "step": 21
+    },
+    {
+      "epoch": 0.005161290322580645,
+      "grad_norm": 5.3665280506949715,
+      "learning_rate": 4.918032786885246e-07,
+      "loss": 1.3485,
+      "step": 22
+    },
+    {
+      "epoch": 0.005395894428152493,
+      "grad_norm": 12.778939105297262,
+      "learning_rate": 5.152224824355973e-07,
+      "loss": 1.3281,
+      "step": 23
+    },
+    {
+      "epoch": 0.00563049853372434,
+      "grad_norm": 3.954317406271846,
+      "learning_rate": 5.386416861826698e-07,
+      "loss": 1.3843,
+      "step": 24
+    },
+    {
+      "epoch": 0.005865102639296188,
+      "grad_norm": 4.116531974224858,
+      "learning_rate": 5.620608899297424e-07,
+      "loss": 1.3211,
+      "step": 25
+    },
+    {
+      "epoch": 0.006099706744868035,
+      "grad_norm": 3.9250460374163683,
+      "learning_rate": 5.85480093676815e-07,
+      "loss": 1.2407,
+      "step": 26
+    },
+    {
+      "epoch": 0.006334310850439883,
+      "grad_norm": 3.519670059389327,
+      "learning_rate": 6.088992974238876e-07,
+      "loss": 1.2983,
+      "step": 27
+    },
+    {
+      "epoch": 0.00656891495601173,
+      "grad_norm": 4.136300157405571,
+      "learning_rate": 6.323185011709602e-07,
+      "loss": 1.3416,
+      "step": 28
+    },
+    {
+      "epoch": 0.006803519061583578,
+      "grad_norm": 10.279347514776052,
+      "learning_rate": 6.557377049180328e-07,
+      "loss": 1.3196,
+      "step": 29
+    },
+    {
+      "epoch": 0.007038123167155425,
+      "grad_norm": 4.765152922147381,
+      "learning_rate": 6.791569086651055e-07,
+      "loss": 1.2835,
+      "step": 30
+    },
+    {
+      "epoch": 0.007272727272727273,
+      "grad_norm": 7.208624944075502,
+      "learning_rate": 7.02576112412178e-07,
+      "loss": 1.3429,
+      "step": 31
+    },
+    {
+      "epoch": 0.00750733137829912,
+      "grad_norm": 18.131451894085693,
+      "learning_rate": 7.259953161592506e-07,
+      "loss": 1.3031,
+      "step": 32
+    },
+    {
+      "epoch": 0.007741935483870968,
+      "grad_norm": 3.251140914430706,
+      "learning_rate": 7.494145199063233e-07,
+      "loss": 1.3911,
+      "step": 33
+    },
+    {
+      "epoch": 0.007976539589442815,
+      "grad_norm": 8.79645269699139,
+      "learning_rate": 7.728337236533959e-07,
+      "loss": 1.3865,
+      "step": 34
+    },
+    {
+      "epoch": 0.008211143695014663,
+      "grad_norm": 5.75165059623209,
+      "learning_rate": 7.962529274004685e-07,
+      "loss": 1.2966,
+      "step": 35
+    },
+    {
+      "epoch": 0.00844574780058651,
+      "grad_norm": 4.8261727607857505,
+      "learning_rate": 8.196721311475409e-07,
+      "loss": 1.3001,
+      "step": 36
+    },
+    {
+      "epoch": 0.008680351906158358,
+      "grad_norm": 3.7899028159194055,
+      "learning_rate": 8.430913348946136e-07,
+      "loss": 1.3233,
+      "step": 37
+    },
+    {
+      "epoch": 0.008914956011730205,
+      "grad_norm": 77.06771197169489,
+      "learning_rate": 8.665105386416862e-07,
+      "loss": 1.4009,
+      "step": 38
+    },
+    {
+      "epoch": 0.009149560117302053,
+      "grad_norm": 5.551275085083891,
+      "learning_rate": 8.899297423887588e-07,
+      "loss": 1.3317,
+      "step": 39
+    },
+    {
+      "epoch": 0.0093841642228739,
+      "grad_norm": 6.607621487321918,
+      "learning_rate": 9.133489461358315e-07,
+      "loss": 1.2016,
+      "step": 40
+    },
+    {
+      "epoch": 0.009618768328445748,
+      "grad_norm": 4.4406064626417985,
+      "learning_rate": 9.367681498829041e-07,
+      "loss": 1.3097,
+      "step": 41
+    },
+    {
+      "epoch": 0.009853372434017595,
+      "grad_norm": 5.952053250291134,
+      "learning_rate": 9.601873536299766e-07,
+      "loss": 1.2593,
+      "step": 42
+    },
+    {
+      "epoch": 0.010087976539589443,
+      "grad_norm": 5.4801646507078505,
+      "learning_rate": 9.836065573770493e-07,
+      "loss": 1.3073,
+      "step": 43
+    },
+    {
+      "epoch": 0.01032258064516129,
+      "grad_norm": 3.277972266250071,
+      "learning_rate": 1.0070257611241219e-06,
+      "loss": 1.3387,
+      "step": 44
+    },
+    {
+      "epoch": 0.010557184750733138,
+      "grad_norm": 12.429665956983847,
+      "learning_rate": 1.0304449648711945e-06,
+      "loss": 1.3176,
+      "step": 45
+    },
+    {
+      "epoch": 0.010791788856304985,
+      "grad_norm": 37.48972536707086,
+      "learning_rate": 1.0538641686182672e-06,
+      "loss": 1.3101,
+      "step": 46
+    },
+    {
+      "epoch": 0.011026392961876833,
+      "grad_norm": 10.426155225949543,
+      "learning_rate": 1.0772833723653396e-06,
+      "loss": 1.2829,
+      "step": 47
+    },
+    {
+      "epoch": 0.01126099706744868,
+      "grad_norm": 6.811979793371468,
+      "learning_rate": 1.1007025761124122e-06,
+      "loss": 1.3173,
+      "step": 48
+    },
+    {
+      "epoch": 0.011495601173020528,
+      "grad_norm": 5.344640324674893,
+      "learning_rate": 1.1241217798594848e-06,
+      "loss": 1.3259,
+      "step": 49
+    },
+    {
+      "epoch": 0.011730205278592375,
+      "grad_norm": 8.884608922039357,
+      "learning_rate": 1.1475409836065575e-06,
+      "loss": 1.3234,
+      "step": 50
+    },
+    {
+      "epoch": 0.011964809384164223,
+      "grad_norm": 3.9335710107098607,
+      "learning_rate": 1.17096018735363e-06,
+      "loss": 1.3233,
+      "step": 51
+    },
+    {
+      "epoch": 0.01219941348973607,
+      "grad_norm": 9.519135441340808,
+      "learning_rate": 1.1943793911007025e-06,
+      "loss": 1.3034,
+      "step": 52
+    },
+    {
+      "epoch": 0.012434017595307918,
+      "grad_norm": 8.035792757540126,
+      "learning_rate": 1.2177985948477752e-06,
+      "loss": 1.2138,
+      "step": 53
+    },
+    {
+      "epoch": 0.012668621700879765,
+      "grad_norm": 4.904948094072176,
+      "learning_rate": 1.2412177985948478e-06,
+      "loss": 1.2059,
+      "step": 54
+    },
+    {
+      "epoch": 0.012903225806451613,
+      "grad_norm": 12.830754470261745,
+      "learning_rate": 1.2646370023419204e-06,
+      "loss": 1.2152,
+      "step": 55
+    },
+    {
+      "epoch": 0.01313782991202346,
+      "grad_norm": 7.256363284150356,
+      "learning_rate": 1.288056206088993e-06,
+      "loss": 1.2153,
+      "step": 56
+    },
+    {
+      "epoch": 0.013372434017595308,
+      "grad_norm": 6.491778580398843,
+      "learning_rate": 1.3114754098360657e-06,
+      "loss": 1.3016,
+      "step": 57
+    },
+    {
+      "epoch": 0.013607038123167155,
+      "grad_norm": 6.033267561982283,
+      "learning_rate": 1.3348946135831383e-06,
+      "loss": 1.2461,
+      "step": 58
+    },
+    {
+      "epoch": 0.013841642228739003,
+      "grad_norm": 3.059255577392171,
+      "learning_rate": 1.358313817330211e-06,
+      "loss": 1.2672,
+      "step": 59
+    },
+    {
+      "epoch": 0.01407624633431085,
+      "grad_norm": 3.3358502094614546,
+      "learning_rate": 1.3817330210772834e-06,
+      "loss": 1.1633,
+      "step": 60
+    },
+    {
+      "epoch": 0.014310850439882698,
+      "grad_norm": 4.572746088964143,
+      "learning_rate": 1.405152224824356e-06,
+      "loss": 1.2182,
+      "step": 61
+    },
+    {
+      "epoch": 0.014545454545454545,
+      "grad_norm": 5.754073739243907,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.1733,
+      "step": 62
+    },
+    {
+      "epoch": 0.014780058651026393,
+      "grad_norm": 4.461948662542373,
+      "learning_rate": 1.4519906323185013e-06,
+      "loss": 1.2704,
+      "step": 63
+    },
+    {
+      "epoch": 0.01501466275659824,
+      "grad_norm": 3.9199423514140728,
+      "learning_rate": 1.4754098360655739e-06,
+      "loss": 1.2226,
+      "step": 64
+    },
+    {
+      "epoch": 0.015249266862170088,
+      "grad_norm": 4.243247745044473,
+      "learning_rate": 1.4988290398126465e-06,
+      "loss": 1.1889,
+      "step": 65
+    },
+    {
+      "epoch": 0.015483870967741935,
+      "grad_norm": 5.197052200876362,
+      "learning_rate": 1.5222482435597192e-06,
+      "loss": 1.1651,
+      "step": 66
+    },
+    {
+      "epoch": 0.015718475073313785,
+      "grad_norm": 5.624190656276907,
+      "learning_rate": 1.5456674473067918e-06,
+      "loss": 1.2036,
+      "step": 67
+    },
+    {
+      "epoch": 0.01595307917888563,
+      "grad_norm": 4.622812813704082,
+      "learning_rate": 1.5690866510538644e-06,
+      "loss": 1.1532,
+      "step": 68
+    },
+    {
+      "epoch": 0.01618768328445748,
+      "grad_norm": 8.192368911444243,
+      "learning_rate": 1.592505854800937e-06,
+      "loss": 1.2101,
+      "step": 69
+    },
+    {
+      "epoch": 0.016422287390029325,
+      "grad_norm": 12.034619839284026,
+      "learning_rate": 1.6159250585480097e-06,
+      "loss": 1.1385,
+      "step": 70
+    },
+    {
+      "epoch": 0.016656891495601175,
+      "grad_norm": 6.585215889253293,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 1.102,
+      "step": 71
+    },
+    {
+      "epoch": 0.01689149560117302,
+      "grad_norm": 6.77462634029134,
+      "learning_rate": 1.6627634660421545e-06,
+      "loss": 1.1775,
+      "step": 72
+    },
+    {
+      "epoch": 0.01712609970674487,
+      "grad_norm": 2.7854183249596143,
+      "learning_rate": 1.6861826697892272e-06,
+      "loss": 1.0731,
+      "step": 73
+    },
+    {
+      "epoch": 0.017360703812316716,
+      "grad_norm": 7.808614997644782,
+      "learning_rate": 1.7096018735362998e-06,
+      "loss": 1.1439,
+      "step": 74
+    },
+    {
+      "epoch": 0.017595307917888565,
+      "grad_norm": 4.510160348409813,
+      "learning_rate": 1.7330210772833724e-06,
+      "loss": 1.104,
+      "step": 75
+    },
+    {
+      "epoch": 0.01782991202346041,
+      "grad_norm": 5.556851183428352,
+      "learning_rate": 1.756440281030445e-06,
+      "loss": 1.1709,
+      "step": 76
+    },
+    {
+      "epoch": 0.01806451612903226,
+      "grad_norm": 3.139324465797069,
+      "learning_rate": 1.7798594847775177e-06,
+      "loss": 1.1606,
+      "step": 77
+    },
+    {
+      "epoch": 0.018299120234604106,
+      "grad_norm": 3.5137450847721152,
+      "learning_rate": 1.8032786885245903e-06,
+      "loss": 1.0535,
+      "step": 78
+    },
+    {
+      "epoch": 0.018533724340175955,
+      "grad_norm": 3.213645933277671,
+      "learning_rate": 1.826697892271663e-06,
+      "loss": 1.0311,
+      "step": 79
+    },
+    {
+      "epoch": 0.0187683284457478,
+      "grad_norm": 4.812998571795422,
+      "learning_rate": 1.8501170960187356e-06,
+      "loss": 1.0272,
+      "step": 80
+    },
+    {
+      "epoch": 0.01900293255131965,
+      "grad_norm": 2.6303156249950534,
+      "learning_rate": 1.8735362997658082e-06,
+      "loss": 1.1085,
+      "step": 81
+    },
+    {
+      "epoch": 0.019237536656891496,
+      "grad_norm": 3.0519113832521416,
+      "learning_rate": 1.8969555035128806e-06,
+      "loss": 0.9947,
+      "step": 82
+    },
+    {
+      "epoch": 0.019472140762463345,
+      "grad_norm": 3.442196083590346,
+      "learning_rate": 1.9203747072599533e-06,
+      "loss": 1.0389,
+      "step": 83
+    },
+    {
+      "epoch": 0.01970674486803519,
+      "grad_norm": 3.451466998541888,
+      "learning_rate": 1.9437939110070257e-06,
+      "loss": 1.0224,
+      "step": 84
+    },
+    {
+      "epoch": 0.01994134897360704,
+      "grad_norm": 6.902751613310287,
+      "learning_rate": 1.9672131147540985e-06,
+      "loss": 0.9854,
+      "step": 85
+    },
+    {
+      "epoch": 0.020175953079178886,
+      "grad_norm": 2.1334082210475227,
+      "learning_rate": 1.990632318501171e-06,
+      "loss": 0.9993,
+      "step": 86
+    },
+    {
+      "epoch": 0.020410557184750735,
+      "grad_norm": 4.796576233245253,
+      "learning_rate": 2.0140515222482438e-06,
+      "loss": 1.085,
+      "step": 87
+    },
+    {
+      "epoch": 0.02064516129032258,
+      "grad_norm": 4.0395890106166,
+      "learning_rate": 2.037470725995316e-06,
+      "loss": 1.0382,
+      "step": 88
+    },
+    {
+      "epoch": 0.02087976539589443,
+      "grad_norm": 3.497753662084459,
+      "learning_rate": 2.060889929742389e-06,
+      "loss": 1.0052,
+      "step": 89
+    },
+    {
+      "epoch": 0.021114369501466276,
+      "grad_norm": 3.375552138252609,
+      "learning_rate": 2.0843091334894615e-06,
+      "loss": 0.9285,
+      "step": 90
+    },
+    {
+      "epoch": 0.021348973607038125,
+      "grad_norm": 4.70530115072386,
+      "learning_rate": 2.1077283372365343e-06,
+      "loss": 1.03,
+      "step": 91
+    },
+    {
+      "epoch": 0.02158357771260997,
+      "grad_norm": 8.331460010851679,
+      "learning_rate": 2.1311475409836067e-06,
+      "loss": 1.0102,
+      "step": 92
+    },
+    {
+      "epoch": 0.02181818181818182,
+      "grad_norm": 3.334930764388384,
+      "learning_rate": 2.154566744730679e-06,
+      "loss": 0.9987,
+      "step": 93
+    },
+    {
+      "epoch": 0.022052785923753666,
+      "grad_norm": 3.4622456034454814,
+      "learning_rate": 2.177985948477752e-06,
+      "loss": 0.9862,
+      "step": 94
+    },
+    {
+      "epoch": 0.022287390029325515,
+      "grad_norm": 2.9998967371732967,
+      "learning_rate": 2.2014051522248244e-06,
+      "loss": 0.9958,
+      "step": 95
+    },
+    {
+      "epoch": 0.02252199413489736,
+      "grad_norm": 23.01438365400511,
+      "learning_rate": 2.2248243559718973e-06,
+      "loss": 0.9896,
+      "step": 96
+    },
+    {
+      "epoch": 0.02275659824046921,
+      "grad_norm": 2.914651920561855,
+      "learning_rate": 2.2482435597189697e-06,
+      "loss": 0.9177,
+      "step": 97
+    },
+    {
+      "epoch": 0.022991202346041056,
+      "grad_norm": 5.57426785924633,
+      "learning_rate": 2.2716627634660425e-06,
+      "loss": 0.9969,
+      "step": 98
+    },
+    {
+      "epoch": 0.023225806451612905,
+      "grad_norm": 3.237900155474082,
+      "learning_rate": 2.295081967213115e-06,
+      "loss": 0.9954,
+      "step": 99
+    },
+    {
+      "epoch": 0.02346041055718475,
+      "grad_norm": 2.291976229975076,
+      "learning_rate": 2.3185011709601878e-06,
+      "loss": 0.9271,
+      "step": 100
+    },
+    {
+      "epoch": 0.0236950146627566,
+      "grad_norm": 3.370251844607358,
+      "learning_rate": 2.34192037470726e-06,
+      "loss": 0.9272,
+      "step": 101
+    },
+    {
+      "epoch": 0.023929618768328446,
+      "grad_norm": 2.480256748243063,
+      "learning_rate": 2.365339578454333e-06,
+      "loss": 0.9075,
+      "step": 102
+    },
+    {
+      "epoch": 0.024164222873900295,
+      "grad_norm": 2.6163399629836546,
+      "learning_rate": 2.388758782201405e-06,
+      "loss": 0.9246,
+      "step": 103
+    },
+    {
+      "epoch": 0.02439882697947214,
+      "grad_norm": 6.052092050155686,
+      "learning_rate": 2.412177985948478e-06,
+      "loss": 1.0002,
+      "step": 104
+    },
+    {
+      "epoch": 0.02463343108504399,
+      "grad_norm": 2.899665607986737,
+      "learning_rate": 2.4355971896955503e-06,
+      "loss": 0.9357,
+      "step": 105
+    },
+    {
+      "epoch": 0.024868035190615836,
+      "grad_norm": 19.84317921055058,
+      "learning_rate": 2.459016393442623e-06,
+      "loss": 0.9558,
+      "step": 106
+    },
+    {
+      "epoch": 0.025102639296187685,
+      "grad_norm": 2.698172065560486,
+      "learning_rate": 2.4824355971896956e-06,
+      "loss": 0.9606,
+      "step": 107
+    },
+    {
+      "epoch": 0.02533724340175953,
+      "grad_norm": 2.433886867758403,
+      "learning_rate": 2.5058548009367684e-06,
+      "loss": 0.9222,
+      "step": 108
+    },
+    {
+      "epoch": 0.02557184750733138,
+      "grad_norm": 2.2816602174764893,
+      "learning_rate": 2.529274004683841e-06,
+      "loss": 0.9257,
+      "step": 109
+    },
+    {
+      "epoch": 0.025806451612903226,
+      "grad_norm": 4.635696388912117,
+      "learning_rate": 2.5526932084309137e-06,
+      "loss": 0.9065,
+      "step": 110
+    },
+    {
+      "epoch": 0.026041055718475075,
+      "grad_norm": 3.0329036230732744,
+      "learning_rate": 2.576112412177986e-06,
+      "loss": 0.895,
+      "step": 111
+    },
+    {
+      "epoch": 0.02627565982404692,
+      "grad_norm": 2.5498245911594104,
+      "learning_rate": 2.599531615925059e-06,
+      "loss": 0.912,
+      "step": 112
+    },
+    {
+      "epoch": 0.02651026392961877,
+      "grad_norm": 2.45877636159667,
+      "learning_rate": 2.6229508196721314e-06,
+      "loss": 0.9333,
+      "step": 113
+    },
+    {
+      "epoch": 0.026744868035190616,
+      "grad_norm": 20.90236117147583,
+      "learning_rate": 2.646370023419204e-06,
+      "loss": 0.8683,
+      "step": 114
+    },
+    {
+      "epoch": 0.026979472140762465,
+      "grad_norm": 2.359657953967608,
+      "learning_rate": 2.6697892271662766e-06,
+      "loss": 0.9192,
+      "step": 115
+    },
+    {
+      "epoch": 0.02721407624633431,
+      "grad_norm": 2.369533504919403,
+      "learning_rate": 2.6932084309133495e-06,
+      "loss": 0.8735,
+      "step": 116
+    },
+    {
+      "epoch": 0.02744868035190616,
+      "grad_norm": 13.020543014935672,
+      "learning_rate": 2.716627634660422e-06,
+      "loss": 0.9399,
+      "step": 117
+    },
+    {
+      "epoch": 0.027683284457478006,
+      "grad_norm": 2.387279361264047,
+      "learning_rate": 2.7400468384074947e-06,
+      "loss": 0.8968,
+      "step": 118
+    },
+    {
+      "epoch": 0.027917888563049855,
+      "grad_norm": 3.089762607316902,
+      "learning_rate": 2.7634660421545667e-06,
+      "loss": 0.8932,
+      "step": 119
+    },
+    {
+      "epoch": 0.0281524926686217,
+      "grad_norm": 3.9455101031754864,
+      "learning_rate": 2.786885245901639e-06,
+      "loss": 0.8669,
+      "step": 120
+    },
+    {
+      "epoch": 0.02838709677419355,
+      "grad_norm": 2.2155994106074215,
+      "learning_rate": 2.810304449648712e-06,
+      "loss": 0.8372,
+      "step": 121
+    },
+    {
+      "epoch": 0.028621700879765396,
+      "grad_norm": 1.9575993824588998,
+      "learning_rate": 2.8337236533957844e-06,
+      "loss": 0.9265,
+      "step": 122
+    },
+    {
+      "epoch": 0.028856304985337245,
+      "grad_norm": 3.235759284809253,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.8679,
+      "step": 123
+    },
+    {
+      "epoch": 0.02909090909090909,
+      "grad_norm": 3.181817998484809,
+      "learning_rate": 2.8805620608899297e-06,
+      "loss": 0.9384,
+      "step": 124
+    },
+    {
+      "epoch": 0.02932551319648094,
+      "grad_norm": 3.6594983240678904,
+      "learning_rate": 2.9039812646370025e-06,
+      "loss": 0.8532,
+      "step": 125
+    },
+    {
+      "epoch": 0.029560117302052786,
+      "grad_norm": 2.3937017582314404,
+      "learning_rate": 2.927400468384075e-06,
+      "loss": 0.8876,
+      "step": 126
+    },
+    {
+      "epoch": 0.029794721407624635,
+      "grad_norm": 2.3734098946643485,
+      "learning_rate": 2.9508196721311478e-06,
+      "loss": 0.874,
+      "step": 127
+    },
+    {
+      "epoch": 0.03002932551319648,
+      "grad_norm": 2.4713426475904448,
+      "learning_rate": 2.97423887587822e-06,
+      "loss": 0.9142,
+      "step": 128
+    },
+    {
+      "epoch": 0.03026392961876833,
+      "grad_norm": 3.397728713320515,
+      "learning_rate": 2.997658079625293e-06,
+      "loss": 0.915,
+      "step": 129
+    },
+    {
+      "epoch": 0.030498533724340176,
+      "grad_norm": 2.0874593525371723,
+      "learning_rate": 3.0210772833723655e-06,
+      "loss": 0.9252,
+      "step": 130
+    },
+    {
+      "epoch": 0.030733137829912025,
+      "grad_norm": 6.3434303421296026,
+      "learning_rate": 3.0444964871194383e-06,
+      "loss": 0.8908,
+      "step": 131
+    },
+    {
+      "epoch": 0.03096774193548387,
+      "grad_norm": 57.7898256836101,
+      "learning_rate": 3.0679156908665107e-06,
+      "loss": 0.8971,
+      "step": 132
+    },
+    {
+      "epoch": 0.03120234604105572,
+      "grad_norm": 1.711252751491627,
+      "learning_rate": 3.0913348946135836e-06,
+      "loss": 0.9249,
+      "step": 133
+    },
+    {
+      "epoch": 0.03143695014662757,
+      "grad_norm": 5.0126693472094646,
+      "learning_rate": 3.114754098360656e-06,
+      "loss": 0.875,
+      "step": 134
+    },
+    {
+      "epoch": 0.03167155425219941,
+      "grad_norm": 5.85541088273383,
+      "learning_rate": 3.138173302107729e-06,
+      "loss": 0.9186,
+      "step": 135
+    },
+    {
+      "epoch": 0.03190615835777126,
+      "grad_norm": 3.0975868523648717,
+      "learning_rate": 3.1615925058548013e-06,
+      "loss": 0.9148,
+      "step": 136
+    },
+    {
+      "epoch": 0.03214076246334311,
+      "grad_norm": 2.606556275041088,
+      "learning_rate": 3.185011709601874e-06,
+      "loss": 0.8415,
+      "step": 137
+    },
+    {
+      "epoch": 0.03237536656891496,
+      "grad_norm": 2.484589455292374,
+      "learning_rate": 3.2084309133489465e-06,
+      "loss": 0.9124,
+      "step": 138
+    },
+    {
+      "epoch": 0.0326099706744868,
+      "grad_norm": 2.6972597309695545,
+      "learning_rate": 3.2318501170960194e-06,
+      "loss": 0.8407,
+      "step": 139
+    },
+    {
+      "epoch": 0.03284457478005865,
+      "grad_norm": 3.375807212236533,
+      "learning_rate": 3.2552693208430914e-06,
+      "loss": 0.8976,
+      "step": 140
+    },
+    {
+      "epoch": 0.0330791788856305,
+      "grad_norm": 20.40922837350589,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 0.7912,
+      "step": 141
+    },
+    {
+      "epoch": 0.03331378299120235,
+      "grad_norm": 3.9405236611888794,
+      "learning_rate": 3.3021077283372366e-06,
+      "loss": 0.8665,
+      "step": 142
+    },
+    {
+      "epoch": 0.03354838709677419,
+      "grad_norm": 2.7508052811782795,
+      "learning_rate": 3.325526932084309e-06,
+      "loss": 0.8238,
+      "step": 143
+    },
+    {
+      "epoch": 0.03378299120234604,
+      "grad_norm": 5.017113290178286,
+      "learning_rate": 3.348946135831382e-06,
+      "loss": 0.8803,
+      "step": 144
+    },
+    {
+      "epoch": 0.03401759530791789,
+      "grad_norm": 2.180732748232138,
+      "learning_rate": 3.3723653395784543e-06,
+      "loss": 0.8724,
+      "step": 145
+    },
+    {
+      "epoch": 0.03425219941348974,
+      "grad_norm": 3.4586882050068692,
+      "learning_rate": 3.395784543325527e-06,
+      "loss": 0.7973,
+      "step": 146
+    },
+    {
+      "epoch": 0.03448680351906158,
+      "grad_norm": 6.8618732473004265,
+      "learning_rate": 3.4192037470725996e-06,
+      "loss": 0.8987,
+      "step": 147
+    },
+    {
+      "epoch": 0.03472140762463343,
+      "grad_norm": 1.6297016411593837,
+      "learning_rate": 3.4426229508196724e-06,
+      "loss": 0.8103,
+      "step": 148
+    },
+    {
+      "epoch": 0.03495601173020528,
+      "grad_norm": 2.7847740603147257,
+      "learning_rate": 3.466042154566745e-06,
+      "loss": 0.8684,
+      "step": 149
+    },
+    {
+      "epoch": 0.03519061583577713,
+      "grad_norm": 1.7603035059230738,
+      "learning_rate": 3.4894613583138177e-06,
+      "loss": 0.8308,
+      "step": 150
+    },
+    {
+      "epoch": 0.03542521994134897,
+      "grad_norm": 1.6040986957520253,
+      "learning_rate": 3.51288056206089e-06,
+      "loss": 0.8349,
+      "step": 151
+    },
+    {
+      "epoch": 0.03565982404692082,
+      "grad_norm": 3.1963778933146907,
+      "learning_rate": 3.536299765807963e-06,
+      "loss": 0.88,
+      "step": 152
+    },
+    {
+      "epoch": 0.03589442815249267,
+      "grad_norm": 3.3496492672952014,
+      "learning_rate": 3.5597189695550354e-06,
+      "loss": 0.909,
+      "step": 153
+    },
+    {
+      "epoch": 0.03612903225806452,
+      "grad_norm": 2.455904571079815,
+      "learning_rate": 3.583138173302108e-06,
+      "loss": 0.8191,
+      "step": 154
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 1.5896943471703224,
+      "learning_rate": 3.6065573770491806e-06,
+      "loss": 0.9028,
+      "step": 155
+    },
+    {
+      "epoch": 0.03659824046920821,
+      "grad_norm": 2.6915939177699078,
+      "learning_rate": 3.6299765807962535e-06,
+      "loss": 0.8393,
+      "step": 156
+    },
+    {
+      "epoch": 0.03683284457478006,
+      "grad_norm": 7.175761101203019,
+      "learning_rate": 3.653395784543326e-06,
+      "loss": 0.8643,
+      "step": 157
+    },
+    {
+      "epoch": 0.03706744868035191,
+      "grad_norm": 2.2229527474502735,
+      "learning_rate": 3.6768149882903987e-06,
+      "loss": 0.8513,
+      "step": 158
+    },
+    {
+      "epoch": 0.03730205278592375,
+      "grad_norm": 2.9865753215500197,
+      "learning_rate": 3.700234192037471e-06,
+      "loss": 0.8868,
+      "step": 159
+    },
+    {
+      "epoch": 0.0375366568914956,
+      "grad_norm": 2.9734325151644554,
+      "learning_rate": 3.723653395784544e-06,
+      "loss": 0.8532,
+      "step": 160
+    },
+    {
+      "epoch": 0.03777126099706745,
+      "grad_norm": 1.8588526791176345,
+      "learning_rate": 3.7470725995316164e-06,
+      "loss": 0.819,
+      "step": 161
+    },
+    {
+      "epoch": 0.0380058651026393,
+      "grad_norm": 8.645485871225635,
+      "learning_rate": 3.7704918032786884e-06,
+      "loss": 0.8521,
+      "step": 162
+    },
+    {
+      "epoch": 0.03824046920821114,
+      "grad_norm": 3.002196055017224,
+      "learning_rate": 3.7939110070257613e-06,
+      "loss": 0.8084,
+      "step": 163
+    },
+    {
+      "epoch": 0.03847507331378299,
+      "grad_norm": 3.2302818373533966,
+      "learning_rate": 3.817330210772834e-06,
+      "loss": 0.9311,
+      "step": 164
+    },
+    {
+      "epoch": 0.03870967741935484,
+      "grad_norm": 3.6213961241692036,
+      "learning_rate": 3.8407494145199065e-06,
+      "loss": 0.844,
+      "step": 165
+    },
+    {
+      "epoch": 0.03894428152492669,
+      "grad_norm": 2.7852526099599224,
+      "learning_rate": 3.864168618266979e-06,
+      "loss": 0.863,
+      "step": 166
+    },
+    {
+      "epoch": 0.03917888563049853,
+      "grad_norm": 4.334673507783282,
+      "learning_rate": 3.887587822014051e-06,
+      "loss": 0.8363,
+      "step": 167
+    },
+    {
+      "epoch": 0.03941348973607038,
+      "grad_norm": 2.048656770606946,
+      "learning_rate": 3.911007025761124e-06,
+      "loss": 0.8315,
+      "step": 168
+    },
+    {
+      "epoch": 0.03964809384164223,
+      "grad_norm": 4.177636179080158,
+      "learning_rate": 3.934426229508197e-06,
+      "loss": 0.7962,
+      "step": 169
+    },
+    {
+      "epoch": 0.03988269794721408,
+      "grad_norm": 1.8523171799760045,
+      "learning_rate": 3.95784543325527e-06,
+      "loss": 0.8058,
+      "step": 170
+    },
+    {
+      "epoch": 0.04011730205278592,
+      "grad_norm": 2.9191774008849567,
+      "learning_rate": 3.981264637002342e-06,
+      "loss": 0.876,
+      "step": 171
+    },
+    {
+      "epoch": 0.04035190615835777,
+      "grad_norm": 2.530425709085943,
+      "learning_rate": 4.004683840749415e-06,
+      "loss": 0.8415,
+      "step": 172
+    },
+    {
+      "epoch": 0.04058651026392962,
+      "grad_norm": 2.918808157433132,
+      "learning_rate": 4.0281030444964876e-06,
+      "loss": 0.8353,
+      "step": 173
+    },
+    {
+      "epoch": 0.04082111436950147,
+      "grad_norm": 2.37263511816868,
+      "learning_rate": 4.05152224824356e-06,
+      "loss": 0.856,
+      "step": 174
+    },
+    {
+      "epoch": 0.04105571847507331,
+      "grad_norm": 2.7199432299385293,
+      "learning_rate": 4.074941451990632e-06,
+      "loss": 0.8177,
+      "step": 175
+    },
+    {
+      "epoch": 0.04129032258064516,
+      "grad_norm": 1.9137699316073797,
+      "learning_rate": 4.098360655737705e-06,
+      "loss": 0.774,
+      "step": 176
+    },
+    {
+      "epoch": 0.04152492668621701,
+      "grad_norm": 1.9969742439162197,
+      "learning_rate": 4.121779859484778e-06,
+      "loss": 0.8587,
+      "step": 177
+    },
+    {
+      "epoch": 0.04175953079178886,
+      "grad_norm": 1.62573462869543,
+      "learning_rate": 4.145199063231851e-06,
+      "loss": 0.8197,
+      "step": 178
+    },
+    {
+      "epoch": 0.0419941348973607,
+      "grad_norm": 2.6789717841087635,
+      "learning_rate": 4.168618266978923e-06,
+      "loss": 0.7843,
+      "step": 179
+    },
+    {
+      "epoch": 0.04222873900293255,
+      "grad_norm": 2.8296926850030437,
+      "learning_rate": 4.192037470725996e-06,
+      "loss": 0.7829,
+      "step": 180
+    },
+    {
+      "epoch": 0.0424633431085044,
+      "grad_norm": 2.9687571177348078,
+      "learning_rate": 4.215456674473069e-06,
+      "loss": 0.8265,
+      "step": 181
+    },
+    {
+      "epoch": 0.04269794721407625,
+      "grad_norm": 8.057262005821913,
+      "learning_rate": 4.2388758782201415e-06,
+      "loss": 0.8063,
+      "step": 182
+    },
+    {
+      "epoch": 0.04293255131964809,
+      "grad_norm": 1.3309377653809522,
+      "learning_rate": 4.2622950819672135e-06,
+      "loss": 0.85,
+      "step": 183
+    },
+    {
+      "epoch": 0.04316715542521994,
+      "grad_norm": 4.796473479193148,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.8535,
+      "step": 184
+    },
+    {
+      "epoch": 0.04340175953079179,
+      "grad_norm": 2.29076687107649,
+      "learning_rate": 4.309133489461358e-06,
+      "loss": 0.8308,
+      "step": 185
+    },
+    {
+      "epoch": 0.04363636363636364,
+      "grad_norm": 1.9668639979537659,
+      "learning_rate": 4.332552693208431e-06,
+      "loss": 0.8164,
+      "step": 186
+    },
+    {
+      "epoch": 0.04387096774193548,
+      "grad_norm": 1.7492988720129186,
+      "learning_rate": 4.355971896955504e-06,
+      "loss": 0.7862,
+      "step": 187
+    },
+    {
+      "epoch": 0.04410557184750733,
+      "grad_norm": 1.902084590499895,
+      "learning_rate": 4.379391100702576e-06,
+      "loss": 0.7948,
+      "step": 188
+    },
+    {
+      "epoch": 0.04434017595307918,
+      "grad_norm": 9.860967065038771,
+      "learning_rate": 4.402810304449649e-06,
+      "loss": 0.8499,
+      "step": 189
+    },
+    {
+      "epoch": 0.04457478005865103,
+      "grad_norm": 1.5904532338079087,
+      "learning_rate": 4.426229508196722e-06,
+      "loss": 0.8706,
+      "step": 190
+    },
+    {
+      "epoch": 0.04480938416422287,
+      "grad_norm": 3.2286150229605584,
+      "learning_rate": 4.4496487119437945e-06,
+      "loss": 0.8434,
+      "step": 191
+    },
+    {
+      "epoch": 0.04504398826979472,
+      "grad_norm": 2.642953811023596,
+      "learning_rate": 4.4730679156908665e-06,
+      "loss": 0.7932,
+      "step": 192
+    },
+    {
+      "epoch": 0.04527859237536657,
+      "grad_norm": 6.708623247161702,
+      "learning_rate": 4.496487119437939e-06,
+      "loss": 0.7557,
+      "step": 193
+    },
+    {
+      "epoch": 0.04551319648093842,
+      "grad_norm": 4.245544259005578,
+      "learning_rate": 4.519906323185012e-06,
+      "loss": 0.7919,
+      "step": 194
+    },
+    {
+      "epoch": 0.04574780058651026,
+      "grad_norm": 2.2209264855596773,
+      "learning_rate": 4.543325526932085e-06,
+      "loss": 0.8008,
+      "step": 195
+    },
+    {
+      "epoch": 0.04598240469208211,
+      "grad_norm": 4.615319317159054,
+      "learning_rate": 4.566744730679157e-06,
+      "loss": 0.7972,
+      "step": 196
+    },
+    {
+      "epoch": 0.04621700879765396,
+      "grad_norm": 2.4016647930607617,
+      "learning_rate": 4.59016393442623e-06,
+      "loss": 0.7921,
+      "step": 197
+    },
+    {
+      "epoch": 0.04645161290322581,
+      "grad_norm": 2.1065053768148054,
+      "learning_rate": 4.613583138173303e-06,
+      "loss": 0.8194,
+      "step": 198
+    },
+    {
+      "epoch": 0.04668621700879765,
+      "grad_norm": 1.3854497678939341,
+      "learning_rate": 4.6370023419203756e-06,
+      "loss": 0.8471,
+      "step": 199
+    },
+    {
+      "epoch": 0.0469208211143695,
+      "grad_norm": 6.859498238314571,
+      "learning_rate": 4.6604215456674476e-06,
+      "loss": 0.8511,
+      "step": 200
+    },
+    {
+      "epoch": 0.04715542521994135,
+      "grad_norm": 2.2870014320129126,
+      "learning_rate": 4.68384074941452e-06,
+      "loss": 0.778,
+      "step": 201
+    },
+    {
+      "epoch": 0.0473900293255132,
+      "grad_norm": 3.412042568878162,
+      "learning_rate": 4.707259953161593e-06,
+      "loss": 0.7838,
+      "step": 202
+    },
+    {
+      "epoch": 0.04762463343108504,
+      "grad_norm": 1.5441510074320814,
+      "learning_rate": 4.730679156908666e-06,
+      "loss": 0.7667,
+      "step": 203
+    },
+    {
+      "epoch": 0.04785923753665689,
+      "grad_norm": 2.4279262506292065,
+      "learning_rate": 4.754098360655738e-06,
+      "loss": 0.7373,
+      "step": 204
+    },
+    {
+      "epoch": 0.04809384164222874,
+      "grad_norm": 2.68155491287197,
+      "learning_rate": 4.77751756440281e-06,
+      "loss": 0.7851,
+      "step": 205
+    },
+    {
+      "epoch": 0.04832844574780059,
+      "grad_norm": 2.5361715948287236,
+      "learning_rate": 4.800936768149883e-06,
+      "loss": 0.7991,
+      "step": 206
+    },
+    {
+      "epoch": 0.04856304985337243,
+      "grad_norm": 1.965596504616627,
+      "learning_rate": 4.824355971896956e-06,
+      "loss": 0.754,
+      "step": 207
+    },
+    {
+      "epoch": 0.04879765395894428,
+      "grad_norm": 5.682372140740957,
+      "learning_rate": 4.847775175644029e-06,
+      "loss": 0.8054,
+      "step": 208
+    },
+    {
+      "epoch": 0.04903225806451613,
+      "grad_norm": 3.5475724826725252,
+      "learning_rate": 4.871194379391101e-06,
+      "loss": 0.8108,
+      "step": 209
+    },
+    {
+      "epoch": 0.04926686217008798,
+      "grad_norm": 2.0753730912698294,
+      "learning_rate": 4.8946135831381735e-06,
+      "loss": 0.7497,
+      "step": 210
+    },
+    {
+      "epoch": 0.04950146627565982,
+      "grad_norm": 4.063420592276077,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 0.8285,
+      "step": 211
+    },
+    {
+      "epoch": 0.04973607038123167,
+      "grad_norm": 1.7307046068411553,
+      "learning_rate": 4.941451990632319e-06,
+      "loss": 0.8044,
+      "step": 212
+    },
+    {
+      "epoch": 0.04997067448680352,
+      "grad_norm": 2.2640298230945226,
+      "learning_rate": 4.964871194379391e-06,
+      "loss": 0.7995,
+      "step": 213
+    },
+    {
+      "epoch": 0.05020527859237537,
+      "grad_norm": 2.3810018614242403,
+      "learning_rate": 4.988290398126464e-06,
+      "loss": 0.786,
+      "step": 214
+    },
+    {
+      "epoch": 0.05043988269794721,
+      "grad_norm": 4.549399873923761,
+      "learning_rate": 5.011709601873537e-06,
+      "loss": 0.7859,
+      "step": 215
+    },
+    {
+      "epoch": 0.05067448680351906,
+      "grad_norm": 2.0135449375550003,
+      "learning_rate": 5.03512880562061e-06,
+      "loss": 0.7344,
+      "step": 216
+    },
+    {
+      "epoch": 0.05090909090909091,
+      "grad_norm": 3.872319715699657,
+      "learning_rate": 5.058548009367682e-06,
+      "loss": 0.7831,
+      "step": 217
+    },
+    {
+      "epoch": 0.05114369501466276,
+      "grad_norm": 1.6872682506797199,
+      "learning_rate": 5.0819672131147545e-06,
+      "loss": 0.8038,
+      "step": 218
+    },
+    {
+      "epoch": 0.0513782991202346,
+      "grad_norm": 2.1830008681001147,
+      "learning_rate": 5.105386416861827e-06,
+      "loss": 0.7964,
+      "step": 219
+    },
+    {
+      "epoch": 0.05161290322580645,
+      "grad_norm": 1.4493570098406372,
+      "learning_rate": 5.1288056206089e-06,
+      "loss": 0.7337,
+      "step": 220
+    },
+    {
+      "epoch": 0.0518475073313783,
+      "grad_norm": 2.5689272205754294,
+      "learning_rate": 5.152224824355972e-06,
+      "loss": 0.7797,
+      "step": 221
+    },
+    {
+      "epoch": 0.05208211143695015,
+      "grad_norm": 3.389990850999057,
+      "learning_rate": 5.175644028103045e-06,
+      "loss": 0.8088,
+      "step": 222
+    },
+    {
+      "epoch": 0.05231671554252199,
+      "grad_norm": 4.273411621435788,
+      "learning_rate": 5.199063231850118e-06,
+      "loss": 0.7198,
+      "step": 223
+    },
+    {
+      "epoch": 0.05255131964809384,
+      "grad_norm": 4.9769025368159685,
+      "learning_rate": 5.222482435597191e-06,
+      "loss": 0.7838,
+      "step": 224
+    },
+    {
+      "epoch": 0.05278592375366569,
+      "grad_norm": 2.167852931175855,
+      "learning_rate": 5.245901639344263e-06,
+      "loss": 0.7534,
+      "step": 225
+    },
+    {
+      "epoch": 0.05302052785923754,
+      "grad_norm": 3.798350568648415,
+      "learning_rate": 5.2693208430913356e-06,
+      "loss": 0.7651,
+      "step": 226
+    },
+    {
+      "epoch": 0.05325513196480938,
+      "grad_norm": 1.299473999291837,
+      "learning_rate": 5.292740046838408e-06,
+      "loss": 0.7829,
+      "step": 227
+    },
+    {
+      "epoch": 0.05348973607038123,
+      "grad_norm": 3.2234237743302283,
+      "learning_rate": 5.316159250585481e-06,
+      "loss": 0.799,
+      "step": 228
+    },
+    {
+      "epoch": 0.05372434017595308,
+      "grad_norm": 8.030276193203514,
+      "learning_rate": 5.339578454332553e-06,
+      "loss": 0.7717,
+      "step": 229
+    },
+    {
+      "epoch": 0.05395894428152493,
+      "grad_norm": 2.5370284044134253,
+      "learning_rate": 5.362997658079626e-06,
+      "loss": 0.7736,
+      "step": 230
+    },
+    {
+      "epoch": 0.05419354838709677,
+      "grad_norm": 3.844656305538049,
+      "learning_rate": 5.386416861826699e-06,
+      "loss": 0.749,
+      "step": 231
+    },
+    {
+      "epoch": 0.05442815249266862,
+      "grad_norm": 1.9427868676296431,
+      "learning_rate": 5.409836065573772e-06,
+      "loss": 0.7908,
+      "step": 232
+    },
+    {
+      "epoch": 0.05466275659824047,
+      "grad_norm": 3.9911344896775107,
+      "learning_rate": 5.433255269320844e-06,
+      "loss": 0.8008,
+      "step": 233
+    },
+    {
+      "epoch": 0.05489736070381232,
+      "grad_norm": 1.9734180350151789,
+      "learning_rate": 5.456674473067917e-06,
+      "loss": 0.7454,
+      "step": 234
+    },
+    {
+      "epoch": 0.05513196480938416,
+      "grad_norm": 4.702544055041282,
+      "learning_rate": 5.4800936768149895e-06,
+      "loss": 0.7252,
+      "step": 235
+    },
+    {
+      "epoch": 0.05536656891495601,
+      "grad_norm": 2.1911754420328022,
+      "learning_rate": 5.503512880562061e-06,
+      "loss": 0.7967,
+      "step": 236
+    },
+    {
+      "epoch": 0.05560117302052786,
+      "grad_norm": 1.875015352817215,
+      "learning_rate": 5.5269320843091335e-06,
+      "loss": 0.807,
+      "step": 237
+    },
+    {
+      "epoch": 0.05583577712609971,
+      "grad_norm": 3.8966209518040555,
+      "learning_rate": 5.550351288056206e-06,
+      "loss": 0.7806,
+      "step": 238
+    },
+    {
+      "epoch": 0.05607038123167155,
+      "grad_norm": 1.4214964022874772,
+      "learning_rate": 5.573770491803278e-06,
+      "loss": 0.7544,
+      "step": 239
+    },
+    {
+      "epoch": 0.0563049853372434,
+      "grad_norm": 3.967817931858348,
+      "learning_rate": 5.597189695550351e-06,
+      "loss": 0.8101,
+      "step": 240
+    },
+    {
+      "epoch": 0.05653958944281525,
+      "grad_norm": 1.8768806953038544,
+      "learning_rate": 5.620608899297424e-06,
+      "loss": 0.7151,
+      "step": 241
+    },
+    {
+      "epoch": 0.0567741935483871,
+      "grad_norm": 1.5456485323537321,
+      "learning_rate": 5.644028103044497e-06,
+      "loss": 0.7259,
+      "step": 242
+    },
+    {
+      "epoch": 0.05700879765395894,
+      "grad_norm": 11.443111964052978,
+      "learning_rate": 5.667447306791569e-06,
+      "loss": 0.752,
+      "step": 243
+    },
+    {
+      "epoch": 0.05724340175953079,
+      "grad_norm": 2.2004989352128472,
+      "learning_rate": 5.690866510538642e-06,
+      "loss": 0.8032,
+      "step": 244
+    },
+    {
+      "epoch": 0.05747800586510264,
+      "grad_norm": 2.509396292931848,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.7954,
+      "step": 245
+    },
+    {
+      "epoch": 0.05771260997067449,
+      "grad_norm": 4.62406672031256,
+      "learning_rate": 5.737704918032787e-06,
+      "loss": 0.6891,
+      "step": 246
+    },
+    {
+      "epoch": 0.05794721407624633,
+      "grad_norm": 1.7023733051611323,
+      "learning_rate": 5.761124121779859e-06,
+      "loss": 0.7457,
+      "step": 247
+    },
+    {
+      "epoch": 0.05818181818181818,
+      "grad_norm": 2.1076480192526903,
+      "learning_rate": 5.784543325526932e-06,
+      "loss": 0.7325,
+      "step": 248
+    },
+    {
+      "epoch": 0.05841642228739003,
+      "grad_norm": 3.3089258745559524,
+      "learning_rate": 5.807962529274005e-06,
+      "loss": 0.7939,
+      "step": 249
+    },
+    {
+      "epoch": 0.05865102639296188,
+      "grad_norm": 3.148213007235479,
+      "learning_rate": 5.831381733021078e-06,
+      "loss": 0.7545,
+      "step": 250
+    },
+    {
+      "epoch": 0.05888563049853372,
+      "grad_norm": 1.6415544135274773,
+      "learning_rate": 5.85480093676815e-06,
+      "loss": 0.8057,
+      "step": 251
+    },
+    {
+      "epoch": 0.05912023460410557,
+      "grad_norm": 2.419528106327016,
+      "learning_rate": 5.878220140515223e-06,
+      "loss": 0.7841,
+      "step": 252
+    },
+    {
+      "epoch": 0.05935483870967742,
+      "grad_norm": 3.1088794443157854,
+      "learning_rate": 5.9016393442622956e-06,
+      "loss": 0.7512,
+      "step": 253
+    },
+    {
+      "epoch": 0.05958944281524927,
+      "grad_norm": 1.9587822334450857,
+      "learning_rate": 5.925058548009368e-06,
+      "loss": 0.7119,
+      "step": 254
+    },
+    {
+      "epoch": 0.05982404692082111,
+      "grad_norm": 3.5891298714306776,
+      "learning_rate": 5.94847775175644e-06,
+      "loss": 0.7205,
+      "step": 255
+    },
+    {
+      "epoch": 0.06005865102639296,
+      "grad_norm": 1.7806666652881635,
+      "learning_rate": 5.971896955503513e-06,
+      "loss": 0.7718,
+      "step": 256
+    },
+    {
+      "epoch": 0.06029325513196481,
+      "grad_norm": 1.5452906290772128,
+      "learning_rate": 5.995316159250586e-06,
+      "loss": 0.7554,
+      "step": 257
+    },
+    {
+      "epoch": 0.06052785923753666,
+      "grad_norm": 1.7300711388008159,
+      "learning_rate": 6.018735362997659e-06,
+      "loss": 0.7274,
+      "step": 258
+    },
+    {
+      "epoch": 0.0607624633431085,
+      "grad_norm": 3.4093157523447273,
+      "learning_rate": 6.042154566744731e-06,
+      "loss": 0.7214,
+      "step": 259
+    },
+    {
+      "epoch": 0.06099706744868035,
+      "grad_norm": 1.6095409898986446,
+      "learning_rate": 6.065573770491804e-06,
+      "loss": 0.7483,
+      "step": 260
+    },
+    {
+      "epoch": 0.0612316715542522,
+      "grad_norm": 2.6284860137387382,
+      "learning_rate": 6.088992974238877e-06,
+      "loss": 0.7856,
+      "step": 261
+    },
+    {
+      "epoch": 0.06146627565982405,
+      "grad_norm": 2.0697469231199914,
+      "learning_rate": 6.1124121779859495e-06,
+      "loss": 0.727,
+      "step": 262
+    },
+    {
+      "epoch": 0.06170087976539589,
+      "grad_norm": 1.7751716037540561,
+      "learning_rate": 6.1358313817330215e-06,
+      "loss": 0.7186,
+      "step": 263
+    },
+    {
+      "epoch": 0.06193548387096774,
+      "grad_norm": 10.737798739408827,
+      "learning_rate": 6.159250585480094e-06,
+      "loss": 0.8159,
+      "step": 264
+    },
+    {
+      "epoch": 0.06217008797653959,
+      "grad_norm": 4.095125238646114,
+      "learning_rate": 6.182669789227167e-06,
+      "loss": 0.7606,
+      "step": 265
+    },
+    {
+      "epoch": 0.06240469208211144,
+      "grad_norm": 2.533566695384937,
+      "learning_rate": 6.20608899297424e-06,
+      "loss": 0.7428,
+      "step": 266
+    },
+    {
+      "epoch": 0.06263929618768328,
+      "grad_norm": 3.9375956780165087,
+      "learning_rate": 6.229508196721312e-06,
+      "loss": 0.7536,
+      "step": 267
+    },
+    {
+      "epoch": 0.06287390029325514,
+      "grad_norm": 3.587867577881465,
+      "learning_rate": 6.252927400468385e-06,
+      "loss": 0.8013,
+      "step": 268
+    },
+    {
+      "epoch": 0.06310850439882698,
+      "grad_norm": 1.9828841441463887,
+      "learning_rate": 6.276346604215458e-06,
+      "loss": 0.6658,
+      "step": 269
+    },
+    {
+      "epoch": 0.06334310850439882,
+      "grad_norm": 2.532887381542946,
+      "learning_rate": 6.2997658079625305e-06,
+      "loss": 0.7835,
+      "step": 270
+    },
+    {
+      "epoch": 0.06357771260997068,
+      "grad_norm": 3.5378104561825925,
+      "learning_rate": 6.3231850117096025e-06,
+      "loss": 0.7792,
+      "step": 271
+    },
+    {
+      "epoch": 0.06381231671554252,
+      "grad_norm": 8.346765497137895,
+      "learning_rate": 6.346604215456675e-06,
+      "loss": 0.6928,
+      "step": 272
+    },
+    {
+      "epoch": 0.06404692082111436,
+      "grad_norm": 1.874277091383251,
+      "learning_rate": 6.370023419203748e-06,
+      "loss": 0.7879,
+      "step": 273
+    },
+    {
+      "epoch": 0.06428152492668622,
+      "grad_norm": 2.5695005602644456,
+      "learning_rate": 6.393442622950821e-06,
+      "loss": 0.7346,
+      "step": 274
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 1.8923461355980111,
+      "learning_rate": 6.416861826697893e-06,
+      "loss": 0.7706,
+      "step": 275
+    },
+    {
+      "epoch": 0.06475073313782992,
+      "grad_norm": 2.345792784341501,
+      "learning_rate": 6.440281030444966e-06,
+      "loss": 0.7925,
+      "step": 276
+    },
+    {
+      "epoch": 0.06498533724340176,
+      "grad_norm": 1.3700649626976735,
+      "learning_rate": 6.463700234192039e-06,
+      "loss": 0.7343,
+      "step": 277
+    },
+    {
+      "epoch": 0.0652199413489736,
+      "grad_norm": 4.2957117515820915,
+      "learning_rate": 6.487119437939111e-06,
+      "loss": 0.7899,
+      "step": 278
+    },
+    {
+      "epoch": 0.06545454545454546,
+      "grad_norm": 1.8088841845847408,
+      "learning_rate": 6.510538641686183e-06,
+      "loss": 0.7432,
+      "step": 279
+    },
+    {
+      "epoch": 0.0656891495601173,
+      "grad_norm": 1.9353168320863243,
+      "learning_rate": 6.5339578454332556e-06,
+      "loss": 0.7195,
+      "step": 280
+    },
+    {
+      "epoch": 0.06592375366568914,
+      "grad_norm": 2.0971904069466185,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 0.797,
+      "step": 281
+    },
+    {
+      "epoch": 0.066158357771261,
+      "grad_norm": 2.9785324231253605,
+      "learning_rate": 6.5807962529274e-06,
+      "loss": 0.7344,
+      "step": 282
+    },
+    {
+      "epoch": 0.06639296187683284,
+      "grad_norm": 3.4524226360583423,
+      "learning_rate": 6.604215456674473e-06,
+      "loss": 0.7101,
+      "step": 283
+    },
+    {
+      "epoch": 0.0666275659824047,
+      "grad_norm": 2.5242898267713128,
+      "learning_rate": 6.627634660421546e-06,
+      "loss": 0.753,
+      "step": 284
+    },
+    {
+      "epoch": 0.06686217008797654,
+      "grad_norm": 1.4934534118028575,
+      "learning_rate": 6.651053864168618e-06,
+      "loss": 0.7951,
+      "step": 285
+    },
+    {
+      "epoch": 0.06709677419354838,
+      "grad_norm": 5.172684338844145,
+      "learning_rate": 6.674473067915691e-06,
+      "loss": 0.7276,
+      "step": 286
+    },
+    {
+      "epoch": 0.06733137829912024,
+      "grad_norm": 13.510287957162044,
+      "learning_rate": 6.697892271662764e-06,
+      "loss": 0.7135,
+      "step": 287
+    },
+    {
+      "epoch": 0.06756598240469208,
+      "grad_norm": 2.9564063625674395,
+      "learning_rate": 6.721311475409837e-06,
+      "loss": 0.8378,
+      "step": 288
+    },
+    {
+      "epoch": 0.06780058651026392,
+      "grad_norm": 1.859065502082555,
+      "learning_rate": 6.744730679156909e-06,
+      "loss": 0.7448,
+      "step": 289
+    },
+    {
+      "epoch": 0.06803519061583578,
+      "grad_norm": 4.68045877292896,
+      "learning_rate": 6.7681498829039815e-06,
+      "loss": 0.6803,
+      "step": 290
+    },
+    {
+      "epoch": 0.06826979472140762,
+      "grad_norm": 3.0530974991612134,
+      "learning_rate": 6.791569086651054e-06,
+      "loss": 0.7392,
+      "step": 291
+    },
+    {
+      "epoch": 0.06850439882697948,
+      "grad_norm": 2.9913463833605274,
+      "learning_rate": 6.814988290398127e-06,
+      "loss": 0.7589,
+      "step": 292
+    },
+    {
+      "epoch": 0.06873900293255132,
+      "grad_norm": 2.1123062226147153,
+      "learning_rate": 6.838407494145199e-06,
+      "loss": 0.7948,
+      "step": 293
+    },
+    {
+      "epoch": 0.06897360703812316,
+      "grad_norm": 12.645487621717868,
+      "learning_rate": 6.861826697892272e-06,
+      "loss": 0.7567,
+      "step": 294
+    },
+    {
+      "epoch": 0.06920821114369502,
+      "grad_norm": 1.465655431460024,
+      "learning_rate": 6.885245901639345e-06,
+      "loss": 0.7775,
+      "step": 295
+    },
+    {
+      "epoch": 0.06944281524926686,
+      "grad_norm": 5.270313079861416,
+      "learning_rate": 6.908665105386418e-06,
+      "loss": 0.7425,
+      "step": 296
+    },
+    {
+      "epoch": 0.0696774193548387,
+      "grad_norm": 11.365986847972149,
+      "learning_rate": 6.93208430913349e-06,
+      "loss": 0.7524,
+      "step": 297
+    },
+    {
+      "epoch": 0.06991202346041056,
+      "grad_norm": 1.788096518614783,
+      "learning_rate": 6.9555035128805625e-06,
+      "loss": 0.676,
+      "step": 298
+    },
+    {
+      "epoch": 0.0701466275659824,
+      "grad_norm": 2.7086246514533583,
+      "learning_rate": 6.978922716627635e-06,
+      "loss": 0.7226,
+      "step": 299
+    },
+    {
+      "epoch": 0.07038123167155426,
+      "grad_norm": 1.3262605972809178,
+      "learning_rate": 7.002341920374708e-06,
+      "loss": 0.7954,
+      "step": 300
+    },
+    {
+      "epoch": 0.0706158357771261,
+      "grad_norm": 15.440608301240019,
+      "learning_rate": 7.02576112412178e-06,
+      "loss": 0.7638,
+      "step": 301
+    },
+    {
+      "epoch": 0.07085043988269794,
+      "grad_norm": 2.14628459786964,
+      "learning_rate": 7.049180327868853e-06,
+      "loss": 0.7971,
+      "step": 302
+    },
+    {
+      "epoch": 0.0710850439882698,
+      "grad_norm": 2.787143616061244,
+      "learning_rate": 7.072599531615926e-06,
+      "loss": 0.7924,
+      "step": 303
+    },
+    {
+      "epoch": 0.07131964809384164,
+      "grad_norm": 6.82002908629792,
+      "learning_rate": 7.096018735362999e-06,
+      "loss": 0.7978,
+      "step": 304
+    },
+    {
+      "epoch": 0.07155425219941348,
+      "grad_norm": 1.8496761797035541,
+      "learning_rate": 7.119437939110071e-06,
+      "loss": 0.7508,
+      "step": 305
+    },
+    {
+      "epoch": 0.07178885630498534,
+      "grad_norm": 4.338350528797399,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.7585,
+      "step": 306
+    },
+    {
+      "epoch": 0.07202346041055718,
+      "grad_norm": 4.857979418553227,
+      "learning_rate": 7.166276346604216e-06,
+      "loss": 0.6679,
+      "step": 307
+    },
+    {
+      "epoch": 0.07225806451612904,
+      "grad_norm": 1.7348608947080373,
+      "learning_rate": 7.189695550351289e-06,
+      "loss": 0.6959,
+      "step": 308
+    },
+    {
+      "epoch": 0.07249266862170088,
+      "grad_norm": 2.2509485892266348,
+      "learning_rate": 7.213114754098361e-06,
+      "loss": 0.7304,
+      "step": 309
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 1.5645365862972824,
+      "learning_rate": 7.236533957845434e-06,
+      "loss": 0.7468,
+      "step": 310
+    },
+    {
+      "epoch": 0.07296187683284458,
+      "grad_norm": 5.243283974944376,
+      "learning_rate": 7.259953161592507e-06,
+      "loss": 0.7103,
+      "step": 311
+    },
+    {
+      "epoch": 0.07319648093841642,
+      "grad_norm": 1.8240644408057207,
+      "learning_rate": 7.28337236533958e-06,
+      "loss": 0.6277,
+      "step": 312
+    },
+    {
+      "epoch": 0.07343108504398826,
+      "grad_norm": 4.009256485400117,
+      "learning_rate": 7.306791569086652e-06,
+      "loss": 0.7491,
+      "step": 313
+    },
+    {
+      "epoch": 0.07366568914956012,
+      "grad_norm": 2.5862860233978204,
+      "learning_rate": 7.330210772833725e-06,
+      "loss": 0.6999,
+      "step": 314
+    },
+    {
+      "epoch": 0.07390029325513196,
+      "grad_norm": 1.95411282236033,
+      "learning_rate": 7.3536299765807975e-06,
+      "loss": 0.7655,
+      "step": 315
+    },
+    {
+      "epoch": 0.07413489736070382,
+      "grad_norm": 2.5324854678602726,
+      "learning_rate": 7.3770491803278695e-06,
+      "loss": 0.7145,
+      "step": 316
+    },
+    {
+      "epoch": 0.07436950146627566,
+      "grad_norm": 2.2048590914639883,
+      "learning_rate": 7.400468384074942e-06,
+      "loss": 0.752,
+      "step": 317
+    },
+    {
+      "epoch": 0.0746041055718475,
+      "grad_norm": 10.802056503356317,
+      "learning_rate": 7.423887587822015e-06,
+      "loss": 0.7544,
+      "step": 318
+    },
+    {
+      "epoch": 0.07483870967741936,
+      "grad_norm": 5.947513346289248,
+      "learning_rate": 7.447306791569088e-06,
+      "loss": 0.7024,
+      "step": 319
+    },
+    {
+      "epoch": 0.0750733137829912,
+      "grad_norm": 2.2835797681082726,
+      "learning_rate": 7.47072599531616e-06,
+      "loss": 0.7484,
+      "step": 320
+    },
+    {
+      "epoch": 0.07530791788856304,
+      "grad_norm": 2.0787193869349947,
+      "learning_rate": 7.494145199063233e-06,
+      "loss": 0.747,
+      "step": 321
+    },
+    {
+      "epoch": 0.0755425219941349,
+      "grad_norm": 1.755869711708477,
+      "learning_rate": 7.517564402810305e-06,
+      "loss": 0.6877,
+      "step": 322
+    },
+    {
+      "epoch": 0.07577712609970674,
+      "grad_norm": 3.1638128679507758,
+      "learning_rate": 7.540983606557377e-06,
+      "loss": 0.7697,
+      "step": 323
+    },
+    {
+      "epoch": 0.0760117302052786,
+      "grad_norm": 1.924419722616588,
+      "learning_rate": 7.56440281030445e-06,
+      "loss": 0.7838,
+      "step": 324
+    },
+    {
+      "epoch": 0.07624633431085044,
+      "grad_norm": 1.784712238212702,
+      "learning_rate": 7.5878220140515225e-06,
+      "loss": 0.7164,
+      "step": 325
+    },
+    {
+      "epoch": 0.07648093841642228,
+      "grad_norm": 1.5556484461805318,
+      "learning_rate": 7.611241217798595e-06,
+      "loss": 0.7584,
+      "step": 326
+    },
+    {
+      "epoch": 0.07671554252199414,
+      "grad_norm": 1.6239563401361596,
+      "learning_rate": 7.634660421545667e-06,
+      "loss": 0.672,
+      "step": 327
+    },
+    {
+      "epoch": 0.07695014662756598,
+      "grad_norm": 2.7172441896685084,
+      "learning_rate": 7.65807962529274e-06,
+      "loss": 0.7238,
+      "step": 328
+    },
+    {
+      "epoch": 0.07718475073313782,
+      "grad_norm": 2.7504509737580323,
+      "learning_rate": 7.681498829039813e-06,
+      "loss": 0.7063,
+      "step": 329
+    },
+    {
+      "epoch": 0.07741935483870968,
+      "grad_norm": 4.209971644960854,
+      "learning_rate": 7.704918032786886e-06,
+      "loss": 0.7075,
+      "step": 330
+    },
+    {
+      "epoch": 0.07765395894428152,
+      "grad_norm": 3.8102324360788793,
+      "learning_rate": 7.728337236533959e-06,
+      "loss": 0.759,
+      "step": 331
+    },
+    {
+      "epoch": 0.07788856304985338,
+      "grad_norm": 2.948409700744492,
+      "learning_rate": 7.751756440281032e-06,
+      "loss": 0.6859,
+      "step": 332
+    },
+    {
+      "epoch": 0.07812316715542522,
+      "grad_norm": 2.12117726157581,
+      "learning_rate": 7.775175644028103e-06,
+      "loss": 0.7335,
+      "step": 333
+    },
+    {
+      "epoch": 0.07835777126099706,
+      "grad_norm": 3.7746062350923677,
+      "learning_rate": 7.798594847775176e-06,
+      "loss": 0.7585,
+      "step": 334
+    },
+    {
+      "epoch": 0.07859237536656892,
+      "grad_norm": 7.485470991883209,
+      "learning_rate": 7.822014051522248e-06,
+      "loss": 0.6875,
+      "step": 335
+    },
+    {
+      "epoch": 0.07882697947214076,
+      "grad_norm": 2.438658899659587,
+      "learning_rate": 7.845433255269321e-06,
+      "loss": 0.7741,
+      "step": 336
+    },
+    {
+      "epoch": 0.0790615835777126,
+      "grad_norm": 2.460708668740363,
+      "learning_rate": 7.868852459016394e-06,
+      "loss": 0.7959,
+      "step": 337
+    },
+    {
+      "epoch": 0.07929618768328446,
+      "grad_norm": 2.074075903721013,
+      "learning_rate": 7.892271662763467e-06,
+      "loss": 0.7462,
+      "step": 338
+    },
+    {
+      "epoch": 0.0795307917888563,
+      "grad_norm": 3.7240911196711077,
+      "learning_rate": 7.91569086651054e-06,
+      "loss": 0.7188,
+      "step": 339
+    },
+    {
+      "epoch": 0.07976539589442816,
+      "grad_norm": 5.283811424972736,
+      "learning_rate": 7.939110070257613e-06,
+      "loss": 0.7185,
+      "step": 340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2104694215673786,
+      "learning_rate": 7.962529274004684e-06,
+      "loss": 0.7172,
+      "step": 341
+    },
+    {
+      "epoch": 0.08023460410557184,
+      "grad_norm": 2.997008713052917,
+      "learning_rate": 7.985948477751757e-06,
+      "loss": 0.7621,
+      "step": 342
+    },
+    {
+      "epoch": 0.0804692082111437,
+      "grad_norm": 2.506302477784963,
+      "learning_rate": 8.00936768149883e-06,
+      "loss": 0.7508,
+      "step": 343
+    },
+    {
+      "epoch": 0.08070381231671554,
+      "grad_norm": 3.1906634786571337,
+      "learning_rate": 8.032786885245902e-06,
+      "loss": 0.6898,
+      "step": 344
+    },
+    {
+      "epoch": 0.08093841642228738,
+      "grad_norm": 1.8579820431446714,
+      "learning_rate": 8.056206088992975e-06,
+      "loss": 0.744,
+      "step": 345
+    },
+    {
+      "epoch": 0.08117302052785924,
+      "grad_norm": 3.014565732438054,
+      "learning_rate": 8.079625292740048e-06,
+      "loss": 0.7069,
+      "step": 346
+    },
+    {
+      "epoch": 0.08140762463343108,
+      "grad_norm": 1.2533412435142142,
+      "learning_rate": 8.10304449648712e-06,
+      "loss": 0.6775,
+      "step": 347
+    },
+    {
+      "epoch": 0.08164222873900294,
+      "grad_norm": 8.653046483399084,
+      "learning_rate": 8.126463700234194e-06,
+      "loss": 0.7725,
+      "step": 348
+    },
+    {
+      "epoch": 0.08187683284457478,
+      "grad_norm": 1.7503738115203018,
+      "learning_rate": 8.149882903981265e-06,
+      "loss": 0.7135,
+      "step": 349
+    },
+    {
+      "epoch": 0.08211143695014662,
+      "grad_norm": 1.9139633404300693,
+      "learning_rate": 8.173302107728338e-06,
+      "loss": 0.7412,
+      "step": 350
+    },
+    {
+      "epoch": 0.08234604105571848,
+      "grad_norm": 1.536727880890203,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 0.7359,
+      "step": 351
+    },
+    {
+      "epoch": 0.08258064516129032,
+      "grad_norm": 2.8660997603136744,
+      "learning_rate": 8.220140515222483e-06,
+      "loss": 0.7667,
+      "step": 352
+    },
+    {
+      "epoch": 0.08281524926686216,
+      "grad_norm": 2.23335924032936,
+      "learning_rate": 8.243559718969556e-06,
+      "loss": 0.7349,
+      "step": 353
+    },
+    {
+      "epoch": 0.08304985337243402,
+      "grad_norm": 2.415132428444925,
+      "learning_rate": 8.266978922716629e-06,
+      "loss": 0.7546,
+      "step": 354
+    },
+    {
+      "epoch": 0.08328445747800586,
+      "grad_norm": 2.0073321543371736,
+      "learning_rate": 8.290398126463702e-06,
+      "loss": 0.7138,
+      "step": 355
+    },
+    {
+      "epoch": 0.08351906158357772,
+      "grad_norm": 3.2944999780355095,
+      "learning_rate": 8.313817330210773e-06,
+      "loss": 0.6752,
+      "step": 356
+    },
+    {
+      "epoch": 0.08375366568914956,
+      "grad_norm": 1.7139802340737562,
+      "learning_rate": 8.337236533957846e-06,
+      "loss": 0.6795,
+      "step": 357
+    },
+    {
+      "epoch": 0.0839882697947214,
+      "grad_norm": 5.315894715413043,
+      "learning_rate": 8.360655737704919e-06,
+      "loss": 0.745,
+      "step": 358
+    },
+    {
+      "epoch": 0.08422287390029326,
+      "grad_norm": 2.060548793865288,
+      "learning_rate": 8.384074941451992e-06,
+      "loss": 0.7033,
+      "step": 359
+    },
+    {
+      "epoch": 0.0844574780058651,
+      "grad_norm": 2.854342221447936,
+      "learning_rate": 8.407494145199064e-06,
+      "loss": 0.7255,
+      "step": 360
+    },
+    {
+      "epoch": 0.08469208211143694,
+      "grad_norm": 1.5296233063364677,
+      "learning_rate": 8.430913348946137e-06,
+      "loss": 0.7143,
+      "step": 361
+    },
+    {
+      "epoch": 0.0849266862170088,
+      "grad_norm": 2.9366860113406483,
+      "learning_rate": 8.45433255269321e-06,
+      "loss": 0.7498,
+      "step": 362
+    },
+    {
+      "epoch": 0.08516129032258064,
+      "grad_norm": 1.4632395156755973,
+      "learning_rate": 8.477751756440283e-06,
+      "loss": 0.6979,
+      "step": 363
+    },
+    {
+      "epoch": 0.0853958944281525,
+      "grad_norm": 4.237013919995425,
+      "learning_rate": 8.501170960187354e-06,
+      "loss": 0.7334,
+      "step": 364
+    },
+    {
+      "epoch": 0.08563049853372434,
+      "grad_norm": 2.173475328218578,
+      "learning_rate": 8.524590163934427e-06,
+      "loss": 0.7177,
+      "step": 365
+    },
+    {
+      "epoch": 0.08586510263929618,
+      "grad_norm": 2.5213528899664968,
+      "learning_rate": 8.5480093676815e-06,
+      "loss": 0.6737,
+      "step": 366
+    },
+    {
+      "epoch": 0.08609970674486804,
+      "grad_norm": 4.2993196772593345,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.7739,
+      "step": 367
+    },
+    {
+      "epoch": 0.08633431085043988,
+      "grad_norm": 2.5155006280989087,
+      "learning_rate": 8.594847775175644e-06,
+      "loss": 0.7361,
+      "step": 368
+    },
+    {
+      "epoch": 0.08656891495601172,
+      "grad_norm": 1.9647822308984213,
+      "learning_rate": 8.618266978922717e-06,
+      "loss": 0.7205,
+      "step": 369
+    },
+    {
+      "epoch": 0.08680351906158358,
+      "grad_norm": 4.562372816515158,
+      "learning_rate": 8.64168618266979e-06,
+      "loss": 0.7631,
+      "step": 370
+    },
+    {
+      "epoch": 0.08703812316715542,
+      "grad_norm": 3.1620928347884623,
+      "learning_rate": 8.665105386416862e-06,
+      "loss": 0.7482,
+      "step": 371
+    },
+    {
+      "epoch": 0.08727272727272728,
+      "grad_norm": 2.901318848524828,
+      "learning_rate": 8.688524590163935e-06,
+      "loss": 0.7036,
+      "step": 372
+    },
+    {
+      "epoch": 0.08750733137829912,
+      "grad_norm": 4.864603743688095,
+      "learning_rate": 8.711943793911008e-06,
+      "loss": 0.763,
+      "step": 373
+    },
+    {
+      "epoch": 0.08774193548387096,
+      "grad_norm": 1.8308419723689138,
+      "learning_rate": 8.73536299765808e-06,
+      "loss": 0.7396,
+      "step": 374
+    },
+    {
+      "epoch": 0.08797653958944282,
+      "grad_norm": 2.411984661545923,
+      "learning_rate": 8.758782201405152e-06,
+      "loss": 0.7045,
+      "step": 375
+    },
+    {
+      "epoch": 0.08821114369501466,
+      "grad_norm": 4.24982232258306,
+      "learning_rate": 8.782201405152225e-06,
+      "loss": 0.7396,
+      "step": 376
+    },
+    {
+      "epoch": 0.0884457478005865,
+      "grad_norm": 3.190757144859846,
+      "learning_rate": 8.805620608899298e-06,
+      "loss": 0.7109,
+      "step": 377
+    },
+    {
+      "epoch": 0.08868035190615836,
+      "grad_norm": 1.5464630154259653,
+      "learning_rate": 8.82903981264637e-06,
+      "loss": 0.7067,
+      "step": 378
+    },
+    {
+      "epoch": 0.0889149560117302,
+      "grad_norm": 1.4056128329371305,
+      "learning_rate": 8.852459016393443e-06,
+      "loss": 0.6872,
+      "step": 379
+    },
+    {
+      "epoch": 0.08914956011730206,
+      "grad_norm": 1.6722723816773548,
+      "learning_rate": 8.875878220140516e-06,
+      "loss": 0.7472,
+      "step": 380
+    },
+    {
+      "epoch": 0.0893841642228739,
+      "grad_norm": 7.751412884958818,
+      "learning_rate": 8.899297423887589e-06,
+      "loss": 0.6917,
+      "step": 381
+    },
+    {
+      "epoch": 0.08961876832844574,
+      "grad_norm": 3.2510202761271447,
+      "learning_rate": 8.922716627634662e-06,
+      "loss": 0.6748,
+      "step": 382
+    },
+    {
+      "epoch": 0.0898533724340176,
+      "grad_norm": 2.5669568293551728,
+      "learning_rate": 8.946135831381733e-06,
+      "loss": 0.725,
+      "step": 383
+    },
+    {
+      "epoch": 0.09008797653958944,
+      "grad_norm": 3.7440576716747813,
+      "learning_rate": 8.969555035128806e-06,
+      "loss": 0.6906,
+      "step": 384
+    },
+    {
+      "epoch": 0.09032258064516129,
+      "grad_norm": 8.296111298322213,
+      "learning_rate": 8.992974238875879e-06,
+      "loss": 0.6661,
+      "step": 385
+    },
+    {
+      "epoch": 0.09055718475073314,
+      "grad_norm": 2.716287161549124,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 0.6638,
+      "step": 386
+    },
+    {
+      "epoch": 0.09079178885630498,
+      "grad_norm": 1.4637051860404733,
+      "learning_rate": 9.039812646370024e-06,
+      "loss": 0.75,
+      "step": 387
+    },
+    {
+      "epoch": 0.09102639296187684,
+      "grad_norm": 1.9054264366067561,
+      "learning_rate": 9.063231850117097e-06,
+      "loss": 0.7222,
+      "step": 388
+    },
+    {
+      "epoch": 0.09126099706744868,
+      "grad_norm": 5.88123789755757,
+      "learning_rate": 9.08665105386417e-06,
+      "loss": 0.7677,
+      "step": 389
+    },
+    {
+      "epoch": 0.09149560117302052,
+      "grad_norm": 1.5836577273520098,
+      "learning_rate": 9.110070257611243e-06,
+      "loss": 0.6886,
+      "step": 390
+    },
+    {
+      "epoch": 0.09173020527859238,
+      "grad_norm": 1.4769405393823147,
+      "learning_rate": 9.133489461358314e-06,
+      "loss": 0.6943,
+      "step": 391
+    },
+    {
+      "epoch": 0.09196480938416422,
+      "grad_norm": 1.5393930672289444,
+      "learning_rate": 9.156908665105387e-06,
+      "loss": 0.6734,
+      "step": 392
+    },
+    {
+      "epoch": 0.09219941348973607,
+      "grad_norm": 2.04117582768801,
+      "learning_rate": 9.18032786885246e-06,
+      "loss": 0.7467,
+      "step": 393
+    },
+    {
+      "epoch": 0.09243401759530792,
+      "grad_norm": 3.3753912706732523,
+      "learning_rate": 9.203747072599533e-06,
+      "loss": 0.7235,
+      "step": 394
+    },
+    {
+      "epoch": 0.09266862170087976,
+      "grad_norm": 1.8420173745879682,
+      "learning_rate": 9.227166276346605e-06,
+      "loss": 0.6881,
+      "step": 395
+    },
+    {
+      "epoch": 0.09290322580645162,
+      "grad_norm": 2.7424273016845655,
+      "learning_rate": 9.250585480093678e-06,
+      "loss": 0.6923,
+      "step": 396
+    },
+    {
+      "epoch": 0.09313782991202346,
+      "grad_norm": 3.345559957565154,
+      "learning_rate": 9.274004683840751e-06,
+      "loss": 0.7716,
+      "step": 397
+    },
+    {
+      "epoch": 0.0933724340175953,
+      "grad_norm": 4.7620075885039945,
+      "learning_rate": 9.297423887587822e-06,
+      "loss": 0.7342,
+      "step": 398
+    },
+    {
+      "epoch": 0.09360703812316716,
+      "grad_norm": 1.8391282182105828,
+      "learning_rate": 9.320843091334895e-06,
+      "loss": 0.6778,
+      "step": 399
+    },
+    {
+      "epoch": 0.093841642228739,
+      "grad_norm": 4.021920739798764,
+      "learning_rate": 9.344262295081968e-06,
+      "loss": 0.7419,
+      "step": 400
+    },
+    {
+      "epoch": 0.09407624633431085,
+      "grad_norm": 1.7430126105824388,
+      "learning_rate": 9.36768149882904e-06,
+      "loss": 0.6887,
+      "step": 401
+    },
+    {
+      "epoch": 0.0943108504398827,
+      "grad_norm": 1.5251332201860028,
+      "learning_rate": 9.391100702576114e-06,
+      "loss": 0.6463,
+      "step": 402
+    },
+    {
+      "epoch": 0.09454545454545454,
+      "grad_norm": 2.5201755219435515,
+      "learning_rate": 9.414519906323187e-06,
+      "loss": 0.69,
+      "step": 403
+    },
+    {
+      "epoch": 0.0947800586510264,
+      "grad_norm": 1.5280311623002627,
+      "learning_rate": 9.43793911007026e-06,
+      "loss": 0.7285,
+      "step": 404
+    },
+    {
+      "epoch": 0.09501466275659824,
+      "grad_norm": 2.447952313998854,
+      "learning_rate": 9.461358313817332e-06,
+      "loss": 0.6911,
+      "step": 405
+    },
+    {
+      "epoch": 0.09524926686217008,
+      "grad_norm": 1.9015531040690363,
+      "learning_rate": 9.484777517564403e-06,
+      "loss": 0.7142,
+      "step": 406
+    },
+    {
+      "epoch": 0.09548387096774194,
+      "grad_norm": 1.9335270760208139,
+      "learning_rate": 9.508196721311476e-06,
+      "loss": 0.7182,
+      "step": 407
+    },
+    {
+      "epoch": 0.09571847507331378,
+      "grad_norm": 2.471047004477268,
+      "learning_rate": 9.531615925058549e-06,
+      "loss": 0.7297,
+      "step": 408
+    },
+    {
+      "epoch": 0.09595307917888563,
+      "grad_norm": 3.904270141650119,
+      "learning_rate": 9.55503512880562e-06,
+      "loss": 0.716,
+      "step": 409
+    },
+    {
+      "epoch": 0.09618768328445748,
+      "grad_norm": 2.3095723128690984,
+      "learning_rate": 9.578454332552693e-06,
+      "loss": 0.6413,
+      "step": 410
+    },
+    {
+      "epoch": 0.09642228739002932,
+      "grad_norm": 11.94533457021786,
+      "learning_rate": 9.601873536299766e-06,
+      "loss": 0.656,
+      "step": 411
+    },
+    {
+      "epoch": 0.09665689149560118,
+      "grad_norm": 3.348994409400017,
+      "learning_rate": 9.625292740046839e-06,
+      "loss": 0.7164,
+      "step": 412
+    },
+    {
+      "epoch": 0.09689149560117302,
+      "grad_norm": 2.5657738183621404,
+      "learning_rate": 9.648711943793912e-06,
+      "loss": 0.659,
+      "step": 413
+    },
+    {
+      "epoch": 0.09712609970674486,
+      "grad_norm": 1.9489336510081479,
+      "learning_rate": 9.672131147540984e-06,
+      "loss": 0.6836,
+      "step": 414
+    },
+    {
+      "epoch": 0.09736070381231672,
+      "grad_norm": 5.068307001090185,
+      "learning_rate": 9.695550351288057e-06,
+      "loss": 0.7652,
+      "step": 415
+    },
+    {
+      "epoch": 0.09759530791788856,
+      "grad_norm": 2.748453987481012,
+      "learning_rate": 9.71896955503513e-06,
+      "loss": 0.7079,
+      "step": 416
+    },
+    {
+      "epoch": 0.0978299120234604,
+      "grad_norm": 2.4709648008841603,
+      "learning_rate": 9.742388758782201e-06,
+      "loss": 0.6939,
+      "step": 417
+    },
+    {
+      "epoch": 0.09806451612903226,
+      "grad_norm": 1.511478089066913,
+      "learning_rate": 9.765807962529274e-06,
+      "loss": 0.6551,
+      "step": 418
+    },
+    {
+      "epoch": 0.0982991202346041,
+      "grad_norm": 2.4978493830642563,
+      "learning_rate": 9.789227166276347e-06,
+      "loss": 0.7331,
+      "step": 419
+    },
+    {
+      "epoch": 0.09853372434017596,
+      "grad_norm": 2.484188148554729,
+      "learning_rate": 9.81264637002342e-06,
+      "loss": 0.7333,
+      "step": 420
+    },
+    {
+      "epoch": 0.0987683284457478,
+      "grad_norm": 2.896283868289636,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 0.6859,
+      "step": 421
+    },
+    {
+      "epoch": 0.09900293255131964,
+      "grad_norm": 2.7543059139679955,
+      "learning_rate": 9.859484777517565e-06,
+      "loss": 0.7014,
+      "step": 422
+    },
+    {
+      "epoch": 0.0992375366568915,
+      "grad_norm": 1.93457550943428,
+      "learning_rate": 9.882903981264638e-06,
+      "loss": 0.6821,
+      "step": 423
+    },
+    {
+      "epoch": 0.09947214076246334,
+      "grad_norm": 3.541421711373985,
+      "learning_rate": 9.906323185011711e-06,
+      "loss": 0.7795,
+      "step": 424
+    },
+    {
+      "epoch": 0.09970674486803519,
+      "grad_norm": 2.3210901497458902,
+      "learning_rate": 9.929742388758782e-06,
+      "loss": 0.7066,
+      "step": 425
+    },
+    {
+      "epoch": 0.09994134897360704,
+      "grad_norm": 2.061503531228248,
+      "learning_rate": 9.953161592505855e-06,
+      "loss": 0.6978,
+      "step": 426
+    },
+    {
+      "epoch": 0.10017595307917888,
+      "grad_norm": 2.881370488654951,
+      "learning_rate": 9.976580796252928e-06,
+      "loss": 0.6827,
+      "step": 427
+    },
+    {
+      "epoch": 0.10041055718475074,
+      "grad_norm": 1.9376756360692144,
+      "learning_rate": 1e-05,
+      "loss": 0.6985,
+      "step": 428
+    },
+    {
+      "epoch": 0.10064516129032258,
+      "grad_norm": 9.185995500420793,
+      "learning_rate": 9.99999832232052e-06,
+      "loss": 0.7061,
+      "step": 429
+    },
+    {
+      "epoch": 0.10087976539589442,
+      "grad_norm": 1.3779257570194963,
+      "learning_rate": 9.999993289283198e-06,
+      "loss": 0.6873,
+      "step": 430
+    },
+    {
+      "epoch": 0.10111436950146628,
+      "grad_norm": 7.567980802908717,
+      "learning_rate": 9.999984900891416e-06,
+      "loss": 0.7555,
+      "step": 431
+    },
+    {
+      "epoch": 0.10134897360703812,
+      "grad_norm": 5.516879792515433,
+      "learning_rate": 9.999973157150802e-06,
+      "loss": 0.7445,
+      "step": 432
+    },
+    {
+      "epoch": 0.10158357771260997,
+      "grad_norm": 2.8992965270273783,
+      "learning_rate": 9.999958058069237e-06,
+      "loss": 0.7679,
+      "step": 433
+    },
+    {
+      "epoch": 0.10181818181818182,
+      "grad_norm": 3.2406301614465884,
+      "learning_rate": 9.999939603656855e-06,
+      "loss": 0.7243,
+      "step": 434
+    },
+    {
+      "epoch": 0.10205278592375366,
+      "grad_norm": 3.4869381136029345,
+      "learning_rate": 9.999917793926037e-06,
+      "loss": 0.6637,
+      "step": 435
+    },
+    {
+      "epoch": 0.10228739002932552,
+      "grad_norm": 1.3184474068296157,
+      "learning_rate": 9.999892628891421e-06,
+      "loss": 0.7128,
+      "step": 436
+    },
+    {
+      "epoch": 0.10252199413489736,
+      "grad_norm": 2.1731601294137515,
+      "learning_rate": 9.999864108569895e-06,
+      "loss": 0.7009,
+      "step": 437
+    },
+    {
+      "epoch": 0.1027565982404692,
+      "grad_norm": 1.5904788438718565,
+      "learning_rate": 9.999832232980598e-06,
+      "loss": 0.6918,
+      "step": 438
+    },
+    {
+      "epoch": 0.10299120234604106,
+      "grad_norm": 2.40384485405665,
+      "learning_rate": 9.99979700214492e-06,
+      "loss": 0.7086,
+      "step": 439
+    },
+    {
+      "epoch": 0.1032258064516129,
+      "grad_norm": 9.0473937275457,
+      "learning_rate": 9.999758416086503e-06,
+      "loss": 0.6649,
+      "step": 440
+    },
+    {
+      "epoch": 0.10346041055718475,
+      "grad_norm": 1.8878111541309193,
+      "learning_rate": 9.999716474831241e-06,
+      "loss": 0.6822,
+      "step": 441
+    },
+    {
+      "epoch": 0.1036950146627566,
+      "grad_norm": 2.01013703192227,
+      "learning_rate": 9.999671178407283e-06,
+      "loss": 0.7049,
+      "step": 442
+    },
+    {
+      "epoch": 0.10392961876832844,
+      "grad_norm": 1.8871464386483952,
+      "learning_rate": 9.99962252684502e-06,
+      "loss": 0.6495,
+      "step": 443
+    },
+    {
+      "epoch": 0.1041642228739003,
+      "grad_norm": 2.715514625300475,
+      "learning_rate": 9.999570520177107e-06,
+      "loss": 0.7042,
+      "step": 444
+    },
+    {
+      "epoch": 0.10439882697947214,
+      "grad_norm": 1.679303542504971,
+      "learning_rate": 9.99951515843844e-06,
+      "loss": 0.7118,
+      "step": 445
+    },
+    {
+      "epoch": 0.10463343108504398,
+      "grad_norm": 3.641101181087701,
+      "learning_rate": 9.999456441666172e-06,
+      "loss": 0.7259,
+      "step": 446
+    },
+    {
+      "epoch": 0.10486803519061584,
+      "grad_norm": 3.341226347625401,
+      "learning_rate": 9.999394369899706e-06,
+      "loss": 0.7389,
+      "step": 447
+    },
+    {
+      "epoch": 0.10510263929618768,
+      "grad_norm": 10.714831824909773,
+      "learning_rate": 9.999328943180696e-06,
+      "loss": 0.6824,
+      "step": 448
+    },
+    {
+      "epoch": 0.10533724340175953,
+      "grad_norm": 20.08655518501293,
+      "learning_rate": 9.999260161553052e-06,
+      "loss": 0.7064,
+      "step": 449
+    },
+    {
+      "epoch": 0.10557184750733138,
+      "grad_norm": 1.2515783765249133,
+      "learning_rate": 9.999188025062925e-06,
+      "loss": 0.6944,
+      "step": 450
+    },
+    {
+      "epoch": 0.10580645161290322,
+      "grad_norm": 1.7578886402867833,
+      "learning_rate": 9.99911253375873e-06,
+      "loss": 0.7077,
+      "step": 451
+    },
+    {
+      "epoch": 0.10604105571847508,
+      "grad_norm": 1.9250104195615887,
+      "learning_rate": 9.999033687691122e-06,
+      "loss": 0.6437,
+      "step": 452
+    },
+    {
+      "epoch": 0.10627565982404692,
+      "grad_norm": 26.687022521364753,
+      "learning_rate": 9.998951486913015e-06,
+      "loss": 0.7095,
+      "step": 453
+    },
+    {
+      "epoch": 0.10651026392961876,
+      "grad_norm": 2.7693857312683225,
+      "learning_rate": 9.998865931479572e-06,
+      "loss": 0.7758,
+      "step": 454
+    },
+    {
+      "epoch": 0.10674486803519062,
+      "grad_norm": 2.615705716385251,
+      "learning_rate": 9.998777021448204e-06,
+      "loss": 0.6635,
+      "step": 455
+    },
+    {
+      "epoch": 0.10697947214076246,
+      "grad_norm": 2.2314457566850345,
+      "learning_rate": 9.99868475687858e-06,
+      "loss": 0.6474,
+      "step": 456
+    },
+    {
+      "epoch": 0.1072140762463343,
+      "grad_norm": 2.821634138231432,
+      "learning_rate": 9.998589137832612e-06,
+      "loss": 0.6494,
+      "step": 457
+    },
+    {
+      "epoch": 0.10744868035190616,
+      "grad_norm": 4.725909833280999,
+      "learning_rate": 9.998490164374472e-06,
+      "loss": 0.7363,
+      "step": 458
+    },
+    {
+      "epoch": 0.107683284457478,
+      "grad_norm": 1.5841167891781551,
+      "learning_rate": 9.998387836570573e-06,
+      "loss": 0.7518,
+      "step": 459
+    },
+    {
+      "epoch": 0.10791788856304986,
+      "grad_norm": 1.7115189418913666,
+      "learning_rate": 9.998282154489588e-06,
+      "loss": 0.6926,
+      "step": 460
+    },
+    {
+      "epoch": 0.1081524926686217,
+      "grad_norm": 2.388898935446305,
+      "learning_rate": 9.998173118202438e-06,
+      "loss": 0.6799,
+      "step": 461
+    },
+    {
+      "epoch": 0.10838709677419354,
+      "grad_norm": 2.9397115202024615,
+      "learning_rate": 9.998060727782291e-06,
+      "loss": 0.6984,
+      "step": 462
+    },
+    {
+      "epoch": 0.1086217008797654,
+      "grad_norm": 1.8706360121603745,
+      "learning_rate": 9.99794498330457e-06,
+      "loss": 0.6952,
+      "step": 463
+    },
+    {
+      "epoch": 0.10885630498533724,
+      "grad_norm": 1.684443764370597,
+      "learning_rate": 9.997825884846947e-06,
+      "loss": 0.7163,
+      "step": 464
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 2.304327777151329,
+      "learning_rate": 9.99770343248935e-06,
+      "loss": 0.7277,
+      "step": 465
+    },
+    {
+      "epoch": 0.10932551319648094,
+      "grad_norm": 2.500667829864606,
+      "learning_rate": 9.997577626313948e-06,
+      "loss": 0.7621,
+      "step": 466
+    },
+    {
+      "epoch": 0.10956011730205278,
+      "grad_norm": 1.377090939725991,
+      "learning_rate": 9.99744846640517e-06,
+      "loss": 0.6609,
+      "step": 467
+    },
+    {
+      "epoch": 0.10979472140762464,
+      "grad_norm": 2.296341610377368,
+      "learning_rate": 9.99731595284969e-06,
+      "loss": 0.7315,
+      "step": 468
+    },
+    {
+      "epoch": 0.11002932551319648,
+      "grad_norm": 4.818232817430393,
+      "learning_rate": 9.997180085736431e-06,
+      "loss": 0.6855,
+      "step": 469
+    },
+    {
+      "epoch": 0.11026392961876832,
+      "grad_norm": 3.6547377355731085,
+      "learning_rate": 9.997040865156575e-06,
+      "loss": 0.6983,
+      "step": 470
+    },
+    {
+      "epoch": 0.11049853372434018,
+      "grad_norm": 5.386859126999124,
+      "learning_rate": 9.996898291203544e-06,
+      "loss": 0.6702,
+      "step": 471
+    },
+    {
+      "epoch": 0.11073313782991202,
+      "grad_norm": 2.8098348542077938,
+      "learning_rate": 9.99675236397302e-06,
+      "loss": 0.7545,
+      "step": 472
+    },
+    {
+      "epoch": 0.11096774193548387,
+      "grad_norm": 2.536389393667118,
+      "learning_rate": 9.996603083562928e-06,
+      "loss": 0.6799,
+      "step": 473
+    },
+    {
+      "epoch": 0.11120234604105572,
+      "grad_norm": 1.3917740827397262,
+      "learning_rate": 9.996450450073447e-06,
+      "loss": 0.6756,
+      "step": 474
+    },
+    {
+      "epoch": 0.11143695014662756,
+      "grad_norm": 2.345795464079582,
+      "learning_rate": 9.996294463607004e-06,
+      "loss": 0.7228,
+      "step": 475
+    },
+    {
+      "epoch": 0.11167155425219942,
+      "grad_norm": 1.5696093579763688,
+      "learning_rate": 9.996135124268276e-06,
+      "loss": 0.6942,
+      "step": 476
+    },
+    {
+      "epoch": 0.11190615835777126,
+      "grad_norm": 1.2610491066110727,
+      "learning_rate": 9.995972432164194e-06,
+      "loss": 0.7187,
+      "step": 477
+    },
+    {
+      "epoch": 0.1121407624633431,
+      "grad_norm": 2.9972015560304666,
+      "learning_rate": 9.995806387403935e-06,
+      "loss": 0.704,
+      "step": 478
+    },
+    {
+      "epoch": 0.11237536656891496,
+      "grad_norm": 2.044081312404094,
+      "learning_rate": 9.995636990098928e-06,
+      "loss": 0.7446,
+      "step": 479
+    },
+    {
+      "epoch": 0.1126099706744868,
+      "grad_norm": 1.7584032388011313,
+      "learning_rate": 9.995464240362847e-06,
+      "loss": 0.7194,
+      "step": 480
+    },
+    {
+      "epoch": 0.11284457478005865,
+      "grad_norm": 3.3210695065864937,
+      "learning_rate": 9.995288138311622e-06,
+      "loss": 0.6865,
+      "step": 481
+    },
+    {
+      "epoch": 0.1130791788856305,
+      "grad_norm": 8.73885391298322,
+      "learning_rate": 9.995108684063431e-06,
+      "loss": 0.724,
+      "step": 482
+    },
+    {
+      "epoch": 0.11331378299120234,
+      "grad_norm": 4.418388663156157,
+      "learning_rate": 9.9949258777387e-06,
+      "loss": 0.7392,
+      "step": 483
+    },
+    {
+      "epoch": 0.1135483870967742,
+      "grad_norm": 1.7514645214551325,
+      "learning_rate": 9.994739719460103e-06,
+      "loss": 0.643,
+      "step": 484
+    },
+    {
+      "epoch": 0.11378299120234604,
+      "grad_norm": 2.34653301290369,
+      "learning_rate": 9.994550209352569e-06,
+      "loss": 0.7112,
+      "step": 485
+    },
+    {
+      "epoch": 0.11401759530791788,
+      "grad_norm": 1.5689601767472114,
+      "learning_rate": 9.99435734754327e-06,
+      "loss": 0.6981,
+      "step": 486
+    },
+    {
+      "epoch": 0.11425219941348974,
+      "grad_norm": 2.7327993867035234,
+      "learning_rate": 9.994161134161635e-06,
+      "loss": 0.703,
+      "step": 487
+    },
+    {
+      "epoch": 0.11448680351906158,
+      "grad_norm": 2.6868476495035014,
+      "learning_rate": 9.99396156933933e-06,
+      "loss": 0.7232,
+      "step": 488
+    },
+    {
+      "epoch": 0.11472140762463343,
+      "grad_norm": 4.318914367666479,
+      "learning_rate": 9.993758653210283e-06,
+      "loss": 0.7436,
+      "step": 489
+    },
+    {
+      "epoch": 0.11495601173020528,
+      "grad_norm": 2.3801491122694087,
+      "learning_rate": 9.993552385910663e-06,
+      "loss": 0.6563,
+      "step": 490
+    },
+    {
+      "epoch": 0.11519061583577712,
+      "grad_norm": 3.3685873283237537,
+      "learning_rate": 9.993342767578888e-06,
+      "loss": 0.6611,
+      "step": 491
+    },
+    {
+      "epoch": 0.11542521994134898,
+      "grad_norm": 9.216519497431854,
+      "learning_rate": 9.993129798355633e-06,
+      "loss": 0.7036,
+      "step": 492
+    },
+    {
+      "epoch": 0.11565982404692082,
+      "grad_norm": 2.3057579250776237,
+      "learning_rate": 9.99291347838381e-06,
+      "loss": 0.7227,
+      "step": 493
+    },
+    {
+      "epoch": 0.11589442815249267,
+      "grad_norm": 5.09132377443943,
+      "learning_rate": 9.992693807808589e-06,
+      "loss": 0.7115,
+      "step": 494
+    },
+    {
+      "epoch": 0.11612903225806452,
+      "grad_norm": 1.784883173027611,
+      "learning_rate": 9.992470786777382e-06,
+      "loss": 0.6987,
+      "step": 495
+    },
+    {
+      "epoch": 0.11636363636363636,
+      "grad_norm": 2.374758365052141,
+      "learning_rate": 9.992244415439853e-06,
+      "loss": 0.6656,
+      "step": 496
+    },
+    {
+      "epoch": 0.1165982404692082,
+      "grad_norm": 3.146138261728129,
+      "learning_rate": 9.992014693947914e-06,
+      "loss": 0.678,
+      "step": 497
+    },
+    {
+      "epoch": 0.11683284457478006,
+      "grad_norm": 7.476451543495553,
+      "learning_rate": 9.991781622455723e-06,
+      "loss": 0.6661,
+      "step": 498
+    },
+    {
+      "epoch": 0.1170674486803519,
+      "grad_norm": 2.072243392022778,
+      "learning_rate": 9.991545201119692e-06,
+      "loss": 0.6763,
+      "step": 499
+    },
+    {
+      "epoch": 0.11730205278592376,
+      "grad_norm": 2.0321025061706512,
+      "learning_rate": 9.991305430098471e-06,
+      "loss": 0.6666,
+      "step": 500
+    },
+    {
+      "epoch": 0.11730205278592376,
+      "eval_loss": 0.6888240575790405,
+      "eval_runtime": 28.5702,
+      "eval_samples_per_second": 19.111,
+      "eval_steps_per_second": 0.175,
+      "step": 500
+    },
+    {
+      "epoch": 0.1175366568914956,
+      "grad_norm": 1.6318127853421331,
+      "learning_rate": 9.991062309552965e-06,
+      "loss": 0.6265,
+      "step": 501
+    },
+    {
+      "epoch": 0.11777126099706745,
+      "grad_norm": 1.7473636157017485,
+      "learning_rate": 9.990815839646329e-06,
+      "loss": 0.655,
+      "step": 502
+    },
+    {
+      "epoch": 0.1180058651026393,
+      "grad_norm": 1.5512715760294424,
+      "learning_rate": 9.99056602054396e-06,
+      "loss": 0.668,
+      "step": 503
+    },
+    {
+      "epoch": 0.11824046920821114,
+      "grad_norm": 2.3950227354199374,
+      "learning_rate": 9.990312852413501e-06,
+      "loss": 0.7096,
+      "step": 504
+    },
+    {
+      "epoch": 0.11847507331378299,
+      "grad_norm": 1.5296372584551887,
+      "learning_rate": 9.990056335424853e-06,
+      "loss": 0.6857,
+      "step": 505
+    },
+    {
+      "epoch": 0.11870967741935484,
+      "grad_norm": 1.867314419724177,
+      "learning_rate": 9.98979646975015e-06,
+      "loss": 0.659,
+      "step": 506
+    },
+    {
+      "epoch": 0.11894428152492668,
+      "grad_norm": 2.2755337401213285,
+      "learning_rate": 9.989533255563786e-06,
+      "loss": 0.7274,
+      "step": 507
+    },
+    {
+      "epoch": 0.11917888563049854,
+      "grad_norm": 6.432226473108074,
+      "learning_rate": 9.989266693042394e-06,
+      "loss": 0.6454,
+      "step": 508
+    },
+    {
+      "epoch": 0.11941348973607038,
+      "grad_norm": 7.977964543167525,
+      "learning_rate": 9.988996782364858e-06,
+      "loss": 0.6986,
+      "step": 509
+    },
+    {
+      "epoch": 0.11964809384164223,
+      "grad_norm": 2.492942939386226,
+      "learning_rate": 9.988723523712304e-06,
+      "loss": 0.7261,
+      "step": 510
+    },
+    {
+      "epoch": 0.11988269794721408,
+      "grad_norm": 13.259547924776628,
+      "learning_rate": 9.988446917268113e-06,
+      "loss": 0.6809,
+      "step": 511
+    },
+    {
+      "epoch": 0.12011730205278592,
+      "grad_norm": 2.11113904634305,
+      "learning_rate": 9.988166963217903e-06,
+      "loss": 0.7064,
+      "step": 512
+    },
+    {
+      "epoch": 0.12035190615835777,
+      "grad_norm": 1.3151089527283533,
+      "learning_rate": 9.987883661749548e-06,
+      "loss": 0.7164,
+      "step": 513
+    },
+    {
+      "epoch": 0.12058651026392962,
+      "grad_norm": 2.733472003363916,
+      "learning_rate": 9.987597013053161e-06,
+      "loss": 0.6691,
+      "step": 514
+    },
+    {
+      "epoch": 0.12082111436950146,
+      "grad_norm": 2.1188076613210702,
+      "learning_rate": 9.987307017321105e-06,
+      "loss": 0.6608,
+      "step": 515
+    },
+    {
+      "epoch": 0.12105571847507332,
+      "grad_norm": 1.413454097880752,
+      "learning_rate": 9.987013674747986e-06,
+      "loss": 0.7397,
+      "step": 516
+    },
+    {
+      "epoch": 0.12129032258064516,
+      "grad_norm": 3.0244674122202584,
+      "learning_rate": 9.986716985530658e-06,
+      "loss": 0.6533,
+      "step": 517
+    },
+    {
+      "epoch": 0.121524926686217,
+      "grad_norm": 2.1969109132092246,
+      "learning_rate": 9.986416949868223e-06,
+      "loss": 0.7449,
+      "step": 518
+    },
+    {
+      "epoch": 0.12175953079178886,
+      "grad_norm": 3.233230646965809,
+      "learning_rate": 9.986113567962025e-06,
+      "loss": 0.7096,
+      "step": 519
+    },
+    {
+      "epoch": 0.1219941348973607,
+      "grad_norm": 2.5578549246896296,
+      "learning_rate": 9.985806840015657e-06,
+      "loss": 0.7025,
+      "step": 520
+    },
+    {
+      "epoch": 0.12222873900293255,
+      "grad_norm": 2.379363537670544,
+      "learning_rate": 9.985496766234952e-06,
+      "loss": 0.7108,
+      "step": 521
+    },
+    {
+      "epoch": 0.1224633431085044,
+      "grad_norm": 1.6069661740113346,
+      "learning_rate": 9.985183346827995e-06,
+      "loss": 0.7185,
+      "step": 522
+    },
+    {
+      "epoch": 0.12269794721407624,
+      "grad_norm": 2.870991700048875,
+      "learning_rate": 9.984866582005112e-06,
+      "loss": 0.7037,
+      "step": 523
+    },
+    {
+      "epoch": 0.1229325513196481,
+      "grad_norm": 1.775199433098271,
+      "learning_rate": 9.984546471978872e-06,
+      "loss": 0.6602,
+      "step": 524
+    },
+    {
+      "epoch": 0.12316715542521994,
+      "grad_norm": 2.1573711551915564,
+      "learning_rate": 9.984223016964096e-06,
+      "loss": 0.6794,
+      "step": 525
+    },
+    {
+      "epoch": 0.12340175953079179,
+      "grad_norm": 1.502186193122687,
+      "learning_rate": 9.983896217177843e-06,
+      "loss": 0.6603,
+      "step": 526
+    },
+    {
+      "epoch": 0.12363636363636364,
+      "grad_norm": 2.0934295229608386,
+      "learning_rate": 9.983566072839423e-06,
+      "loss": 0.6766,
+      "step": 527
+    },
+    {
+      "epoch": 0.12387096774193548,
+      "grad_norm": 1.6941035922061842,
+      "learning_rate": 9.98323258417038e-06,
+      "loss": 0.66,
+      "step": 528
+    },
+    {
+      "epoch": 0.12410557184750733,
+      "grad_norm": 2.165000453568371,
+      "learning_rate": 9.982895751394515e-06,
+      "loss": 0.7038,
+      "step": 529
+    },
+    {
+      "epoch": 0.12434017595307918,
+      "grad_norm": 2.043357882178577,
+      "learning_rate": 9.982555574737862e-06,
+      "loss": 0.6648,
+      "step": 530
+    },
+    {
+      "epoch": 0.12457478005865102,
+      "grad_norm": 1.895714786932282,
+      "learning_rate": 9.982212054428708e-06,
+      "loss": 0.7147,
+      "step": 531
+    },
+    {
+      "epoch": 0.12480938416422288,
+      "grad_norm": 2.654530541516175,
+      "learning_rate": 9.981865190697577e-06,
+      "loss": 0.7089,
+      "step": 532
+    },
+    {
+      "epoch": 0.1250439882697947,
+      "grad_norm": 1.597405747196785,
+      "learning_rate": 9.98151498377724e-06,
+      "loss": 0.7359,
+      "step": 533
+    },
+    {
+      "epoch": 0.12527859237536657,
+      "grad_norm": 2.9749176981598384,
+      "learning_rate": 9.981161433902712e-06,
+      "loss": 0.6893,
+      "step": 534
+    },
+    {
+      "epoch": 0.12551319648093842,
+      "grad_norm": 2.4159566695813104,
+      "learning_rate": 9.980804541311252e-06,
+      "loss": 0.7106,
+      "step": 535
+    },
+    {
+      "epoch": 0.12574780058651028,
+      "grad_norm": 22.842593769056897,
+      "learning_rate": 9.980444306242355e-06,
+      "loss": 0.7076,
+      "step": 536
+    },
+    {
+      "epoch": 0.1259824046920821,
+      "grad_norm": 2.666099139107497,
+      "learning_rate": 9.98008072893777e-06,
+      "loss": 0.6791,
+      "step": 537
+    },
+    {
+      "epoch": 0.12621700879765396,
+      "grad_norm": 1.6269471981229704,
+      "learning_rate": 9.979713809641482e-06,
+      "loss": 0.6798,
+      "step": 538
+    },
+    {
+      "epoch": 0.12645161290322582,
+      "grad_norm": 4.202791295534201,
+      "learning_rate": 9.97934354859972e-06,
+      "loss": 0.6882,
+      "step": 539
+    },
+    {
+      "epoch": 0.12668621700879765,
+      "grad_norm": 1.5103425993192718,
+      "learning_rate": 9.978969946060954e-06,
+      "loss": 0.6643,
+      "step": 540
+    },
+    {
+      "epoch": 0.1269208211143695,
+      "grad_norm": 1.596861789287424,
+      "learning_rate": 9.9785930022759e-06,
+      "loss": 0.697,
+      "step": 541
+    },
+    {
+      "epoch": 0.12715542521994136,
+      "grad_norm": 1.1667616716218816,
+      "learning_rate": 9.978212717497515e-06,
+      "loss": 0.7428,
+      "step": 542
+    },
+    {
+      "epoch": 0.1273900293255132,
+      "grad_norm": 13.354063943546418,
+      "learning_rate": 9.977829091980996e-06,
+      "loss": 0.6302,
+      "step": 543
+    },
+    {
+      "epoch": 0.12762463343108504,
+      "grad_norm": 2.3558509951518074,
+      "learning_rate": 9.977442125983781e-06,
+      "loss": 0.6839,
+      "step": 544
+    },
+    {
+      "epoch": 0.1278592375366569,
+      "grad_norm": 3.1414319250785487,
+      "learning_rate": 9.977051819765558e-06,
+      "loss": 0.7044,
+      "step": 545
+    },
+    {
+      "epoch": 0.12809384164222873,
+      "grad_norm": 3.1115702760435413,
+      "learning_rate": 9.976658173588244e-06,
+      "loss": 0.7383,
+      "step": 546
+    },
+    {
+      "epoch": 0.12832844574780058,
+      "grad_norm": 3.5567349181056316,
+      "learning_rate": 9.976261187716008e-06,
+      "loss": 0.6763,
+      "step": 547
+    },
+    {
+      "epoch": 0.12856304985337244,
+      "grad_norm": 2.3978902826675954,
+      "learning_rate": 9.975860862415255e-06,
+      "loss": 0.7546,
+      "step": 548
+    },
+    {
+      "epoch": 0.12879765395894427,
+      "grad_norm": 29.013203676935372,
+      "learning_rate": 9.975457197954632e-06,
+      "loss": 0.7034,
+      "step": 549
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 2.0575848372214307,
+      "learning_rate": 9.975050194605027e-06,
+      "loss": 0.6951,
+      "step": 550
+    },
+    {
+      "epoch": 0.12926686217008798,
+      "grad_norm": 1.4098369621927678,
+      "learning_rate": 9.974639852639568e-06,
+      "loss": 0.6922,
+      "step": 551
+    },
+    {
+      "epoch": 0.12950146627565984,
+      "grad_norm": 2.897930701241758,
+      "learning_rate": 9.974226172333623e-06,
+      "loss": 0.7343,
+      "step": 552
+    },
+    {
+      "epoch": 0.12973607038123167,
+      "grad_norm": 2.59537067559108,
+      "learning_rate": 9.973809153964803e-06,
+      "loss": 0.6689,
+      "step": 553
+    },
+    {
+      "epoch": 0.12997067448680352,
+      "grad_norm": 3.027373892989576,
+      "learning_rate": 9.973388797812958e-06,
+      "loss": 0.6718,
+      "step": 554
+    },
+    {
+      "epoch": 0.13020527859237538,
+      "grad_norm": 2.113493209589616,
+      "learning_rate": 9.972965104160175e-06,
+      "loss": 0.6639,
+      "step": 555
+    },
+    {
+      "epoch": 0.1304398826979472,
+      "grad_norm": 1.3623863166221668,
+      "learning_rate": 9.972538073290783e-06,
+      "loss": 0.6278,
+      "step": 556
+    },
+    {
+      "epoch": 0.13067448680351906,
+      "grad_norm": 2.5295207930233685,
+      "learning_rate": 9.972107705491351e-06,
+      "loss": 0.6006,
+      "step": 557
+    },
+    {
+      "epoch": 0.13090909090909092,
+      "grad_norm": 1.6585614514698723,
+      "learning_rate": 9.971674001050687e-06,
+      "loss": 0.6385,
+      "step": 558
+    },
+    {
+      "epoch": 0.13114369501466275,
+      "grad_norm": 6.160387588977871,
+      "learning_rate": 9.971236960259836e-06,
+      "loss": 0.6728,
+      "step": 559
+    },
+    {
+      "epoch": 0.1313782991202346,
+      "grad_norm": 2.4013303577531806,
+      "learning_rate": 9.970796583412088e-06,
+      "loss": 0.6698,
+      "step": 560
+    },
+    {
+      "epoch": 0.13161290322580646,
+      "grad_norm": 1.8285169637361147,
+      "learning_rate": 9.970352870802962e-06,
+      "loss": 0.6668,
+      "step": 561
+    },
+    {
+      "epoch": 0.1318475073313783,
+      "grad_norm": 1.303920615682513,
+      "learning_rate": 9.969905822730226e-06,
+      "loss": 0.6264,
+      "step": 562
+    },
+    {
+      "epoch": 0.13208211143695014,
+      "grad_norm": 1.9978814715093303,
+      "learning_rate": 9.969455439493877e-06,
+      "loss": 0.6987,
+      "step": 563
+    },
+    {
+      "epoch": 0.132316715542522,
+      "grad_norm": 7.972031582327647,
+      "learning_rate": 9.969001721396157e-06,
+      "loss": 0.7122,
+      "step": 564
+    },
+    {
+      "epoch": 0.13255131964809383,
+      "grad_norm": 2.1939819423538207,
+      "learning_rate": 9.968544668741544e-06,
+      "loss": 0.7235,
+      "step": 565
+    },
+    {
+      "epoch": 0.13278592375366569,
+      "grad_norm": 2.111594241775808,
+      "learning_rate": 9.968084281836752e-06,
+      "loss": 0.7514,
+      "step": 566
+    },
+    {
+      "epoch": 0.13302052785923754,
+      "grad_norm": 1.6540698615973652,
+      "learning_rate": 9.967620560990733e-06,
+      "loss": 0.6209,
+      "step": 567
+    },
+    {
+      "epoch": 0.1332551319648094,
+      "grad_norm": 1.734001307851314,
+      "learning_rate": 9.967153506514677e-06,
+      "loss": 0.6821,
+      "step": 568
+    },
+    {
+      "epoch": 0.13348973607038123,
+      "grad_norm": 2.898214053367665,
+      "learning_rate": 9.966683118722014e-06,
+      "loss": 0.6357,
+      "step": 569
+    },
+    {
+      "epoch": 0.13372434017595308,
+      "grad_norm": 2.8199846284828034,
+      "learning_rate": 9.966209397928404e-06,
+      "loss": 0.669,
+      "step": 570
+    },
+    {
+      "epoch": 0.13395894428152494,
+      "grad_norm": 3.209199942211371,
+      "learning_rate": 9.965732344451751e-06,
+      "loss": 0.683,
+      "step": 571
+    },
+    {
+      "epoch": 0.13419354838709677,
+      "grad_norm": 5.56090036713883,
+      "learning_rate": 9.96525195861219e-06,
+      "loss": 0.7191,
+      "step": 572
+    },
+    {
+      "epoch": 0.13442815249266862,
+      "grad_norm": 2.5431957689833666,
+      "learning_rate": 9.964768240732094e-06,
+      "loss": 0.692,
+      "step": 573
+    },
+    {
+      "epoch": 0.13466275659824048,
+      "grad_norm": 1.4805250152975054,
+      "learning_rate": 9.964281191136073e-06,
+      "loss": 0.7041,
+      "step": 574
+    },
+    {
+      "epoch": 0.1348973607038123,
+      "grad_norm": 3.560789408514976,
+      "learning_rate": 9.963790810150975e-06,
+      "loss": 0.653,
+      "step": 575
+    },
+    {
+      "epoch": 0.13513196480938416,
+      "grad_norm": 3.372954282563982,
+      "learning_rate": 9.963297098105877e-06,
+      "loss": 0.6656,
+      "step": 576
+    },
+    {
+      "epoch": 0.13536656891495602,
+      "grad_norm": 1.716873774016931,
+      "learning_rate": 9.962800055332098e-06,
+      "loss": 0.6824,
+      "step": 577
+    },
+    {
+      "epoch": 0.13560117302052785,
+      "grad_norm": 1.5706571563062237,
+      "learning_rate": 9.962299682163185e-06,
+      "loss": 0.6542,
+      "step": 578
+    },
+    {
+      "epoch": 0.1358357771260997,
+      "grad_norm": 4.257196400759658,
+      "learning_rate": 9.961795978934932e-06,
+      "loss": 0.6593,
+      "step": 579
+    },
+    {
+      "epoch": 0.13607038123167156,
+      "grad_norm": 1.6639609965913367,
+      "learning_rate": 9.961288945985352e-06,
+      "loss": 0.647,
+      "step": 580
+    },
+    {
+      "epoch": 0.1363049853372434,
+      "grad_norm": 25.232129147256217,
+      "learning_rate": 9.960778583654704e-06,
+      "loss": 0.6564,
+      "step": 581
+    },
+    {
+      "epoch": 0.13653958944281525,
+      "grad_norm": 1.2901397862502486,
+      "learning_rate": 9.960264892285481e-06,
+      "loss": 0.658,
+      "step": 582
+    },
+    {
+      "epoch": 0.1367741935483871,
+      "grad_norm": 1.8257065820782765,
+      "learning_rate": 9.9597478722224e-06,
+      "loss": 0.7135,
+      "step": 583
+    },
+    {
+      "epoch": 0.13700879765395896,
+      "grad_norm": 1.1847585123600912,
+      "learning_rate": 9.959227523812423e-06,
+      "loss": 0.6541,
+      "step": 584
+    },
+    {
+      "epoch": 0.1372434017595308,
+      "grad_norm": 1.6699141702923748,
+      "learning_rate": 9.958703847404742e-06,
+      "loss": 0.7057,
+      "step": 585
+    },
+    {
+      "epoch": 0.13747800586510264,
+      "grad_norm": 1.2962710057457882,
+      "learning_rate": 9.958176843350778e-06,
+      "loss": 0.6709,
+      "step": 586
+    },
+    {
+      "epoch": 0.1377126099706745,
+      "grad_norm": 2.457488711215677,
+      "learning_rate": 9.957646512004191e-06,
+      "loss": 0.6516,
+      "step": 587
+    },
+    {
+      "epoch": 0.13794721407624633,
+      "grad_norm": 1.8892307626675284,
+      "learning_rate": 9.95711285372087e-06,
+      "loss": 0.6908,
+      "step": 588
+    },
+    {
+      "epoch": 0.13818181818181818,
+      "grad_norm": 4.763454832394259,
+      "learning_rate": 9.95657586885894e-06,
+      "loss": 0.6708,
+      "step": 589
+    },
+    {
+      "epoch": 0.13841642228739004,
+      "grad_norm": 8.020713129219255,
+      "learning_rate": 9.956035557778753e-06,
+      "loss": 0.6752,
+      "step": 590
+    },
+    {
+      "epoch": 0.13865102639296187,
+      "grad_norm": 3.4908925487869467,
+      "learning_rate": 9.955491920842902e-06,
+      "loss": 0.6909,
+      "step": 591
+    },
+    {
+      "epoch": 0.13888563049853372,
+      "grad_norm": 1.776193850374824,
+      "learning_rate": 9.9549449584162e-06,
+      "loss": 0.6748,
+      "step": 592
+    },
+    {
+      "epoch": 0.13912023460410558,
+      "grad_norm": 3.1813693527805396,
+      "learning_rate": 9.954394670865701e-06,
+      "loss": 0.66,
+      "step": 593
+    },
+    {
+      "epoch": 0.1393548387096774,
+      "grad_norm": 1.342747330197621,
+      "learning_rate": 9.95384105856069e-06,
+      "loss": 0.6375,
+      "step": 594
+    },
+    {
+      "epoch": 0.13958944281524927,
+      "grad_norm": 1.897633631785078,
+      "learning_rate": 9.953284121872676e-06,
+      "loss": 0.6633,
+      "step": 595
+    },
+    {
+      "epoch": 0.13982404692082112,
+      "grad_norm": 1.5089958808746635,
+      "learning_rate": 9.952723861175406e-06,
+      "loss": 0.6554,
+      "step": 596
+    },
+    {
+      "epoch": 0.14005865102639295,
+      "grad_norm": 2.07875619918608,
+      "learning_rate": 9.952160276844855e-06,
+      "loss": 0.7027,
+      "step": 597
+    },
+    {
+      "epoch": 0.1402932551319648,
+      "grad_norm": 19.9773501877953,
+      "learning_rate": 9.95159336925923e-06,
+      "loss": 0.7008,
+      "step": 598
+    },
+    {
+      "epoch": 0.14052785923753666,
+      "grad_norm": 3.149299323527888,
+      "learning_rate": 9.951023138798964e-06,
+      "loss": 0.7071,
+      "step": 599
+    },
+    {
+      "epoch": 0.14076246334310852,
+      "grad_norm": 1.662461198147052,
+      "learning_rate": 9.950449585846722e-06,
+      "loss": 0.6658,
+      "step": 600
+    },
+    {
+      "epoch": 0.14099706744868035,
+      "grad_norm": 1.533295748310548,
+      "learning_rate": 9.949872710787402e-06,
+      "loss": 0.684,
+      "step": 601
+    },
+    {
+      "epoch": 0.1412316715542522,
+      "grad_norm": 1.7816342527457778,
+      "learning_rate": 9.949292514008127e-06,
+      "loss": 0.6591,
+      "step": 602
+    },
+    {
+      "epoch": 0.14146627565982406,
+      "grad_norm": 3.1159357600480035,
+      "learning_rate": 9.948708995898251e-06,
+      "loss": 0.6505,
+      "step": 603
+    },
+    {
+      "epoch": 0.1417008797653959,
+      "grad_norm": 6.279947901045703,
+      "learning_rate": 9.948122156849357e-06,
+      "loss": 0.7105,
+      "step": 604
+    },
+    {
+      "epoch": 0.14193548387096774,
+      "grad_norm": 1.4514443812808602,
+      "learning_rate": 9.947531997255256e-06,
+      "loss": 0.6555,
+      "step": 605
+    },
+    {
+      "epoch": 0.1421700879765396,
+      "grad_norm": 16.123057899871036,
+      "learning_rate": 9.946938517511988e-06,
+      "loss": 0.6481,
+      "step": 606
+    },
+    {
+      "epoch": 0.14240469208211143,
+      "grad_norm": 1.8485177049511416,
+      "learning_rate": 9.946341718017817e-06,
+      "loss": 0.6444,
+      "step": 607
+    },
+    {
+      "epoch": 0.14263929618768328,
+      "grad_norm": 1.381903063248314,
+      "learning_rate": 9.945741599173244e-06,
+      "loss": 0.6847,
+      "step": 608
+    },
+    {
+      "epoch": 0.14287390029325514,
+      "grad_norm": 1.9733424925107392,
+      "learning_rate": 9.945138161380987e-06,
+      "loss": 0.6566,
+      "step": 609
+    },
+    {
+      "epoch": 0.14310850439882697,
+      "grad_norm": 1.4736215812750697,
+      "learning_rate": 9.944531405046001e-06,
+      "loss": 0.6727,
+      "step": 610
+    },
+    {
+      "epoch": 0.14334310850439883,
+      "grad_norm": 1.674427950216094,
+      "learning_rate": 9.943921330575457e-06,
+      "loss": 0.6949,
+      "step": 611
+    },
+    {
+      "epoch": 0.14357771260997068,
+      "grad_norm": 1.353555356656059,
+      "learning_rate": 9.943307938378762e-06,
+      "loss": 0.6239,
+      "step": 612
+    },
+    {
+      "epoch": 0.1438123167155425,
+      "grad_norm": 6.462783821793124,
+      "learning_rate": 9.942691228867548e-06,
+      "loss": 0.6735,
+      "step": 613
+    },
+    {
+      "epoch": 0.14404692082111437,
+      "grad_norm": 1.8066326048928583,
+      "learning_rate": 9.942071202455667e-06,
+      "loss": 0.6635,
+      "step": 614
+    },
+    {
+      "epoch": 0.14428152492668622,
+      "grad_norm": 7.481709632140974,
+      "learning_rate": 9.941447859559204e-06,
+      "loss": 0.7008,
+      "step": 615
+    },
+    {
+      "epoch": 0.14451612903225808,
+      "grad_norm": 1.868185126323605,
+      "learning_rate": 9.940821200596466e-06,
+      "loss": 0.7114,
+      "step": 616
+    },
+    {
+      "epoch": 0.1447507331378299,
+      "grad_norm": 10.616230292276052,
+      "learning_rate": 9.94019122598799e-06,
+      "loss": 0.6753,
+      "step": 617
+    },
+    {
+      "epoch": 0.14498533724340176,
+      "grad_norm": 1.6889062877523384,
+      "learning_rate": 9.939557936156527e-06,
+      "loss": 0.6832,
+      "step": 618
+    },
+    {
+      "epoch": 0.14521994134897362,
+      "grad_norm": 1.3246938175496756,
+      "learning_rate": 9.938921331527066e-06,
+      "loss": 0.6755,
+      "step": 619
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 1.5267349547650975,
+      "learning_rate": 9.938281412526811e-06,
+      "loss": 0.7326,
+      "step": 620
+    },
+    {
+      "epoch": 0.1456891495601173,
+      "grad_norm": 1.3327513478934103,
+      "learning_rate": 9.937638179585195e-06,
+      "loss": 0.621,
+      "step": 621
+    },
+    {
+      "epoch": 0.14592375366568916,
+      "grad_norm": 3.3950719442877033,
+      "learning_rate": 9.936991633133872e-06,
+      "loss": 0.6918,
+      "step": 622
+    },
+    {
+      "epoch": 0.146158357771261,
+      "grad_norm": 2.281479467961296,
+      "learning_rate": 9.936341773606723e-06,
+      "loss": 0.6954,
+      "step": 623
+    },
+    {
+      "epoch": 0.14639296187683284,
+      "grad_norm": 1.1802407354424511,
+      "learning_rate": 9.93568860143985e-06,
+      "loss": 0.7372,
+      "step": 624
+    },
+    {
+      "epoch": 0.1466275659824047,
+      "grad_norm": 2.315080643869437,
+      "learning_rate": 9.93503211707158e-06,
+      "loss": 0.6501,
+      "step": 625
+    },
+    {
+      "epoch": 0.14686217008797653,
+      "grad_norm": 5.117637311109828,
+      "learning_rate": 9.934372320942455e-06,
+      "loss": 0.6691,
+      "step": 626
+    },
+    {
+      "epoch": 0.14709677419354839,
+      "grad_norm": 7.696774629839776,
+      "learning_rate": 9.933709213495253e-06,
+      "loss": 0.7046,
+      "step": 627
+    },
+    {
+      "epoch": 0.14733137829912024,
+      "grad_norm": 1.4226681215442163,
+      "learning_rate": 9.933042795174964e-06,
+      "loss": 0.684,
+      "step": 628
+    },
+    {
+      "epoch": 0.14756598240469207,
+      "grad_norm": 2.984733020473841,
+      "learning_rate": 9.9323730664288e-06,
+      "loss": 0.6815,
+      "step": 629
+    },
+    {
+      "epoch": 0.14780058651026393,
+      "grad_norm": 3.935812255521951,
+      "learning_rate": 9.931700027706199e-06,
+      "loss": 0.7096,
+      "step": 630
+    },
+    {
+      "epoch": 0.14803519061583578,
+      "grad_norm": 2.4457763134413084,
+      "learning_rate": 9.93102367945882e-06,
+      "loss": 0.6311,
+      "step": 631
+    },
+    {
+      "epoch": 0.14826979472140764,
+      "grad_norm": 1.4359062113524443,
+      "learning_rate": 9.930344022140539e-06,
+      "loss": 0.6773,
+      "step": 632
+    },
+    {
+      "epoch": 0.14850439882697947,
+      "grad_norm": 1.2100526324980292,
+      "learning_rate": 9.929661056207455e-06,
+      "loss": 0.6622,
+      "step": 633
+    },
+    {
+      "epoch": 0.14873900293255132,
+      "grad_norm": 1.0190753291928116,
+      "learning_rate": 9.928974782117889e-06,
+      "loss": 0.6981,
+      "step": 634
+    },
+    {
+      "epoch": 0.14897360703812318,
+      "grad_norm": 2.799854932121048,
+      "learning_rate": 9.928285200332377e-06,
+      "loss": 0.6984,
+      "step": 635
+    },
+    {
+      "epoch": 0.149208211143695,
+      "grad_norm": 2.905366083267552,
+      "learning_rate": 9.927592311313681e-06,
+      "loss": 0.6738,
+      "step": 636
+    },
+    {
+      "epoch": 0.14944281524926686,
+      "grad_norm": 2.0090448417142257,
+      "learning_rate": 9.926896115526777e-06,
+      "loss": 0.6518,
+      "step": 637
+    },
+    {
+      "epoch": 0.14967741935483872,
+      "grad_norm": 3.414729484354847,
+      "learning_rate": 9.926196613438866e-06,
+      "loss": 0.6537,
+      "step": 638
+    },
+    {
+      "epoch": 0.14991202346041055,
+      "grad_norm": 2.6622638389494115,
+      "learning_rate": 9.925493805519359e-06,
+      "loss": 0.6597,
+      "step": 639
+    },
+    {
+      "epoch": 0.1501466275659824,
+      "grad_norm": 2.7737320980633533,
+      "learning_rate": 9.924787692239892e-06,
+      "loss": 0.6643,
+      "step": 640
+    },
+    {
+      "epoch": 0.15038123167155426,
+      "grad_norm": 1.5644793983992447,
+      "learning_rate": 9.92407827407432e-06,
+      "loss": 0.6744,
+      "step": 641
+    },
+    {
+      "epoch": 0.1506158357771261,
+      "grad_norm": 1.6708319503739981,
+      "learning_rate": 9.923365551498713e-06,
+      "loss": 0.7092,
+      "step": 642
+    },
+    {
+      "epoch": 0.15085043988269795,
+      "grad_norm": 2.016621110195836,
+      "learning_rate": 9.922649524991359e-06,
+      "loss": 0.7289,
+      "step": 643
+    },
+    {
+      "epoch": 0.1510850439882698,
+      "grad_norm": 2.504724838609121,
+      "learning_rate": 9.921930195032763e-06,
+      "loss": 0.7298,
+      "step": 644
+    },
+    {
+      "epoch": 0.15131964809384163,
+      "grad_norm": 1.4473345607651253,
+      "learning_rate": 9.921207562105643e-06,
+      "loss": 0.6649,
+      "step": 645
+    },
+    {
+      "epoch": 0.1515542521994135,
+      "grad_norm": 1.3709821954549388,
+      "learning_rate": 9.920481626694945e-06,
+      "loss": 0.6696,
+      "step": 646
+    },
+    {
+      "epoch": 0.15178885630498534,
+      "grad_norm": 1.8706584965395907,
+      "learning_rate": 9.919752389287818e-06,
+      "loss": 0.6705,
+      "step": 647
+    },
+    {
+      "epoch": 0.1520234604105572,
+      "grad_norm": 2.249129643505635,
+      "learning_rate": 9.919019850373635e-06,
+      "loss": 0.6147,
+      "step": 648
+    },
+    {
+      "epoch": 0.15225806451612903,
+      "grad_norm": 2.097343552522159,
+      "learning_rate": 9.918284010443981e-06,
+      "loss": 0.6334,
+      "step": 649
+    },
+    {
+      "epoch": 0.15249266862170088,
+      "grad_norm": 1.8768058707443809,
+      "learning_rate": 9.91754486999266e-06,
+      "loss": 0.7041,
+      "step": 650
+    },
+    {
+      "epoch": 0.15272727272727274,
+      "grad_norm": 4.491233851879105,
+      "learning_rate": 9.916802429515685e-06,
+      "loss": 0.681,
+      "step": 651
+    },
+    {
+      "epoch": 0.15296187683284457,
+      "grad_norm": 1.533031987182923,
+      "learning_rate": 9.916056689511287e-06,
+      "loss": 0.6536,
+      "step": 652
+    },
+    {
+      "epoch": 0.15319648093841642,
+      "grad_norm": 1.759911764487037,
+      "learning_rate": 9.915307650479915e-06,
+      "loss": 0.6293,
+      "step": 653
+    },
+    {
+      "epoch": 0.15343108504398828,
+      "grad_norm": 1.2374310964542206,
+      "learning_rate": 9.914555312924223e-06,
+      "loss": 0.7126,
+      "step": 654
+    },
+    {
+      "epoch": 0.1536656891495601,
+      "grad_norm": 1.8037339536830177,
+      "learning_rate": 9.913799677349087e-06,
+      "loss": 0.6788,
+      "step": 655
+    },
+    {
+      "epoch": 0.15390029325513196,
+      "grad_norm": 2.2873799899386493,
+      "learning_rate": 9.913040744261594e-06,
+      "loss": 0.6878,
+      "step": 656
+    },
+    {
+      "epoch": 0.15413489736070382,
+      "grad_norm": 8.552252314012465,
+      "learning_rate": 9.912278514171037e-06,
+      "loss": 0.6558,
+      "step": 657
+    },
+    {
+      "epoch": 0.15436950146627565,
+      "grad_norm": 4.043452822531138,
+      "learning_rate": 9.911512987588932e-06,
+      "loss": 0.6074,
+      "step": 658
+    },
+    {
+      "epoch": 0.1546041055718475,
+      "grad_norm": 1.5929166007016708,
+      "learning_rate": 9.910744165028998e-06,
+      "loss": 0.6456,
+      "step": 659
+    },
+    {
+      "epoch": 0.15483870967741936,
+      "grad_norm": 1.322345644985473,
+      "learning_rate": 9.909972047007176e-06,
+      "loss": 0.6527,
+      "step": 660
+    },
+    {
+      "epoch": 0.1550733137829912,
+      "grad_norm": 1.9134269361860932,
+      "learning_rate": 9.909196634041607e-06,
+      "loss": 0.6721,
+      "step": 661
+    },
+    {
+      "epoch": 0.15530791788856305,
+      "grad_norm": 3.2279892391112743,
+      "learning_rate": 9.908417926652653e-06,
+      "loss": 0.7016,
+      "step": 662
+    },
+    {
+      "epoch": 0.1555425219941349,
+      "grad_norm": 4.605703097073935,
+      "learning_rate": 9.90763592536288e-06,
+      "loss": 0.7173,
+      "step": 663
+    },
+    {
+      "epoch": 0.15577712609970676,
+      "grad_norm": 2.2844915719419006,
+      "learning_rate": 9.906850630697068e-06,
+      "loss": 0.684,
+      "step": 664
+    },
+    {
+      "epoch": 0.1560117302052786,
+      "grad_norm": 1.3671944830965983,
+      "learning_rate": 9.906062043182205e-06,
+      "loss": 0.7538,
+      "step": 665
+    },
+    {
+      "epoch": 0.15624633431085044,
+      "grad_norm": 17.315048588685254,
+      "learning_rate": 9.905270163347491e-06,
+      "loss": 0.6221,
+      "step": 666
+    },
+    {
+      "epoch": 0.1564809384164223,
+      "grad_norm": 1.1715675455606274,
+      "learning_rate": 9.904474991724335e-06,
+      "loss": 0.6396,
+      "step": 667
+    },
+    {
+      "epoch": 0.15671554252199413,
+      "grad_norm": 1.7500860284386124,
+      "learning_rate": 9.903676528846353e-06,
+      "loss": 0.6947,
+      "step": 668
+    },
+    {
+      "epoch": 0.15695014662756598,
+      "grad_norm": 1.3131765793711139,
+      "learning_rate": 9.90287477524937e-06,
+      "loss": 0.7223,
+      "step": 669
+    },
+    {
+      "epoch": 0.15718475073313784,
+      "grad_norm": 5.544168444593145,
+      "learning_rate": 9.902069731471422e-06,
+      "loss": 0.6371,
+      "step": 670
+    },
+    {
+      "epoch": 0.15741935483870967,
+      "grad_norm": 1.5030904405800865,
+      "learning_rate": 9.90126139805275e-06,
+      "loss": 0.6581,
+      "step": 671
+    },
+    {
+      "epoch": 0.15765395894428152,
+      "grad_norm": 2.211411041740967,
+      "learning_rate": 9.900449775535802e-06,
+      "loss": 0.6776,
+      "step": 672
+    },
+    {
+      "epoch": 0.15788856304985338,
+      "grad_norm": 101.66172969743218,
+      "learning_rate": 9.899634864465242e-06,
+      "loss": 0.63,
+      "step": 673
+    },
+    {
+      "epoch": 0.1581231671554252,
+      "grad_norm": 1.6269899042718914,
+      "learning_rate": 9.898816665387924e-06,
+      "loss": 0.6985,
+      "step": 674
+    },
+    {
+      "epoch": 0.15835777126099707,
+      "grad_norm": 25.473981421672807,
+      "learning_rate": 9.897995178852927e-06,
+      "loss": 0.6415,
+      "step": 675
+    },
+    {
+      "epoch": 0.15859237536656892,
+      "grad_norm": 2.7339580672894845,
+      "learning_rate": 9.897170405411522e-06,
+      "loss": 0.6727,
+      "step": 676
+    },
+    {
+      "epoch": 0.15882697947214075,
+      "grad_norm": 4.514988151322233,
+      "learning_rate": 9.896342345617193e-06,
+      "loss": 0.6098,
+      "step": 677
+    },
+    {
+      "epoch": 0.1590615835777126,
+      "grad_norm": 7.419269783067423,
+      "learning_rate": 9.89551100002563e-06,
+      "loss": 0.7198,
+      "step": 678
+    },
+    {
+      "epoch": 0.15929618768328446,
+      "grad_norm": 2.6347918463592412,
+      "learning_rate": 9.89467636919472e-06,
+      "loss": 0.6904,
+      "step": 679
+    },
+    {
+      "epoch": 0.15953079178885632,
+      "grad_norm": 1.6903424361035653,
+      "learning_rate": 9.893838453684566e-06,
+      "loss": 0.6073,
+      "step": 680
+    },
+    {
+      "epoch": 0.15976539589442815,
+      "grad_norm": 4.683500265566674,
+      "learning_rate": 9.892997254057463e-06,
+      "loss": 0.6521,
+      "step": 681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.898678756553585,
+      "learning_rate": 9.892152770877923e-06,
+      "loss": 0.7192,
+      "step": 682
+    },
+    {
+      "epoch": 0.16023460410557186,
+      "grad_norm": 4.566830073066921,
+      "learning_rate": 9.891305004712652e-06,
+      "loss": 0.6814,
+      "step": 683
+    },
+    {
+      "epoch": 0.1604692082111437,
+      "grad_norm": 1.5554337393038395,
+      "learning_rate": 9.890453956130562e-06,
+      "loss": 0.6378,
+      "step": 684
+    },
+    {
+      "epoch": 0.16070381231671554,
+      "grad_norm": 1.6568419750114889,
+      "learning_rate": 9.889599625702765e-06,
+      "loss": 0.6609,
+      "step": 685
+    },
+    {
+      "epoch": 0.1609384164222874,
+      "grad_norm": 1.6968141200340794,
+      "learning_rate": 9.888742014002583e-06,
+      "loss": 0.6211,
+      "step": 686
+    },
+    {
+      "epoch": 0.16117302052785923,
+      "grad_norm": 1.1858462092018027,
+      "learning_rate": 9.887881121605532e-06,
+      "loss": 0.677,
+      "step": 687
+    },
+    {
+      "epoch": 0.16140762463343108,
+      "grad_norm": 1.7694121781659031,
+      "learning_rate": 9.887016949089334e-06,
+      "loss": 0.6674,
+      "step": 688
+    },
+    {
+      "epoch": 0.16164222873900294,
+      "grad_norm": 2.193857961379748,
+      "learning_rate": 9.88614949703391e-06,
+      "loss": 0.676,
+      "step": 689
+    },
+    {
+      "epoch": 0.16187683284457477,
+      "grad_norm": 1.0564631760371397,
+      "learning_rate": 9.885278766021383e-06,
+      "loss": 0.68,
+      "step": 690
+    },
+    {
+      "epoch": 0.16211143695014663,
+      "grad_norm": 4.093060650971836,
+      "learning_rate": 9.884404756636074e-06,
+      "loss": 0.6445,
+      "step": 691
+    },
+    {
+      "epoch": 0.16234604105571848,
+      "grad_norm": 3.4529279760731915,
+      "learning_rate": 9.88352746946451e-06,
+      "loss": 0.6595,
+      "step": 692
+    },
+    {
+      "epoch": 0.1625806451612903,
+      "grad_norm": 1.7772841785991123,
+      "learning_rate": 9.88264690509541e-06,
+      "loss": 0.649,
+      "step": 693
+    },
+    {
+      "epoch": 0.16281524926686217,
+      "grad_norm": 2.349712763219154,
+      "learning_rate": 9.881763064119697e-06,
+      "loss": 0.6299,
+      "step": 694
+    },
+    {
+      "epoch": 0.16304985337243402,
+      "grad_norm": 1.954528333429281,
+      "learning_rate": 9.880875947130492e-06,
+      "loss": 0.6594,
+      "step": 695
+    },
+    {
+      "epoch": 0.16328445747800588,
+      "grad_norm": 1.588185836220011,
+      "learning_rate": 9.879985554723116e-06,
+      "loss": 0.6469,
+      "step": 696
+    },
+    {
+      "epoch": 0.1635190615835777,
+      "grad_norm": 1.774096724121723,
+      "learning_rate": 9.879091887495083e-06,
+      "loss": 0.6966,
+      "step": 697
+    },
+    {
+      "epoch": 0.16375366568914956,
+      "grad_norm": 1.328322615162832,
+      "learning_rate": 9.87819494604611e-06,
+      "loss": 0.6754,
+      "step": 698
+    },
+    {
+      "epoch": 0.16398826979472142,
+      "grad_norm": 2.221119835330553,
+      "learning_rate": 9.877294730978108e-06,
+      "loss": 0.6782,
+      "step": 699
+    },
+    {
+      "epoch": 0.16422287390029325,
+      "grad_norm": 1.4702304509045119,
+      "learning_rate": 9.876391242895188e-06,
+      "loss": 0.691,
+      "step": 700
+    },
+    {
+      "epoch": 0.1644574780058651,
+      "grad_norm": 1.9739830661007176,
+      "learning_rate": 9.875484482403653e-06,
+      "loss": 0.6662,
+      "step": 701
+    },
+    {
+      "epoch": 0.16469208211143696,
+      "grad_norm": 2.2302699298247317,
+      "learning_rate": 9.874574450112005e-06,
+      "loss": 0.608,
+      "step": 702
+    },
+    {
+      "epoch": 0.1649266862170088,
+      "grad_norm": 1.8007165172025146,
+      "learning_rate": 9.87366114663094e-06,
+      "loss": 0.6431,
+      "step": 703
+    },
+    {
+      "epoch": 0.16516129032258065,
+      "grad_norm": 1.4797559173895927,
+      "learning_rate": 9.872744572573352e-06,
+      "loss": 0.5899,
+      "step": 704
+    },
+    {
+      "epoch": 0.1653958944281525,
+      "grad_norm": 2.371687957119544,
+      "learning_rate": 9.871824728554329e-06,
+      "loss": 0.7164,
+      "step": 705
+    },
+    {
+      "epoch": 0.16563049853372433,
+      "grad_norm": 3.6577081342450564,
+      "learning_rate": 9.87090161519115e-06,
+      "loss": 0.6789,
+      "step": 706
+    },
+    {
+      "epoch": 0.16586510263929619,
+      "grad_norm": 1.3932301047626219,
+      "learning_rate": 9.869975233103291e-06,
+      "loss": 0.6866,
+      "step": 707
+    },
+    {
+      "epoch": 0.16609970674486804,
+      "grad_norm": 2.0554334361093876,
+      "learning_rate": 9.869045582912419e-06,
+      "loss": 0.6793,
+      "step": 708
+    },
+    {
+      "epoch": 0.16633431085043987,
+      "grad_norm": 2.224053361060325,
+      "learning_rate": 9.868112665242401e-06,
+      "loss": 0.6999,
+      "step": 709
+    },
+    {
+      "epoch": 0.16656891495601173,
+      "grad_norm": 1.6551641825130865,
+      "learning_rate": 9.867176480719287e-06,
+      "loss": 0.7042,
+      "step": 710
+    },
+    {
+      "epoch": 0.16680351906158358,
+      "grad_norm": 1.6116484588806546,
+      "learning_rate": 9.866237029971327e-06,
+      "loss": 0.6283,
+      "step": 711
+    },
+    {
+      "epoch": 0.16703812316715544,
+      "grad_norm": 1.794097210603171,
+      "learning_rate": 9.865294313628959e-06,
+      "loss": 0.7035,
+      "step": 712
+    },
+    {
+      "epoch": 0.16727272727272727,
+      "grad_norm": 1.8271196863853747,
+      "learning_rate": 9.864348332324811e-06,
+      "loss": 0.611,
+      "step": 713
+    },
+    {
+      "epoch": 0.16750733137829912,
+      "grad_norm": 1.3309702239327663,
+      "learning_rate": 9.863399086693709e-06,
+      "loss": 0.6366,
+      "step": 714
+    },
+    {
+      "epoch": 0.16774193548387098,
+      "grad_norm": 3.331889826028672,
+      "learning_rate": 9.86244657737266e-06,
+      "loss": 0.6506,
+      "step": 715
+    },
+    {
+      "epoch": 0.1679765395894428,
+      "grad_norm": 4.773871403810453,
+      "learning_rate": 9.861490805000868e-06,
+      "loss": 0.657,
+      "step": 716
+    },
+    {
+      "epoch": 0.16821114369501466,
+      "grad_norm": 1.6248289349019442,
+      "learning_rate": 9.860531770219729e-06,
+      "loss": 0.6469,
+      "step": 717
+    },
+    {
+      "epoch": 0.16844574780058652,
+      "grad_norm": 1.1194824056463804,
+      "learning_rate": 9.859569473672816e-06,
+      "loss": 0.6715,
+      "step": 718
+    },
+    {
+      "epoch": 0.16868035190615835,
+      "grad_norm": 3.5798125313775007,
+      "learning_rate": 9.858603916005907e-06,
+      "loss": 0.6333,
+      "step": 719
+    },
+    {
+      "epoch": 0.1689149560117302,
+      "grad_norm": 8.283484339955482,
+      "learning_rate": 9.857635097866955e-06,
+      "loss": 0.6114,
+      "step": 720
+    },
+    {
+      "epoch": 0.16914956011730206,
+      "grad_norm": 2.0421584535823474,
+      "learning_rate": 9.856663019906112e-06,
+      "loss": 0.6806,
+      "step": 721
+    },
+    {
+      "epoch": 0.1693841642228739,
+      "grad_norm": 1.8570662954694708,
+      "learning_rate": 9.855687682775706e-06,
+      "loss": 0.6636,
+      "step": 722
+    },
+    {
+      "epoch": 0.16961876832844575,
+      "grad_norm": 2.166770834539674,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.6291,
+      "step": 723
+    },
+    {
+      "epoch": 0.1698533724340176,
+      "grad_norm": 3.239680976176979,
+      "learning_rate": 9.853727233626485e-06,
+      "loss": 0.6652,
+      "step": 724
+    },
+    {
+      "epoch": 0.17008797653958943,
+      "grad_norm": 1.0323438162265144,
+      "learning_rate": 9.852742122923274e-06,
+      "loss": 0.6699,
+      "step": 725
+    },
+    {
+      "epoch": 0.1703225806451613,
+      "grad_norm": 3.40075441491087,
+      "learning_rate": 9.851753755681703e-06,
+      "loss": 0.6995,
+      "step": 726
+    },
+    {
+      "epoch": 0.17055718475073314,
+      "grad_norm": 5.241387160101221,
+      "learning_rate": 9.850762132565044e-06,
+      "loss": 0.6687,
+      "step": 727
+    },
+    {
+      "epoch": 0.170791788856305,
+      "grad_norm": 2.5134466364367536,
+      "learning_rate": 9.849767254238741e-06,
+      "loss": 0.687,
+      "step": 728
+    },
+    {
+      "epoch": 0.17102639296187683,
+      "grad_norm": 3.2450867951663804,
+      "learning_rate": 9.848769121370434e-06,
+      "loss": 0.7159,
+      "step": 729
+    },
+    {
+      "epoch": 0.17126099706744868,
+      "grad_norm": 1.8679718244097496,
+      "learning_rate": 9.847767734629938e-06,
+      "loss": 0.6633,
+      "step": 730
+    },
+    {
+      "epoch": 0.17149560117302054,
+      "grad_norm": 1.989276653817671,
+      "learning_rate": 9.846763094689255e-06,
+      "loss": 0.6178,
+      "step": 731
+    },
+    {
+      "epoch": 0.17173020527859237,
+      "grad_norm": 1.2513787228924333,
+      "learning_rate": 9.845755202222575e-06,
+      "loss": 0.6619,
+      "step": 732
+    },
+    {
+      "epoch": 0.17196480938416422,
+      "grad_norm": 2.136836891169221,
+      "learning_rate": 9.844744057906263e-06,
+      "loss": 0.6295,
+      "step": 733
+    },
+    {
+      "epoch": 0.17219941348973608,
+      "grad_norm": 2.4852757416980555,
+      "learning_rate": 9.84372966241887e-06,
+      "loss": 0.6192,
+      "step": 734
+    },
+    {
+      "epoch": 0.1724340175953079,
+      "grad_norm": 2.4380261568528034,
+      "learning_rate": 9.842712016441129e-06,
+      "loss": 0.7001,
+      "step": 735
+    },
+    {
+      "epoch": 0.17266862170087977,
+      "grad_norm": 1.7994954992881838,
+      "learning_rate": 9.841691120655952e-06,
+      "loss": 0.7044,
+      "step": 736
+    },
+    {
+      "epoch": 0.17290322580645162,
+      "grad_norm": 2.332061873888857,
+      "learning_rate": 9.840666975748432e-06,
+      "loss": 0.7257,
+      "step": 737
+    },
+    {
+      "epoch": 0.17313782991202345,
+      "grad_norm": 1.7137094658452967,
+      "learning_rate": 9.83963958240585e-06,
+      "loss": 0.6831,
+      "step": 738
+    },
+    {
+      "epoch": 0.1733724340175953,
+      "grad_norm": 2.715564110665805,
+      "learning_rate": 9.838608941317653e-06,
+      "loss": 0.6834,
+      "step": 739
+    },
+    {
+      "epoch": 0.17360703812316716,
+      "grad_norm": 1.4855822273183557,
+      "learning_rate": 9.837575053175479e-06,
+      "loss": 0.6729,
+      "step": 740
+    },
+    {
+      "epoch": 0.173841642228739,
+      "grad_norm": 1.7370975706002307,
+      "learning_rate": 9.836537918673142e-06,
+      "loss": 0.6588,
+      "step": 741
+    },
+    {
+      "epoch": 0.17407624633431085,
+      "grad_norm": 1.3785306739940393,
+      "learning_rate": 9.835497538506634e-06,
+      "loss": 0.7089,
+      "step": 742
+    },
+    {
+      "epoch": 0.1743108504398827,
+      "grad_norm": 3.8083059374134676,
+      "learning_rate": 9.83445391337412e-06,
+      "loss": 0.6702,
+      "step": 743
+    },
+    {
+      "epoch": 0.17454545454545456,
+      "grad_norm": 1.8835346626639433,
+      "learning_rate": 9.833407043975952e-06,
+      "loss": 0.6741,
+      "step": 744
+    },
+    {
+      "epoch": 0.1747800586510264,
+      "grad_norm": 2.524266850602126,
+      "learning_rate": 9.832356931014653e-06,
+      "loss": 0.6245,
+      "step": 745
+    },
+    {
+      "epoch": 0.17501466275659824,
+      "grad_norm": 1.6516777113546257,
+      "learning_rate": 9.831303575194926e-06,
+      "loss": 0.6318,
+      "step": 746
+    },
+    {
+      "epoch": 0.1752492668621701,
+      "grad_norm": 1.1086329070239551,
+      "learning_rate": 9.830246977223645e-06,
+      "loss": 0.7131,
+      "step": 747
+    },
+    {
+      "epoch": 0.17548387096774193,
+      "grad_norm": 18.679637874630913,
+      "learning_rate": 9.829187137809865e-06,
+      "loss": 0.6033,
+      "step": 748
+    },
+    {
+      "epoch": 0.17571847507331378,
+      "grad_norm": 3.023618669662723,
+      "learning_rate": 9.828124057664816e-06,
+      "loss": 0.6851,
+      "step": 749
+    },
+    {
+      "epoch": 0.17595307917888564,
+      "grad_norm": 1.6983262840922475,
+      "learning_rate": 9.827057737501897e-06,
+      "loss": 0.6404,
+      "step": 750
+    },
+    {
+      "epoch": 0.17618768328445747,
+      "grad_norm": 1.2721188731290216,
+      "learning_rate": 9.825988178036689e-06,
+      "loss": 0.6785,
+      "step": 751
+    },
+    {
+      "epoch": 0.17642228739002933,
+      "grad_norm": 1.1787327418597884,
+      "learning_rate": 9.824915379986943e-06,
+      "loss": 0.6468,
+      "step": 752
+    },
+    {
+      "epoch": 0.17665689149560118,
+      "grad_norm": 3.2793541700112976,
+      "learning_rate": 9.823839344072582e-06,
+      "loss": 0.6879,
+      "step": 753
+    },
+    {
+      "epoch": 0.176891495601173,
+      "grad_norm": 8.160211428506143,
+      "learning_rate": 9.822760071015702e-06,
+      "loss": 0.6498,
+      "step": 754
+    },
+    {
+      "epoch": 0.17712609970674487,
+      "grad_norm": 1.6675553229919307,
+      "learning_rate": 9.821677561540575e-06,
+      "loss": 0.6875,
+      "step": 755
+    },
+    {
+      "epoch": 0.17736070381231672,
+      "grad_norm": 7.943861265740587,
+      "learning_rate": 9.820591816373642e-06,
+      "loss": 0.6838,
+      "step": 756
+    },
+    {
+      "epoch": 0.17759530791788855,
+      "grad_norm": 2.4811259708091935,
+      "learning_rate": 9.819502836243518e-06,
+      "loss": 0.6246,
+      "step": 757
+    },
+    {
+      "epoch": 0.1778299120234604,
+      "grad_norm": 1.8298941304912442,
+      "learning_rate": 9.818410621880983e-06,
+      "loss": 0.6349,
+      "step": 758
+    },
+    {
+      "epoch": 0.17806451612903226,
+      "grad_norm": 2.0813385082191806,
+      "learning_rate": 9.817315174018993e-06,
+      "loss": 0.6656,
+      "step": 759
+    },
+    {
+      "epoch": 0.17829912023460412,
+      "grad_norm": 2.1358142132141134,
+      "learning_rate": 9.816216493392673e-06,
+      "loss": 0.7075,
+      "step": 760
+    },
+    {
+      "epoch": 0.17853372434017595,
+      "grad_norm": 1.427783318018125,
+      "learning_rate": 9.815114580739316e-06,
+      "loss": 0.6801,
+      "step": 761
+    },
+    {
+      "epoch": 0.1787683284457478,
+      "grad_norm": 1.3619377798420857,
+      "learning_rate": 9.814009436798383e-06,
+      "loss": 0.6703,
+      "step": 762
+    },
+    {
+      "epoch": 0.17900293255131966,
+      "grad_norm": 1.4922700171855006,
+      "learning_rate": 9.812901062311508e-06,
+      "loss": 0.6294,
+      "step": 763
+    },
+    {
+      "epoch": 0.1792375366568915,
+      "grad_norm": 1.1267813344280002,
+      "learning_rate": 9.811789458022487e-06,
+      "loss": 0.659,
+      "step": 764
+    },
+    {
+      "epoch": 0.17947214076246334,
+      "grad_norm": 1.058523036042874,
+      "learning_rate": 9.810674624677288e-06,
+      "loss": 0.6694,
+      "step": 765
+    },
+    {
+      "epoch": 0.1797067448680352,
+      "grad_norm": 1.4117029808866475,
+      "learning_rate": 9.809556563024043e-06,
+      "loss": 0.6793,
+      "step": 766
+    },
+    {
+      "epoch": 0.17994134897360703,
+      "grad_norm": 2.5544433600431193,
+      "learning_rate": 9.808435273813053e-06,
+      "loss": 0.6806,
+      "step": 767
+    },
+    {
+      "epoch": 0.18017595307917889,
+      "grad_norm": 1.3183795596647168,
+      "learning_rate": 9.807310757796782e-06,
+      "loss": 0.6803,
+      "step": 768
+    },
+    {
+      "epoch": 0.18041055718475074,
+      "grad_norm": 1.7218640838577965,
+      "learning_rate": 9.806183015729862e-06,
+      "loss": 0.7058,
+      "step": 769
+    },
+    {
+      "epoch": 0.18064516129032257,
+      "grad_norm": 1.5752529981894803,
+      "learning_rate": 9.805052048369089e-06,
+      "loss": 0.7531,
+      "step": 770
+    },
+    {
+      "epoch": 0.18087976539589443,
+      "grad_norm": 3.1420200938798337,
+      "learning_rate": 9.803917856473423e-06,
+      "loss": 0.6798,
+      "step": 771
+    },
+    {
+      "epoch": 0.18111436950146628,
+      "grad_norm": 3.7302991907088736,
+      "learning_rate": 9.802780440803986e-06,
+      "loss": 0.6542,
+      "step": 772
+    },
+    {
+      "epoch": 0.1813489736070381,
+      "grad_norm": 1.7749633842591563,
+      "learning_rate": 9.801639802124073e-06,
+      "loss": 0.6246,
+      "step": 773
+    },
+    {
+      "epoch": 0.18158357771260997,
+      "grad_norm": 2.541138158664817,
+      "learning_rate": 9.800495941199125e-06,
+      "loss": 0.6606,
+      "step": 774
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 1.6776874386735878,
+      "learning_rate": 9.79934885879676e-06,
+      "loss": 0.6663,
+      "step": 775
+    },
+    {
+      "epoch": 0.18205278592375368,
+      "grad_norm": 1.0805462857796522,
+      "learning_rate": 9.798198555686751e-06,
+      "loss": 0.6555,
+      "step": 776
+    },
+    {
+      "epoch": 0.1822873900293255,
+      "grad_norm": 3.0300715489113785,
+      "learning_rate": 9.797045032641036e-06,
+      "loss": 0.7376,
+      "step": 777
+    },
+    {
+      "epoch": 0.18252199413489736,
+      "grad_norm": 1.0673645067009438,
+      "learning_rate": 9.795888290433709e-06,
+      "loss": 0.6737,
+      "step": 778
+    },
+    {
+      "epoch": 0.18275659824046922,
+      "grad_norm": 10.552559588259303,
+      "learning_rate": 9.79472832984103e-06,
+      "loss": 0.6717,
+      "step": 779
+    },
+    {
+      "epoch": 0.18299120234604105,
+      "grad_norm": 4.436167680881757,
+      "learning_rate": 9.793565151641411e-06,
+      "loss": 0.6501,
+      "step": 780
+    },
+    {
+      "epoch": 0.1832258064516129,
+      "grad_norm": 2.055463005158563,
+      "learning_rate": 9.792398756615435e-06,
+      "loss": 0.6546,
+      "step": 781
+    },
+    {
+      "epoch": 0.18346041055718476,
+      "grad_norm": 1.9264261961298847,
+      "learning_rate": 9.791229145545832e-06,
+      "loss": 0.6632,
+      "step": 782
+    },
+    {
+      "epoch": 0.1836950146627566,
+      "grad_norm": 1.5882634716208426,
+      "learning_rate": 9.790056319217496e-06,
+      "loss": 0.6561,
+      "step": 783
+    },
+    {
+      "epoch": 0.18392961876832845,
+      "grad_norm": 1.7661044756680175,
+      "learning_rate": 9.788880278417479e-06,
+      "loss": 0.6875,
+      "step": 784
+    },
+    {
+      "epoch": 0.1841642228739003,
+      "grad_norm": 1.6240516754366172,
+      "learning_rate": 9.787701023934985e-06,
+      "loss": 0.6501,
+      "step": 785
+    },
+    {
+      "epoch": 0.18439882697947213,
+      "grad_norm": 8.357259392450423,
+      "learning_rate": 9.786518556561383e-06,
+      "loss": 0.6823,
+      "step": 786
+    },
+    {
+      "epoch": 0.184633431085044,
+      "grad_norm": 1.4589649174263475,
+      "learning_rate": 9.78533287709019e-06,
+      "loss": 0.6444,
+      "step": 787
+    },
+    {
+      "epoch": 0.18486803519061584,
+      "grad_norm": 1.968470742577988,
+      "learning_rate": 9.784143986317084e-06,
+      "loss": 0.6723,
+      "step": 788
+    },
+    {
+      "epoch": 0.18510263929618767,
+      "grad_norm": 1.2035726817002486,
+      "learning_rate": 9.782951885039897e-06,
+      "loss": 0.6581,
+      "step": 789
+    },
+    {
+      "epoch": 0.18533724340175953,
+      "grad_norm": 1.0962656011799101,
+      "learning_rate": 9.781756574058614e-06,
+      "loss": 0.666,
+      "step": 790
+    },
+    {
+      "epoch": 0.18557184750733138,
+      "grad_norm": 2.2196841349727054,
+      "learning_rate": 9.780558054175373e-06,
+      "loss": 0.7009,
+      "step": 791
+    },
+    {
+      "epoch": 0.18580645161290324,
+      "grad_norm": 1.2976460256622553,
+      "learning_rate": 9.779356326194466e-06,
+      "loss": 0.6668,
+      "step": 792
+    },
+    {
+      "epoch": 0.18604105571847507,
+      "grad_norm": 1.4330922832281572,
+      "learning_rate": 9.77815139092234e-06,
+      "loss": 0.6858,
+      "step": 793
+    },
+    {
+      "epoch": 0.18627565982404692,
+      "grad_norm": 1.1778095642510038,
+      "learning_rate": 9.776943249167595e-06,
+      "loss": 0.6462,
+      "step": 794
+    },
+    {
+      "epoch": 0.18651026392961878,
+      "grad_norm": 2.084625097304774,
+      "learning_rate": 9.77573190174098e-06,
+      "loss": 0.6666,
+      "step": 795
+    },
+    {
+      "epoch": 0.1867448680351906,
+      "grad_norm": 1.3980507530165778,
+      "learning_rate": 9.774517349455393e-06,
+      "loss": 0.694,
+      "step": 796
+    },
+    {
+      "epoch": 0.18697947214076246,
+      "grad_norm": 2.6094745942939963,
+      "learning_rate": 9.77329959312589e-06,
+      "loss": 0.6127,
+      "step": 797
+    },
+    {
+      "epoch": 0.18721407624633432,
+      "grad_norm": 1.8460884345290358,
+      "learning_rate": 9.772078633569671e-06,
+      "loss": 0.7003,
+      "step": 798
+    },
+    {
+      "epoch": 0.18744868035190615,
+      "grad_norm": 1.9135578730776541,
+      "learning_rate": 9.770854471606086e-06,
+      "loss": 0.6962,
+      "step": 799
+    },
+    {
+      "epoch": 0.187683284457478,
+      "grad_norm": 1.2722374112691082,
+      "learning_rate": 9.769627108056638e-06,
+      "loss": 0.6594,
+      "step": 800
+    },
+    {
+      "epoch": 0.18791788856304986,
+      "grad_norm": 1.9797278719422926,
+      "learning_rate": 9.768396543744976e-06,
+      "loss": 0.6509,
+      "step": 801
+    },
+    {
+      "epoch": 0.1881524926686217,
+      "grad_norm": 1.3958306309948436,
+      "learning_rate": 9.767162779496894e-06,
+      "loss": 0.5916,
+      "step": 802
+    },
+    {
+      "epoch": 0.18838709677419355,
+      "grad_norm": 5.267892629337839,
+      "learning_rate": 9.76592581614034e-06,
+      "loss": 0.6734,
+      "step": 803
+    },
+    {
+      "epoch": 0.1886217008797654,
+      "grad_norm": 1.605813259416675,
+      "learning_rate": 9.764685654505405e-06,
+      "loss": 0.6719,
+      "step": 804
+    },
+    {
+      "epoch": 0.18885630498533723,
+      "grad_norm": 2.5451781138228275,
+      "learning_rate": 9.763442295424323e-06,
+      "loss": 0.6623,
+      "step": 805
+    },
+    {
+      "epoch": 0.1890909090909091,
+      "grad_norm": 3.475430507902762,
+      "learning_rate": 9.762195739731482e-06,
+      "loss": 0.6365,
+      "step": 806
+    },
+    {
+      "epoch": 0.18932551319648094,
+      "grad_norm": 6.680299483411663,
+      "learning_rate": 9.760945988263406e-06,
+      "loss": 0.6101,
+      "step": 807
+    },
+    {
+      "epoch": 0.1895601173020528,
+      "grad_norm": 5.3399812265684865,
+      "learning_rate": 9.75969304185877e-06,
+      "loss": 0.6359,
+      "step": 808
+    },
+    {
+      "epoch": 0.18979472140762463,
+      "grad_norm": 2.407122041291883,
+      "learning_rate": 9.758436901358391e-06,
+      "loss": 0.6736,
+      "step": 809
+    },
+    {
+      "epoch": 0.19002932551319648,
+      "grad_norm": 2.9490572761081375,
+      "learning_rate": 9.757177567605229e-06,
+      "loss": 0.6364,
+      "step": 810
+    },
+    {
+      "epoch": 0.19026392961876834,
+      "grad_norm": 6.217157809007754,
+      "learning_rate": 9.755915041444389e-06,
+      "loss": 0.671,
+      "step": 811
+    },
+    {
+      "epoch": 0.19049853372434017,
+      "grad_norm": 1.9173672087643958,
+      "learning_rate": 9.754649323723113e-06,
+      "loss": 0.6805,
+      "step": 812
+    },
+    {
+      "epoch": 0.19073313782991203,
+      "grad_norm": 1.7146151689127684,
+      "learning_rate": 9.753380415290793e-06,
+      "loss": 0.6029,
+      "step": 813
+    },
+    {
+      "epoch": 0.19096774193548388,
+      "grad_norm": 3.5549682056714245,
+      "learning_rate": 9.752108316998955e-06,
+      "loss": 0.6708,
+      "step": 814
+    },
+    {
+      "epoch": 0.1912023460410557,
+      "grad_norm": 4.602087148937077,
+      "learning_rate": 9.750833029701268e-06,
+      "loss": 0.6749,
+      "step": 815
+    },
+    {
+      "epoch": 0.19143695014662757,
+      "grad_norm": 1.3035557920997363,
+      "learning_rate": 9.749554554253542e-06,
+      "loss": 0.6647,
+      "step": 816
+    },
+    {
+      "epoch": 0.19167155425219942,
+      "grad_norm": 1.2639233011244417,
+      "learning_rate": 9.748272891513727e-06,
+      "loss": 0.6321,
+      "step": 817
+    },
+    {
+      "epoch": 0.19190615835777125,
+      "grad_norm": 1.8385936360209119,
+      "learning_rate": 9.746988042341907e-06,
+      "loss": 0.6579,
+      "step": 818
+    },
+    {
+      "epoch": 0.1921407624633431,
+      "grad_norm": 1.6376446434078913,
+      "learning_rate": 9.745700007600312e-06,
+      "loss": 0.6757,
+      "step": 819
+    },
+    {
+      "epoch": 0.19237536656891496,
+      "grad_norm": 2.9475770551988254,
+      "learning_rate": 9.744408788153305e-06,
+      "loss": 0.6732,
+      "step": 820
+    },
+    {
+      "epoch": 0.1926099706744868,
+      "grad_norm": 1.172714115730443,
+      "learning_rate": 9.743114384867387e-06,
+      "loss": 0.6162,
+      "step": 821
+    },
+    {
+      "epoch": 0.19284457478005865,
+      "grad_norm": 2.1645302554961603,
+      "learning_rate": 9.741816798611193e-06,
+      "loss": 0.6618,
+      "step": 822
+    },
+    {
+      "epoch": 0.1930791788856305,
+      "grad_norm": 2.7584078984953635,
+      "learning_rate": 9.7405160302555e-06,
+      "loss": 0.6776,
+      "step": 823
+    },
+    {
+      "epoch": 0.19331378299120236,
+      "grad_norm": 2.817753347077862,
+      "learning_rate": 9.739212080673215e-06,
+      "loss": 0.6836,
+      "step": 824
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 1.3924384398039422,
+      "learning_rate": 9.73790495073938e-06,
+      "loss": 0.6815,
+      "step": 825
+    },
+    {
+      "epoch": 0.19378299120234604,
+      "grad_norm": 2.7466754222436913,
+      "learning_rate": 9.736594641331178e-06,
+      "loss": 0.6706,
+      "step": 826
+    },
+    {
+      "epoch": 0.1940175953079179,
+      "grad_norm": 1.4690068934712832,
+      "learning_rate": 9.735281153327916e-06,
+      "loss": 0.6652,
+      "step": 827
+    },
+    {
+      "epoch": 0.19425219941348973,
+      "grad_norm": 0.9912171222047818,
+      "learning_rate": 9.733964487611044e-06,
+      "loss": 0.6923,
+      "step": 828
+    },
+    {
+      "epoch": 0.19448680351906159,
+      "grad_norm": 2.3940548177009275,
+      "learning_rate": 9.73264464506413e-06,
+      "loss": 0.6826,
+      "step": 829
+    },
+    {
+      "epoch": 0.19472140762463344,
+      "grad_norm": 5.982181001504478,
+      "learning_rate": 9.731321626572894e-06,
+      "loss": 0.6929,
+      "step": 830
+    },
+    {
+      "epoch": 0.19495601173020527,
+      "grad_norm": 1.3475966109935826,
+      "learning_rate": 9.729995433025169e-06,
+      "loss": 0.7073,
+      "step": 831
+    },
+    {
+      "epoch": 0.19519061583577713,
+      "grad_norm": 1.043564031113599,
+      "learning_rate": 9.728666065310929e-06,
+      "loss": 0.6182,
+      "step": 832
+    },
+    {
+      "epoch": 0.19542521994134898,
+      "grad_norm": 1.7350303914688778,
+      "learning_rate": 9.727333524322274e-06,
+      "loss": 0.668,
+      "step": 833
+    },
+    {
+      "epoch": 0.1956598240469208,
+      "grad_norm": 2.0000050782238272,
+      "learning_rate": 9.725997810953436e-06,
+      "loss": 0.6098,
+      "step": 834
+    },
+    {
+      "epoch": 0.19589442815249267,
+      "grad_norm": 1.291321907822395,
+      "learning_rate": 9.724658926100775e-06,
+      "loss": 0.6728,
+      "step": 835
+    },
+    {
+      "epoch": 0.19612903225806452,
+      "grad_norm": 1.763048699714866,
+      "learning_rate": 9.723316870662775e-06,
+      "loss": 0.6375,
+      "step": 836
+    },
+    {
+      "epoch": 0.19636363636363635,
+      "grad_norm": 2.3531051325550982,
+      "learning_rate": 9.721971645540055e-06,
+      "loss": 0.6648,
+      "step": 837
+    },
+    {
+      "epoch": 0.1965982404692082,
+      "grad_norm": 2.1961219586654317,
+      "learning_rate": 9.720623251635358e-06,
+      "loss": 0.6738,
+      "step": 838
+    },
+    {
+      "epoch": 0.19683284457478006,
+      "grad_norm": 3.7024000150930774,
+      "learning_rate": 9.71927168985355e-06,
+      "loss": 0.6566,
+      "step": 839
+    },
+    {
+      "epoch": 0.19706744868035192,
+      "grad_norm": 2.624254973613114,
+      "learning_rate": 9.717916961101632e-06,
+      "loss": 0.6579,
+      "step": 840
+    },
+    {
+      "epoch": 0.19730205278592375,
+      "grad_norm": 2.1037066238923123,
+      "learning_rate": 9.716559066288716e-06,
+      "loss": 0.6406,
+      "step": 841
+    },
+    {
+      "epoch": 0.1975366568914956,
+      "grad_norm": 1.1942497303256887,
+      "learning_rate": 9.715198006326053e-06,
+      "loss": 0.7186,
+      "step": 842
+    },
+    {
+      "epoch": 0.19777126099706746,
+      "grad_norm": 2.45649961745645,
+      "learning_rate": 9.713833782127008e-06,
+      "loss": 0.7026,
+      "step": 843
+    },
+    {
+      "epoch": 0.1980058651026393,
+      "grad_norm": 1.2871905539089525,
+      "learning_rate": 9.712466394607079e-06,
+      "loss": 0.6928,
+      "step": 844
+    },
+    {
+      "epoch": 0.19824046920821115,
+      "grad_norm": 1.5272448501545626,
+      "learning_rate": 9.711095844683875e-06,
+      "loss": 0.645,
+      "step": 845
+    },
+    {
+      "epoch": 0.198475073313783,
+      "grad_norm": 1.5635761481172537,
+      "learning_rate": 9.709722133277137e-06,
+      "loss": 0.6785,
+      "step": 846
+    },
+    {
+      "epoch": 0.19870967741935483,
+      "grad_norm": 0.9778959639272299,
+      "learning_rate": 9.708345261308724e-06,
+      "loss": 0.6714,
+      "step": 847
+    },
+    {
+      "epoch": 0.19894428152492669,
+      "grad_norm": 1.7864809366163665,
+      "learning_rate": 9.706965229702614e-06,
+      "loss": 0.6575,
+      "step": 848
+    },
+    {
+      "epoch": 0.19917888563049854,
+      "grad_norm": 1.2864871417843071,
+      "learning_rate": 9.705582039384907e-06,
+      "loss": 0.6633,
+      "step": 849
+    },
+    {
+      "epoch": 0.19941348973607037,
+      "grad_norm": 2.7203996971022395,
+      "learning_rate": 9.704195691283826e-06,
+      "loss": 0.6957,
+      "step": 850
+    },
+    {
+      "epoch": 0.19964809384164223,
+      "grad_norm": 1.730215994334602,
+      "learning_rate": 9.702806186329707e-06,
+      "loss": 0.7061,
+      "step": 851
+    },
+    {
+      "epoch": 0.19988269794721408,
+      "grad_norm": 4.388761434832178,
+      "learning_rate": 9.70141352545501e-06,
+      "loss": 0.6612,
+      "step": 852
+    },
+    {
+      "epoch": 0.2001173020527859,
+      "grad_norm": 4.265248714766533,
+      "learning_rate": 9.70001770959431e-06,
+      "loss": 0.644,
+      "step": 853
+    },
+    {
+      "epoch": 0.20035190615835777,
+      "grad_norm": 1.068965180050243,
+      "learning_rate": 9.698618739684298e-06,
+      "loss": 0.6442,
+      "step": 854
+    },
+    {
+      "epoch": 0.20058651026392962,
+      "grad_norm": 1.3756426647548368,
+      "learning_rate": 9.697216616663782e-06,
+      "loss": 0.6561,
+      "step": 855
+    },
+    {
+      "epoch": 0.20082111436950148,
+      "grad_norm": 1.766210240227299,
+      "learning_rate": 9.695811341473694e-06,
+      "loss": 0.6247,
+      "step": 856
+    },
+    {
+      "epoch": 0.2010557184750733,
+      "grad_norm": 2.160118229146718,
+      "learning_rate": 9.694402915057066e-06,
+      "loss": 0.6571,
+      "step": 857
+    },
+    {
+      "epoch": 0.20129032258064516,
+      "grad_norm": 2.469393549137364,
+      "learning_rate": 9.69299133835906e-06,
+      "loss": 0.5879,
+      "step": 858
+    },
+    {
+      "epoch": 0.20152492668621702,
+      "grad_norm": 1.5273736674661544,
+      "learning_rate": 9.691576612326941e-06,
+      "loss": 0.6743,
+      "step": 859
+    },
+    {
+      "epoch": 0.20175953079178885,
+      "grad_norm": 1.0668532110943905,
+      "learning_rate": 9.690158737910093e-06,
+      "loss": 0.6445,
+      "step": 860
+    },
+    {
+      "epoch": 0.2019941348973607,
+      "grad_norm": 2.890483837414929,
+      "learning_rate": 9.688737716060014e-06,
+      "loss": 0.6298,
+      "step": 861
+    },
+    {
+      "epoch": 0.20222873900293256,
+      "grad_norm": 1.843801407096205,
+      "learning_rate": 9.687313547730308e-06,
+      "loss": 0.6854,
+      "step": 862
+    },
+    {
+      "epoch": 0.2024633431085044,
+      "grad_norm": 0.9286428173977231,
+      "learning_rate": 9.685886233876696e-06,
+      "loss": 0.6231,
+      "step": 863
+    },
+    {
+      "epoch": 0.20269794721407625,
+      "grad_norm": 1.4125851665859233,
+      "learning_rate": 9.684455775457009e-06,
+      "loss": 0.6245,
+      "step": 864
+    },
+    {
+      "epoch": 0.2029325513196481,
+      "grad_norm": 6.073805403505272,
+      "learning_rate": 9.683022173431184e-06,
+      "loss": 0.6709,
+      "step": 865
+    },
+    {
+      "epoch": 0.20316715542521993,
+      "grad_norm": 0.9691315830576945,
+      "learning_rate": 9.681585428761272e-06,
+      "loss": 0.6798,
+      "step": 866
+    },
+    {
+      "epoch": 0.2034017595307918,
+      "grad_norm": 2.5907417889006443,
+      "learning_rate": 9.680145542411436e-06,
+      "loss": 0.6738,
+      "step": 867
+    },
+    {
+      "epoch": 0.20363636363636364,
+      "grad_norm": 11.722014166762376,
+      "learning_rate": 9.678702515347937e-06,
+      "loss": 0.5994,
+      "step": 868
+    },
+    {
+      "epoch": 0.20387096774193547,
+      "grad_norm": 1.6480056607813116,
+      "learning_rate": 9.677256348539154e-06,
+      "loss": 0.6621,
+      "step": 869
+    },
+    {
+      "epoch": 0.20410557184750733,
+      "grad_norm": 2.1007903859154027,
+      "learning_rate": 9.675807042955567e-06,
+      "loss": 0.7475,
+      "step": 870
+    },
+    {
+      "epoch": 0.20434017595307918,
+      "grad_norm": 1.0714730702482032,
+      "learning_rate": 9.674354599569767e-06,
+      "loss": 0.6802,
+      "step": 871
+    },
+    {
+      "epoch": 0.20457478005865104,
+      "grad_norm": 1.5835331150662713,
+      "learning_rate": 9.672899019356442e-06,
+      "loss": 0.6558,
+      "step": 872
+    },
+    {
+      "epoch": 0.20480938416422287,
+      "grad_norm": 1.2478724145092908,
+      "learning_rate": 9.671440303292395e-06,
+      "loss": 0.6264,
+      "step": 873
+    },
+    {
+      "epoch": 0.20504398826979472,
+      "grad_norm": 1.8847929441323756,
+      "learning_rate": 9.669978452356528e-06,
+      "loss": 0.6572,
+      "step": 874
+    },
+    {
+      "epoch": 0.20527859237536658,
+      "grad_norm": 1.1087069245802912,
+      "learning_rate": 9.66851346752985e-06,
+      "loss": 0.7131,
+      "step": 875
+    },
+    {
+      "epoch": 0.2055131964809384,
+      "grad_norm": 1.6276937142263947,
+      "learning_rate": 9.667045349795468e-06,
+      "loss": 0.647,
+      "step": 876
+    },
+    {
+      "epoch": 0.20574780058651027,
+      "grad_norm": 2.9117212268666752,
+      "learning_rate": 9.665574100138595e-06,
+      "loss": 0.674,
+      "step": 877
+    },
+    {
+      "epoch": 0.20598240469208212,
+      "grad_norm": 1.802950641271152,
+      "learning_rate": 9.664099719546547e-06,
+      "loss": 0.643,
+      "step": 878
+    },
+    {
+      "epoch": 0.20621700879765395,
+      "grad_norm": 2.04554378713996,
+      "learning_rate": 9.662622209008735e-06,
+      "loss": 0.6304,
+      "step": 879
+    },
+    {
+      "epoch": 0.2064516129032258,
+      "grad_norm": 2.089329297788112,
+      "learning_rate": 9.661141569516682e-06,
+      "loss": 0.6362,
+      "step": 880
+    },
+    {
+      "epoch": 0.20668621700879766,
+      "grad_norm": 1.1748100618556383,
+      "learning_rate": 9.659657802063996e-06,
+      "loss": 0.6586,
+      "step": 881
+    },
+    {
+      "epoch": 0.2069208211143695,
+      "grad_norm": 2.1537283721903457,
+      "learning_rate": 9.658170907646396e-06,
+      "loss": 0.6507,
+      "step": 882
+    },
+    {
+      "epoch": 0.20715542521994135,
+      "grad_norm": 1.1763604556671938,
+      "learning_rate": 9.656680887261693e-06,
+      "loss": 0.6495,
+      "step": 883
+    },
+    {
+      "epoch": 0.2073900293255132,
+      "grad_norm": 1.760765827261432,
+      "learning_rate": 9.655187741909797e-06,
+      "loss": 0.6522,
+      "step": 884
+    },
+    {
+      "epoch": 0.20762463343108503,
+      "grad_norm": 2.101601249166311,
+      "learning_rate": 9.653691472592718e-06,
+      "loss": 0.6592,
+      "step": 885
+    },
+    {
+      "epoch": 0.2078592375366569,
+      "grad_norm": 1.5337659374549815,
+      "learning_rate": 9.65219208031456e-06,
+      "loss": 0.6481,
+      "step": 886
+    },
+    {
+      "epoch": 0.20809384164222874,
+      "grad_norm": 1.4401051067751012,
+      "learning_rate": 9.65068956608152e-06,
+      "loss": 0.6675,
+      "step": 887
+    },
+    {
+      "epoch": 0.2083284457478006,
+      "grad_norm": 1.2219105521679687,
+      "learning_rate": 9.649183930901895e-06,
+      "loss": 0.6642,
+      "step": 888
+    },
+    {
+      "epoch": 0.20856304985337243,
+      "grad_norm": 1.201737906274961,
+      "learning_rate": 9.647675175786075e-06,
+      "loss": 0.6724,
+      "step": 889
+    },
+    {
+      "epoch": 0.20879765395894428,
+      "grad_norm": 3.073136252336065,
+      "learning_rate": 9.64616330174654e-06,
+      "loss": 0.6887,
+      "step": 890
+    },
+    {
+      "epoch": 0.20903225806451614,
+      "grad_norm": 1.7337964061237228,
+      "learning_rate": 9.644648309797871e-06,
+      "loss": 0.6434,
+      "step": 891
+    },
+    {
+      "epoch": 0.20926686217008797,
+      "grad_norm": 2.140464167952869,
+      "learning_rate": 9.643130200956732e-06,
+      "loss": 0.6565,
+      "step": 892
+    },
+    {
+      "epoch": 0.20950146627565983,
+      "grad_norm": 1.7400320702775463,
+      "learning_rate": 9.641608976241883e-06,
+      "loss": 0.6312,
+      "step": 893
+    },
+    {
+      "epoch": 0.20973607038123168,
+      "grad_norm": 3.1966431953328693,
+      "learning_rate": 9.640084636674179e-06,
+      "loss": 0.6036,
+      "step": 894
+    },
+    {
+      "epoch": 0.2099706744868035,
+      "grad_norm": 0.9554403450244524,
+      "learning_rate": 9.638557183276556e-06,
+      "loss": 0.7085,
+      "step": 895
+    },
+    {
+      "epoch": 0.21020527859237537,
+      "grad_norm": 1.498413589018548,
+      "learning_rate": 9.63702661707405e-06,
+      "loss": 0.6753,
+      "step": 896
+    },
+    {
+      "epoch": 0.21043988269794722,
+      "grad_norm": 1.8207602804392655,
+      "learning_rate": 9.635492939093778e-06,
+      "loss": 0.7039,
+      "step": 897
+    },
+    {
+      "epoch": 0.21067448680351905,
+      "grad_norm": 1.631255264135716,
+      "learning_rate": 9.633956150364948e-06,
+      "loss": 0.6373,
+      "step": 898
+    },
+    {
+      "epoch": 0.2109090909090909,
+      "grad_norm": 2.109503092325333,
+      "learning_rate": 9.632416251918855e-06,
+      "loss": 0.6666,
+      "step": 899
+    },
+    {
+      "epoch": 0.21114369501466276,
+      "grad_norm": 1.4734913865053512,
+      "learning_rate": 9.630873244788884e-06,
+      "loss": 0.6793,
+      "step": 900
+    },
+    {
+      "epoch": 0.2113782991202346,
+      "grad_norm": 1.4997857924939262,
+      "learning_rate": 9.629327130010501e-06,
+      "loss": 0.6083,
+      "step": 901
+    },
+    {
+      "epoch": 0.21161290322580645,
+      "grad_norm": 2.307097896233073,
+      "learning_rate": 9.62777790862126e-06,
+      "loss": 0.6057,
+      "step": 902
+    },
+    {
+      "epoch": 0.2118475073313783,
+      "grad_norm": 1.1774544123047836,
+      "learning_rate": 9.626225581660802e-06,
+      "loss": 0.656,
+      "step": 903
+    },
+    {
+      "epoch": 0.21208211143695016,
+      "grad_norm": 0.9239882041669124,
+      "learning_rate": 9.62467015017085e-06,
+      "loss": 0.6021,
+      "step": 904
+    },
+    {
+      "epoch": 0.212316715542522,
+      "grad_norm": 3.663284148439789,
+      "learning_rate": 9.623111615195206e-06,
+      "loss": 0.6102,
+      "step": 905
+    },
+    {
+      "epoch": 0.21255131964809384,
+      "grad_norm": 2.22539971046485,
+      "learning_rate": 9.621549977779765e-06,
+      "loss": 0.6936,
+      "step": 906
+    },
+    {
+      "epoch": 0.2127859237536657,
+      "grad_norm": 1.901468590290738,
+      "learning_rate": 9.61998523897249e-06,
+      "loss": 0.6553,
+      "step": 907
+    },
+    {
+      "epoch": 0.21302052785923753,
+      "grad_norm": 1.6609483168416135,
+      "learning_rate": 9.618417399823442e-06,
+      "loss": 0.6995,
+      "step": 908
+    },
+    {
+      "epoch": 0.21325513196480939,
+      "grad_norm": 1.7380882151027832,
+      "learning_rate": 9.616846461384748e-06,
+      "loss": 0.6501,
+      "step": 909
+    },
+    {
+      "epoch": 0.21348973607038124,
+      "grad_norm": 1.8347208426109232,
+      "learning_rate": 9.61527242471062e-06,
+      "loss": 0.6866,
+      "step": 910
+    },
+    {
+      "epoch": 0.21372434017595307,
+      "grad_norm": 3.5676223816730133,
+      "learning_rate": 9.613695290857352e-06,
+      "loss": 0.669,
+      "step": 911
+    },
+    {
+      "epoch": 0.21395894428152493,
+      "grad_norm": 1.191285952404426,
+      "learning_rate": 9.612115060883315e-06,
+      "loss": 0.653,
+      "step": 912
+    },
+    {
+      "epoch": 0.21419354838709678,
+      "grad_norm": 1.6788989812175175,
+      "learning_rate": 9.610531735848953e-06,
+      "loss": 0.6226,
+      "step": 913
+    },
+    {
+      "epoch": 0.2144281524926686,
+      "grad_norm": 1.4586256760129428,
+      "learning_rate": 9.608945316816794e-06,
+      "loss": 0.6539,
+      "step": 914
+    },
+    {
+      "epoch": 0.21466275659824047,
+      "grad_norm": 1.403774897238066,
+      "learning_rate": 9.607355804851437e-06,
+      "loss": 0.6669,
+      "step": 915
+    },
+    {
+      "epoch": 0.21489736070381232,
+      "grad_norm": 2.0700943384283157,
+      "learning_rate": 9.60576320101956e-06,
+      "loss": 0.6422,
+      "step": 916
+    },
+    {
+      "epoch": 0.21513196480938415,
+      "grad_norm": 1.2489560665864599,
+      "learning_rate": 9.604167506389915e-06,
+      "loss": 0.6755,
+      "step": 917
+    },
+    {
+      "epoch": 0.215366568914956,
+      "grad_norm": 1.1549235647914915,
+      "learning_rate": 9.602568722033325e-06,
+      "loss": 0.6623,
+      "step": 918
+    },
+    {
+      "epoch": 0.21560117302052786,
+      "grad_norm": 1.8360283762120344,
+      "learning_rate": 9.600966849022692e-06,
+      "loss": 0.6891,
+      "step": 919
+    },
+    {
+      "epoch": 0.21583577712609972,
+      "grad_norm": 1.0393580997748637,
+      "learning_rate": 9.599361888432987e-06,
+      "loss": 0.6497,
+      "step": 920
+    },
+    {
+      "epoch": 0.21607038123167155,
+      "grad_norm": 1.7945581650248654,
+      "learning_rate": 9.597753841341253e-06,
+      "loss": 0.6318,
+      "step": 921
+    },
+    {
+      "epoch": 0.2163049853372434,
+      "grad_norm": 1.6188242066426033,
+      "learning_rate": 9.596142708826604e-06,
+      "loss": 0.6192,
+      "step": 922
+    },
+    {
+      "epoch": 0.21653958944281526,
+      "grad_norm": 2.8364396280552726,
+      "learning_rate": 9.594528491970228e-06,
+      "loss": 0.6287,
+      "step": 923
+    },
+    {
+      "epoch": 0.2167741935483871,
+      "grad_norm": 1.317938874984234,
+      "learning_rate": 9.592911191855381e-06,
+      "loss": 0.614,
+      "step": 924
+    },
+    {
+      "epoch": 0.21700879765395895,
+      "grad_norm": 1.2506757855251625,
+      "learning_rate": 9.591290809567384e-06,
+      "loss": 0.6462,
+      "step": 925
+    },
+    {
+      "epoch": 0.2172434017595308,
+      "grad_norm": 1.5113552648070603,
+      "learning_rate": 9.589667346193632e-06,
+      "loss": 0.69,
+      "step": 926
+    },
+    {
+      "epoch": 0.21747800586510263,
+      "grad_norm": 3.007533355217859,
+      "learning_rate": 9.588040802823585e-06,
+      "loss": 0.6632,
+      "step": 927
+    },
+    {
+      "epoch": 0.2177126099706745,
+      "grad_norm": 2.2544121182751016,
+      "learning_rate": 9.586411180548771e-06,
+      "loss": 0.698,
+      "step": 928
+    },
+    {
+      "epoch": 0.21794721407624634,
+      "grad_norm": 1.6804815608054733,
+      "learning_rate": 9.584778480462785e-06,
+      "loss": 0.6661,
+      "step": 929
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 1.8591404754299143,
+      "learning_rate": 9.58314270366128e-06,
+      "loss": 0.6638,
+      "step": 930
+    },
+    {
+      "epoch": 0.21841642228739003,
+      "grad_norm": 1.7913063387462433,
+      "learning_rate": 9.581503851241987e-06,
+      "loss": 0.6493,
+      "step": 931
+    },
+    {
+      "epoch": 0.21865102639296188,
+      "grad_norm": 1.296486739763903,
+      "learning_rate": 9.57986192430469e-06,
+      "loss": 0.6721,
+      "step": 932
+    },
+    {
+      "epoch": 0.2188856304985337,
+      "grad_norm": 1.458842281608999,
+      "learning_rate": 9.57821692395124e-06,
+      "loss": 0.6794,
+      "step": 933
+    },
+    {
+      "epoch": 0.21912023460410557,
+      "grad_norm": 1.8068416750746825,
+      "learning_rate": 9.57656885128555e-06,
+      "loss": 0.6618,
+      "step": 934
+    },
+    {
+      "epoch": 0.21935483870967742,
+      "grad_norm": 1.4064094091161847,
+      "learning_rate": 9.574917707413596e-06,
+      "loss": 0.618,
+      "step": 935
+    },
+    {
+      "epoch": 0.21958944281524928,
+      "grad_norm": 4.204221566331364,
+      "learning_rate": 9.573263493443415e-06,
+      "loss": 0.6702,
+      "step": 936
+    },
+    {
+      "epoch": 0.2198240469208211,
+      "grad_norm": 3.8912619628950247,
+      "learning_rate": 9.5716062104851e-06,
+      "loss": 0.685,
+      "step": 937
+    },
+    {
+      "epoch": 0.22005865102639297,
+      "grad_norm": 4.04575953654595,
+      "learning_rate": 9.56994585965081e-06,
+      "loss": 0.6739,
+      "step": 938
+    },
+    {
+      "epoch": 0.22029325513196482,
+      "grad_norm": 1.0417380395271678,
+      "learning_rate": 9.568282442054759e-06,
+      "loss": 0.6571,
+      "step": 939
+    },
+    {
+      "epoch": 0.22052785923753665,
+      "grad_norm": 1.1622667689308102,
+      "learning_rate": 9.566615958813219e-06,
+      "loss": 0.6343,
+      "step": 940
+    },
+    {
+      "epoch": 0.2207624633431085,
+      "grad_norm": 2.272194661714964,
+      "learning_rate": 9.56494641104452e-06,
+      "loss": 0.6957,
+      "step": 941
+    },
+    {
+      "epoch": 0.22099706744868036,
+      "grad_norm": 4.689768435648827,
+      "learning_rate": 9.563273799869048e-06,
+      "loss": 0.5986,
+      "step": 942
+    },
+    {
+      "epoch": 0.2212316715542522,
+      "grad_norm": 2.322851221802614,
+      "learning_rate": 9.561598126409247e-06,
+      "loss": 0.6414,
+      "step": 943
+    },
+    {
+      "epoch": 0.22146627565982405,
+      "grad_norm": 1.2832930768450737,
+      "learning_rate": 9.559919391789611e-06,
+      "loss": 0.6687,
+      "step": 944
+    },
+    {
+      "epoch": 0.2217008797653959,
+      "grad_norm": 1.3127832267230448,
+      "learning_rate": 9.558237597136695e-06,
+      "loss": 0.667,
+      "step": 945
+    },
+    {
+      "epoch": 0.22193548387096773,
+      "grad_norm": 1.061136215661653,
+      "learning_rate": 9.556552743579102e-06,
+      "loss": 0.6333,
+      "step": 946
+    },
+    {
+      "epoch": 0.2221700879765396,
+      "grad_norm": 1.310962500440387,
+      "learning_rate": 9.554864832247488e-06,
+      "loss": 0.6293,
+      "step": 947
+    },
+    {
+      "epoch": 0.22240469208211144,
+      "grad_norm": 1.847743699079574,
+      "learning_rate": 9.553173864274567e-06,
+      "loss": 0.6147,
+      "step": 948
+    },
+    {
+      "epoch": 0.22263929618768327,
+      "grad_norm": 1.2710568543733551,
+      "learning_rate": 9.551479840795098e-06,
+      "loss": 0.6437,
+      "step": 949
+    },
+    {
+      "epoch": 0.22287390029325513,
+      "grad_norm": 7.601474393023154,
+      "learning_rate": 9.54978276294589e-06,
+      "loss": 0.6349,
+      "step": 950
+    },
+    {
+      "epoch": 0.22310850439882698,
+      "grad_norm": 1.415799715852848,
+      "learning_rate": 9.548082631865806e-06,
+      "loss": 0.6329,
+      "step": 951
+    },
+    {
+      "epoch": 0.22334310850439884,
+      "grad_norm": 1.503562399006367,
+      "learning_rate": 9.546379448695758e-06,
+      "loss": 0.676,
+      "step": 952
+    },
+    {
+      "epoch": 0.22357771260997067,
+      "grad_norm": 1.5623571065914645,
+      "learning_rate": 9.544673214578699e-06,
+      "loss": 0.7151,
+      "step": 953
+    },
+    {
+      "epoch": 0.22381231671554253,
+      "grad_norm": 4.279176198980748,
+      "learning_rate": 9.54296393065964e-06,
+      "loss": 0.6018,
+      "step": 954
+    },
+    {
+      "epoch": 0.22404692082111438,
+      "grad_norm": 8.779017143517844,
+      "learning_rate": 9.541251598085631e-06,
+      "loss": 0.6721,
+      "step": 955
+    },
+    {
+      "epoch": 0.2242815249266862,
+      "grad_norm": 2.644174124377122,
+      "learning_rate": 9.53953621800577e-06,
+      "loss": 0.5881,
+      "step": 956
+    },
+    {
+      "epoch": 0.22451612903225807,
+      "grad_norm": 1.7114965928788233,
+      "learning_rate": 9.5378177915712e-06,
+      "loss": 0.6229,
+      "step": 957
+    },
+    {
+      "epoch": 0.22475073313782992,
+      "grad_norm": 1.1050818447192206,
+      "learning_rate": 9.536096319935109e-06,
+      "loss": 0.6163,
+      "step": 958
+    },
+    {
+      "epoch": 0.22498533724340175,
+      "grad_norm": 1.6257590344746873,
+      "learning_rate": 9.534371804252727e-06,
+      "loss": 0.6568,
+      "step": 959
+    },
+    {
+      "epoch": 0.2252199413489736,
+      "grad_norm": 1.0705841536133995,
+      "learning_rate": 9.53264424568133e-06,
+      "loss": 0.6444,
+      "step": 960
+    },
+    {
+      "epoch": 0.22545454545454546,
+      "grad_norm": 1.6502909334600855,
+      "learning_rate": 9.530913645380233e-06,
+      "loss": 0.693,
+      "step": 961
+    },
+    {
+      "epoch": 0.2256891495601173,
+      "grad_norm": 5.6394366718967515,
+      "learning_rate": 9.529180004510791e-06,
+      "loss": 0.6618,
+      "step": 962
+    },
+    {
+      "epoch": 0.22592375366568915,
+      "grad_norm": 8.228928112712278,
+      "learning_rate": 9.527443324236403e-06,
+      "loss": 0.7084,
+      "step": 963
+    },
+    {
+      "epoch": 0.226158357771261,
+      "grad_norm": 1.7709964191958698,
+      "learning_rate": 9.525703605722508e-06,
+      "loss": 0.5879,
+      "step": 964
+    },
+    {
+      "epoch": 0.22639296187683283,
+      "grad_norm": 0.9300113928923643,
+      "learning_rate": 9.52396085013658e-06,
+      "loss": 0.6626,
+      "step": 965
+    },
+    {
+      "epoch": 0.2266275659824047,
+      "grad_norm": 1.3345040668724075,
+      "learning_rate": 9.522215058648132e-06,
+      "loss": 0.668,
+      "step": 966
+    },
+    {
+      "epoch": 0.22686217008797654,
+      "grad_norm": 6.219237350890805,
+      "learning_rate": 9.520466232428717e-06,
+      "loss": 0.6748,
+      "step": 967
+    },
+    {
+      "epoch": 0.2270967741935484,
+      "grad_norm": 1.9923763181200986,
+      "learning_rate": 9.518714372651922e-06,
+      "loss": 0.6387,
+      "step": 968
+    },
+    {
+      "epoch": 0.22733137829912023,
+      "grad_norm": 1.4539746482582692,
+      "learning_rate": 9.516959480493373e-06,
+      "loss": 0.6638,
+      "step": 969
+    },
+    {
+      "epoch": 0.22756598240469209,
+      "grad_norm": 1.6238750637770847,
+      "learning_rate": 9.515201557130726e-06,
+      "loss": 0.6723,
+      "step": 970
+    },
+    {
+      "epoch": 0.22780058651026394,
+      "grad_norm": 1.2739388924616808,
+      "learning_rate": 9.513440603743676e-06,
+      "loss": 0.6362,
+      "step": 971
+    },
+    {
+      "epoch": 0.22803519061583577,
+      "grad_norm": 1.6651793574871914,
+      "learning_rate": 9.511676621513947e-06,
+      "loss": 0.6363,
+      "step": 972
+    },
+    {
+      "epoch": 0.22826979472140763,
+      "grad_norm": 1.2091741125938216,
+      "learning_rate": 9.509909611625298e-06,
+      "loss": 0.6081,
+      "step": 973
+    },
+    {
+      "epoch": 0.22850439882697948,
+      "grad_norm": 1.2741025241549604,
+      "learning_rate": 9.508139575263522e-06,
+      "loss": 0.6316,
+      "step": 974
+    },
+    {
+      "epoch": 0.2287390029325513,
+      "grad_norm": 2.724056489379569,
+      "learning_rate": 9.506366513616439e-06,
+      "loss": 0.7075,
+      "step": 975
+    },
+    {
+      "epoch": 0.22897360703812317,
+      "grad_norm": 3.217519450192786,
+      "learning_rate": 9.504590427873897e-06,
+      "loss": 0.6302,
+      "step": 976
+    },
+    {
+      "epoch": 0.22920821114369502,
+      "grad_norm": 6.242868733111834,
+      "learning_rate": 9.502811319227783e-06,
+      "loss": 0.6743,
+      "step": 977
+    },
+    {
+      "epoch": 0.22944281524926685,
+      "grad_norm": 1.3589476990754552,
+      "learning_rate": 9.501029188872005e-06,
+      "loss": 0.684,
+      "step": 978
+    },
+    {
+      "epoch": 0.2296774193548387,
+      "grad_norm": 1.3256733080498688,
+      "learning_rate": 9.499244038002495e-06,
+      "loss": 0.6038,
+      "step": 979
+    },
+    {
+      "epoch": 0.22991202346041056,
+      "grad_norm": 1.2609101984988522,
+      "learning_rate": 9.497455867817225e-06,
+      "loss": 0.6731,
+      "step": 980
+    },
+    {
+      "epoch": 0.2301466275659824,
+      "grad_norm": 1.4483857272646994,
+      "learning_rate": 9.495664679516183e-06,
+      "loss": 0.6333,
+      "step": 981
+    },
+    {
+      "epoch": 0.23038123167155425,
+      "grad_norm": 1.2304459998771808,
+      "learning_rate": 9.493870474301383e-06,
+      "loss": 0.6334,
+      "step": 982
+    },
+    {
+      "epoch": 0.2306158357771261,
+      "grad_norm": 1.0767086812601114,
+      "learning_rate": 9.492073253376865e-06,
+      "loss": 0.6207,
+      "step": 983
+    },
+    {
+      "epoch": 0.23085043988269796,
+      "grad_norm": 1.1457537977191967,
+      "learning_rate": 9.490273017948698e-06,
+      "loss": 0.6675,
+      "step": 984
+    },
+    {
+      "epoch": 0.2310850439882698,
+      "grad_norm": 1.6217273167197999,
+      "learning_rate": 9.488469769224966e-06,
+      "loss": 0.6492,
+      "step": 985
+    },
+    {
+      "epoch": 0.23131964809384165,
+      "grad_norm": 2.947889514848731,
+      "learning_rate": 9.486663508415777e-06,
+      "loss": 0.62,
+      "step": 986
+    },
+    {
+      "epoch": 0.2315542521994135,
+      "grad_norm": 1.5930045829986972,
+      "learning_rate": 9.484854236733265e-06,
+      "loss": 0.645,
+      "step": 987
+    },
+    {
+      "epoch": 0.23178885630498533,
+      "grad_norm": 1.0411270862368243,
+      "learning_rate": 9.48304195539158e-06,
+      "loss": 0.6569,
+      "step": 988
+    },
+    {
+      "epoch": 0.2320234604105572,
+      "grad_norm": 1.8714760360067142,
+      "learning_rate": 9.48122666560689e-06,
+      "loss": 0.5927,
+      "step": 989
+    },
+    {
+      "epoch": 0.23225806451612904,
+      "grad_norm": 1.2536983547748748,
+      "learning_rate": 9.479408368597386e-06,
+      "loss": 0.611,
+      "step": 990
+    },
+    {
+      "epoch": 0.23249266862170087,
+      "grad_norm": 2.198810965849353,
+      "learning_rate": 9.477587065583281e-06,
+      "loss": 0.6586,
+      "step": 991
+    },
+    {
+      "epoch": 0.23272727272727273,
+      "grad_norm": 2.1986534808557616,
+      "learning_rate": 9.475762757786796e-06,
+      "loss": 0.6528,
+      "step": 992
+    },
+    {
+      "epoch": 0.23296187683284458,
+      "grad_norm": 1.0833049104621884,
+      "learning_rate": 9.473935446432169e-06,
+      "loss": 0.6475,
+      "step": 993
+    },
+    {
+      "epoch": 0.2331964809384164,
+      "grad_norm": 2.964486398105646,
+      "learning_rate": 9.472105132745664e-06,
+      "loss": 0.624,
+      "step": 994
+    },
+    {
+      "epoch": 0.23343108504398827,
+      "grad_norm": 1.0466438463264922,
+      "learning_rate": 9.470271817955547e-06,
+      "loss": 0.6104,
+      "step": 995
+    },
+    {
+      "epoch": 0.23366568914956012,
+      "grad_norm": 3.852799335158689,
+      "learning_rate": 9.468435503292108e-06,
+      "loss": 0.6377,
+      "step": 996
+    },
+    {
+      "epoch": 0.23390029325513195,
+      "grad_norm": 1.220217041154978,
+      "learning_rate": 9.466596189987644e-06,
+      "loss": 0.6733,
+      "step": 997
+    },
+    {
+      "epoch": 0.2341348973607038,
+      "grad_norm": 1.0496610362694465,
+      "learning_rate": 9.464753879276467e-06,
+      "loss": 0.6215,
+      "step": 998
+    },
+    {
+      "epoch": 0.23436950146627566,
+      "grad_norm": 2.2817838781315336,
+      "learning_rate": 9.4629085723949e-06,
+      "loss": 0.6116,
+      "step": 999
+    },
+    {
+      "epoch": 0.23460410557184752,
+      "grad_norm": 1.3680657229056437,
+      "learning_rate": 9.461060270581276e-06,
+      "loss": 0.6267,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23460410557184752,
+      "eval_loss": 0.6492986679077148,
+      "eval_runtime": 25.3774,
+      "eval_samples_per_second": 21.515,
+      "eval_steps_per_second": 0.197,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23483870967741935,
+      "grad_norm": 5.649880362698411,
+      "learning_rate": 9.459208975075936e-06,
+      "loss": 0.5947,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2350733137829912,
+      "grad_norm": 1.3704636130767294,
+      "learning_rate": 9.457354687121234e-06,
+      "loss": 0.6535,
+      "step": 1002
+    },
+    {
+      "epoch": 0.23530791788856306,
+      "grad_norm": 1.5531555837354978,
+      "learning_rate": 9.455497407961533e-06,
+      "loss": 0.6578,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2355425219941349,
+      "grad_norm": 2.385298625756661,
+      "learning_rate": 9.453637138843196e-06,
+      "loss": 0.6898,
+      "step": 1004
+    },
+    {
+      "epoch": 0.23577712609970675,
+      "grad_norm": 2.18241034809193,
+      "learning_rate": 9.4517738810146e-06,
+      "loss": 0.6549,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2360117302052786,
+      "grad_norm": 1.1336483483024176,
+      "learning_rate": 9.449907635726125e-06,
+      "loss": 0.6367,
+      "step": 1006
+    },
+    {
+      "epoch": 0.23624633431085043,
+      "grad_norm": 1.0691181136009236,
+      "learning_rate": 9.448038404230156e-06,
+      "loss": 0.6423,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2364809384164223,
+      "grad_norm": 1.0553117097822624,
+      "learning_rate": 9.446166187781077e-06,
+      "loss": 0.6984,
+      "step": 1008
+    },
+    {
+      "epoch": 0.23671554252199414,
+      "grad_norm": 1.5341499013376068,
+      "learning_rate": 9.444290987635285e-06,
+      "loss": 0.6761,
+      "step": 1009
+    },
+    {
+      "epoch": 0.23695014662756597,
+      "grad_norm": 1.3608123842594628,
+      "learning_rate": 9.442412805051171e-06,
+      "loss": 0.6885,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23718475073313783,
+      "grad_norm": 2.4917367720298387,
+      "learning_rate": 9.440531641289133e-06,
+      "loss": 0.5706,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23741935483870968,
+      "grad_norm": 1.303283510097461,
+      "learning_rate": 9.438647497611562e-06,
+      "loss": 0.6636,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2376539589442815,
+      "grad_norm": 1.0013777791433367,
+      "learning_rate": 9.436760375282858e-06,
+      "loss": 0.664,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23788856304985337,
+      "grad_norm": 1.2638124086460538,
+      "learning_rate": 9.434870275569416e-06,
+      "loss": 0.6067,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23812316715542522,
+      "grad_norm": 2.9285589828455536,
+      "learning_rate": 9.432977199739625e-06,
+      "loss": 0.642,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23835777126099708,
+      "grad_norm": 2.208856498844545,
+      "learning_rate": 9.431081149063877e-06,
+      "loss": 0.6834,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2385923753665689,
+      "grad_norm": 1.5849986325533076,
+      "learning_rate": 9.429182124814559e-06,
+      "loss": 0.5901,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23882697947214077,
+      "grad_norm": 1.257929746202979,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.6423,
+      "step": 1018
+    },
+    {
+      "epoch": 0.23906158357771262,
+      "grad_norm": 2.2476085484775834,
+      "learning_rate": 9.425375160694728e-06,
+      "loss": 0.6493,
+      "step": 1019
+    },
+    {
+      "epoch": 0.23929618768328445,
+      "grad_norm": 1.0074820565121343,
+      "learning_rate": 9.423467223378963e-06,
+      "loss": 0.6378,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2395307917888563,
+      "grad_norm": 1.4286848351206902,
+      "learning_rate": 9.42155631759912e-06,
+      "loss": 0.6528,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23976539589442816,
+      "grad_norm": 1.605810190143817,
+      "learning_rate": 9.419642444637547e-06,
+      "loss": 0.6155,
+      "step": 1022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.124106366515897,
+      "learning_rate": 9.417725605778599e-06,
+      "loss": 0.6336,
+      "step": 1023
+    },
+    {
+      "epoch": 0.24023460410557185,
+      "grad_norm": 1.1513761949498305,
+      "learning_rate": 9.415805802308606e-06,
+      "loss": 0.6878,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2404692082111437,
+      "grad_norm": 1.684519474529015,
+      "learning_rate": 9.413883035515897e-06,
+      "loss": 0.6533,
+      "step": 1025
+    },
+    {
+      "epoch": 0.24070381231671553,
+      "grad_norm": 3.3090245316020157,
+      "learning_rate": 9.411957306690784e-06,
+      "loss": 0.5911,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2409384164222874,
+      "grad_norm": 1.400329534624225,
+      "learning_rate": 9.410028617125572e-06,
+      "loss": 0.6988,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24117302052785924,
+      "grad_norm": 1.1639084569386746,
+      "learning_rate": 9.40809696811455e-06,
+      "loss": 0.6551,
+      "step": 1028
+    },
+    {
+      "epoch": 0.24140762463343107,
+      "grad_norm": 1.3969320801825622,
+      "learning_rate": 9.406162360953993e-06,
+      "loss": 0.6635,
+      "step": 1029
+    },
+    {
+      "epoch": 0.24164222873900293,
+      "grad_norm": 1.5563103547874322,
+      "learning_rate": 9.40422479694216e-06,
+      "loss": 0.5857,
+      "step": 1030
+    },
+    {
+      "epoch": 0.24187683284457479,
+      "grad_norm": 3.31361637878754,
+      "learning_rate": 9.402284277379295e-06,
+      "loss": 0.6511,
+      "step": 1031
+    },
+    {
+      "epoch": 0.24211143695014664,
+      "grad_norm": 1.575052667811003,
+      "learning_rate": 9.400340803567628e-06,
+      "loss": 0.6414,
+      "step": 1032
+    },
+    {
+      "epoch": 0.24234604105571847,
+      "grad_norm": 4.850669535283172,
+      "learning_rate": 9.398394376811368e-06,
+      "loss": 0.6273,
+      "step": 1033
+    },
+    {
+      "epoch": 0.24258064516129033,
+      "grad_norm": 1.8310303846415128,
+      "learning_rate": 9.39644499841671e-06,
+      "loss": 0.6402,
+      "step": 1034
+    },
+    {
+      "epoch": 0.24281524926686218,
+      "grad_norm": 1.9977695614564166,
+      "learning_rate": 9.394492669691822e-06,
+      "loss": 0.5988,
+      "step": 1035
+    },
+    {
+      "epoch": 0.243049853372434,
+      "grad_norm": 1.2847567253839351,
+      "learning_rate": 9.39253739194686e-06,
+      "loss": 0.6792,
+      "step": 1036
+    },
+    {
+      "epoch": 0.24328445747800587,
+      "grad_norm": 1.4476041025422224,
+      "learning_rate": 9.390579166493955e-06,
+      "loss": 0.6865,
+      "step": 1037
+    },
+    {
+      "epoch": 0.24351906158357772,
+      "grad_norm": 2.0841697799025276,
+      "learning_rate": 9.388617994647217e-06,
+      "loss": 0.6343,
+      "step": 1038
+    },
+    {
+      "epoch": 0.24375366568914955,
+      "grad_norm": 1.19250423254534,
+      "learning_rate": 9.386653877722734e-06,
+      "loss": 0.6307,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2439882697947214,
+      "grad_norm": 2.4696550859126396,
+      "learning_rate": 9.384686817038569e-06,
+      "loss": 0.6656,
+      "step": 1040
+    },
+    {
+      "epoch": 0.24422287390029326,
+      "grad_norm": 1.7291518224078675,
+      "learning_rate": 9.382716813914758e-06,
+      "loss": 0.5661,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2444574780058651,
+      "grad_norm": 1.671832316453634,
+      "learning_rate": 9.380743869673319e-06,
+      "loss": 0.6366,
+      "step": 1042
+    },
+    {
+      "epoch": 0.24469208211143695,
+      "grad_norm": 1.9079638634802105,
+      "learning_rate": 9.378767985638236e-06,
+      "loss": 0.6393,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2449266862170088,
+      "grad_norm": 1.4178979802273974,
+      "learning_rate": 9.37678916313547e-06,
+      "loss": 0.621,
+      "step": 1044
+    },
+    {
+      "epoch": 0.24516129032258063,
+      "grad_norm": 4.628465027913799,
+      "learning_rate": 9.374807403492953e-06,
+      "loss": 0.6233,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2453958944281525,
+      "grad_norm": 1.4082937621006473,
+      "learning_rate": 9.37282270804059e-06,
+      "loss": 0.6288,
+      "step": 1046
+    },
+    {
+      "epoch": 0.24563049853372435,
+      "grad_norm": 1.3120202178320575,
+      "learning_rate": 9.370835078110252e-06,
+      "loss": 0.6206,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2458651026392962,
+      "grad_norm": 3.760185749341159,
+      "learning_rate": 9.36884451503578e-06,
+      "loss": 0.6435,
+      "step": 1048
+    },
+    {
+      "epoch": 0.24609970674486803,
+      "grad_norm": 1.6052202436924854,
+      "learning_rate": 9.366851020152988e-06,
+      "loss": 0.6317,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24633431085043989,
+      "grad_norm": 1.3160901614252933,
+      "learning_rate": 9.364854594799653e-06,
+      "loss": 0.6246,
+      "step": 1050
+    },
+    {
+      "epoch": 0.24656891495601174,
+      "grad_norm": 6.051988896786264,
+      "learning_rate": 9.362855240315519e-06,
+      "loss": 0.6584,
+      "step": 1051
+    },
+    {
+      "epoch": 0.24680351906158357,
+      "grad_norm": 1.4054889298572484,
+      "learning_rate": 9.360852958042294e-06,
+      "loss": 0.7053,
+      "step": 1052
+    },
+    {
+      "epoch": 0.24703812316715543,
+      "grad_norm": 1.3057230010347616,
+      "learning_rate": 9.35884774932366e-06,
+      "loss": 0.584,
+      "step": 1053
+    },
+    {
+      "epoch": 0.24727272727272728,
+      "grad_norm": 1.960286991708724,
+      "learning_rate": 9.35683961550525e-06,
+      "loss": 0.6803,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2475073313782991,
+      "grad_norm": 1.4665550768481745,
+      "learning_rate": 9.354828557934667e-06,
+      "loss": 0.6629,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24774193548387097,
+      "grad_norm": 1.4036753887323576,
+      "learning_rate": 9.352814577961478e-06,
+      "loss": 0.6403,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24797653958944282,
+      "grad_norm": 1.2706243050107302,
+      "learning_rate": 9.350797676937204e-06,
+      "loss": 0.6162,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24821114369501465,
+      "grad_norm": 1.0365410635812358,
+      "learning_rate": 9.348777856215335e-06,
+      "loss": 0.633,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2484457478005865,
+      "grad_norm": 4.164437790587883,
+      "learning_rate": 9.346755117151311e-06,
+      "loss": 0.6338,
+      "step": 1059
+    },
+    {
+      "epoch": 0.24868035190615836,
+      "grad_norm": 2.0340855529569044,
+      "learning_rate": 9.344729461102537e-06,
+      "loss": 0.7013,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2489149560117302,
+      "grad_norm": 1.1934887503823248,
+      "learning_rate": 9.342700889428376e-06,
+      "loss": 0.6704,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24914956011730205,
+      "grad_norm": 1.2040869613771354,
+      "learning_rate": 9.340669403490142e-06,
+      "loss": 0.6119,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2493841642228739,
+      "grad_norm": 6.798485147523802,
+      "learning_rate": 9.338635004651108e-06,
+      "loss": 0.6574,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24961876832844576,
+      "grad_norm": 1.2031946135972638,
+      "learning_rate": 9.336597694276505e-06,
+      "loss": 0.6772,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2498533724340176,
+      "grad_norm": 2.7896772865947974,
+      "learning_rate": 9.33455747373351e-06,
+      "loss": 0.6685,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2500879765395894,
+      "grad_norm": 1.434718978079452,
+      "learning_rate": 9.332514344391261e-06,
+      "loss": 0.6707,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2503225806451613,
+      "grad_norm": 1.649597067527315,
+      "learning_rate": 9.330468307620844e-06,
+      "loss": 0.6524,
+      "step": 1067
+    },
+    {
+      "epoch": 0.25055718475073313,
+      "grad_norm": 0.9706807975515751,
+      "learning_rate": 9.328419364795295e-06,
+      "loss": 0.6286,
+      "step": 1068
+    },
+    {
+      "epoch": 0.250791788856305,
+      "grad_norm": 2.8391301835048055,
+      "learning_rate": 9.326367517289601e-06,
+      "loss": 0.6309,
+      "step": 1069
+    },
+    {
+      "epoch": 0.25102639296187684,
+      "grad_norm": 10.149338307211673,
+      "learning_rate": 9.324312766480703e-06,
+      "loss": 0.6086,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2512609970674487,
+      "grad_norm": 3.332973642515985,
+      "learning_rate": 9.322255113747483e-06,
+      "loss": 0.668,
+      "step": 1071
+    },
+    {
+      "epoch": 0.25149560117302056,
+      "grad_norm": 1.6289507101115803,
+      "learning_rate": 9.320194560470774e-06,
+      "loss": 0.6529,
+      "step": 1072
+    },
+    {
+      "epoch": 0.25173020527859236,
+      "grad_norm": 1.057748663764242,
+      "learning_rate": 9.318131108033355e-06,
+      "loss": 0.7172,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2519648093841642,
+      "grad_norm": 5.911103910329838,
+      "learning_rate": 9.316064757819951e-06,
+      "loss": 0.6615,
+      "step": 1074
+    },
+    {
+      "epoch": 0.25219941348973607,
+      "grad_norm": 0.8361519820241816,
+      "learning_rate": 9.313995511217234e-06,
+      "loss": 0.647,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2524340175953079,
+      "grad_norm": 1.430006543360786,
+      "learning_rate": 9.311923369613814e-06,
+      "loss": 0.5779,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2526686217008798,
+      "grad_norm": 1.3705999959998239,
+      "learning_rate": 9.309848334400247e-06,
+      "loss": 0.6185,
+      "step": 1077
+    },
+    {
+      "epoch": 0.25290322580645164,
+      "grad_norm": 1.8027003953341156,
+      "learning_rate": 9.307770406969032e-06,
+      "loss": 0.7027,
+      "step": 1078
+    },
+    {
+      "epoch": 0.25313782991202344,
+      "grad_norm": 1.147881489060793,
+      "learning_rate": 9.305689588714607e-06,
+      "loss": 0.6871,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2533724340175953,
+      "grad_norm": 1.2640670079055658,
+      "learning_rate": 9.303605881033347e-06,
+      "loss": 0.6705,
+      "step": 1080
+    },
+    {
+      "epoch": 0.25360703812316715,
+      "grad_norm": 2.0806876631568647,
+      "learning_rate": 9.301519285323574e-06,
+      "loss": 0.6329,
+      "step": 1081
+    },
+    {
+      "epoch": 0.253841642228739,
+      "grad_norm": 1.0093463389901316,
+      "learning_rate": 9.299429802985544e-06,
+      "loss": 0.6615,
+      "step": 1082
+    },
+    {
+      "epoch": 0.25407624633431086,
+      "grad_norm": 0.9052612031399299,
+      "learning_rate": 9.297337435421447e-06,
+      "loss": 0.621,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2543108504398827,
+      "grad_norm": 2.2904057629841077,
+      "learning_rate": 9.295242184035411e-06,
+      "loss": 0.706,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 3.3432158091490893,
+      "learning_rate": 9.293144050233502e-06,
+      "loss": 0.6377,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2547800586510264,
+      "grad_norm": 1.3541565854115583,
+      "learning_rate": 9.29104303542372e-06,
+      "loss": 0.6255,
+      "step": 1086
+    },
+    {
+      "epoch": 0.25501466275659823,
+      "grad_norm": 1.36594944692631,
+      "learning_rate": 9.288939141015993e-06,
+      "loss": 0.602,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2552492668621701,
+      "grad_norm": 1.5060651999404175,
+      "learning_rate": 9.286832368422188e-06,
+      "loss": 0.6483,
+      "step": 1088
+    },
+    {
+      "epoch": 0.25548387096774194,
+      "grad_norm": 14.08745174413511,
+      "learning_rate": 9.2847227190561e-06,
+      "loss": 0.6518,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2557184750733138,
+      "grad_norm": 1.0285700361179184,
+      "learning_rate": 9.282610194333453e-06,
+      "loss": 0.6561,
+      "step": 1090
+    },
+    {
+      "epoch": 0.25595307917888566,
+      "grad_norm": 3.3255615306981947,
+      "learning_rate": 9.280494795671906e-06,
+      "loss": 0.6508,
+      "step": 1091
+    },
+    {
+      "epoch": 0.25618768328445746,
+      "grad_norm": 1.2084681719142458,
+      "learning_rate": 9.278376524491041e-06,
+      "loss": 0.7036,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2564222873900293,
+      "grad_norm": 2.6036419960007904,
+      "learning_rate": 9.276255382212374e-06,
+      "loss": 0.6228,
+      "step": 1093
+    },
+    {
+      "epoch": 0.25665689149560117,
+      "grad_norm": 27.743670008252177,
+      "learning_rate": 9.274131370259338e-06,
+      "loss": 0.64,
+      "step": 1094
+    },
+    {
+      "epoch": 0.256891495601173,
+      "grad_norm": 1.4044619403265397,
+      "learning_rate": 9.272004490057301e-06,
+      "loss": 0.608,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2571260997067449,
+      "grad_norm": 2.356120977313629,
+      "learning_rate": 9.269874743033552e-06,
+      "loss": 0.6457,
+      "step": 1096
+    },
+    {
+      "epoch": 0.25736070381231674,
+      "grad_norm": 1.2596090390033359,
+      "learning_rate": 9.267742130617303e-06,
+      "loss": 0.657,
+      "step": 1097
+    },
+    {
+      "epoch": 0.25759530791788854,
+      "grad_norm": 1.3946092978167728,
+      "learning_rate": 9.265606654239691e-06,
+      "loss": 0.6597,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2578299120234604,
+      "grad_norm": 1.4432119584318348,
+      "learning_rate": 9.263468315333774e-06,
+      "loss": 0.6312,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 1.5515541825522112,
+      "learning_rate": 9.261327115334531e-06,
+      "loss": 0.599,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2582991202346041,
+      "grad_norm": 1.4007693081450738,
+      "learning_rate": 9.259183055678862e-06,
+      "loss": 0.6489,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25853372434017596,
+      "grad_norm": 0.9103206016592846,
+      "learning_rate": 9.257036137805582e-06,
+      "loss": 0.6335,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2587683284457478,
+      "grad_norm": 2.888495959965812,
+      "learning_rate": 9.254886363155429e-06,
+      "loss": 0.6071,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2590029325513197,
+      "grad_norm": 1.3821034837192627,
+      "learning_rate": 9.252733733171056e-06,
+      "loss": 0.6486,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2592375366568915,
+      "grad_norm": 1.4320609764115106,
+      "learning_rate": 9.25057824929703e-06,
+      "loss": 0.6374,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25947214076246333,
+      "grad_norm": 1.119572011338813,
+      "learning_rate": 9.248419912979839e-06,
+      "loss": 0.6845,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2597067448680352,
+      "grad_norm": 1.2163580289103197,
+      "learning_rate": 9.246258725667881e-06,
+      "loss": 0.6264,
+      "step": 1107
+    },
+    {
+      "epoch": 0.25994134897360704,
+      "grad_norm": 1.535185463904399,
+      "learning_rate": 9.244094688811464e-06,
+      "loss": 0.6681,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2601759530791789,
+      "grad_norm": 1.7914866375488323,
+      "learning_rate": 9.241927803862818e-06,
+      "loss": 0.6775,
+      "step": 1109
+    },
+    {
+      "epoch": 0.26041055718475076,
+      "grad_norm": 1.3248249659555873,
+      "learning_rate": 9.239758072276074e-06,
+      "loss": 0.6521,
+      "step": 1110
+    },
+    {
+      "epoch": 0.26064516129032256,
+      "grad_norm": 2.295648656574757,
+      "learning_rate": 9.23758549550728e-06,
+      "loss": 0.6337,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2608797653958944,
+      "grad_norm": 11.75891681296019,
+      "learning_rate": 9.235410075014388e-06,
+      "loss": 0.6349,
+      "step": 1112
+    },
+    {
+      "epoch": 0.26111436950146627,
+      "grad_norm": 1.6452283269390608,
+      "learning_rate": 9.233231812257266e-06,
+      "loss": 0.5934,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2613489736070381,
+      "grad_norm": 1.4467978833872515,
+      "learning_rate": 9.23105070869768e-06,
+      "loss": 0.6278,
+      "step": 1114
+    },
+    {
+      "epoch": 0.26158357771261,
+      "grad_norm": 1.113625485020427,
+      "learning_rate": 9.228866765799309e-06,
+      "loss": 0.6643,
+      "step": 1115
+    },
+    {
+      "epoch": 0.26181818181818184,
+      "grad_norm": 10.773666725709122,
+      "learning_rate": 9.226679985027737e-06,
+      "loss": 0.6215,
+      "step": 1116
+    },
+    {
+      "epoch": 0.26205278592375364,
+      "grad_norm": 1.148086209338217,
+      "learning_rate": 9.22449036785045e-06,
+      "loss": 0.7284,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2622873900293255,
+      "grad_norm": 4.126390930082524,
+      "learning_rate": 9.222297915736835e-06,
+      "loss": 0.635,
+      "step": 1118
+    },
+    {
+      "epoch": 0.26252199413489735,
+      "grad_norm": 3.2049462664895296,
+      "learning_rate": 9.22010263015819e-06,
+      "loss": 0.6171,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2627565982404692,
+      "grad_norm": 21.67792858284106,
+      "learning_rate": 9.217904512587705e-06,
+      "loss": 0.5676,
+      "step": 1120
+    },
+    {
+      "epoch": 0.26299120234604106,
+      "grad_norm": 12.556299898852679,
+      "learning_rate": 9.215703564500478e-06,
+      "loss": 0.6152,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2632258064516129,
+      "grad_norm": 1.1779982705559455,
+      "learning_rate": 9.213499787373501e-06,
+      "loss": 0.6658,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2634604105571848,
+      "grad_norm": 1.3696518802678466,
+      "learning_rate": 9.211293182685668e-06,
+      "loss": 0.5802,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2636950146627566,
+      "grad_norm": 1.8874730792051249,
+      "learning_rate": 9.209083751917767e-06,
+      "loss": 0.6727,
+      "step": 1124
+    },
+    {
+      "epoch": 0.26392961876832843,
+      "grad_norm": 0.958902915394563,
+      "learning_rate": 9.206871496552488e-06,
+      "loss": 0.6486,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2641642228739003,
+      "grad_norm": 6.635419383316604,
+      "learning_rate": 9.20465641807441e-06,
+      "loss": 0.7071,
+      "step": 1126
+    },
+    {
+      "epoch": 0.26439882697947215,
+      "grad_norm": 1.4702756870641926,
+      "learning_rate": 9.202438517970011e-06,
+      "loss": 0.6239,
+      "step": 1127
+    },
+    {
+      "epoch": 0.264633431085044,
+      "grad_norm": 1.7660617836374857,
+      "learning_rate": 9.200217797727663e-06,
+      "loss": 0.6057,
+      "step": 1128
+    },
+    {
+      "epoch": 0.26486803519061586,
+      "grad_norm": 1.4327848284209121,
+      "learning_rate": 9.197994258837627e-06,
+      "loss": 0.6537,
+      "step": 1129
+    },
+    {
+      "epoch": 0.26510263929618766,
+      "grad_norm": 2.2357305825059925,
+      "learning_rate": 9.195767902792055e-06,
+      "loss": 0.6545,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2653372434017595,
+      "grad_norm": 1.115617574359435,
+      "learning_rate": 9.193538731084994e-06,
+      "loss": 0.6483,
+      "step": 1131
+    },
+    {
+      "epoch": 0.26557184750733137,
+      "grad_norm": 2.1127768529842204,
+      "learning_rate": 9.19130674521238e-06,
+      "loss": 0.6839,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2658064516129032,
+      "grad_norm": 0.9313613458147902,
+      "learning_rate": 9.189071946672034e-06,
+      "loss": 0.6875,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2660410557184751,
+      "grad_norm": 0.8318000531860501,
+      "learning_rate": 9.186834336963665e-06,
+      "loss": 0.6093,
+      "step": 1134
+    },
+    {
+      "epoch": 0.26627565982404694,
+      "grad_norm": 1.1986092453350174,
+      "learning_rate": 9.184593917588872e-06,
+      "loss": 0.6201,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2665102639296188,
+      "grad_norm": 1.7391576367935673,
+      "learning_rate": 9.182350690051134e-06,
+      "loss": 0.5965,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2667448680351906,
+      "grad_norm": 2.0279189111500484,
+      "learning_rate": 9.180104655855822e-06,
+      "loss": 0.7121,
+      "step": 1137
+    },
+    {
+      "epoch": 0.26697947214076245,
+      "grad_norm": 1.341125336561769,
+      "learning_rate": 9.177855816510184e-06,
+      "loss": 0.6235,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2672140762463343,
+      "grad_norm": 1.540030271635298,
+      "learning_rate": 9.17560417352335e-06,
+      "loss": 0.6698,
+      "step": 1139
+    },
+    {
+      "epoch": 0.26744868035190617,
+      "grad_norm": 1.5842640964603802,
+      "learning_rate": 9.173349728406342e-06,
+      "loss": 0.6461,
+      "step": 1140
+    },
+    {
+      "epoch": 0.267683284457478,
+      "grad_norm": 1.279593804742548,
+      "learning_rate": 9.171092482672046e-06,
+      "loss": 0.7069,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2679178885630499,
+      "grad_norm": 4.766485979159509,
+      "learning_rate": 9.16883243783524e-06,
+      "loss": 0.6368,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2681524926686217,
+      "grad_norm": 3.4148221767290217,
+      "learning_rate": 9.166569595412576e-06,
+      "loss": 0.6509,
+      "step": 1143
+    },
+    {
+      "epoch": 0.26838709677419353,
+      "grad_norm": 1.3395580626279817,
+      "learning_rate": 9.164303956922582e-06,
+      "loss": 0.6265,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2686217008797654,
+      "grad_norm": 1.9671916289784808,
+      "learning_rate": 9.162035523885669e-06,
+      "loss": 0.6307,
+      "step": 1145
+    },
+    {
+      "epoch": 0.26885630498533725,
+      "grad_norm": 1.1231349124358356,
+      "learning_rate": 9.159764297824111e-06,
+      "loss": 0.5958,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2690909090909091,
+      "grad_norm": 1.19635176688707,
+      "learning_rate": 9.15749028026207e-06,
+      "loss": 0.6029,
+      "step": 1147
+    },
+    {
+      "epoch": 0.26932551319648096,
+      "grad_norm": 1.0063746732596204,
+      "learning_rate": 9.155213472725572e-06,
+      "loss": 0.6571,
+      "step": 1148
+    },
+    {
+      "epoch": 0.26956011730205276,
+      "grad_norm": 1.2724902231266055,
+      "learning_rate": 9.152933876742518e-06,
+      "loss": 0.6915,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2697947214076246,
+      "grad_norm": 1.093749239254766,
+      "learning_rate": 9.150651493842683e-06,
+      "loss": 0.6117,
+      "step": 1150
+    },
+    {
+      "epoch": 0.27002932551319647,
+      "grad_norm": 13.259899934384775,
+      "learning_rate": 9.148366325557707e-06,
+      "loss": 0.6972,
+      "step": 1151
+    },
+    {
+      "epoch": 0.27026392961876833,
+      "grad_norm": 0.7952876762386353,
+      "learning_rate": 9.146078373421106e-06,
+      "loss": 0.5975,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2704985337243402,
+      "grad_norm": 1.077868856066063,
+      "learning_rate": 9.143787638968255e-06,
+      "loss": 0.6165,
+      "step": 1153
+    },
+    {
+      "epoch": 0.27073313782991204,
+      "grad_norm": 1.1436430969707327,
+      "learning_rate": 9.141494123736403e-06,
+      "loss": 0.6669,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2709677419354839,
+      "grad_norm": 1.441281901569953,
+      "learning_rate": 9.139197829264668e-06,
+      "loss": 0.6202,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2712023460410557,
+      "grad_norm": 2.2442609787801726,
+      "learning_rate": 9.136898757094022e-06,
+      "loss": 0.6579,
+      "step": 1156
+    },
+    {
+      "epoch": 0.27143695014662755,
+      "grad_norm": 2.083902766981121,
+      "learning_rate": 9.13459690876731e-06,
+      "loss": 0.6156,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2716715542521994,
+      "grad_norm": 1.4164132517497274,
+      "learning_rate": 9.132292285829237e-06,
+      "loss": 0.6866,
+      "step": 1158
+    },
+    {
+      "epoch": 0.27190615835777127,
+      "grad_norm": 1.3221074842967184,
+      "learning_rate": 9.129984889826373e-06,
+      "loss": 0.6511,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2721407624633431,
+      "grad_norm": 1.452528344416446,
+      "learning_rate": 9.127674722307142e-06,
+      "loss": 0.7298,
+      "step": 1160
+    },
+    {
+      "epoch": 0.272375366568915,
+      "grad_norm": 1.2918319576786386,
+      "learning_rate": 9.125361784821837e-06,
+      "loss": 0.6537,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2726099706744868,
+      "grad_norm": 0.9875594783287066,
+      "learning_rate": 9.123046078922601e-06,
+      "loss": 0.6512,
+      "step": 1162
+    },
+    {
+      "epoch": 0.27284457478005864,
+      "grad_norm": 1.6672133667940412,
+      "learning_rate": 9.120727606163443e-06,
+      "loss": 0.641,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2730791788856305,
+      "grad_norm": 1.2912257331658266,
+      "learning_rate": 9.11840636810022e-06,
+      "loss": 0.6634,
+      "step": 1164
+    },
+    {
+      "epoch": 0.27331378299120235,
+      "grad_norm": 1.8275919353703045,
+      "learning_rate": 9.116082366290653e-06,
+      "loss": 0.6561,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2735483870967742,
+      "grad_norm": 2.0492504335429915,
+      "learning_rate": 9.113755602294311e-06,
+      "loss": 0.6155,
+      "step": 1166
+    },
+    {
+      "epoch": 0.27378299120234606,
+      "grad_norm": 23.51595392877474,
+      "learning_rate": 9.111426077672623e-06,
+      "loss": 0.6691,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2740175953079179,
+      "grad_norm": 1.336879483997251,
+      "learning_rate": 9.109093793988866e-06,
+      "loss": 0.6681,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2742521994134897,
+      "grad_norm": 1.4331122193508863,
+      "learning_rate": 9.10675875280817e-06,
+      "loss": 0.6119,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2744868035190616,
+      "grad_norm": 1.847249001964482,
+      "learning_rate": 9.104420955697513e-06,
+      "loss": 0.6777,
+      "step": 1170
+    },
+    {
+      "epoch": 0.27472140762463343,
+      "grad_norm": 1.1832291339880137,
+      "learning_rate": 9.102080404225726e-06,
+      "loss": 0.6798,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2749560117302053,
+      "grad_norm": 4.932581301158991,
+      "learning_rate": 9.09973709996349e-06,
+      "loss": 0.5892,
+      "step": 1172
+    },
+    {
+      "epoch": 0.27519061583577714,
+      "grad_norm": 0.9293782191072257,
+      "learning_rate": 9.097391044483325e-06,
+      "loss": 0.672,
+      "step": 1173
+    },
+    {
+      "epoch": 0.275425219941349,
+      "grad_norm": 1.4098630567348378,
+      "learning_rate": 9.095042239359608e-06,
+      "loss": 0.6194,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2756598240469208,
+      "grad_norm": 1.701552206764051,
+      "learning_rate": 9.092690686168552e-06,
+      "loss": 0.7233,
+      "step": 1175
+    },
+    {
+      "epoch": 0.27589442815249265,
+      "grad_norm": 2.0167746179023855,
+      "learning_rate": 9.090336386488222e-06,
+      "loss": 0.6088,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2761290322580645,
+      "grad_norm": 1.4990342339704141,
+      "learning_rate": 9.087979341898518e-06,
+      "loss": 0.6503,
+      "step": 1177
+    },
+    {
+      "epoch": 0.27636363636363637,
+      "grad_norm": 3.8467707530438044,
+      "learning_rate": 9.085619553981186e-06,
+      "loss": 0.62,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2765982404692082,
+      "grad_norm": 1.320421580962417,
+      "learning_rate": 9.083257024319817e-06,
+      "loss": 0.617,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2768328445747801,
+      "grad_norm": 2.4487808473763626,
+      "learning_rate": 9.080891754499836e-06,
+      "loss": 0.6253,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2770674486803519,
+      "grad_norm": 2.4803890794677357,
+      "learning_rate": 9.078523746108508e-06,
+      "loss": 0.6302,
+      "step": 1181
+    },
+    {
+      "epoch": 0.27730205278592374,
+      "grad_norm": 1.3205711383390113,
+      "learning_rate": 9.076153000734938e-06,
+      "loss": 0.6563,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2775366568914956,
+      "grad_norm": 2.767032869268495,
+      "learning_rate": 9.073779519970066e-06,
+      "loss": 0.6445,
+      "step": 1183
+    },
+    {
+      "epoch": 0.27777126099706745,
+      "grad_norm": 1.2609096823864498,
+      "learning_rate": 9.071403305406664e-06,
+      "loss": 0.693,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2780058651026393,
+      "grad_norm": 1.161706749463485,
+      "learning_rate": 9.069024358639352e-06,
+      "loss": 0.709,
+      "step": 1185
+    },
+    {
+      "epoch": 0.27824046920821116,
+      "grad_norm": 1.4075792283668007,
+      "learning_rate": 9.066642681264565e-06,
+      "loss": 0.67,
+      "step": 1186
+    },
+    {
+      "epoch": 0.278475073313783,
+      "grad_norm": 1.400424726965094,
+      "learning_rate": 9.064258274880583e-06,
+      "loss": 0.6217,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2787096774193548,
+      "grad_norm": 1.488864332901986,
+      "learning_rate": 9.061871141087514e-06,
+      "loss": 0.6405,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2789442815249267,
+      "grad_norm": 2.41257517341978,
+      "learning_rate": 9.059481281487297e-06,
+      "loss": 0.6633,
+      "step": 1189
+    },
+    {
+      "epoch": 0.27917888563049853,
+      "grad_norm": 1.1878581629498228,
+      "learning_rate": 9.057088697683695e-06,
+      "loss": 0.655,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2794134897360704,
+      "grad_norm": 1.6685399189300316,
+      "learning_rate": 9.05469339128231e-06,
+      "loss": 0.6261,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27964809384164224,
+      "grad_norm": 2.304310182312456,
+      "learning_rate": 9.052295363890561e-06,
+      "loss": 0.6444,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2798826979472141,
+      "grad_norm": 1.3241671634056527,
+      "learning_rate": 9.049894617117697e-06,
+      "loss": 0.65,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2801173020527859,
+      "grad_norm": 1.26071567796767,
+      "learning_rate": 9.04749115257479e-06,
+      "loss": 0.6223,
+      "step": 1194
+    },
+    {
+      "epoch": 0.28035190615835776,
+      "grad_norm": 3.4060533850546073,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.6414,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2805865102639296,
+      "grad_norm": 1.0797584781386456,
+      "learning_rate": 9.042676076632262e-06,
+      "loss": 0.6103,
+      "step": 1196
+    },
+    {
+      "epoch": 0.28082111436950147,
+      "grad_norm": 2.3500412110531226,
+      "learning_rate": 9.040264468463905e-06,
+      "loss": 0.6658,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2810557184750733,
+      "grad_norm": 2.5888536931024158,
+      "learning_rate": 9.037850148988025e-06,
+      "loss": 0.6292,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2812903225806452,
+      "grad_norm": 1.2721951957423576,
+      "learning_rate": 9.035433119824808e-06,
+      "loss": 0.6561,
+      "step": 1199
+    },
+    {
+      "epoch": 0.28152492668621704,
+      "grad_norm": 1.6362863863423596,
+      "learning_rate": 9.033013382596251e-06,
+      "loss": 0.6251,
+      "step": 1200
+    },
+    {
+      "epoch": 0.28175953079178884,
+      "grad_norm": 1.6474849766998976,
+      "learning_rate": 9.030590938926173e-06,
+      "loss": 0.673,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2819941348973607,
+      "grad_norm": 1.6701199617740936,
+      "learning_rate": 9.028165790440206e-06,
+      "loss": 0.6441,
+      "step": 1202
+    },
+    {
+      "epoch": 0.28222873900293255,
+      "grad_norm": 1.2187041648785473,
+      "learning_rate": 9.025737938765803e-06,
+      "loss": 0.6457,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2824633431085044,
+      "grad_norm": 2.5892033320650727,
+      "learning_rate": 9.023307385532221e-06,
+      "loss": 0.6626,
+      "step": 1204
+    },
+    {
+      "epoch": 0.28269794721407626,
+      "grad_norm": 1.6534979371731078,
+      "learning_rate": 9.020874132370539e-06,
+      "loss": 0.6619,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2829325513196481,
+      "grad_norm": 2.3196392706709608,
+      "learning_rate": 9.018438180913644e-06,
+      "loss": 0.6608,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2831671554252199,
+      "grad_norm": 1.2616125476534032,
+      "learning_rate": 9.015999532796237e-06,
+      "loss": 0.6394,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2834017595307918,
+      "grad_norm": 1.079851047458475,
+      "learning_rate": 9.013558189654819e-06,
+      "loss": 0.6734,
+      "step": 1208
+    },
+    {
+      "epoch": 0.28363636363636363,
+      "grad_norm": 0.945963473909306,
+      "learning_rate": 9.011114153127713e-06,
+      "loss": 0.6595,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2838709677419355,
+      "grad_norm": 1.2202345279339784,
+      "learning_rate": 9.00866742485504e-06,
+      "loss": 0.6094,
+      "step": 1210
+    },
+    {
+      "epoch": 0.28410557184750734,
+      "grad_norm": 1.6632063032700166,
+      "learning_rate": 9.006218006478733e-06,
+      "loss": 0.6195,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2843401759530792,
+      "grad_norm": 1.543917239991805,
+      "learning_rate": 9.003765899642522e-06,
+      "loss": 0.6378,
+      "step": 1212
+    },
+    {
+      "epoch": 0.284574780058651,
+      "grad_norm": 1.2988372928219616,
+      "learning_rate": 9.001311105991955e-06,
+      "loss": 0.6478,
+      "step": 1213
+    },
+    {
+      "epoch": 0.28480938416422286,
+      "grad_norm": 1.5211791632950824,
+      "learning_rate": 8.998853627174366e-06,
+      "loss": 0.5813,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2850439882697947,
+      "grad_norm": 1.764491215207445,
+      "learning_rate": 8.996393464838906e-06,
+      "loss": 0.6872,
+      "step": 1215
+    },
+    {
+      "epoch": 0.28527859237536657,
+      "grad_norm": 1.457599821694021,
+      "learning_rate": 8.99393062063652e-06,
+      "loss": 0.6764,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2855131964809384,
+      "grad_norm": 7.9843204821114915,
+      "learning_rate": 8.991465096219949e-06,
+      "loss": 0.6762,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2857478005865103,
+      "grad_norm": 2.281328672609779,
+      "learning_rate": 8.988996893243742e-06,
+      "loss": 0.6836,
+      "step": 1218
+    },
+    {
+      "epoch": 0.28598240469208214,
+      "grad_norm": 34.42511778676702,
+      "learning_rate": 8.986526013364236e-06,
+      "loss": 0.6762,
+      "step": 1219
+    },
+    {
+      "epoch": 0.28621700879765394,
+      "grad_norm": 1.782916570198515,
+      "learning_rate": 8.98405245823957e-06,
+      "loss": 0.5941,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2864516129032258,
+      "grad_norm": 1.2283119944541914,
+      "learning_rate": 8.98157622952968e-06,
+      "loss": 0.6419,
+      "step": 1221
+    },
+    {
+      "epoch": 0.28668621700879765,
+      "grad_norm": 3.963871646741732,
+      "learning_rate": 8.979097328896292e-06,
+      "loss": 0.6615,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2869208211143695,
+      "grad_norm": 1.8606785533696573,
+      "learning_rate": 8.976615758002924e-06,
+      "loss": 0.6159,
+      "step": 1223
+    },
+    {
+      "epoch": 0.28715542521994136,
+      "grad_norm": 11.427896013323235,
+      "learning_rate": 8.97413151851489e-06,
+      "loss": 0.6047,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2873900293255132,
+      "grad_norm": 2.3284593765249832,
+      "learning_rate": 8.971644612099292e-06,
+      "loss": 0.6466,
+      "step": 1225
+    },
+    {
+      "epoch": 0.287624633431085,
+      "grad_norm": 1.6781674773194468,
+      "learning_rate": 8.969155040425024e-06,
+      "loss": 0.5746,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2878592375366569,
+      "grad_norm": 1.6035318347922205,
+      "learning_rate": 8.966662805162766e-06,
+      "loss": 0.6735,
+      "step": 1227
+    },
+    {
+      "epoch": 0.28809384164222873,
+      "grad_norm": 1.5828439617406542,
+      "learning_rate": 8.964167907984989e-06,
+      "loss": 0.6923,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2883284457478006,
+      "grad_norm": 1.1751048929274226,
+      "learning_rate": 8.961670350565945e-06,
+      "loss": 0.6525,
+      "step": 1229
+    },
+    {
+      "epoch": 0.28856304985337244,
+      "grad_norm": 6.7059065740479875,
+      "learning_rate": 8.959170134581677e-06,
+      "loss": 0.5547,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2887976539589443,
+      "grad_norm": 1.4608480268772748,
+      "learning_rate": 8.956667261710007e-06,
+      "loss": 0.5718,
+      "step": 1231
+    },
+    {
+      "epoch": 0.28903225806451616,
+      "grad_norm": 1.6912842066364013,
+      "learning_rate": 8.954161733630546e-06,
+      "loss": 0.6642,
+      "step": 1232
+    },
+    {
+      "epoch": 0.28926686217008796,
+      "grad_norm": 1.116960226766822,
+      "learning_rate": 8.951653552024681e-06,
+      "loss": 0.6187,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2895014662756598,
+      "grad_norm": 1.9517351039051145,
+      "learning_rate": 8.949142718575582e-06,
+      "loss": 0.667,
+      "step": 1234
+    },
+    {
+      "epoch": 0.28973607038123167,
+      "grad_norm": 2.4300227400850214,
+      "learning_rate": 8.946629234968196e-06,
+      "loss": 0.6293,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2899706744868035,
+      "grad_norm": 1.1995657881689836,
+      "learning_rate": 8.944113102889257e-06,
+      "loss": 0.6431,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2902052785923754,
+      "grad_norm": 1.6908230703481364,
+      "learning_rate": 8.941594324027263e-06,
+      "loss": 0.6384,
+      "step": 1237
+    },
+    {
+      "epoch": 0.29043988269794724,
+      "grad_norm": 2.6265952904076557,
+      "learning_rate": 8.939072900072501e-06,
+      "loss": 0.6417,
+      "step": 1238
+    },
+    {
+      "epoch": 0.29067448680351904,
+      "grad_norm": 2.0392197385607904,
+      "learning_rate": 8.936548832717024e-06,
+      "loss": 0.6557,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 2.0104281934582895,
+      "learning_rate": 8.934022123654667e-06,
+      "loss": 0.6026,
+      "step": 1240
+    },
+    {
+      "epoch": 0.29114369501466275,
+      "grad_norm": 1.2199453887680525,
+      "learning_rate": 8.931492774581028e-06,
+      "loss": 0.6262,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2913782991202346,
+      "grad_norm": 1.1753234947448552,
+      "learning_rate": 8.928960787193486e-06,
+      "loss": 0.6915,
+      "step": 1242
+    },
+    {
+      "epoch": 0.29161290322580646,
+      "grad_norm": 1.5561084137983587,
+      "learning_rate": 8.926426163191181e-06,
+      "loss": 0.6373,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2918475073313783,
+      "grad_norm": 1.362777603513328,
+      "learning_rate": 8.923888904275036e-06,
+      "loss": 0.6325,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2920821114369501,
+      "grad_norm": 2.8333130460272145,
+      "learning_rate": 8.921349012147724e-06,
+      "loss": 0.6566,
+      "step": 1245
+    },
+    {
+      "epoch": 0.292316715542522,
+      "grad_norm": 1.1212176544259231,
+      "learning_rate": 8.918806488513702e-06,
+      "loss": 0.6217,
+      "step": 1246
+    },
+    {
+      "epoch": 0.29255131964809383,
+      "grad_norm": 1.3257534046241175,
+      "learning_rate": 8.916261335079185e-06,
+      "loss": 0.6553,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2927859237536657,
+      "grad_norm": 2.101574963760381,
+      "learning_rate": 8.91371355355215e-06,
+      "loss": 0.58,
+      "step": 1248
+    },
+    {
+      "epoch": 0.29302052785923755,
+      "grad_norm": 1.1721164736904788,
+      "learning_rate": 8.911163145642346e-06,
+      "loss": 0.6542,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2932551319648094,
+      "grad_norm": 1.8243554113686955,
+      "learning_rate": 8.908610113061276e-06,
+      "loss": 0.603,
+      "step": 1250
+    },
+    {
+      "epoch": 0.29348973607038126,
+      "grad_norm": 9.405338704027788,
+      "learning_rate": 8.90605445752221e-06,
+      "loss": 0.6416,
+      "step": 1251
+    },
+    {
+      "epoch": 0.29372434017595306,
+      "grad_norm": 1.2324339661112436,
+      "learning_rate": 8.903496180740176e-06,
+      "loss": 0.5899,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2939589442815249,
+      "grad_norm": 2.05393213429527,
+      "learning_rate": 8.900935284431962e-06,
+      "loss": 0.6328,
+      "step": 1253
+    },
+    {
+      "epoch": 0.29419354838709677,
+      "grad_norm": 2.229622561102952,
+      "learning_rate": 8.898371770316113e-06,
+      "loss": 0.6716,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2944281524926686,
+      "grad_norm": 1.0086112497124473,
+      "learning_rate": 8.89580564011293e-06,
+      "loss": 0.6624,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2946627565982405,
+      "grad_norm": 2.0998846289072532,
+      "learning_rate": 8.893236895544472e-06,
+      "loss": 0.6717,
+      "step": 1256
+    },
+    {
+      "epoch": 0.29489736070381234,
+      "grad_norm": 2.7024253872943578,
+      "learning_rate": 8.89066553833455e-06,
+      "loss": 0.6527,
+      "step": 1257
+    },
+    {
+      "epoch": 0.29513196480938414,
+      "grad_norm": 2.1374557550177475,
+      "learning_rate": 8.88809157020873e-06,
+      "loss": 0.6144,
+      "step": 1258
+    },
+    {
+      "epoch": 0.295366568914956,
+      "grad_norm": 1.387422227052281,
+      "learning_rate": 8.885514992894328e-06,
+      "loss": 0.6455,
+      "step": 1259
+    },
+    {
+      "epoch": 0.29560117302052785,
+      "grad_norm": 1.4483659812274923,
+      "learning_rate": 8.882935808120415e-06,
+      "loss": 0.6259,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2958357771260997,
+      "grad_norm": 2.4045389283475775,
+      "learning_rate": 8.880354017617806e-06,
+      "loss": 0.6515,
+      "step": 1261
+    },
+    {
+      "epoch": 0.29607038123167156,
+      "grad_norm": 1.2142610769997242,
+      "learning_rate": 8.87776962311907e-06,
+      "loss": 0.6213,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2963049853372434,
+      "grad_norm": 1.1932169412093907,
+      "learning_rate": 8.87518262635852e-06,
+      "loss": 0.6601,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2965395894428153,
+      "grad_norm": 1.3103600311840702,
+      "learning_rate": 8.872593029072218e-06,
+      "loss": 0.6008,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2967741935483871,
+      "grad_norm": 12.806866833231975,
+      "learning_rate": 8.870000832997969e-06,
+      "loss": 0.6514,
+      "step": 1265
+    },
+    {
+      "epoch": 0.29700879765395893,
+      "grad_norm": 1.8867732638569539,
+      "learning_rate": 8.867406039875322e-06,
+      "loss": 0.6072,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2972434017595308,
+      "grad_norm": 1.7776538825058188,
+      "learning_rate": 8.864808651445571e-06,
+      "loss": 0.6305,
+      "step": 1267
+    },
+    {
+      "epoch": 0.29747800586510265,
+      "grad_norm": 2.5356146132409836,
+      "learning_rate": 8.862208669451748e-06,
+      "loss": 0.6274,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2977126099706745,
+      "grad_norm": 1.1405033241200686,
+      "learning_rate": 8.85960609563863e-06,
+      "loss": 0.6065,
+      "step": 1269
+    },
+    {
+      "epoch": 0.29794721407624636,
+      "grad_norm": 5.247737154694485,
+      "learning_rate": 8.857000931752727e-06,
+      "loss": 0.6206,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29818181818181816,
+      "grad_norm": 1.6192986589450444,
+      "learning_rate": 8.854393179542296e-06,
+      "loss": 0.649,
+      "step": 1271
+    },
+    {
+      "epoch": 0.29841642228739,
+      "grad_norm": 0.955327285671442,
+      "learning_rate": 8.851782840757322e-06,
+      "loss": 0.6322,
+      "step": 1272
+    },
+    {
+      "epoch": 0.29865102639296187,
+      "grad_norm": 10.205085139887778,
+      "learning_rate": 8.849169917149532e-06,
+      "loss": 0.6312,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2988856304985337,
+      "grad_norm": 2.0419819861702027,
+      "learning_rate": 8.846554410472385e-06,
+      "loss": 0.6312,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2991202346041056,
+      "grad_norm": 2.539642348519771,
+      "learning_rate": 8.843936322481073e-06,
+      "loss": 0.6329,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29935483870967744,
+      "grad_norm": 1.9786206541515448,
+      "learning_rate": 8.841315654932521e-06,
+      "loss": 0.6982,
+      "step": 1276
+    },
+    {
+      "epoch": 0.29958944281524924,
+      "grad_norm": 2.714725438430519,
+      "learning_rate": 8.838692409585386e-06,
+      "loss": 0.6069,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2998240469208211,
+      "grad_norm": 1.614491519495665,
+      "learning_rate": 8.836066588200052e-06,
+      "loss": 0.7046,
+      "step": 1278
+    },
+    {
+      "epoch": 0.30005865102639295,
+      "grad_norm": 2.255260228506507,
+      "learning_rate": 8.833438192538635e-06,
+      "loss": 0.6133,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3002932551319648,
+      "grad_norm": 2.280776635454184,
+      "learning_rate": 8.830807224364978e-06,
+      "loss": 0.6639,
+      "step": 1280
+    },
+    {
+      "epoch": 0.30052785923753667,
+      "grad_norm": 1.1117201617882997,
+      "learning_rate": 8.82817368544465e-06,
+      "loss": 0.6253,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3007624633431085,
+      "grad_norm": 1.048458771972502,
+      "learning_rate": 8.825537577544941e-06,
+      "loss": 0.5812,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3009970674486804,
+      "grad_norm": 0.8928909363340157,
+      "learning_rate": 8.822898902434873e-06,
+      "loss": 0.6959,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3012316715542522,
+      "grad_norm": 1.700575200742987,
+      "learning_rate": 8.820257661885184e-06,
+      "loss": 0.6136,
+      "step": 1284
+    },
+    {
+      "epoch": 0.30146627565982403,
+      "grad_norm": 0.9843673483434341,
+      "learning_rate": 8.817613857668336e-06,
+      "loss": 0.6366,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3017008797653959,
+      "grad_norm": 1.065147000843776,
+      "learning_rate": 8.814967491558513e-06,
+      "loss": 0.6487,
+      "step": 1286
+    },
+    {
+      "epoch": 0.30193548387096775,
+      "grad_norm": 2.0813320709130085,
+      "learning_rate": 8.812318565331615e-06,
+      "loss": 0.6214,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3021700879765396,
+      "grad_norm": 2.780332471849667,
+      "learning_rate": 8.809667080765262e-06,
+      "loss": 0.5831,
+      "step": 1288
+    },
+    {
+      "epoch": 0.30240469208211146,
+      "grad_norm": 1.1904230973336396,
+      "learning_rate": 8.807013039638792e-06,
+      "loss": 0.5899,
+      "step": 1289
+    },
+    {
+      "epoch": 0.30263929618768326,
+      "grad_norm": 0.9612417248794958,
+      "learning_rate": 8.804356443733252e-06,
+      "loss": 0.6273,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3028739002932551,
+      "grad_norm": 1.4060937168302299,
+      "learning_rate": 8.801697294831418e-06,
+      "loss": 0.6508,
+      "step": 1291
+    },
+    {
+      "epoch": 0.303108504398827,
+      "grad_norm": 1.192759901327455,
+      "learning_rate": 8.79903559471776e-06,
+      "loss": 0.6359,
+      "step": 1292
+    },
+    {
+      "epoch": 0.30334310850439883,
+      "grad_norm": 1.9201539193558765,
+      "learning_rate": 8.796371345178477e-06,
+      "loss": 0.6493,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3035777126099707,
+      "grad_norm": 2.1214585599349567,
+      "learning_rate": 8.793704548001468e-06,
+      "loss": 0.5816,
+      "step": 1294
+    },
+    {
+      "epoch": 0.30381231671554254,
+      "grad_norm": 0.8599548018777288,
+      "learning_rate": 8.791035204976346e-06,
+      "loss": 0.6127,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3040469208211144,
+      "grad_norm": 1.5129429147192395,
+      "learning_rate": 8.788363317894432e-06,
+      "loss": 0.6351,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3042815249266862,
+      "grad_norm": 2.6109877591111403,
+      "learning_rate": 8.785688888548756e-06,
+      "loss": 0.6309,
+      "step": 1297
+    },
+    {
+      "epoch": 0.30451612903225805,
+      "grad_norm": 1.0230297416784169,
+      "learning_rate": 8.783011918734048e-06,
+      "loss": 0.5925,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3047507331378299,
+      "grad_norm": 2.186467330141591,
+      "learning_rate": 8.780332410246751e-06,
+      "loss": 0.5888,
+      "step": 1299
+    },
+    {
+      "epoch": 0.30498533724340177,
+      "grad_norm": 1.67491301133915,
+      "learning_rate": 8.777650364885004e-06,
+      "loss": 0.6724,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3052199413489736,
+      "grad_norm": 1.1812740966549162,
+      "learning_rate": 8.774965784448655e-06,
+      "loss": 0.6233,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3054545454545455,
+      "grad_norm": 1.3653435440565702,
+      "learning_rate": 8.772278670739249e-06,
+      "loss": 0.646,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3056891495601173,
+      "grad_norm": 0.974001510579673,
+      "learning_rate": 8.76958902556003e-06,
+      "loss": 0.6553,
+      "step": 1303
+    },
+    {
+      "epoch": 0.30592375366568914,
+      "grad_norm": 1.8805764437627832,
+      "learning_rate": 8.766896850715946e-06,
+      "loss": 0.5921,
+      "step": 1304
+    },
+    {
+      "epoch": 0.306158357771261,
+      "grad_norm": 1.3919439552419068,
+      "learning_rate": 8.764202148013641e-06,
+      "loss": 0.6436,
+      "step": 1305
+    },
+    {
+      "epoch": 0.30639296187683285,
+      "grad_norm": 1.1462932005939608,
+      "learning_rate": 8.761504919261449e-06,
+      "loss": 0.6637,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3066275659824047,
+      "grad_norm": 1.1892967885593082,
+      "learning_rate": 8.758805166269408e-06,
+      "loss": 0.5987,
+      "step": 1307
+    },
+    {
+      "epoch": 0.30686217008797656,
+      "grad_norm": 1.1052793782042585,
+      "learning_rate": 8.756102890849246e-06,
+      "loss": 0.6807,
+      "step": 1308
+    },
+    {
+      "epoch": 0.30709677419354836,
+      "grad_norm": 6.883279038259657,
+      "learning_rate": 8.753398094814382e-06,
+      "loss": 0.7119,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3073313782991202,
+      "grad_norm": 1.081980986043162,
+      "learning_rate": 8.750690779979929e-06,
+      "loss": 0.5951,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3075659824046921,
+      "grad_norm": 2.0675410988875154,
+      "learning_rate": 8.74798094816269e-06,
+      "loss": 0.6512,
+      "step": 1311
+    },
+    {
+      "epoch": 0.30780058651026393,
+      "grad_norm": 1.3202120008001563,
+      "learning_rate": 8.745268601181156e-06,
+      "loss": 0.6496,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3080351906158358,
+      "grad_norm": 1.0247827042515312,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.6551,
+      "step": 1313
+    },
+    {
+      "epoch": 0.30826979472140764,
+      "grad_norm": 5.786819309028735,
+      "learning_rate": 8.739836369007609e-06,
+      "loss": 0.6285,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3085043988269795,
+      "grad_norm": 60.92674679982773,
+      "learning_rate": 8.737116487461012e-06,
+      "loss": 0.6267,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3087390029325513,
+      "grad_norm": 1.2774203798381183,
+      "learning_rate": 8.734394098040958e-06,
+      "loss": 0.5971,
+      "step": 1316
+    },
+    {
+      "epoch": 0.30897360703812315,
+      "grad_norm": 2.163502894715537,
+      "learning_rate": 8.731669202574358e-06,
+      "loss": 0.6197,
+      "step": 1317
+    },
+    {
+      "epoch": 0.309208211143695,
+      "grad_norm": 2.152718603403912,
+      "learning_rate": 8.728941802889816e-06,
+      "loss": 0.6498,
+      "step": 1318
+    },
+    {
+      "epoch": 0.30944281524926687,
+      "grad_norm": 1.4893222425762045,
+      "learning_rate": 8.726211900817615e-06,
+      "loss": 0.627,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3096774193548387,
+      "grad_norm": 1.249555665489977,
+      "learning_rate": 8.723479498189709e-06,
+      "loss": 0.6245,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3099120234604106,
+      "grad_norm": 1.1178881942453802,
+      "learning_rate": 8.720744596839743e-06,
+      "loss": 0.7129,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3101466275659824,
+      "grad_norm": 1.3964290043601861,
+      "learning_rate": 8.718007198603028e-06,
+      "loss": 0.5847,
+      "step": 1322
+    },
+    {
+      "epoch": 0.31038123167155424,
+      "grad_norm": 1.6111423346882032,
+      "learning_rate": 8.715267305316559e-06,
+      "loss": 0.6642,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3106158357771261,
+      "grad_norm": 0.9612562831354406,
+      "learning_rate": 8.712524918818997e-06,
+      "loss": 0.6563,
+      "step": 1324
+    },
+    {
+      "epoch": 0.31085043988269795,
+      "grad_norm": 1.1038307346379574,
+      "learning_rate": 8.70978004095068e-06,
+      "loss": 0.5907,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3110850439882698,
+      "grad_norm": 0.766408183617374,
+      "learning_rate": 8.70703267355362e-06,
+      "loss": 0.5443,
+      "step": 1326
+    },
+    {
+      "epoch": 0.31131964809384166,
+      "grad_norm": 1.478909949815166,
+      "learning_rate": 8.704282818471501e-06,
+      "loss": 0.6915,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3115542521994135,
+      "grad_norm": 1.195723486093403,
+      "learning_rate": 8.701530477549666e-06,
+      "loss": 0.6195,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3117888563049853,
+      "grad_norm": 0.7850915533063527,
+      "learning_rate": 8.698775652635139e-06,
+      "loss": 0.6113,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3120234604105572,
+      "grad_norm": 1.3491574479114687,
+      "learning_rate": 8.696018345576602e-06,
+      "loss": 0.705,
+      "step": 1330
+    },
+    {
+      "epoch": 0.31225806451612903,
+      "grad_norm": 1.724935421468019,
+      "learning_rate": 8.693258558224406e-06,
+      "loss": 0.6542,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3124926686217009,
+      "grad_norm": 1.8955182529789312,
+      "learning_rate": 8.690496292430571e-06,
+      "loss": 0.6627,
+      "step": 1332
+    },
+    {
+      "epoch": 0.31272727272727274,
+      "grad_norm": 2.1000158286295925,
+      "learning_rate": 8.68773155004877e-06,
+      "loss": 0.6359,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3129618768328446,
+      "grad_norm": 1.2343201375497233,
+      "learning_rate": 8.684964332934347e-06,
+      "loss": 0.63,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3131964809384164,
+      "grad_norm": 1.1670954778337763,
+      "learning_rate": 8.682194642944304e-06,
+      "loss": 0.635,
+      "step": 1335
+    },
+    {
+      "epoch": 0.31343108504398826,
+      "grad_norm": 2.1456346002897657,
+      "learning_rate": 8.679422481937297e-06,
+      "loss": 0.6245,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3136656891495601,
+      "grad_norm": 1.9460672497958593,
+      "learning_rate": 8.67664785177365e-06,
+      "loss": 0.6102,
+      "step": 1337
+    },
+    {
+      "epoch": 0.31390029325513197,
+      "grad_norm": 1.755780007413827,
+      "learning_rate": 8.673870754315336e-06,
+      "loss": 0.6433,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3141348973607038,
+      "grad_norm": 2.120790636904061,
+      "learning_rate": 8.671091191425991e-06,
+      "loss": 0.6888,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3143695014662757,
+      "grad_norm": 0.8886934532473196,
+      "learning_rate": 8.668309164970896e-06,
+      "loss": 0.5888,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3146041055718475,
+      "grad_norm": 2.4991745759717463,
+      "learning_rate": 8.665524676816992e-06,
+      "loss": 0.6203,
+      "step": 1341
+    },
+    {
+      "epoch": 0.31483870967741934,
+      "grad_norm": 2.598062493726064,
+      "learning_rate": 8.662737728832873e-06,
+      "loss": 0.6058,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3150733137829912,
+      "grad_norm": 0.9442172449383684,
+      "learning_rate": 8.659948322888778e-06,
+      "loss": 0.5829,
+      "step": 1343
+    },
+    {
+      "epoch": 0.31530791788856305,
+      "grad_norm": 2.691296024074173,
+      "learning_rate": 8.657156460856602e-06,
+      "loss": 0.6585,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3155425219941349,
+      "grad_norm": 1.0442090423796575,
+      "learning_rate": 8.654362144609882e-06,
+      "loss": 0.6217,
+      "step": 1345
+    },
+    {
+      "epoch": 0.31577712609970676,
+      "grad_norm": 0.9727296807581846,
+      "learning_rate": 8.651565376023805e-06,
+      "loss": 0.6,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3160117302052786,
+      "grad_norm": 1.550040656766428,
+      "learning_rate": 8.648766156975207e-06,
+      "loss": 0.658,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3162463343108504,
+      "grad_norm": 2.026571473268327,
+      "learning_rate": 8.645964489342558e-06,
+      "loss": 0.5898,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3164809384164223,
+      "grad_norm": 0.9336579973671697,
+      "learning_rate": 8.643160375005985e-06,
+      "loss": 0.6419,
+      "step": 1349
+    },
+    {
+      "epoch": 0.31671554252199413,
+      "grad_norm": 2.1600425115092596,
+      "learning_rate": 8.640353815847247e-06,
+      "loss": 0.6264,
+      "step": 1350
+    },
+    {
+      "epoch": 0.316950146627566,
+      "grad_norm": 1.2313394543528944,
+      "learning_rate": 8.637544813749747e-06,
+      "loss": 0.6003,
+      "step": 1351
+    },
+    {
+      "epoch": 0.31718475073313784,
+      "grad_norm": 1.007341668974774,
+      "learning_rate": 8.634733370598526e-06,
+      "loss": 0.6332,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3174193548387097,
+      "grad_norm": 1.914352211483566,
+      "learning_rate": 8.631919488280267e-06,
+      "loss": 0.6323,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3176539589442815,
+      "grad_norm": 0.92833216095689,
+      "learning_rate": 8.629103168683284e-06,
+      "loss": 0.6215,
+      "step": 1354
+    },
+    {
+      "epoch": 0.31788856304985336,
+      "grad_norm": 1.2857911810777851,
+      "learning_rate": 8.626284413697532e-06,
+      "loss": 0.6565,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3181231671554252,
+      "grad_norm": 1.555815894161518,
+      "learning_rate": 8.623463225214597e-06,
+      "loss": 0.6787,
+      "step": 1356
+    },
+    {
+      "epoch": 0.31835777126099707,
+      "grad_norm": 2.0162124628364015,
+      "learning_rate": 8.620639605127698e-06,
+      "loss": 0.6358,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3185923753665689,
+      "grad_norm": 0.8076909991521062,
+      "learning_rate": 8.61781355533169e-06,
+      "loss": 0.6373,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3188269794721408,
+      "grad_norm": 2.0354027172745,
+      "learning_rate": 8.61498507772305e-06,
+      "loss": 0.6122,
+      "step": 1359
+    },
+    {
+      "epoch": 0.31906158357771264,
+      "grad_norm": 1.245336316860784,
+      "learning_rate": 8.612154174199896e-06,
+      "loss": 0.5933,
+      "step": 1360
+    },
+    {
+      "epoch": 0.31929618768328444,
+      "grad_norm": 3.5947107493839967,
+      "learning_rate": 8.60932084666196e-06,
+      "loss": 0.6302,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3195307917888563,
+      "grad_norm": 3.39452963397432,
+      "learning_rate": 8.606485097010616e-06,
+      "loss": 0.6256,
+      "step": 1362
+    },
+    {
+      "epoch": 0.31976539589442815,
+      "grad_norm": 1.5234498781472734,
+      "learning_rate": 8.60364692714885e-06,
+      "loss": 0.6682,
+      "step": 1363
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5146001755726328,
+      "learning_rate": 8.60080633898128e-06,
+      "loss": 0.6822,
+      "step": 1364
+    },
+    {
+      "epoch": 0.32023460410557186,
+      "grad_norm": 2.0777388506099834,
+      "learning_rate": 8.597963334414146e-06,
+      "loss": 0.6053,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3204692082111437,
+      "grad_norm": 2.8714871255271497,
+      "learning_rate": 8.595117915355304e-06,
+      "loss": 0.7123,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3207038123167155,
+      "grad_norm": 0.9002279050923622,
+      "learning_rate": 8.592270083714238e-06,
+      "loss": 0.658,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3209384164222874,
+      "grad_norm": 1.2911358484741762,
+      "learning_rate": 8.589419841402046e-06,
+      "loss": 0.6094,
+      "step": 1368
+    },
+    {
+      "epoch": 0.32117302052785923,
+      "grad_norm": 25.665582483281256,
+      "learning_rate": 8.586567190331446e-06,
+      "loss": 0.6069,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3214076246334311,
+      "grad_norm": 1.537281364347152,
+      "learning_rate": 8.58371213241677e-06,
+      "loss": 0.5976,
+      "step": 1370
+    },
+    {
+      "epoch": 0.32164222873900294,
+      "grad_norm": 1.3845179576829354,
+      "learning_rate": 8.58085466957397e-06,
+      "loss": 0.6334,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3218768328445748,
+      "grad_norm": 1.1386389309429865,
+      "learning_rate": 8.577994803720605e-06,
+      "loss": 0.7086,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3221114369501466,
+      "grad_norm": 1.3254556899668493,
+      "learning_rate": 8.575132536775854e-06,
+      "loss": 0.5968,
+      "step": 1373
+    },
+    {
+      "epoch": 0.32234604105571846,
+      "grad_norm": 1.123212271302889,
+      "learning_rate": 8.572267870660502e-06,
+      "loss": 0.6763,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 1.930857001417153,
+      "learning_rate": 8.569400807296944e-06,
+      "loss": 0.6183,
+      "step": 1375
+    },
+    {
+      "epoch": 0.32281524926686217,
+      "grad_norm": 2.9025633792142815,
+      "learning_rate": 8.566531348609186e-06,
+      "loss": 0.6276,
+      "step": 1376
+    },
+    {
+      "epoch": 0.323049853372434,
+      "grad_norm": 1.3048918897875166,
+      "learning_rate": 8.563659496522843e-06,
+      "loss": 0.6092,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3232844574780059,
+      "grad_norm": 10.034757630145917,
+      "learning_rate": 8.560785252965131e-06,
+      "loss": 0.6447,
+      "step": 1378
+    },
+    {
+      "epoch": 0.32351906158357774,
+      "grad_norm": 1.0461685899146798,
+      "learning_rate": 8.557908619864877e-06,
+      "loss": 0.6641,
+      "step": 1379
+    },
+    {
+      "epoch": 0.32375366568914954,
+      "grad_norm": 1.786751820197665,
+      "learning_rate": 8.555029599152504e-06,
+      "loss": 0.6361,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3239882697947214,
+      "grad_norm": 1.2934920307732916,
+      "learning_rate": 8.552148192760047e-06,
+      "loss": 0.6363,
+      "step": 1381
+    },
+    {
+      "epoch": 0.32422287390029325,
+      "grad_norm": 1.0223073535206437,
+      "learning_rate": 8.549264402621133e-06,
+      "loss": 0.6551,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3244574780058651,
+      "grad_norm": 3.149917754690008,
+      "learning_rate": 8.546378230670992e-06,
+      "loss": 0.5889,
+      "step": 1383
+    },
+    {
+      "epoch": 0.32469208211143696,
+      "grad_norm": 1.1804970129191603,
+      "learning_rate": 8.543489678846453e-06,
+      "loss": 0.6581,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3249266862170088,
+      "grad_norm": 1.5082960885509378,
+      "learning_rate": 8.540598749085944e-06,
+      "loss": 0.6824,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3251612903225806,
+      "grad_norm": 1.1248000804124532,
+      "learning_rate": 8.537705443329482e-06,
+      "loss": 0.6595,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3253958944281525,
+      "grad_norm": 1.3529893497939975,
+      "learning_rate": 8.534809763518688e-06,
+      "loss": 0.643,
+      "step": 1387
+    },
+    {
+      "epoch": 0.32563049853372433,
+      "grad_norm": 1.4854156645555927,
+      "learning_rate": 8.531911711596767e-06,
+      "loss": 0.6337,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3258651026392962,
+      "grad_norm": 1.4938945083987845,
+      "learning_rate": 8.529011289508522e-06,
+      "loss": 0.6729,
+      "step": 1389
+    },
+    {
+      "epoch": 0.32609970674486805,
+      "grad_norm": 2.454648325683254,
+      "learning_rate": 8.526108499200344e-06,
+      "loss": 0.6601,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3263343108504399,
+      "grad_norm": 3.7530756452385896,
+      "learning_rate": 8.523203342620212e-06,
+      "loss": 0.6357,
+      "step": 1391
+    },
+    {
+      "epoch": 0.32656891495601176,
+      "grad_norm": 1.2082065032805829,
+      "learning_rate": 8.520295821717697e-06,
+      "loss": 0.6231,
+      "step": 1392
+    },
+    {
+      "epoch": 0.32680351906158356,
+      "grad_norm": 1.4691056719546882,
+      "learning_rate": 8.517385938443955e-06,
+      "loss": 0.6318,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3270381231671554,
+      "grad_norm": 1.3820520549209452,
+      "learning_rate": 8.514473694751723e-06,
+      "loss": 0.6629,
+      "step": 1394
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 1.8581106263157248,
+      "learning_rate": 8.511559092595329e-06,
+      "loss": 0.5865,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3275073313782991,
+      "grad_norm": 1.5517032644123216,
+      "learning_rate": 8.508642133930678e-06,
+      "loss": 0.6412,
+      "step": 1396
+    },
+    {
+      "epoch": 0.327741935483871,
+      "grad_norm": 0.9314890362404621,
+      "learning_rate": 8.50572282071526e-06,
+      "loss": 0.6346,
+      "step": 1397
+    },
+    {
+      "epoch": 0.32797653958944284,
+      "grad_norm": 4.83121516223516,
+      "learning_rate": 8.502801154908142e-06,
+      "loss": 0.6453,
+      "step": 1398
+    },
+    {
+      "epoch": 0.32821114369501464,
+      "grad_norm": 1.3559041143950146,
+      "learning_rate": 8.499877138469973e-06,
+      "loss": 0.6618,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3284457478005865,
+      "grad_norm": 4.184967102088001,
+      "learning_rate": 8.49695077336298e-06,
+      "loss": 0.6231,
+      "step": 1400
+    },
+    {
+      "epoch": 0.32868035190615835,
+      "grad_norm": 1.0389706936140033,
+      "learning_rate": 8.49402206155096e-06,
+      "loss": 0.6215,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3289149560117302,
+      "grad_norm": 1.0812713256722006,
+      "learning_rate": 8.491091004999293e-06,
+      "loss": 0.6327,
+      "step": 1402
+    },
+    {
+      "epoch": 0.32914956011730206,
+      "grad_norm": 1.9312825000734952,
+      "learning_rate": 8.488157605674924e-06,
+      "loss": 0.5882,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3293841642228739,
+      "grad_norm": 1.357776340775692,
+      "learning_rate": 8.485221865546377e-06,
+      "loss": 0.6269,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3296187683284457,
+      "grad_norm": 0.9904610631410483,
+      "learning_rate": 8.482283786583745e-06,
+      "loss": 0.6261,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3298533724340176,
+      "grad_norm": 1.3350484411071832,
+      "learning_rate": 8.479343370758687e-06,
+      "loss": 0.6327,
+      "step": 1406
+    },
+    {
+      "epoch": 0.33008797653958943,
+      "grad_norm": 1.0406124091175053,
+      "learning_rate": 8.476400620044439e-06,
+      "loss": 0.6286,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3303225806451613,
+      "grad_norm": 1.2423300727646929,
+      "learning_rate": 8.47345553641579e-06,
+      "loss": 0.6257,
+      "step": 1408
+    },
+    {
+      "epoch": 0.33055718475073315,
+      "grad_norm": 2.5098220016498884,
+      "learning_rate": 8.470508121849108e-06,
+      "loss": 0.6509,
+      "step": 1409
+    },
+    {
+      "epoch": 0.330791788856305,
+      "grad_norm": 1.7740633957474148,
+      "learning_rate": 8.46755837832232e-06,
+      "loss": 0.6194,
+      "step": 1410
+    },
+    {
+      "epoch": 0.33102639296187686,
+      "grad_norm": 1.1350055293410866,
+      "learning_rate": 8.46460630781491e-06,
+      "loss": 0.6068,
+      "step": 1411
+    },
+    {
+      "epoch": 0.33126099706744866,
+      "grad_norm": 0.9013414198812099,
+      "learning_rate": 8.461651912307937e-06,
+      "loss": 0.7072,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3314956011730205,
+      "grad_norm": 1.4451292309717894,
+      "learning_rate": 8.45869519378401e-06,
+      "loss": 0.5997,
+      "step": 1413
+    },
+    {
+      "epoch": 0.33173020527859237,
+      "grad_norm": 1.042894858513527,
+      "learning_rate": 8.455736154227294e-06,
+      "loss": 0.6516,
+      "step": 1414
+    },
+    {
+      "epoch": 0.33196480938416423,
+      "grad_norm": 1.0817335775599264,
+      "learning_rate": 8.452774795623525e-06,
+      "loss": 0.5674,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3321994134897361,
+      "grad_norm": 0.9306528221739203,
+      "learning_rate": 8.449811119959982e-06,
+      "loss": 0.5637,
+      "step": 1416
+    },
+    {
+      "epoch": 0.33243401759530794,
+      "grad_norm": 2.9259924025571524,
+      "learning_rate": 8.446845129225505e-06,
+      "loss": 0.6871,
+      "step": 1417
+    },
+    {
+      "epoch": 0.33266862170087974,
+      "grad_norm": 1.3116069619316164,
+      "learning_rate": 8.443876825410488e-06,
+      "loss": 0.6283,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3329032258064516,
+      "grad_norm": 1.1151155004404432,
+      "learning_rate": 8.440906210506875e-06,
+      "loss": 0.6526,
+      "step": 1419
+    },
+    {
+      "epoch": 0.33313782991202345,
+      "grad_norm": 1.006631697880105,
+      "learning_rate": 8.437933286508162e-06,
+      "loss": 0.6675,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3333724340175953,
+      "grad_norm": 2.812784011260699,
+      "learning_rate": 8.434958055409395e-06,
+      "loss": 0.5972,
+      "step": 1421
+    },
+    {
+      "epoch": 0.33360703812316717,
+      "grad_norm": 1.3126577072563264,
+      "learning_rate": 8.431980519207168e-06,
+      "loss": 0.5983,
+      "step": 1422
+    },
+    {
+      "epoch": 0.333841642228739,
+      "grad_norm": 2.6915182599167955,
+      "learning_rate": 8.42900067989962e-06,
+      "loss": 0.6569,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3340762463343109,
+      "grad_norm": 1.2287047288297095,
+      "learning_rate": 8.426018539486437e-06,
+      "loss": 0.6355,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3343108504398827,
+      "grad_norm": 1.4258173378604266,
+      "learning_rate": 8.423034099968851e-06,
+      "loss": 0.6191,
+      "step": 1425
+    },
+    {
+      "epoch": 0.33454545454545453,
+      "grad_norm": 1.4937674624905826,
+      "learning_rate": 8.420047363349636e-06,
+      "loss": 0.6695,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3347800586510264,
+      "grad_norm": 1.0141852252485175,
+      "learning_rate": 8.417058331633102e-06,
+      "loss": 0.681,
+      "step": 1427
+    },
+    {
+      "epoch": 0.33501466275659825,
+      "grad_norm": 3.0511373158997768,
+      "learning_rate": 8.414067006825108e-06,
+      "loss": 0.657,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3352492668621701,
+      "grad_norm": 5.300541328028981,
+      "learning_rate": 8.411073390933048e-06,
+      "loss": 0.5846,
+      "step": 1429
+    },
+    {
+      "epoch": 0.33548387096774196,
+      "grad_norm": 1.4936228127988724,
+      "learning_rate": 8.408077485965852e-06,
+      "loss": 0.6645,
+      "step": 1430
+    },
+    {
+      "epoch": 0.33571847507331376,
+      "grad_norm": 0.9734727974080785,
+      "learning_rate": 8.405079293933986e-06,
+      "loss": 0.6686,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3359530791788856,
+      "grad_norm": 1.4529755582105603,
+      "learning_rate": 8.402078816849454e-06,
+      "loss": 0.6269,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3361876832844575,
+      "grad_norm": 1.3421411041498725,
+      "learning_rate": 8.39907605672579e-06,
+      "loss": 0.6578,
+      "step": 1433
+    },
+    {
+      "epoch": 0.33642228739002933,
+      "grad_norm": 3.239032135699637,
+      "learning_rate": 8.396071015578064e-06,
+      "loss": 0.6525,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3366568914956012,
+      "grad_norm": 1.9505286799182235,
+      "learning_rate": 8.393063695422872e-06,
+      "loss": 0.7155,
+      "step": 1435
+    },
+    {
+      "epoch": 0.33689149560117304,
+      "grad_norm": 1.6183739710877414,
+      "learning_rate": 8.390054098278341e-06,
+      "loss": 0.6401,
+      "step": 1436
+    },
+    {
+      "epoch": 0.33712609970674484,
+      "grad_norm": 2.4434525002884087,
+      "learning_rate": 8.38704222616413e-06,
+      "loss": 0.6546,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3373607038123167,
+      "grad_norm": 2.9339990370893885,
+      "learning_rate": 8.384028081101419e-06,
+      "loss": 0.656,
+      "step": 1438
+    },
+    {
+      "epoch": 0.33759530791788855,
+      "grad_norm": 1.1336278724835074,
+      "learning_rate": 8.381011665112918e-06,
+      "loss": 0.6501,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3378299120234604,
+      "grad_norm": 3.213719727493025,
+      "learning_rate": 8.377992980222856e-06,
+      "loss": 0.6522,
+      "step": 1440
+    },
+    {
+      "epoch": 0.33806451612903227,
+      "grad_norm": 0.892289646256161,
+      "learning_rate": 8.374972028456987e-06,
+      "loss": 0.6515,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3382991202346041,
+      "grad_norm": 1.226685613357517,
+      "learning_rate": 8.371948811842589e-06,
+      "loss": 0.6187,
+      "step": 1442
+    },
+    {
+      "epoch": 0.338533724340176,
+      "grad_norm": 1.5404695990366168,
+      "learning_rate": 8.368923332408458e-06,
+      "loss": 0.6378,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3387683284457478,
+      "grad_norm": 3.0407503667494975,
+      "learning_rate": 8.365895592184905e-06,
+      "loss": 0.6314,
+      "step": 1444
+    },
+    {
+      "epoch": 0.33900293255131964,
+      "grad_norm": 1.3866293223711708,
+      "learning_rate": 8.362865593203761e-06,
+      "loss": 0.6249,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3392375366568915,
+      "grad_norm": 2.8038931976209054,
+      "learning_rate": 8.359833337498375e-06,
+      "loss": 0.6405,
+      "step": 1446
+    },
+    {
+      "epoch": 0.33947214076246335,
+      "grad_norm": 3.7934641081707556,
+      "learning_rate": 8.356798827103608e-06,
+      "loss": 0.5757,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3397067448680352,
+      "grad_norm": 1.0649556642285551,
+      "learning_rate": 8.353762064055832e-06,
+      "loss": 0.618,
+      "step": 1448
+    },
+    {
+      "epoch": 0.33994134897360706,
+      "grad_norm": 0.9222634812989948,
+      "learning_rate": 8.350723050392937e-06,
+      "loss": 0.6281,
+      "step": 1449
+    },
+    {
+      "epoch": 0.34017595307917886,
+      "grad_norm": 1.5902921327280632,
+      "learning_rate": 8.347681788154316e-06,
+      "loss": 0.6306,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3404105571847507,
+      "grad_norm": 1.1261301785362399,
+      "learning_rate": 8.344638279380875e-06,
+      "loss": 0.6863,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3406451612903226,
+      "grad_norm": 7.724626258384448,
+      "learning_rate": 8.341592526115028e-06,
+      "loss": 0.5977,
+      "step": 1452
+    },
+    {
+      "epoch": 0.34087976539589443,
+      "grad_norm": 1.1253873255951106,
+      "learning_rate": 8.338544530400693e-06,
+      "loss": 0.6226,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3411143695014663,
+      "grad_norm": 2.387143074516115,
+      "learning_rate": 8.335494294283295e-06,
+      "loss": 0.6078,
+      "step": 1454
+    },
+    {
+      "epoch": 0.34134897360703814,
+      "grad_norm": 1.0063514817449035,
+      "learning_rate": 8.33244181980976e-06,
+      "loss": 0.6459,
+      "step": 1455
+    },
+    {
+      "epoch": 0.34158357771261,
+      "grad_norm": 1.4947287816472277,
+      "learning_rate": 8.329387109028519e-06,
+      "loss": 0.607,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3418181818181818,
+      "grad_norm": 1.0172627081928454,
+      "learning_rate": 8.326330163989501e-06,
+      "loss": 0.6815,
+      "step": 1457
+    },
+    {
+      "epoch": 0.34205278592375365,
+      "grad_norm": 2.27189937321112,
+      "learning_rate": 8.323270986744137e-06,
+      "loss": 0.6376,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3422873900293255,
+      "grad_norm": 3.1671117988610327,
+      "learning_rate": 8.320209579345354e-06,
+      "loss": 0.6316,
+      "step": 1459
+    },
+    {
+      "epoch": 0.34252199413489737,
+      "grad_norm": 1.7703890703140879,
+      "learning_rate": 8.317145943847573e-06,
+      "loss": 0.6067,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3427565982404692,
+      "grad_norm": 2.8224082887446182,
+      "learning_rate": 8.314080082306718e-06,
+      "loss": 0.6491,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3429912023460411,
+      "grad_norm": 1.645283338929989,
+      "learning_rate": 8.3110119967802e-06,
+      "loss": 0.6197,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3432258064516129,
+      "grad_norm": 3.3933543977296905,
+      "learning_rate": 8.307941689326926e-06,
+      "loss": 0.6727,
+      "step": 1463
+    },
+    {
+      "epoch": 0.34346041055718474,
+      "grad_norm": 1.1657356603464364,
+      "learning_rate": 8.30486916200729e-06,
+      "loss": 0.6484,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3436950146627566,
+      "grad_norm": 3.8712679730499007,
+      "learning_rate": 8.301794416883182e-06,
+      "loss": 0.6499,
+      "step": 1465
+    },
+    {
+      "epoch": 0.34392961876832845,
+      "grad_norm": 2.070657817367473,
+      "learning_rate": 8.298717456017975e-06,
+      "loss": 0.6211,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3441642228739003,
+      "grad_norm": 1.3603966816757713,
+      "learning_rate": 8.295638281476529e-06,
+      "loss": 0.6804,
+      "step": 1467
+    },
+    {
+      "epoch": 0.34439882697947216,
+      "grad_norm": 1.2553303332731176,
+      "learning_rate": 8.292556895325195e-06,
+      "loss": 0.5872,
+      "step": 1468
+    },
+    {
+      "epoch": 0.34463343108504396,
+      "grad_norm": 1.6785132104367784,
+      "learning_rate": 8.2894732996318e-06,
+      "loss": 0.6962,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3448680351906158,
+      "grad_norm": 2.350729036974717,
+      "learning_rate": 8.28638749646566e-06,
+      "loss": 0.7025,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3451026392961877,
+      "grad_norm": 1.08608432429923,
+      "learning_rate": 8.28329948789757e-06,
+      "loss": 0.5597,
+      "step": 1471
+    },
+    {
+      "epoch": 0.34533724340175953,
+      "grad_norm": 2.5023569034921884,
+      "learning_rate": 8.28020927599981e-06,
+      "loss": 0.6566,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3455718475073314,
+      "grad_norm": 1.803515300768271,
+      "learning_rate": 8.277116862846127e-06,
+      "loss": 0.6396,
+      "step": 1473
+    },
+    {
+      "epoch": 0.34580645161290324,
+      "grad_norm": 1.7119582483094469,
+      "learning_rate": 8.274022250511755e-06,
+      "loss": 0.6224,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3460410557184751,
+      "grad_norm": 1.7639707272829974,
+      "learning_rate": 8.270925441073402e-06,
+      "loss": 0.5942,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3462756598240469,
+      "grad_norm": 1.1373456186692137,
+      "learning_rate": 8.26782643660925e-06,
+      "loss": 0.5606,
+      "step": 1476
+    },
+    {
+      "epoch": 0.34651026392961876,
+      "grad_norm": 3.2584651090339376,
+      "learning_rate": 8.264725239198952e-06,
+      "loss": 0.648,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3467448680351906,
+      "grad_norm": 0.9893722075990332,
+      "learning_rate": 8.261621850923634e-06,
+      "loss": 0.6523,
+      "step": 1478
+    },
+    {
+      "epoch": 0.34697947214076247,
+      "grad_norm": 1.6595170976168387,
+      "learning_rate": 8.258516273865894e-06,
+      "loss": 0.6635,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3472140762463343,
+      "grad_norm": 1.568496810851159,
+      "learning_rate": 8.255408510109793e-06,
+      "loss": 0.6594,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3474486803519062,
+      "grad_norm": 0.9773769390727363,
+      "learning_rate": 8.252298561740868e-06,
+      "loss": 0.5897,
+      "step": 1481
+    },
+    {
+      "epoch": 0.347683284457478,
+      "grad_norm": 4.976287138083255,
+      "learning_rate": 8.249186430846116e-06,
+      "loss": 0.6633,
+      "step": 1482
+    },
+    {
+      "epoch": 0.34791788856304984,
+      "grad_norm": 1.0167181553948246,
+      "learning_rate": 8.246072119514002e-06,
+      "loss": 0.6101,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3481524926686217,
+      "grad_norm": 1.4218757823242925,
+      "learning_rate": 8.242955629834449e-06,
+      "loss": 0.6224,
+      "step": 1484
+    },
+    {
+      "epoch": 0.34838709677419355,
+      "grad_norm": 1.1119203866797485,
+      "learning_rate": 8.239836963898849e-06,
+      "loss": 0.623,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3486217008797654,
+      "grad_norm": 11.640160565456098,
+      "learning_rate": 8.236716123800048e-06,
+      "loss": 0.6668,
+      "step": 1486
+    },
+    {
+      "epoch": 0.34885630498533726,
+      "grad_norm": 1.6163206052113017,
+      "learning_rate": 8.233593111632354e-06,
+      "loss": 0.5945,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3490909090909091,
+      "grad_norm": 1.4197707509539244,
+      "learning_rate": 8.230467929491533e-06,
+      "loss": 0.6855,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3493255131964809,
+      "grad_norm": 1.5358864085512611,
+      "learning_rate": 8.227340579474808e-06,
+      "loss": 0.6359,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3495601173020528,
+      "grad_norm": 1.1372195244716454,
+      "learning_rate": 8.224211063680854e-06,
+      "loss": 0.6406,
+      "step": 1490
+    },
+    {
+      "epoch": 0.34979472140762463,
+      "grad_norm": 1.2161043134269158,
+      "learning_rate": 8.2210793842098e-06,
+      "loss": 0.6546,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3500293255131965,
+      "grad_norm": 2.4354306159481762,
+      "learning_rate": 8.217945543163231e-06,
+      "loss": 0.6596,
+      "step": 1492
+    },
+    {
+      "epoch": 0.35026392961876834,
+      "grad_norm": 1.1174853950270742,
+      "learning_rate": 8.214809542644173e-06,
+      "loss": 0.6939,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3504985337243402,
+      "grad_norm": 1.6374231300300062,
+      "learning_rate": 8.211671384757114e-06,
+      "loss": 0.6197,
+      "step": 1494
+    },
+    {
+      "epoch": 0.350733137829912,
+      "grad_norm": 2.7910351323346734,
+      "learning_rate": 8.20853107160798e-06,
+      "loss": 0.6833,
+      "step": 1495
+    },
+    {
+      "epoch": 0.35096774193548386,
+      "grad_norm": 3.969565659824886,
+      "learning_rate": 8.205388605304147e-06,
+      "loss": 0.6587,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3512023460410557,
+      "grad_norm": 1.4218637984942035,
+      "learning_rate": 8.202243987954435e-06,
+      "loss": 0.601,
+      "step": 1497
+    },
+    {
+      "epoch": 0.35143695014662757,
+      "grad_norm": 1.4265802208832241,
+      "learning_rate": 8.19909722166911e-06,
+      "loss": 0.6044,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3516715542521994,
+      "grad_norm": 1.3131746883241227,
+      "learning_rate": 8.195948308559878e-06,
+      "loss": 0.6793,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3519061583577713,
+      "grad_norm": 1.3038815650068798,
+      "learning_rate": 8.192797250739883e-06,
+      "loss": 0.6392,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3519061583577713,
+      "eval_loss": 0.6344643235206604,
+      "eval_runtime": 25.3359,
+      "eval_samples_per_second": 21.55,
+      "eval_steps_per_second": 0.197,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3521407624633431,
+      "grad_norm": 1.1798503529087203,
+      "learning_rate": 8.189644050323712e-06,
+      "loss": 0.657,
+      "step": 1501
+    },
+    {
+      "epoch": 0.35237536656891494,
+      "grad_norm": 2.438736384051315,
+      "learning_rate": 8.18648870942739e-06,
+      "loss": 0.676,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3526099706744868,
+      "grad_norm": 1.3476462061493253,
+      "learning_rate": 8.183331230168376e-06,
+      "loss": 0.6299,
+      "step": 1503
+    },
+    {
+      "epoch": 0.35284457478005865,
+      "grad_norm": 1.4625706667923106,
+      "learning_rate": 8.180171614665568e-06,
+      "loss": 0.7143,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3530791788856305,
+      "grad_norm": 3.1193412580407145,
+      "learning_rate": 8.177009865039293e-06,
+      "loss": 0.6125,
+      "step": 1505
+    },
+    {
+      "epoch": 0.35331378299120236,
+      "grad_norm": 2.8272425724286316,
+      "learning_rate": 8.173845983411309e-06,
+      "loss": 0.6497,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3535483870967742,
+      "grad_norm": 2.982948695040381,
+      "learning_rate": 8.170679971904813e-06,
+      "loss": 0.6559,
+      "step": 1507
+    },
+    {
+      "epoch": 0.353782991202346,
+      "grad_norm": 9.668237178820712,
+      "learning_rate": 8.167511832644423e-06,
+      "loss": 0.6642,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3540175953079179,
+      "grad_norm": 1.0092042078247905,
+      "learning_rate": 8.164341567756189e-06,
+      "loss": 0.6357,
+      "step": 1509
+    },
+    {
+      "epoch": 0.35425219941348973,
+      "grad_norm": 1.6381656479825193,
+      "learning_rate": 8.161169179367584e-06,
+      "loss": 0.6943,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3544868035190616,
+      "grad_norm": 1.4139566760854536,
+      "learning_rate": 8.157994669607511e-06,
+      "loss": 0.6653,
+      "step": 1511
+    },
+    {
+      "epoch": 0.35472140762463344,
+      "grad_norm": 1.54224733346737,
+      "learning_rate": 8.154818040606293e-06,
+      "loss": 0.6183,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3549560117302053,
+      "grad_norm": 1.1796702378019979,
+      "learning_rate": 8.151639294495678e-06,
+      "loss": 0.6607,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3551906158357771,
+      "grad_norm": 1.1342868420745096,
+      "learning_rate": 8.148458433408831e-06,
+      "loss": 0.5884,
+      "step": 1514
+    },
+    {
+      "epoch": 0.35542521994134896,
+      "grad_norm": 5.80310803368655,
+      "learning_rate": 8.145275459480337e-06,
+      "loss": 0.628,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3556598240469208,
+      "grad_norm": 2.0126247757965974,
+      "learning_rate": 8.142090374846202e-06,
+      "loss": 0.6126,
+      "step": 1516
+    },
+    {
+      "epoch": 0.35589442815249267,
+      "grad_norm": 1.0570035225393455,
+      "learning_rate": 8.138903181643843e-06,
+      "loss": 0.6033,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3561290322580645,
+      "grad_norm": 2.355084585207633,
+      "learning_rate": 8.135713882012102e-06,
+      "loss": 0.6349,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3563636363636364,
+      "grad_norm": 1.1347525348726937,
+      "learning_rate": 8.13252247809122e-06,
+      "loss": 0.6547,
+      "step": 1519
+    },
+    {
+      "epoch": 0.35659824046920824,
+      "grad_norm": 2.9768106128415597,
+      "learning_rate": 8.129328972022866e-06,
+      "loss": 0.6146,
+      "step": 1520
+    },
+    {
+      "epoch": 0.35683284457478004,
+      "grad_norm": 2.2894577730384627,
+      "learning_rate": 8.126133365950107e-06,
+      "loss": 0.5932,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3570674486803519,
+      "grad_norm": 0.8144632935696673,
+      "learning_rate": 8.122935662017427e-06,
+      "loss": 0.6158,
+      "step": 1522
+    },
+    {
+      "epoch": 0.35730205278592375,
+      "grad_norm": 1.5392163258347284,
+      "learning_rate": 8.11973586237071e-06,
+      "loss": 0.6243,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3575366568914956,
+      "grad_norm": 1.374231087459184,
+      "learning_rate": 8.116533969157256e-06,
+      "loss": 0.5796,
+      "step": 1524
+    },
+    {
+      "epoch": 0.35777126099706746,
+      "grad_norm": 1.0180417289839192,
+      "learning_rate": 8.113329984525765e-06,
+      "loss": 0.6423,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3580058651026393,
+      "grad_norm": 4.482669849828746,
+      "learning_rate": 8.110123910626338e-06,
+      "loss": 0.6423,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3582404692082111,
+      "grad_norm": 0.9880813861443274,
+      "learning_rate": 8.106915749610482e-06,
+      "loss": 0.6009,
+      "step": 1527
+    },
+    {
+      "epoch": 0.358475073313783,
+      "grad_norm": 0.9870708942427716,
+      "learning_rate": 8.103705503631103e-06,
+      "loss": 0.6134,
+      "step": 1528
+    },
+    {
+      "epoch": 0.35870967741935483,
+      "grad_norm": 1.014592280258817,
+      "learning_rate": 8.100493174842509e-06,
+      "loss": 0.6116,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3589442815249267,
+      "grad_norm": 4.905569634351141,
+      "learning_rate": 8.0972787654004e-06,
+      "loss": 0.6067,
+      "step": 1530
+    },
+    {
+      "epoch": 0.35917888563049855,
+      "grad_norm": 2.0453894867981592,
+      "learning_rate": 8.094062277461878e-06,
+      "loss": 0.568,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3594134897360704,
+      "grad_norm": 1.1057508767783957,
+      "learning_rate": 8.090843713185434e-06,
+      "loss": 0.617,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3596480938416422,
+      "grad_norm": 1.4426599564301568,
+      "learning_rate": 8.08762307473096e-06,
+      "loss": 0.6444,
+      "step": 1533
+    },
+    {
+      "epoch": 0.35988269794721406,
+      "grad_norm": 1.4925289383756144,
+      "learning_rate": 8.084400364259733e-06,
+      "loss": 0.6422,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3601173020527859,
+      "grad_norm": 1.288728080259212,
+      "learning_rate": 8.081175583934424e-06,
+      "loss": 0.6437,
+      "step": 1535
+    },
+    {
+      "epoch": 0.36035190615835777,
+      "grad_norm": 1.4617074887411599,
+      "learning_rate": 8.07794873591909e-06,
+      "loss": 0.6511,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3605865102639296,
+      "grad_norm": 1.6179956711475658,
+      "learning_rate": 8.07471982237918e-06,
+      "loss": 0.6262,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3608211143695015,
+      "grad_norm": 1.372859733146479,
+      "learning_rate": 8.071488845481528e-06,
+      "loss": 0.6267,
+      "step": 1538
+    },
+    {
+      "epoch": 0.36105571847507334,
+      "grad_norm": 2.163325222536646,
+      "learning_rate": 8.068255807394347e-06,
+      "loss": 0.5942,
+      "step": 1539
+    },
+    {
+      "epoch": 0.36129032258064514,
+      "grad_norm": 3.2258330889492237,
+      "learning_rate": 8.065020710287241e-06,
+      "loss": 0.6163,
+      "step": 1540
+    },
+    {
+      "epoch": 0.361524926686217,
+      "grad_norm": 0.8290572913019899,
+      "learning_rate": 8.061783556331194e-06,
+      "loss": 0.5965,
+      "step": 1541
+    },
+    {
+      "epoch": 0.36175953079178885,
+      "grad_norm": 2.5995923724601924,
+      "learning_rate": 8.058544347698564e-06,
+      "loss": 0.6257,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3619941348973607,
+      "grad_norm": 9.87904279997343,
+      "learning_rate": 8.055303086563095e-06,
+      "loss": 0.6675,
+      "step": 1543
+    },
+    {
+      "epoch": 0.36222873900293256,
+      "grad_norm": 1.6409661285489316,
+      "learning_rate": 8.052059775099907e-06,
+      "loss": 0.6791,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3624633431085044,
+      "grad_norm": 1.4620368708476654,
+      "learning_rate": 8.048814415485493e-06,
+      "loss": 0.6084,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3626979472140762,
+      "grad_norm": 1.2519698876941296,
+      "learning_rate": 8.045567009897723e-06,
+      "loss": 0.6483,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3629325513196481,
+      "grad_norm": 3.564124817180489,
+      "learning_rate": 8.042317560515839e-06,
+      "loss": 0.6583,
+      "step": 1547
+    },
+    {
+      "epoch": 0.36316715542521993,
+      "grad_norm": 3.4346833491965683,
+      "learning_rate": 8.039066069520455e-06,
+      "loss": 0.6741,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3634017595307918,
+      "grad_norm": 1.379115222500377,
+      "learning_rate": 8.035812539093557e-06,
+      "loss": 0.6442,
+      "step": 1549
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.9001860505314034,
+      "learning_rate": 8.032556971418495e-06,
+      "loss": 0.6556,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3638709677419355,
+      "grad_norm": 1.991770848217314,
+      "learning_rate": 8.029299368679988e-06,
+      "loss": 0.6216,
+      "step": 1551
+    },
+    {
+      "epoch": 0.36410557184750736,
+      "grad_norm": 2.623481501550087,
+      "learning_rate": 8.026039733064121e-06,
+      "loss": 0.5739,
+      "step": 1552
+    },
+    {
+      "epoch": 0.36434017595307916,
+      "grad_norm": 1.0861583032719995,
+      "learning_rate": 8.022778066758348e-06,
+      "loss": 0.6348,
+      "step": 1553
+    },
+    {
+      "epoch": 0.364574780058651,
+      "grad_norm": 1.1200382803600062,
+      "learning_rate": 8.019514371951476e-06,
+      "loss": 0.6584,
+      "step": 1554
+    },
+    {
+      "epoch": 0.36480938416422287,
+      "grad_norm": 1.4113035236318554,
+      "learning_rate": 8.016248650833682e-06,
+      "loss": 0.6396,
+      "step": 1555
+    },
+    {
+      "epoch": 0.36504398826979473,
+      "grad_norm": 0.9922164964676499,
+      "learning_rate": 8.012980905596498e-06,
+      "loss": 0.5767,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3652785923753666,
+      "grad_norm": 3.00133939894505,
+      "learning_rate": 8.009711138432816e-06,
+      "loss": 0.6543,
+      "step": 1557
+    },
+    {
+      "epoch": 0.36551319648093844,
+      "grad_norm": 2.011321682364616,
+      "learning_rate": 8.006439351536884e-06,
+      "loss": 0.6958,
+      "step": 1558
+    },
+    {
+      "epoch": 0.36574780058651024,
+      "grad_norm": 1.148729790715129,
+      "learning_rate": 8.003165547104306e-06,
+      "loss": 0.6776,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3659824046920821,
+      "grad_norm": 1.355040178949832,
+      "learning_rate": 7.99988972733204e-06,
+      "loss": 0.6119,
+      "step": 1560
+    },
+    {
+      "epoch": 0.36621700879765395,
+      "grad_norm": 1.5592633596308831,
+      "learning_rate": 7.996611894418395e-06,
+      "loss": 0.6365,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3664516129032258,
+      "grad_norm": 1.112529982670453,
+      "learning_rate": 7.993332050563035e-06,
+      "loss": 0.6412,
+      "step": 1562
+    },
+    {
+      "epoch": 0.36668621700879767,
+      "grad_norm": 2.148567458438155,
+      "learning_rate": 7.99005019796697e-06,
+      "loss": 0.6052,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3669208211143695,
+      "grad_norm": 2.4879674832876835,
+      "learning_rate": 7.986766338832557e-06,
+      "loss": 0.644,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3671554252199413,
+      "grad_norm": 1.1792942166911304,
+      "learning_rate": 7.9834804753635e-06,
+      "loss": 0.65,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3673900293255132,
+      "grad_norm": 1.9843919849905691,
+      "learning_rate": 7.980192609764853e-06,
+      "loss": 0.6528,
+      "step": 1566
+    },
+    {
+      "epoch": 0.36762463343108503,
+      "grad_norm": 1.095690126906237,
+      "learning_rate": 7.976902744243009e-06,
+      "loss": 0.6004,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3678592375366569,
+      "grad_norm": 2.2214188096716363,
+      "learning_rate": 7.973610881005702e-06,
+      "loss": 0.6527,
+      "step": 1568
+    },
+    {
+      "epoch": 0.36809384164222875,
+      "grad_norm": 1.1484385025704285,
+      "learning_rate": 7.97031702226201e-06,
+      "loss": 0.6095,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3683284457478006,
+      "grad_norm": 0.9056842858340077,
+      "learning_rate": 7.96702117022235e-06,
+      "loss": 0.6455,
+      "step": 1570
+    },
+    {
+      "epoch": 0.36856304985337246,
+      "grad_norm": 1.2334406714572204,
+      "learning_rate": 7.963723327098472e-06,
+      "loss": 0.588,
+      "step": 1571
+    },
+    {
+      "epoch": 0.36879765395894426,
+      "grad_norm": 2.2468236956839895,
+      "learning_rate": 7.960423495103467e-06,
+      "loss": 0.6049,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3690322580645161,
+      "grad_norm": 1.0812342497919107,
+      "learning_rate": 7.95712167645176e-06,
+      "loss": 0.6584,
+      "step": 1573
+    },
+    {
+      "epoch": 0.369266862170088,
+      "grad_norm": 1.4252390275384497,
+      "learning_rate": 7.953817873359107e-06,
+      "loss": 0.6417,
+      "step": 1574
+    },
+    {
+      "epoch": 0.36950146627565983,
+      "grad_norm": 2.5435387288312916,
+      "learning_rate": 7.950512088042599e-06,
+      "loss": 0.6015,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3697360703812317,
+      "grad_norm": 4.8083615188042375,
+      "learning_rate": 7.947204322720654e-06,
+      "loss": 0.6651,
+      "step": 1576
+    },
+    {
+      "epoch": 0.36997067448680354,
+      "grad_norm": 1.8753561698859949,
+      "learning_rate": 7.94389457961302e-06,
+      "loss": 0.6538,
+      "step": 1577
+    },
+    {
+      "epoch": 0.37020527859237534,
+      "grad_norm": 1.808277657054618,
+      "learning_rate": 7.940582860940771e-06,
+      "loss": 0.6536,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3704398826979472,
+      "grad_norm": 1.285649125514911,
+      "learning_rate": 7.937269168926311e-06,
+      "loss": 0.611,
+      "step": 1579
+    },
+    {
+      "epoch": 0.37067448680351905,
+      "grad_norm": 1.0337780987198115,
+      "learning_rate": 7.933953505793363e-06,
+      "loss": 0.6397,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3709090909090909,
+      "grad_norm": 1.4439969739824046,
+      "learning_rate": 7.930635873766976e-06,
+      "loss": 0.6172,
+      "step": 1581
+    },
+    {
+      "epoch": 0.37114369501466277,
+      "grad_norm": 3.425596650950812,
+      "learning_rate": 7.92731627507352e-06,
+      "loss": 0.6754,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3713782991202346,
+      "grad_norm": 1.1319720004327498,
+      "learning_rate": 7.923994711940682e-06,
+      "loss": 0.6371,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3716129032258065,
+      "grad_norm": 1.943792921836966,
+      "learning_rate": 7.92067118659747e-06,
+      "loss": 0.6757,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3718475073313783,
+      "grad_norm": 1.1546751688129961,
+      "learning_rate": 7.917345701274208e-06,
+      "loss": 0.6136,
+      "step": 1585
+    },
+    {
+      "epoch": 0.37208211143695014,
+      "grad_norm": 1.341873182140012,
+      "learning_rate": 7.914018258202536e-06,
+      "loss": 0.6147,
+      "step": 1586
+    },
+    {
+      "epoch": 0.372316715542522,
+      "grad_norm": 1.8473105605190465,
+      "learning_rate": 7.910688859615407e-06,
+      "loss": 0.6845,
+      "step": 1587
+    },
+    {
+      "epoch": 0.37255131964809385,
+      "grad_norm": 1.1531135465554283,
+      "learning_rate": 7.907357507747087e-06,
+      "loss": 0.6211,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3727859237536657,
+      "grad_norm": 1.4230491981622588,
+      "learning_rate": 7.904024204833151e-06,
+      "loss": 0.6462,
+      "step": 1589
+    },
+    {
+      "epoch": 0.37302052785923756,
+      "grad_norm": 1.0823600490098797,
+      "learning_rate": 7.900688953110484e-06,
+      "loss": 0.7231,
+      "step": 1590
+    },
+    {
+      "epoch": 0.37325513196480936,
+      "grad_norm": 2.3049618171224924,
+      "learning_rate": 7.897351754817283e-06,
+      "loss": 0.5903,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3734897360703812,
+      "grad_norm": 2.3037301234266225,
+      "learning_rate": 7.894012612193044e-06,
+      "loss": 0.6455,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3737243401759531,
+      "grad_norm": 1.4791819697214155,
+      "learning_rate": 7.890671527478575e-06,
+      "loss": 0.6001,
+      "step": 1593
+    },
+    {
+      "epoch": 0.37395894428152493,
+      "grad_norm": 7.0775849390416905,
+      "learning_rate": 7.887328502915978e-06,
+      "loss": 0.6076,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3741935483870968,
+      "grad_norm": 0.9013415258414034,
+      "learning_rate": 7.883983540748667e-06,
+      "loss": 0.6262,
+      "step": 1595
+    },
+    {
+      "epoch": 0.37442815249266864,
+      "grad_norm": 1.5643594979656377,
+      "learning_rate": 7.880636643221352e-06,
+      "loss": 0.5712,
+      "step": 1596
+    },
+    {
+      "epoch": 0.37466275659824044,
+      "grad_norm": 0.9350115266763178,
+      "learning_rate": 7.877287812580039e-06,
+      "loss": 0.6307,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3748973607038123,
+      "grad_norm": 1.4591698399746185,
+      "learning_rate": 7.873937051072037e-06,
+      "loss": 0.6154,
+      "step": 1598
+    },
+    {
+      "epoch": 0.37513196480938416,
+      "grad_norm": 4.371467307462957,
+      "learning_rate": 7.870584360945942e-06,
+      "loss": 0.6159,
+      "step": 1599
+    },
+    {
+      "epoch": 0.375366568914956,
+      "grad_norm": 0.9708133122154972,
+      "learning_rate": 7.867229744451656e-06,
+      "loss": 0.6283,
+      "step": 1600
+    },
+    {
+      "epoch": 0.37560117302052787,
+      "grad_norm": 6.681577787074119,
+      "learning_rate": 7.863873203840361e-06,
+      "loss": 0.5997,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3758357771260997,
+      "grad_norm": 1.1144910197496753,
+      "learning_rate": 7.860514741364544e-06,
+      "loss": 0.6308,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3760703812316716,
+      "grad_norm": 1.0968215471761782,
+      "learning_rate": 7.857154359277972e-06,
+      "loss": 0.65,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3763049853372434,
+      "grad_norm": 1.1814685587763383,
+      "learning_rate": 7.853792059835699e-06,
+      "loss": 0.6059,
+      "step": 1604
+    },
+    {
+      "epoch": 0.37653958944281524,
+      "grad_norm": 1.4699600186850428,
+      "learning_rate": 7.850427845294072e-06,
+      "loss": 0.6263,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3767741935483871,
+      "grad_norm": 2.9306878918291726,
+      "learning_rate": 7.847061717910721e-06,
+      "loss": 0.5915,
+      "step": 1606
+    },
+    {
+      "epoch": 0.37700879765395895,
+      "grad_norm": 2.4936041636247546,
+      "learning_rate": 7.84369367994456e-06,
+      "loss": 0.6232,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3772434017595308,
+      "grad_norm": 1.1124911836527542,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.6242,
+      "step": 1608
+    },
+    {
+      "epoch": 0.37747800586510266,
+      "grad_norm": 1.8757080299496558,
+      "learning_rate": 7.836951881305861e-06,
+      "loss": 0.6032,
+      "step": 1609
+    },
+    {
+      "epoch": 0.37771260997067446,
+      "grad_norm": 1.382837018463846,
+      "learning_rate": 7.833578125157557e-06,
+      "loss": 0.6367,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3779472140762463,
+      "grad_norm": 1.4112673324084646,
+      "learning_rate": 7.830202467474899e-06,
+      "loss": 0.598,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3781818181818182,
+      "grad_norm": 1.531733847545577,
+      "learning_rate": 7.826824910523198e-06,
+      "loss": 0.6557,
+      "step": 1612
+    },
+    {
+      "epoch": 0.37841642228739003,
+      "grad_norm": 0.9857359431378977,
+      "learning_rate": 7.823445456569036e-06,
+      "loss": 0.6731,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3786510263929619,
+      "grad_norm": 1.201691042717365,
+      "learning_rate": 7.820064107880269e-06,
+      "loss": 0.6205,
+      "step": 1614
+    },
+    {
+      "epoch": 0.37888563049853374,
+      "grad_norm": 1.2901505532607962,
+      "learning_rate": 7.816680866726026e-06,
+      "loss": 0.6209,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3791202346041056,
+      "grad_norm": 2.330451785881743,
+      "learning_rate": 7.813295735376703e-06,
+      "loss": 0.6262,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3793548387096774,
+      "grad_norm": 1.6877055527703941,
+      "learning_rate": 7.809908716103966e-06,
+      "loss": 0.5826,
+      "step": 1617
+    },
+    {
+      "epoch": 0.37958944281524926,
+      "grad_norm": 1.138528263641915,
+      "learning_rate": 7.80651981118075e-06,
+      "loss": 0.6413,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3798240469208211,
+      "grad_norm": 1.2468604622779627,
+      "learning_rate": 7.803129022881253e-06,
+      "loss": 0.6271,
+      "step": 1619
+    },
+    {
+      "epoch": 0.38005865102639297,
+      "grad_norm": 5.764192010514738,
+      "learning_rate": 7.799736353480936e-06,
+      "loss": 0.6412,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3802932551319648,
+      "grad_norm": 2.198130873342226,
+      "learning_rate": 7.796341805256526e-06,
+      "loss": 0.647,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3805278592375367,
+      "grad_norm": 1.3737147381989245,
+      "learning_rate": 7.792945380486005e-06,
+      "loss": 0.6073,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3807624633431085,
+      "grad_norm": 1.3465757089659696,
+      "learning_rate": 7.789547081448622e-06,
+      "loss": 0.5948,
+      "step": 1623
+    },
+    {
+      "epoch": 0.38099706744868034,
+      "grad_norm": 2.4384398498095523,
+      "learning_rate": 7.786146910424877e-06,
+      "loss": 0.6782,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3812316715542522,
+      "grad_norm": 1.0005342089499756,
+      "learning_rate": 7.78274486969653e-06,
+      "loss": 0.6777,
+      "step": 1625
+    },
+    {
+      "epoch": 0.38146627565982405,
+      "grad_norm": 1.0195375728278713,
+      "learning_rate": 7.779340961546594e-06,
+      "loss": 0.6448,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3817008797653959,
+      "grad_norm": 4.2991384194269715,
+      "learning_rate": 7.775935188259336e-06,
+      "loss": 0.6258,
+      "step": 1627
+    },
+    {
+      "epoch": 0.38193548387096776,
+      "grad_norm": 3.1122548758651725,
+      "learning_rate": 7.772527552120274e-06,
+      "loss": 0.5896,
+      "step": 1628
+    },
+    {
+      "epoch": 0.38217008797653956,
+      "grad_norm": 1.1934580454028738,
+      "learning_rate": 7.769118055416176e-06,
+      "loss": 0.6182,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3824046920821114,
+      "grad_norm": 1.1892450568789514,
+      "learning_rate": 7.76570670043506e-06,
+      "loss": 0.6554,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3826392961876833,
+      "grad_norm": 1.7316044437097646,
+      "learning_rate": 7.76229348946619e-06,
+      "loss": 0.5997,
+      "step": 1631
+    },
+    {
+      "epoch": 0.38287390029325513,
+      "grad_norm": 4.740697767650133,
+      "learning_rate": 7.758878424800079e-06,
+      "loss": 0.656,
+      "step": 1632
+    },
+    {
+      "epoch": 0.383108504398827,
+      "grad_norm": 2.197652151884163,
+      "learning_rate": 7.755461508728473e-06,
+      "loss": 0.6249,
+      "step": 1633
+    },
+    {
+      "epoch": 0.38334310850439884,
+      "grad_norm": 2.0028422669624204,
+      "learning_rate": 7.752042743544372e-06,
+      "loss": 0.6537,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3835777126099707,
+      "grad_norm": 1.4258853071807955,
+      "learning_rate": 7.748622131542015e-06,
+      "loss": 0.6232,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3838123167155425,
+      "grad_norm": 1.1532078913814092,
+      "learning_rate": 7.745199675016877e-06,
+      "loss": 0.6394,
+      "step": 1636
+    },
+    {
+      "epoch": 0.38404692082111436,
+      "grad_norm": 1.5845757651003243,
+      "learning_rate": 7.741775376265668e-06,
+      "loss": 0.6369,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3842815249266862,
+      "grad_norm": 1.6104569009493135,
+      "learning_rate": 7.738349237586344e-06,
+      "loss": 0.6562,
+      "step": 1638
+    },
+    {
+      "epoch": 0.38451612903225807,
+      "grad_norm": 1.9615611064832639,
+      "learning_rate": 7.734921261278087e-06,
+      "loss": 0.6212,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3847507331378299,
+      "grad_norm": 0.9429567174903976,
+      "learning_rate": 7.731491449641315e-06,
+      "loss": 0.6223,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3849853372434018,
+      "grad_norm": 1.8340961559064877,
+      "learning_rate": 7.72805980497768e-06,
+      "loss": 0.6296,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3852199413489736,
+      "grad_norm": 1.1021759982037063,
+      "learning_rate": 7.724626329590059e-06,
+      "loss": 0.5946,
+      "step": 1642
+    },
+    {
+      "epoch": 0.38545454545454544,
+      "grad_norm": 1.4884445571627638,
+      "learning_rate": 7.721191025782563e-06,
+      "loss": 0.6015,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3856891495601173,
+      "grad_norm": 3.226927547681297,
+      "learning_rate": 7.717753895860526e-06,
+      "loss": 0.6333,
+      "step": 1644
+    },
+    {
+      "epoch": 0.38592375366568915,
+      "grad_norm": 2.472600552120965,
+      "learning_rate": 7.71431494213051e-06,
+      "loss": 0.6093,
+      "step": 1645
+    },
+    {
+      "epoch": 0.386158357771261,
+      "grad_norm": 10.70624866986832,
+      "learning_rate": 7.7108741669003e-06,
+      "loss": 0.6136,
+      "step": 1646
+    },
+    {
+      "epoch": 0.38639296187683286,
+      "grad_norm": 1.2892356614800278,
+      "learning_rate": 7.707431572478901e-06,
+      "loss": 0.6365,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3866275659824047,
+      "grad_norm": 6.168935108805046,
+      "learning_rate": 7.703987161176545e-06,
+      "loss": 0.6352,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3868621700879765,
+      "grad_norm": 3.8592196027463257,
+      "learning_rate": 7.700540935304675e-06,
+      "loss": 0.6521,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.9715944124891208,
+      "learning_rate": 7.697092897175957e-06,
+      "loss": 0.6265,
+      "step": 1650
+    },
+    {
+      "epoch": 0.38733137829912023,
+      "grad_norm": 2.5991287683663247,
+      "learning_rate": 7.693643049104273e-06,
+      "loss": 0.5708,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3875659824046921,
+      "grad_norm": 1.9448631815409265,
+      "learning_rate": 7.69019139340472e-06,
+      "loss": 0.6598,
+      "step": 1652
+    },
+    {
+      "epoch": 0.38780058651026394,
+      "grad_norm": 1.6160675968543425,
+      "learning_rate": 7.686737932393606e-06,
+      "loss": 0.7182,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3880351906158358,
+      "grad_norm": 2.391412577308862,
+      "learning_rate": 7.68328266838845e-06,
+      "loss": 0.5988,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3882697947214076,
+      "grad_norm": 3.369471487906246,
+      "learning_rate": 7.679825603707982e-06,
+      "loss": 0.6186,
+      "step": 1655
+    },
+    {
+      "epoch": 0.38850439882697946,
+      "grad_norm": 1.7192333949644023,
+      "learning_rate": 7.676366740672142e-06,
+      "loss": 0.6218,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3887390029325513,
+      "grad_norm": 2.3388067836375774,
+      "learning_rate": 7.672906081602074e-06,
+      "loss": 0.709,
+      "step": 1657
+    },
+    {
+      "epoch": 0.38897360703812317,
+      "grad_norm": 1.4123309427471646,
+      "learning_rate": 7.66944362882013e-06,
+      "loss": 0.6599,
+      "step": 1658
+    },
+    {
+      "epoch": 0.389208211143695,
+      "grad_norm": 1.213105225305302,
+      "learning_rate": 7.665979384649864e-06,
+      "loss": 0.6336,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3894428152492669,
+      "grad_norm": 1.873253318955163,
+      "learning_rate": 7.662513351416034e-06,
+      "loss": 0.6167,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3896774193548387,
+      "grad_norm": 1.4229359040353953,
+      "learning_rate": 7.659045531444596e-06,
+      "loss": 0.6538,
+      "step": 1661
+    },
+    {
+      "epoch": 0.38991202346041054,
+      "grad_norm": 3.2815731166981763,
+      "learning_rate": 7.655575927062704e-06,
+      "loss": 0.6587,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3901466275659824,
+      "grad_norm": 1.5959250089933368,
+      "learning_rate": 7.652104540598712e-06,
+      "loss": 0.6943,
+      "step": 1663
+    },
+    {
+      "epoch": 0.39038123167155425,
+      "grad_norm": 1.4753200296899938,
+      "learning_rate": 7.648631374382174e-06,
+      "loss": 0.6219,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3906158357771261,
+      "grad_norm": 1.1684194218999984,
+      "learning_rate": 7.64515643074383e-06,
+      "loss": 0.6596,
+      "step": 1665
+    },
+    {
+      "epoch": 0.39085043988269796,
+      "grad_norm": 3.957667986106087,
+      "learning_rate": 7.641679712015617e-06,
+      "loss": 0.6477,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3910850439882698,
+      "grad_norm": 1.4101555559607137,
+      "learning_rate": 7.638201220530664e-06,
+      "loss": 0.6176,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3913196480938416,
+      "grad_norm": 2.0687281157254707,
+      "learning_rate": 7.634720958623287e-06,
+      "loss": 0.6814,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3915542521994135,
+      "grad_norm": 1.7861152334922592,
+      "learning_rate": 7.631238928628994e-06,
+      "loss": 0.6347,
+      "step": 1669
+    },
+    {
+      "epoch": 0.39178885630498533,
+      "grad_norm": 0.8241628931940701,
+      "learning_rate": 7.6277551328844755e-06,
+      "loss": 0.562,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3920234604105572,
+      "grad_norm": 6.179110968857805,
+      "learning_rate": 7.6242695737276086e-06,
+      "loss": 0.6288,
+      "step": 1671
+    },
+    {
+      "epoch": 0.39225806451612905,
+      "grad_norm": 1.6377758236012427,
+      "learning_rate": 7.620782253497453e-06,
+      "loss": 0.6214,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3924926686217009,
+      "grad_norm": 1.4804539440002633,
+      "learning_rate": 7.617293174534252e-06,
+      "loss": 0.6045,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3927272727272727,
+      "grad_norm": 1.1820224477276267,
+      "learning_rate": 7.613802339179428e-06,
+      "loss": 0.6143,
+      "step": 1674
+    },
+    {
+      "epoch": 0.39296187683284456,
+      "grad_norm": 1.6663101843072536,
+      "learning_rate": 7.610309749775583e-06,
+      "loss": 0.6225,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3931964809384164,
+      "grad_norm": 1.2223395736441116,
+      "learning_rate": 7.606815408666493e-06,
+      "loss": 0.5776,
+      "step": 1676
+    },
+    {
+      "epoch": 0.39343108504398827,
+      "grad_norm": 6.289964741746519,
+      "learning_rate": 7.6033193181971136e-06,
+      "loss": 0.7023,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3936656891495601,
+      "grad_norm": 1.0330790797924478,
+      "learning_rate": 7.599821480713571e-06,
+      "loss": 0.6286,
+      "step": 1678
+    },
+    {
+      "epoch": 0.393900293255132,
+      "grad_norm": 1.89908810726443,
+      "learning_rate": 7.596321898563167e-06,
+      "loss": 0.6415,
+      "step": 1679
+    },
+    {
+      "epoch": 0.39413489736070384,
+      "grad_norm": 4.35097376095719,
+      "learning_rate": 7.59282057409437e-06,
+      "loss": 0.5727,
+      "step": 1680
+    },
+    {
+      "epoch": 0.39436950146627564,
+      "grad_norm": 2.2722518194853736,
+      "learning_rate": 7.589317509656824e-06,
+      "loss": 0.6787,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3946041055718475,
+      "grad_norm": 1.0161605344772082,
+      "learning_rate": 7.585812707601333e-06,
+      "loss": 0.6062,
+      "step": 1682
+    },
+    {
+      "epoch": 0.39483870967741935,
+      "grad_norm": 1.1379485114464876,
+      "learning_rate": 7.582306170279873e-06,
+      "loss": 0.6193,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3950733137829912,
+      "grad_norm": 5.590600647341276,
+      "learning_rate": 7.57879790004558e-06,
+      "loss": 0.6291,
+      "step": 1684
+    },
+    {
+      "epoch": 0.39530791788856307,
+      "grad_norm": 1.2093191931980196,
+      "learning_rate": 7.575287899252759e-06,
+      "loss": 0.6698,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3955425219941349,
+      "grad_norm": 0.7656633162018557,
+      "learning_rate": 7.571776170256869e-06,
+      "loss": 0.5728,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3957771260997067,
+      "grad_norm": 1.3366721896457685,
+      "learning_rate": 7.568262715414534e-06,
+      "loss": 0.6556,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3960117302052786,
+      "grad_norm": 1.8218187634970278,
+      "learning_rate": 7.564747537083533e-06,
+      "loss": 0.6241,
+      "step": 1688
+    },
+    {
+      "epoch": 0.39624633431085043,
+      "grad_norm": 1.2066453474454495,
+      "learning_rate": 7.561230637622805e-06,
+      "loss": 0.6306,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3964809384164223,
+      "grad_norm": 1.7231063124864543,
+      "learning_rate": 7.5577120193924415e-06,
+      "loss": 0.6729,
+      "step": 1690
+    },
+    {
+      "epoch": 0.39671554252199415,
+      "grad_norm": 0.8481410402095703,
+      "learning_rate": 7.554191684753687e-06,
+      "loss": 0.7079,
+      "step": 1691
+    },
+    {
+      "epoch": 0.396950146627566,
+      "grad_norm": 1.1626186427349046,
+      "learning_rate": 7.550669636068939e-06,
+      "loss": 0.6614,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3971847507331378,
+      "grad_norm": 1.0690244274077698,
+      "learning_rate": 7.547145875701744e-06,
+      "loss": 0.6198,
+      "step": 1693
+    },
+    {
+      "epoch": 0.39741935483870966,
+      "grad_norm": 1.0249960620447252,
+      "learning_rate": 7.543620406016801e-06,
+      "loss": 0.5707,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3976539589442815,
+      "grad_norm": 19.10349514928793,
+      "learning_rate": 7.540093229379951e-06,
+      "loss": 0.656,
+      "step": 1695
+    },
+    {
+      "epoch": 0.39788856304985337,
+      "grad_norm": 2.6427431748274235,
+      "learning_rate": 7.536564348158184e-06,
+      "loss": 0.676,
+      "step": 1696
+    },
+    {
+      "epoch": 0.39812316715542523,
+      "grad_norm": 1.0658332220420659,
+      "learning_rate": 7.5330337647196305e-06,
+      "loss": 0.6751,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3983577712609971,
+      "grad_norm": 1.2692894945661328,
+      "learning_rate": 7.52950148143357e-06,
+      "loss": 0.6647,
+      "step": 1698
+    },
+    {
+      "epoch": 0.39859237536656894,
+      "grad_norm": 1.66387099040758,
+      "learning_rate": 7.525967500670413e-06,
+      "loss": 0.6346,
+      "step": 1699
+    },
+    {
+      "epoch": 0.39882697947214074,
+      "grad_norm": 1.3709588863996383,
+      "learning_rate": 7.522431824801717e-06,
+      "loss": 0.6329,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3990615835777126,
+      "grad_norm": 3.474680034045354,
+      "learning_rate": 7.5188944562001746e-06,
+      "loss": 0.5926,
+      "step": 1701
+    },
+    {
+      "epoch": 0.39929618768328445,
+      "grad_norm": 1.1517654120841698,
+      "learning_rate": 7.515355397239613e-06,
+      "loss": 0.5686,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3995307917888563,
+      "grad_norm": 5.005503902366467,
+      "learning_rate": 7.511814650294994e-06,
+      "loss": 0.5656,
+      "step": 1703
+    },
+    {
+      "epoch": 0.39976539589442817,
+      "grad_norm": 2.7049150211166135,
+      "learning_rate": 7.508272217742416e-06,
+      "loss": 0.6121,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9908828597574337,
+      "learning_rate": 7.504728101959103e-06,
+      "loss": 0.6302,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4002346041055718,
+      "grad_norm": 0.919251116851586,
+      "learning_rate": 7.5011823053234115e-06,
+      "loss": 0.6456,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4004692082111437,
+      "grad_norm": 1.4389194955101958,
+      "learning_rate": 7.497634830214826e-06,
+      "loss": 0.5995,
+      "step": 1707
+    },
+    {
+      "epoch": 0.40070381231671554,
+      "grad_norm": 5.164560163062019,
+      "learning_rate": 7.494085679013959e-06,
+      "loss": 0.5942,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4009384164222874,
+      "grad_norm": 1.1325687584621202,
+      "learning_rate": 7.490534854102543e-06,
+      "loss": 0.6018,
+      "step": 1709
+    },
+    {
+      "epoch": 0.40117302052785925,
+      "grad_norm": 0.7988283948081473,
+      "learning_rate": 7.486982357863435e-06,
+      "loss": 0.6365,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4014076246334311,
+      "grad_norm": 1.75834369524383,
+      "learning_rate": 7.48342819268062e-06,
+      "loss": 0.6335,
+      "step": 1711
+    },
+    {
+      "epoch": 0.40164222873900296,
+      "grad_norm": 1.1967795275195026,
+      "learning_rate": 7.479872360939193e-06,
+      "loss": 0.5808,
+      "step": 1712
+    },
+    {
+      "epoch": 0.40187683284457476,
+      "grad_norm": 1.8206629863153243,
+      "learning_rate": 7.476314865025376e-06,
+      "loss": 0.6394,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4021114369501466,
+      "grad_norm": 1.4687031946397808,
+      "learning_rate": 7.472755707326503e-06,
+      "loss": 0.5859,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4023460410557185,
+      "grad_norm": 1.3555501632175697,
+      "learning_rate": 7.469194890231022e-06,
+      "loss": 0.6619,
+      "step": 1715
+    },
+    {
+      "epoch": 0.40258064516129033,
+      "grad_norm": 0.887119286137143,
+      "learning_rate": 7.465632416128501e-06,
+      "loss": 0.6456,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4028152492668622,
+      "grad_norm": 1.309577536114152,
+      "learning_rate": 7.462068287409614e-06,
+      "loss": 0.6242,
+      "step": 1717
+    },
+    {
+      "epoch": 0.40304985337243404,
+      "grad_norm": 1.914591925810589,
+      "learning_rate": 7.458502506466146e-06,
+      "loss": 0.6502,
+      "step": 1718
+    },
+    {
+      "epoch": 0.40328445747800584,
+      "grad_norm": 1.277094433107499,
+      "learning_rate": 7.454935075690994e-06,
+      "loss": 0.6106,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4035190615835777,
+      "grad_norm": 22.012098844809557,
+      "learning_rate": 7.451365997478159e-06,
+      "loss": 0.5783,
+      "step": 1720
+    },
+    {
+      "epoch": 0.40375366568914955,
+      "grad_norm": 1.7601628159246012,
+      "learning_rate": 7.44779527422275e-06,
+      "loss": 0.6083,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4039882697947214,
+      "grad_norm": 8.431429472241842,
+      "learning_rate": 7.444222908320976e-06,
+      "loss": 0.6239,
+      "step": 1722
+    },
+    {
+      "epoch": 0.40422287390029327,
+      "grad_norm": 2.78628788759552,
+      "learning_rate": 7.440648902170153e-06,
+      "loss": 0.6533,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4044574780058651,
+      "grad_norm": 1.2379131876459923,
+      "learning_rate": 7.4370732581686965e-06,
+      "loss": 0.6407,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4046920821114369,
+      "grad_norm": 1.6940358989013258,
+      "learning_rate": 7.433495978716119e-06,
+      "loss": 0.625,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4049266862170088,
+      "grad_norm": 78.39486316367483,
+      "learning_rate": 7.42991706621303e-06,
+      "loss": 0.6083,
+      "step": 1726
+    },
+    {
+      "epoch": 0.40516129032258064,
+      "grad_norm": 2.4661684430165214,
+      "learning_rate": 7.426336523061141e-06,
+      "loss": 0.7001,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4053958944281525,
+      "grad_norm": 3.1100131190120863,
+      "learning_rate": 7.422754351663252e-06,
+      "loss": 0.6638,
+      "step": 1728
+    },
+    {
+      "epoch": 0.40563049853372435,
+      "grad_norm": 2.2924349765517467,
+      "learning_rate": 7.419170554423255e-06,
+      "loss": 0.6654,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4058651026392962,
+      "grad_norm": 2.6430893925925685,
+      "learning_rate": 7.415585133746137e-06,
+      "loss": 0.5866,
+      "step": 1730
+    },
+    {
+      "epoch": 0.40609970674486806,
+      "grad_norm": 1.14584465992698,
+      "learning_rate": 7.411998092037973e-06,
+      "loss": 0.6111,
+      "step": 1731
+    },
+    {
+      "epoch": 0.40633431085043986,
+      "grad_norm": 1.631849211332541,
+      "learning_rate": 7.408409431705924e-06,
+      "loss": 0.5961,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4065689149560117,
+      "grad_norm": 1.2145397669954725,
+      "learning_rate": 7.40481915515824e-06,
+      "loss": 0.6909,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4068035190615836,
+      "grad_norm": 1.5507848021981117,
+      "learning_rate": 7.401227264804254e-06,
+      "loss": 0.6403,
+      "step": 1734
+    },
+    {
+      "epoch": 0.40703812316715543,
+      "grad_norm": 2.0558491775949235,
+      "learning_rate": 7.397633763054384e-06,
+      "loss": 0.6264,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4072727272727273,
+      "grad_norm": 2.551712659314444,
+      "learning_rate": 7.3940386523201224e-06,
+      "loss": 0.6564,
+      "step": 1736
+    },
+    {
+      "epoch": 0.40750733137829914,
+      "grad_norm": 1.5990076488320568,
+      "learning_rate": 7.390441935014052e-06,
+      "loss": 0.6421,
+      "step": 1737
+    },
+    {
+      "epoch": 0.40774193548387094,
+      "grad_norm": 1.6883610257539423,
+      "learning_rate": 7.386843613549828e-06,
+      "loss": 0.599,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4079765395894428,
+      "grad_norm": 3.512645700209129,
+      "learning_rate": 7.383243690342179e-06,
+      "loss": 0.6161,
+      "step": 1739
+    },
+    {
+      "epoch": 0.40821114369501466,
+      "grad_norm": 1.10750732274883,
+      "learning_rate": 7.379642167806913e-06,
+      "loss": 0.5812,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4084457478005865,
+      "grad_norm": 3.480727119118568,
+      "learning_rate": 7.3760390483609135e-06,
+      "loss": 0.6216,
+      "step": 1741
+    },
+    {
+      "epoch": 0.40868035190615837,
+      "grad_norm": 1.276969992900176,
+      "learning_rate": 7.372434334422128e-06,
+      "loss": 0.6176,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4089149560117302,
+      "grad_norm": 1.2086539896123187,
+      "learning_rate": 7.368828028409581e-06,
+      "loss": 0.5926,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4091495601173021,
+      "grad_norm": 1.4121012202709036,
+      "learning_rate": 7.365220132743363e-06,
+      "loss": 0.6611,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4093841642228739,
+      "grad_norm": 1.195850026147483,
+      "learning_rate": 7.361610649844628e-06,
+      "loss": 0.6325,
+      "step": 1745
+    },
+    {
+      "epoch": 0.40961876832844574,
+      "grad_norm": 1.502712220490566,
+      "learning_rate": 7.357999582135601e-06,
+      "loss": 0.5779,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4098533724340176,
+      "grad_norm": 1.6392808213341148,
+      "learning_rate": 7.354386932039566e-06,
+      "loss": 0.614,
+      "step": 1747
+    },
+    {
+      "epoch": 0.41008797653958945,
+      "grad_norm": 1.0440133915247312,
+      "learning_rate": 7.350772701980873e-06,
+      "loss": 0.6974,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4103225806451613,
+      "grad_norm": 0.86568552994266,
+      "learning_rate": 7.347156894384926e-06,
+      "loss": 0.6348,
+      "step": 1749
+    },
+    {
+      "epoch": 0.41055718475073316,
+      "grad_norm": 4.2340316305038,
+      "learning_rate": 7.343539511678195e-06,
+      "loss": 0.6565,
+      "step": 1750
+    },
+    {
+      "epoch": 0.41079178885630496,
+      "grad_norm": 5.570019801840745,
+      "learning_rate": 7.339920556288202e-06,
+      "loss": 0.6476,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4110263929618768,
+      "grad_norm": 0.9975936348153364,
+      "learning_rate": 7.336300030643526e-06,
+      "loss": 0.6302,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4112609970674487,
+      "grad_norm": 1.1040732239633317,
+      "learning_rate": 7.3326779371738e-06,
+      "loss": 0.6035,
+      "step": 1753
+    },
+    {
+      "epoch": 0.41149560117302053,
+      "grad_norm": 2.047008469580855,
+      "learning_rate": 7.3290542783097085e-06,
+      "loss": 0.5845,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4117302052785924,
+      "grad_norm": 1.161806113312015,
+      "learning_rate": 7.325429056482986e-06,
+      "loss": 0.7006,
+      "step": 1755
+    },
+    {
+      "epoch": 0.41196480938416424,
+      "grad_norm": 1.353184403882004,
+      "learning_rate": 7.321802274126418e-06,
+      "loss": 0.6609,
+      "step": 1756
+    },
+    {
+      "epoch": 0.41219941348973604,
+      "grad_norm": 9.1821190062482,
+      "learning_rate": 7.318173933673836e-06,
+      "loss": 0.6799,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4124340175953079,
+      "grad_norm": 1.057692012937856,
+      "learning_rate": 7.314544037560115e-06,
+      "loss": 0.6297,
+      "step": 1758
+    },
+    {
+      "epoch": 0.41266862170087976,
+      "grad_norm": 0.9493457381189597,
+      "learning_rate": 7.3109125882211765e-06,
+      "loss": 0.5545,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4129032258064516,
+      "grad_norm": 0.8696548362478198,
+      "learning_rate": 7.3072795880939855e-06,
+      "loss": 0.6757,
+      "step": 1760
+    },
+    {
+      "epoch": 0.41313782991202347,
+      "grad_norm": 1.800753236041734,
+      "learning_rate": 7.303645039616544e-06,
+      "loss": 0.6521,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4133724340175953,
+      "grad_norm": 1.0588304341242767,
+      "learning_rate": 7.300008945227895e-06,
+      "loss": 0.5957,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4136070381231672,
+      "grad_norm": 2.683854835153098,
+      "learning_rate": 7.29637130736812e-06,
+      "loss": 0.6379,
+      "step": 1763
+    },
+    {
+      "epoch": 0.413841642228739,
+      "grad_norm": 1.5219693335922593,
+      "learning_rate": 7.292732128478336e-06,
+      "loss": 0.6232,
+      "step": 1764
+    },
+    {
+      "epoch": 0.41407624633431084,
+      "grad_norm": 1.1486892255099441,
+      "learning_rate": 7.289091411000691e-06,
+      "loss": 0.6536,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4143108504398827,
+      "grad_norm": 1.0375978962914074,
+      "learning_rate": 7.285449157378367e-06,
+      "loss": 0.6259,
+      "step": 1766
+    },
+    {
+      "epoch": 0.41454545454545455,
+      "grad_norm": 0.9513297266866592,
+      "learning_rate": 7.2818053700555816e-06,
+      "loss": 0.6523,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4147800586510264,
+      "grad_norm": 1.036511892569362,
+      "learning_rate": 7.278160051477574e-06,
+      "loss": 0.6158,
+      "step": 1768
+    },
+    {
+      "epoch": 0.41501466275659826,
+      "grad_norm": 1.6331785566147172,
+      "learning_rate": 7.274513204090618e-06,
+      "loss": 0.6815,
+      "step": 1769
+    },
+    {
+      "epoch": 0.41524926686217006,
+      "grad_norm": 1.424146712111036,
+      "learning_rate": 7.270864830342006e-06,
+      "loss": 0.6556,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4154838709677419,
+      "grad_norm": 2.2628392752024844,
+      "learning_rate": 7.267214932680062e-06,
+      "loss": 0.5884,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4157184750733138,
+      "grad_norm": 2.5412924077674792,
+      "learning_rate": 7.263563513554127e-06,
+      "loss": 0.613,
+      "step": 1772
+    },
+    {
+      "epoch": 0.41595307917888563,
+      "grad_norm": 7.339083716560217,
+      "learning_rate": 7.259910575414569e-06,
+      "loss": 0.6219,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4161876832844575,
+      "grad_norm": 0.8028217769264021,
+      "learning_rate": 7.256256120712768e-06,
+      "loss": 0.6696,
+      "step": 1774
+    },
+    {
+      "epoch": 0.41642228739002934,
+      "grad_norm": 1.1624684075460954,
+      "learning_rate": 7.252600151901127e-06,
+      "loss": 0.6268,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4166568914956012,
+      "grad_norm": 1.7903872594102144,
+      "learning_rate": 7.248942671433064e-06,
+      "loss": 0.6555,
+      "step": 1776
+    },
+    {
+      "epoch": 0.416891495601173,
+      "grad_norm": 1.1582237563805369,
+      "learning_rate": 7.24528368176301e-06,
+      "loss": 0.6252,
+      "step": 1777
+    },
+    {
+      "epoch": 0.41712609970674486,
+      "grad_norm": 1.0337615573317804,
+      "learning_rate": 7.241623185346409e-06,
+      "loss": 0.5977,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4173607038123167,
+      "grad_norm": 1.7127157102460435,
+      "learning_rate": 7.237961184639719e-06,
+      "loss": 0.6615,
+      "step": 1779
+    },
+    {
+      "epoch": 0.41759530791788857,
+      "grad_norm": 3.469324251654291,
+      "learning_rate": 7.234297682100404e-06,
+      "loss": 0.631,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4178299120234604,
+      "grad_norm": 0.9311411157448523,
+      "learning_rate": 7.23063268018694e-06,
+      "loss": 0.6595,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4180645161290323,
+      "grad_norm": 1.0342309935883347,
+      "learning_rate": 7.2269661813588e-06,
+      "loss": 0.615,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4182991202346041,
+      "grad_norm": 3.127875807541969,
+      "learning_rate": 7.223298188076475e-06,
+      "loss": 0.6298,
+      "step": 1783
+    },
+    {
+      "epoch": 0.41853372434017594,
+      "grad_norm": 1.2357145994413388,
+      "learning_rate": 7.219628702801448e-06,
+      "loss": 0.6833,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4187683284457478,
+      "grad_norm": 4.844067072728044,
+      "learning_rate": 7.215957727996208e-06,
+      "loss": 0.5995,
+      "step": 1785
+    },
+    {
+      "epoch": 0.41900293255131965,
+      "grad_norm": 3.37805068344963,
+      "learning_rate": 7.212285266124241e-06,
+      "loss": 0.5597,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4192375366568915,
+      "grad_norm": 1.9953706043378174,
+      "learning_rate": 7.208611319650036e-06,
+      "loss": 0.6702,
+      "step": 1787
+    },
+    {
+      "epoch": 0.41947214076246336,
+      "grad_norm": 1.108643586188879,
+      "learning_rate": 7.204935891039071e-06,
+      "loss": 0.6511,
+      "step": 1788
+    },
+    {
+      "epoch": 0.41970674486803516,
+      "grad_norm": 1.3078284617761626,
+      "learning_rate": 7.201258982757827e-06,
+      "loss": 0.6233,
+      "step": 1789
+    },
+    {
+      "epoch": 0.419941348973607,
+      "grad_norm": 1.2135587701586805,
+      "learning_rate": 7.197580597273769e-06,
+      "loss": 0.6245,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4201759530791789,
+      "grad_norm": 1.6408345206379327,
+      "learning_rate": 7.1939007370553615e-06,
+      "loss": 0.6422,
+      "step": 1791
+    },
+    {
+      "epoch": 0.42041055718475073,
+      "grad_norm": 2.0834935401659784,
+      "learning_rate": 7.1902194045720495e-06,
+      "loss": 0.6616,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4206451612903226,
+      "grad_norm": 2.0433575326148974,
+      "learning_rate": 7.186536602294278e-06,
+      "loss": 0.6567,
+      "step": 1793
+    },
+    {
+      "epoch": 0.42087976539589445,
+      "grad_norm": 5.513189436713361,
+      "learning_rate": 7.182852332693467e-06,
+      "loss": 0.6242,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4211143695014663,
+      "grad_norm": 1.061612820536488,
+      "learning_rate": 7.179166598242028e-06,
+      "loss": 0.6371,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4213489736070381,
+      "grad_norm": 0.832673904971619,
+      "learning_rate": 7.175479401413353e-06,
+      "loss": 0.6085,
+      "step": 1796
+    },
+    {
+      "epoch": 0.42158357771260996,
+      "grad_norm": 4.1329770556502154,
+      "learning_rate": 7.171790744681816e-06,
+      "loss": 0.679,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4218181818181818,
+      "grad_norm": 1.5563224417692414,
+      "learning_rate": 7.168100630522769e-06,
+      "loss": 0.6288,
+      "step": 1798
+    },
+    {
+      "epoch": 0.42205278592375367,
+      "grad_norm": 1.5160883197367536,
+      "learning_rate": 7.164409061412545e-06,
+      "loss": 0.6729,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4222873900293255,
+      "grad_norm": 0.9237689283235428,
+      "learning_rate": 7.1607160398284534e-06,
+      "loss": 0.6053,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4225219941348974,
+      "grad_norm": 1.5504678306160298,
+      "learning_rate": 7.1570215682487724e-06,
+      "loss": 0.6143,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4227565982404692,
+      "grad_norm": 1.9120918187104634,
+      "learning_rate": 7.153325649152762e-06,
+      "loss": 0.5863,
+      "step": 1802
+    },
+    {
+      "epoch": 0.42299120234604104,
+      "grad_norm": 1.062769322777713,
+      "learning_rate": 7.149628285020647e-06,
+      "loss": 0.6312,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4232258064516129,
+      "grad_norm": 1.5545749260716493,
+      "learning_rate": 7.145929478333626e-06,
+      "loss": 0.662,
+      "step": 1804
+    },
+    {
+      "epoch": 0.42346041055718475,
+      "grad_norm": 1.3440830970119046,
+      "learning_rate": 7.14222923157386e-06,
+      "loss": 0.6323,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4236950146627566,
+      "grad_norm": 1.167036885068974,
+      "learning_rate": 7.138527547224485e-06,
+      "loss": 0.6267,
+      "step": 1806
+    },
+    {
+      "epoch": 0.42392961876832846,
+      "grad_norm": 1.0240846072595606,
+      "learning_rate": 7.134824427769594e-06,
+      "loss": 0.6067,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4241642228739003,
+      "grad_norm": 9.220119696794047,
+      "learning_rate": 7.131119875694246e-06,
+      "loss": 0.7045,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4243988269794721,
+      "grad_norm": 3.2151891902847325,
+      "learning_rate": 7.127413893484463e-06,
+      "loss": 0.6602,
+      "step": 1809
+    },
+    {
+      "epoch": 0.424633431085044,
+      "grad_norm": 4.2408333696511376,
+      "learning_rate": 7.123706483627225e-06,
+      "loss": 0.6672,
+      "step": 1810
+    },
+    {
+      "epoch": 0.42486803519061583,
+      "grad_norm": 1.9624585179056808,
+      "learning_rate": 7.119997648610468e-06,
+      "loss": 0.5959,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4251026392961877,
+      "grad_norm": 4.19021221348873,
+      "learning_rate": 7.116287390923089e-06,
+      "loss": 0.6691,
+      "step": 1812
+    },
+    {
+      "epoch": 0.42533724340175955,
+      "grad_norm": 0.8446859960795003,
+      "learning_rate": 7.112575713054937e-06,
+      "loss": 0.5998,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4255718475073314,
+      "grad_norm": 1.9035935400381916,
+      "learning_rate": 7.108862617496812e-06,
+      "loss": 0.6158,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4258064516129032,
+      "grad_norm": 0.8992856659183355,
+      "learning_rate": 7.105148106740469e-06,
+      "loss": 0.6244,
+      "step": 1815
+    },
+    {
+      "epoch": 0.42604105571847506,
+      "grad_norm": 1.2487981931558807,
+      "learning_rate": 7.1014321832786135e-06,
+      "loss": 0.635,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4262756598240469,
+      "grad_norm": 1.7941136467653054,
+      "learning_rate": 7.097714849604894e-06,
+      "loss": 0.6618,
+      "step": 1817
+    },
+    {
+      "epoch": 0.42651026392961877,
+      "grad_norm": 2.338348486538766,
+      "learning_rate": 7.09399610821391e-06,
+      "loss": 0.6712,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4267448680351906,
+      "grad_norm": 1.1272183144060923,
+      "learning_rate": 7.090275961601203e-06,
+      "loss": 0.5519,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4269794721407625,
+      "grad_norm": 1.2745355777155003,
+      "learning_rate": 7.086554412263259e-06,
+      "loss": 0.6893,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4272140762463343,
+      "grad_norm": 3.4926617128552557,
+      "learning_rate": 7.082831462697505e-06,
+      "loss": 0.5682,
+      "step": 1821
+    },
+    {
+      "epoch": 0.42744868035190614,
+      "grad_norm": 4.027086586041726,
+      "learning_rate": 7.079107115402306e-06,
+      "loss": 0.6657,
+      "step": 1822
+    },
+    {
+      "epoch": 0.427683284457478,
+      "grad_norm": 1.3598308965958918,
+      "learning_rate": 7.075381372876969e-06,
+      "loss": 0.6063,
+      "step": 1823
+    },
+    {
+      "epoch": 0.42791788856304985,
+      "grad_norm": 1.8096930022472564,
+      "learning_rate": 7.07165423762173e-06,
+      "loss": 0.6326,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4281524926686217,
+      "grad_norm": 1.7350170547741561,
+      "learning_rate": 7.0679257121377706e-06,
+      "loss": 0.6305,
+      "step": 1825
+    },
+    {
+      "epoch": 0.42838709677419357,
+      "grad_norm": 1.6282545357478986,
+      "learning_rate": 7.0641957989271956e-06,
+      "loss": 0.577,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4286217008797654,
+      "grad_norm": 2.3377677378479005,
+      "learning_rate": 7.060464500493044e-06,
+      "loss": 0.6412,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4288563049853372,
+      "grad_norm": 1.8744231423038658,
+      "learning_rate": 7.056731819339287e-06,
+      "loss": 0.6191,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4290909090909091,
+      "grad_norm": 1.7219629262070664,
+      "learning_rate": 7.0529977579708195e-06,
+      "loss": 0.6175,
+      "step": 1829
+    },
+    {
+      "epoch": 0.42932551319648093,
+      "grad_norm": 0.8982095927587648,
+      "learning_rate": 7.049262318893466e-06,
+      "loss": 0.6053,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4295601173020528,
+      "grad_norm": 6.44110777874587,
+      "learning_rate": 7.0455255046139734e-06,
+      "loss": 0.5827,
+      "step": 1831
+    },
+    {
+      "epoch": 0.42979472140762465,
+      "grad_norm": 1.5661883166939536,
+      "learning_rate": 7.041787317640014e-06,
+      "loss": 0.6156,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4300293255131965,
+      "grad_norm": 1.2968226800114315,
+      "learning_rate": 7.0380477604801786e-06,
+      "loss": 0.592,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4302639296187683,
+      "grad_norm": 1.1718379659905511,
+      "learning_rate": 7.034306835643978e-06,
+      "loss": 0.6717,
+      "step": 1834
+    },
+    {
+      "epoch": 0.43049853372434016,
+      "grad_norm": 15.81144946578598,
+      "learning_rate": 7.030564545641842e-06,
+      "loss": 0.6009,
+      "step": 1835
+    },
+    {
+      "epoch": 0.430733137829912,
+      "grad_norm": 1.4409313698600532,
+      "learning_rate": 7.0268208929851164e-06,
+      "loss": 0.5781,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4309677419354839,
+      "grad_norm": 9.964109243112334,
+      "learning_rate": 7.023075880186061e-06,
+      "loss": 0.6207,
+      "step": 1837
+    },
+    {
+      "epoch": 0.43120234604105573,
+      "grad_norm": 1.27302283818698,
+      "learning_rate": 7.019329509757845e-06,
+      "loss": 0.6312,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4314369501466276,
+      "grad_norm": 1.7943943144512091,
+      "learning_rate": 7.015581784214557e-06,
+      "loss": 0.5983,
+      "step": 1839
+    },
+    {
+      "epoch": 0.43167155425219944,
+      "grad_norm": 1.008068563518117,
+      "learning_rate": 7.0118327060711865e-06,
+      "loss": 0.5751,
+      "step": 1840
+    },
+    {
+      "epoch": 0.43190615835777124,
+      "grad_norm": 1.0117839893130047,
+      "learning_rate": 7.0080822778436345e-06,
+      "loss": 0.6601,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4321407624633431,
+      "grad_norm": 0.8513872014823226,
+      "learning_rate": 7.004330502048708e-06,
+      "loss": 0.6341,
+      "step": 1842
+    },
+    {
+      "epoch": 0.43237536656891495,
+      "grad_norm": 1.1295110913790627,
+      "learning_rate": 7.000577381204119e-06,
+      "loss": 0.6495,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4326099706744868,
+      "grad_norm": 2.159755829167205,
+      "learning_rate": 6.9968229178284775e-06,
+      "loss": 0.641,
+      "step": 1844
+    },
+    {
+      "epoch": 0.43284457478005867,
+      "grad_norm": 1.2883920798235482,
+      "learning_rate": 6.993067114441303e-06,
+      "loss": 0.6853,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4330791788856305,
+      "grad_norm": 1.592180880850343,
+      "learning_rate": 6.989309973563006e-06,
+      "loss": 0.585,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4333137829912023,
+      "grad_norm": 1.761137530752373,
+      "learning_rate": 6.985551497714898e-06,
+      "loss": 0.6437,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4335483870967742,
+      "grad_norm": 1.542809103262987,
+      "learning_rate": 6.981791689419187e-06,
+      "loss": 0.5708,
+      "step": 1848
+    },
+    {
+      "epoch": 0.43378299120234604,
+      "grad_norm": 1.336389712715548,
+      "learning_rate": 6.978030551198974e-06,
+      "loss": 0.5904,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4340175953079179,
+      "grad_norm": 0.9220463053323381,
+      "learning_rate": 6.974268085578252e-06,
+      "loss": 0.5911,
+      "step": 1850
+    },
+    {
+      "epoch": 0.43425219941348975,
+      "grad_norm": 0.9193301306151851,
+      "learning_rate": 6.970504295081907e-06,
+      "loss": 0.6083,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4344868035190616,
+      "grad_norm": 1.7350690801122728,
+      "learning_rate": 6.966739182235711e-06,
+      "loss": 0.6572,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4347214076246334,
+      "grad_norm": 0.8732511062010674,
+      "learning_rate": 6.9629727495663265e-06,
+      "loss": 0.6557,
+      "step": 1853
+    },
+    {
+      "epoch": 0.43495601173020526,
+      "grad_norm": 1.5657251389423612,
+      "learning_rate": 6.959204999601299e-06,
+      "loss": 0.6372,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4351906158357771,
+      "grad_norm": 1.4136590535535574,
+      "learning_rate": 6.955435934869059e-06,
+      "loss": 0.6115,
+      "step": 1855
+    },
+    {
+      "epoch": 0.435425219941349,
+      "grad_norm": 7.720537086803013,
+      "learning_rate": 6.951665557898923e-06,
+      "loss": 0.6,
+      "step": 1856
+    },
+    {
+      "epoch": 0.43565982404692083,
+      "grad_norm": 1.275708259139756,
+      "learning_rate": 6.9478938712210785e-06,
+      "loss": 0.6539,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4358944281524927,
+      "grad_norm": 0.9695794688862226,
+      "learning_rate": 6.944120877366605e-06,
+      "loss": 0.6149,
+      "step": 1858
+    },
+    {
+      "epoch": 0.43612903225806454,
+      "grad_norm": 1.4441436339665144,
+      "learning_rate": 6.940346578867446e-06,
+      "loss": 0.684,
+      "step": 1859
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 1.5328480561291888,
+      "learning_rate": 6.936570978256431e-06,
+      "loss": 0.6793,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4365982404692082,
+      "grad_norm": 1.0910453658059502,
+      "learning_rate": 6.932794078067255e-06,
+      "loss": 0.6445,
+      "step": 1861
+    },
+    {
+      "epoch": 0.43683284457478005,
+      "grad_norm": 1.4701931397340777,
+      "learning_rate": 6.929015880834495e-06,
+      "loss": 0.6575,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4370674486803519,
+      "grad_norm": 2.09698745681287,
+      "learning_rate": 6.925236389093588e-06,
+      "loss": 0.6485,
+      "step": 1863
+    },
+    {
+      "epoch": 0.43730205278592377,
+      "grad_norm": 1.400083097458855,
+      "learning_rate": 6.921455605380846e-06,
+      "loss": 0.6558,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4375366568914956,
+      "grad_norm": 1.6664198387372906,
+      "learning_rate": 6.917673532233446e-06,
+      "loss": 0.5903,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4377712609970674,
+      "grad_norm": 1.4324783819949398,
+      "learning_rate": 6.91389017218943e-06,
+      "loss": 0.6542,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4380058651026393,
+      "grad_norm": 1.1126814108154655,
+      "learning_rate": 6.910105527787704e-06,
+      "loss": 0.6752,
+      "step": 1867
+    },
+    {
+      "epoch": 0.43824046920821114,
+      "grad_norm": 1.4189156561800234,
+      "learning_rate": 6.906319601568039e-06,
+      "loss": 0.5561,
+      "step": 1868
+    },
+    {
+      "epoch": 0.438475073313783,
+      "grad_norm": 1.2961684619205245,
+      "learning_rate": 6.90253239607106e-06,
+      "loss": 0.6477,
+      "step": 1869
+    },
+    {
+      "epoch": 0.43870967741935485,
+      "grad_norm": 1.8044032786362236,
+      "learning_rate": 6.8987439138382535e-06,
+      "loss": 0.6138,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4389442815249267,
+      "grad_norm": 1.2435404874519989,
+      "learning_rate": 6.894954157411966e-06,
+      "loss": 0.5673,
+      "step": 1871
+    },
+    {
+      "epoch": 0.43917888563049856,
+      "grad_norm": 3.916524373943254,
+      "learning_rate": 6.891163129335397e-06,
+      "loss": 0.6542,
+      "step": 1872
+    },
+    {
+      "epoch": 0.43941348973607036,
+      "grad_norm": 1.5737588998343983,
+      "learning_rate": 6.8873708321525925e-06,
+      "loss": 0.574,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4396480938416422,
+      "grad_norm": 1.6666671112739668,
+      "learning_rate": 6.883577268408461e-06,
+      "loss": 0.6151,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4398826979472141,
+      "grad_norm": 1.1243560789780351,
+      "learning_rate": 6.879782440648755e-06,
+      "loss": 0.6281,
+      "step": 1875
+    },
+    {
+      "epoch": 0.44011730205278593,
+      "grad_norm": 0.8403049964959327,
+      "learning_rate": 6.875986351420077e-06,
+      "loss": 0.6382,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4403519061583578,
+      "grad_norm": 4.859535510978899,
+      "learning_rate": 6.872189003269873e-06,
+      "loss": 0.6218,
+      "step": 1877
+    },
+    {
+      "epoch": 0.44058651026392964,
+      "grad_norm": 1.719034341935976,
+      "learning_rate": 6.86839039874644e-06,
+      "loss": 0.5686,
+      "step": 1878
+    },
+    {
+      "epoch": 0.44082111436950144,
+      "grad_norm": 2.946347376780906,
+      "learning_rate": 6.86459054039891e-06,
+      "loss": 0.62,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4410557184750733,
+      "grad_norm": 1.4181908330571877,
+      "learning_rate": 6.860789430777265e-06,
+      "loss": 0.626,
+      "step": 1880
+    },
+    {
+      "epoch": 0.44129032258064516,
+      "grad_norm": 1.2326386214731289,
+      "learning_rate": 6.856987072432319e-06,
+      "loss": 0.624,
+      "step": 1881
+    },
+    {
+      "epoch": 0.441524926686217,
+      "grad_norm": 5.601743472600724,
+      "learning_rate": 6.853183467915731e-06,
+      "loss": 0.6212,
+      "step": 1882
+    },
+    {
+      "epoch": 0.44175953079178887,
+      "grad_norm": 1.0190872995517646,
+      "learning_rate": 6.849378619779989e-06,
+      "loss": 0.6313,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4419941348973607,
+      "grad_norm": 1.2085972895816905,
+      "learning_rate": 6.845572530578421e-06,
+      "loss": 0.651,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4422287390029325,
+      "grad_norm": 2.0497945874809655,
+      "learning_rate": 6.8417652028651875e-06,
+      "loss": 0.5852,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4424633431085044,
+      "grad_norm": 4.2798005446657275,
+      "learning_rate": 6.837956639195275e-06,
+      "loss": 0.6515,
+      "step": 1886
+    },
+    {
+      "epoch": 0.44269794721407624,
+      "grad_norm": 1.4509746558933565,
+      "learning_rate": 6.834146842124507e-06,
+      "loss": 0.5608,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4429325513196481,
+      "grad_norm": 1.1763189917933676,
+      "learning_rate": 6.830335814209527e-06,
+      "loss": 0.6106,
+      "step": 1888
+    },
+    {
+      "epoch": 0.44316715542521995,
+      "grad_norm": 1.4659870368579202,
+      "learning_rate": 6.826523558007814e-06,
+      "loss": 0.6492,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4434017595307918,
+      "grad_norm": 2.2671677086421096,
+      "learning_rate": 6.822710076077659e-06,
+      "loss": 0.6662,
+      "step": 1890
+    },
+    {
+      "epoch": 0.44363636363636366,
+      "grad_norm": 1.7039536255816103,
+      "learning_rate": 6.818895370978188e-06,
+      "loss": 0.6008,
+      "step": 1891
+    },
+    {
+      "epoch": 0.44387096774193546,
+      "grad_norm": 1.6415167246381555,
+      "learning_rate": 6.815079445269338e-06,
+      "loss": 0.6233,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4441055718475073,
+      "grad_norm": 1.338188601406221,
+      "learning_rate": 6.8112623015118695e-06,
+      "loss": 0.6031,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4443401759530792,
+      "grad_norm": 2.6741621302652048,
+      "learning_rate": 6.807443942267362e-06,
+      "loss": 0.6411,
+      "step": 1894
+    },
+    {
+      "epoch": 0.44457478005865103,
+      "grad_norm": 1.5318056407523253,
+      "learning_rate": 6.803624370098209e-06,
+      "loss": 0.6743,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4448093841642229,
+      "grad_norm": 1.1799940735591201,
+      "learning_rate": 6.799803587567615e-06,
+      "loss": 0.6484,
+      "step": 1896
+    },
+    {
+      "epoch": 0.44504398826979474,
+      "grad_norm": 1.571435534048894,
+      "learning_rate": 6.795981597239599e-06,
+      "loss": 0.6537,
+      "step": 1897
+    },
+    {
+      "epoch": 0.44527859237536654,
+      "grad_norm": 1.1736703928865846,
+      "learning_rate": 6.7921584016789945e-06,
+      "loss": 0.6344,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4455131964809384,
+      "grad_norm": 1.0637970914899377,
+      "learning_rate": 6.788334003451438e-06,
+      "loss": 0.6682,
+      "step": 1899
+    },
+    {
+      "epoch": 0.44574780058651026,
+      "grad_norm": 1.653129961619469,
+      "learning_rate": 6.784508405123375e-06,
+      "loss": 0.5866,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4459824046920821,
+      "grad_norm": 5.0057267635873455,
+      "learning_rate": 6.780681609262057e-06,
+      "loss": 0.6171,
+      "step": 1901
+    },
+    {
+      "epoch": 0.44621700879765397,
+      "grad_norm": 2.2243428911712737,
+      "learning_rate": 6.77685361843554e-06,
+      "loss": 0.6155,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4464516129032258,
+      "grad_norm": 1.963880510484772,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.5808,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4466862170087977,
+      "grad_norm": 1.693426729557211,
+      "learning_rate": 6.769194062163131e-06,
+      "loss": 0.6428,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4469208211143695,
+      "grad_norm": 1.5342204357267053,
+      "learning_rate": 6.765362501857353e-06,
+      "loss": 0.6314,
+      "step": 1905
+    },
+    {
+      "epoch": 0.44715542521994134,
+      "grad_norm": 1.2381517597715204,
+      "learning_rate": 6.761529756866594e-06,
+      "loss": 0.6471,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4473900293255132,
+      "grad_norm": 1.1549886641994709,
+      "learning_rate": 6.757695829762904e-06,
+      "loss": 0.6586,
+      "step": 1907
+    },
+    {
+      "epoch": 0.44762463343108505,
+      "grad_norm": 1.2646696572098253,
+      "learning_rate": 6.753860723119122e-06,
+      "loss": 0.647,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4478592375366569,
+      "grad_norm": 0.8025274153279193,
+      "learning_rate": 6.75002443950888e-06,
+      "loss": 0.5708,
+      "step": 1909
+    },
+    {
+      "epoch": 0.44809384164222876,
+      "grad_norm": 1.4513313803057928,
+      "learning_rate": 6.746186981506599e-06,
+      "loss": 0.6032,
+      "step": 1910
+    },
+    {
+      "epoch": 0.44832844574780056,
+      "grad_norm": 2.9255562450471393,
+      "learning_rate": 6.742348351687491e-06,
+      "loss": 0.6325,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4485630498533724,
+      "grad_norm": 1.3300259933782252,
+      "learning_rate": 6.73850855262755e-06,
+      "loss": 0.6635,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4487976539589443,
+      "grad_norm": 1.3287364286680272,
+      "learning_rate": 6.734667586903557e-06,
+      "loss": 0.5662,
+      "step": 1913
+    },
+    {
+      "epoch": 0.44903225806451613,
+      "grad_norm": 0.996842436056555,
+      "learning_rate": 6.730825457093077e-06,
+      "loss": 0.571,
+      "step": 1914
+    },
+    {
+      "epoch": 0.449266862170088,
+      "grad_norm": 1.4331781076351444,
+      "learning_rate": 6.726982165774455e-06,
+      "loss": 0.6802,
+      "step": 1915
+    },
+    {
+      "epoch": 0.44950146627565984,
+      "grad_norm": 1.5706689804414935,
+      "learning_rate": 6.7231377155268144e-06,
+      "loss": 0.6271,
+      "step": 1916
+    },
+    {
+      "epoch": 0.44973607038123165,
+      "grad_norm": 2.0970147647815707,
+      "learning_rate": 6.7192921089300565e-06,
+      "loss": 0.6063,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4499706744868035,
+      "grad_norm": 1.3010993871155538,
+      "learning_rate": 6.715445348564863e-06,
+      "loss": 0.6537,
+      "step": 1918
+    },
+    {
+      "epoch": 0.45020527859237536,
+      "grad_norm": 1.324428395427468,
+      "learning_rate": 6.711597437012683e-06,
+      "loss": 0.6545,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4504398826979472,
+      "grad_norm": 1.570637482520361,
+      "learning_rate": 6.707748376855741e-06,
+      "loss": 0.642,
+      "step": 1920
+    },
+    {
+      "epoch": 0.45067448680351907,
+      "grad_norm": 1.3174788360455836,
+      "learning_rate": 6.703898170677034e-06,
+      "loss": 0.6057,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4509090909090909,
+      "grad_norm": 1.8110082415582873,
+      "learning_rate": 6.700046821060328e-06,
+      "loss": 0.6131,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4511436950146628,
+      "grad_norm": 1.4819775481519541,
+      "learning_rate": 6.6961943305901515e-06,
+      "loss": 0.5847,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4513782991202346,
+      "grad_norm": 1.4897821818625845,
+      "learning_rate": 6.692340701851807e-06,
+      "loss": 0.613,
+      "step": 1924
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 1.5774010639815104,
+      "learning_rate": 6.688485937431353e-06,
+      "loss": 0.5808,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4518475073313783,
+      "grad_norm": 2.1623584390962485,
+      "learning_rate": 6.684630039915612e-06,
+      "loss": 0.5834,
+      "step": 1926
+    },
+    {
+      "epoch": 0.45208211143695015,
+      "grad_norm": 2.460836104660879,
+      "learning_rate": 6.6807730118921706e-06,
+      "loss": 0.6964,
+      "step": 1927
+    },
+    {
+      "epoch": 0.452316715542522,
+      "grad_norm": 1.3067138443317345,
+      "learning_rate": 6.676914855949372e-06,
+      "loss": 0.6404,
+      "step": 1928
+    },
+    {
+      "epoch": 0.45255131964809386,
+      "grad_norm": 2.6420519994884537,
+      "learning_rate": 6.673055574676314e-06,
+      "loss": 0.6682,
+      "step": 1929
+    },
+    {
+      "epoch": 0.45278592375366566,
+      "grad_norm": 2.717267506649316,
+      "learning_rate": 6.669195170662851e-06,
+      "loss": 0.6462,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4530205278592375,
+      "grad_norm": 1.1087273171939926,
+      "learning_rate": 6.6653336464995926e-06,
+      "loss": 0.6379,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4532551319648094,
+      "grad_norm": 4.247345100435981,
+      "learning_rate": 6.661471004777899e-06,
+      "loss": 0.6277,
+      "step": 1932
+    },
+    {
+      "epoch": 0.45348973607038123,
+      "grad_norm": 2.1968507584563413,
+      "learning_rate": 6.657607248089879e-06,
+      "loss": 0.6248,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4537243401759531,
+      "grad_norm": 2.9111655451005336,
+      "learning_rate": 6.653742379028389e-06,
+      "loss": 0.6352,
+      "step": 1934
+    },
+    {
+      "epoch": 0.45395894428152495,
+      "grad_norm": 1.7338832754085414,
+      "learning_rate": 6.64987640018704e-06,
+      "loss": 0.6329,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4541935483870968,
+      "grad_norm": 0.9310750333439333,
+      "learning_rate": 6.646009314160173e-06,
+      "loss": 0.6333,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4544281524926686,
+      "grad_norm": 3.667141975361051,
+      "learning_rate": 6.642141123542887e-06,
+      "loss": 0.6196,
+      "step": 1937
+    },
+    {
+      "epoch": 0.45466275659824046,
+      "grad_norm": 1.641892619971768,
+      "learning_rate": 6.638271830931011e-06,
+      "loss": 0.6684,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4548973607038123,
+      "grad_norm": 1.3234013033641583,
+      "learning_rate": 6.634401438921122e-06,
+      "loss": 0.6746,
+      "step": 1939
+    },
+    {
+      "epoch": 0.45513196480938417,
+      "grad_norm": 1.0470469732467738,
+      "learning_rate": 6.630529950110528e-06,
+      "loss": 0.7021,
+      "step": 1940
+    },
+    {
+      "epoch": 0.455366568914956,
+      "grad_norm": 0.9471030586762847,
+      "learning_rate": 6.626657367097278e-06,
+      "loss": 0.6298,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4556011730205279,
+      "grad_norm": 0.8134550564494365,
+      "learning_rate": 6.62278369248015e-06,
+      "loss": 0.6649,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4558357771260997,
+      "grad_norm": 1.9352155842283765,
+      "learning_rate": 6.618908928858663e-06,
+      "loss": 0.6156,
+      "step": 1943
+    },
+    {
+      "epoch": 0.45607038123167154,
+      "grad_norm": 2.980139483293199,
+      "learning_rate": 6.615033078833058e-06,
+      "loss": 0.6108,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4563049853372434,
+      "grad_norm": 1.1365315130543463,
+      "learning_rate": 6.611156145004308e-06,
+      "loss": 0.6523,
+      "step": 1945
+    },
+    {
+      "epoch": 0.45653958944281525,
+      "grad_norm": 11.077330773976612,
+      "learning_rate": 6.607278129974115e-06,
+      "loss": 0.6289,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4567741935483871,
+      "grad_norm": 2.329158201272649,
+      "learning_rate": 6.603399036344906e-06,
+      "loss": 0.7082,
+      "step": 1947
+    },
+    {
+      "epoch": 0.45700879765395896,
+      "grad_norm": 2.0870534137036825,
+      "learning_rate": 6.599518866719832e-06,
+      "loss": 0.6364,
+      "step": 1948
+    },
+    {
+      "epoch": 0.45724340175953077,
+      "grad_norm": 2.263397559926266,
+      "learning_rate": 6.595637623702763e-06,
+      "loss": 0.5984,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4574780058651026,
+      "grad_norm": 1.6725044388192607,
+      "learning_rate": 6.591755309898293e-06,
+      "loss": 0.5779,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4577126099706745,
+      "grad_norm": 2.2161112880528253,
+      "learning_rate": 6.587871927911736e-06,
+      "loss": 0.6248,
+      "step": 1951
+    },
+    {
+      "epoch": 0.45794721407624633,
+      "grad_norm": 1.2433596954618322,
+      "learning_rate": 6.583987480349114e-06,
+      "loss": 0.6325,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4581818181818182,
+      "grad_norm": 1.2299981012283725,
+      "learning_rate": 6.580101969817176e-06,
+      "loss": 0.6154,
+      "step": 1953
+    },
+    {
+      "epoch": 0.45841642228739005,
+      "grad_norm": 1.5773499642534672,
+      "learning_rate": 6.576215398923374e-06,
+      "loss": 0.6567,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4586510263929619,
+      "grad_norm": 0.9403302159770113,
+      "learning_rate": 6.5723277702758815e-06,
+      "loss": 0.6432,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4588856304985337,
+      "grad_norm": 1.128632972579098,
+      "learning_rate": 6.568439086483571e-06,
+      "loss": 0.6316,
+      "step": 1956
+    },
+    {
+      "epoch": 0.45912023460410556,
+      "grad_norm": 1.2429754863244453,
+      "learning_rate": 6.564549350156031e-06,
+      "loss": 0.681,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4593548387096774,
+      "grad_norm": 1.355363076347527,
+      "learning_rate": 6.560658563903552e-06,
+      "loss": 0.6611,
+      "step": 1958
+    },
+    {
+      "epoch": 0.45958944281524927,
+      "grad_norm": 1.4509475026824954,
+      "learning_rate": 6.556766730337134e-06,
+      "loss": 0.6673,
+      "step": 1959
+    },
+    {
+      "epoch": 0.45982404692082113,
+      "grad_norm": 1.5079323246745762,
+      "learning_rate": 6.552873852068477e-06,
+      "loss": 0.608,
+      "step": 1960
+    },
+    {
+      "epoch": 0.460058651026393,
+      "grad_norm": 1.4692177513176758,
+      "learning_rate": 6.548979931709977e-06,
+      "loss": 0.6185,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4602932551319648,
+      "grad_norm": 0.9131837214988069,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.6479,
+      "step": 1962
+    },
+    {
+      "epoch": 0.46052785923753664,
+      "grad_norm": 1.1917507594725938,
+      "learning_rate": 6.541188975176557e-06,
+      "loss": 0.5956,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4607624633431085,
+      "grad_norm": 1.1443475410357604,
+      "learning_rate": 6.537291944229928e-06,
+      "loss": 0.6449,
+      "step": 1964
+    },
+    {
+      "epoch": 0.46099706744868035,
+      "grad_norm": 3.0987968141202886,
+      "learning_rate": 6.533393881650035e-06,
+      "loss": 0.6057,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4612316715542522,
+      "grad_norm": 0.9226564047019269,
+      "learning_rate": 6.529494790052764e-06,
+      "loss": 0.588,
+      "step": 1966
+    },
+    {
+      "epoch": 0.46146627565982407,
+      "grad_norm": 1.020077010181788,
+      "learning_rate": 6.52559467205468e-06,
+      "loss": 0.6508,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4617008797653959,
+      "grad_norm": 1.1013146406559506,
+      "learning_rate": 6.521693530273046e-06,
+      "loss": 0.6576,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4619354838709677,
+      "grad_norm": 2.1982949825750056,
+      "learning_rate": 6.517791367325803e-06,
+      "loss": 0.6539,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4621700879765396,
+      "grad_norm": 1.245381337261458,
+      "learning_rate": 6.513888185831589e-06,
+      "loss": 0.6192,
+      "step": 1970
+    },
+    {
+      "epoch": 0.46240469208211143,
+      "grad_norm": 1.6278032653959793,
+      "learning_rate": 6.509983988409715e-06,
+      "loss": 0.6307,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4626392961876833,
+      "grad_norm": 0.9983515455059903,
+      "learning_rate": 6.506078777680178e-06,
+      "loss": 0.5629,
+      "step": 1972
+    },
+    {
+      "epoch": 0.46287390029325515,
+      "grad_norm": 0.9447709125469961,
+      "learning_rate": 6.502172556263656e-06,
+      "loss": 0.6282,
+      "step": 1973
+    },
+    {
+      "epoch": 0.463108504398827,
+      "grad_norm": 1.679620294931306,
+      "learning_rate": 6.4982653267815055e-06,
+      "loss": 0.6206,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4633431085043988,
+      "grad_norm": 0.9613639374971835,
+      "learning_rate": 6.494357091855752e-06,
+      "loss": 0.6221,
+      "step": 1975
+    },
+    {
+      "epoch": 0.46357771260997066,
+      "grad_norm": 2.238864881041864,
+      "learning_rate": 6.490447854109109e-06,
+      "loss": 0.6239,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4638123167155425,
+      "grad_norm": 1.6213560232591244,
+      "learning_rate": 6.486537616164951e-06,
+      "loss": 0.6991,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4640469208211144,
+      "grad_norm": 6.759059559850818,
+      "learning_rate": 6.48262638064733e-06,
+      "loss": 0.6261,
+      "step": 1978
+    },
+    {
+      "epoch": 0.46428152492668623,
+      "grad_norm": 2.1463681214030945,
+      "learning_rate": 6.478714150180964e-06,
+      "loss": 0.5846,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4645161290322581,
+      "grad_norm": 1.7239668052266666,
+      "learning_rate": 6.4748009273912436e-06,
+      "loss": 0.6363,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4647507331378299,
+      "grad_norm": 1.6135701667408295,
+      "learning_rate": 6.470886714904221e-06,
+      "loss": 0.6749,
+      "step": 1981
+    },
+    {
+      "epoch": 0.46498533724340174,
+      "grad_norm": 8.984122051071271,
+      "learning_rate": 6.466971515346612e-06,
+      "loss": 0.6738,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4652199413489736,
+      "grad_norm": 0.812893794337249,
+      "learning_rate": 6.4630553313457985e-06,
+      "loss": 0.6484,
+      "step": 1983
+    },
+    {
+      "epoch": 0.46545454545454545,
+      "grad_norm": 2.7612121859292578,
+      "learning_rate": 6.4591381655298225e-06,
+      "loss": 0.6734,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4656891495601173,
+      "grad_norm": 0.8238126963025542,
+      "learning_rate": 6.4552200205273795e-06,
+      "loss": 0.6092,
+      "step": 1985
+    },
+    {
+      "epoch": 0.46592375366568917,
+      "grad_norm": 1.2316750855753464,
+      "learning_rate": 6.451300898967829e-06,
+      "loss": 0.6208,
+      "step": 1986
+    },
+    {
+      "epoch": 0.466158357771261,
+      "grad_norm": 1.702594986920346,
+      "learning_rate": 6.4473808034811836e-06,
+      "loss": 0.6237,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4663929618768328,
+      "grad_norm": 2.9301433173037084,
+      "learning_rate": 6.443459736698106e-06,
+      "loss": 0.6545,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4666275659824047,
+      "grad_norm": 3.0236575786668065,
+      "learning_rate": 6.439537701249915e-06,
+      "loss": 0.6496,
+      "step": 1989
+    },
+    {
+      "epoch": 0.46686217008797654,
+      "grad_norm": 1.1340598218212603,
+      "learning_rate": 6.435614699768577e-06,
+      "loss": 0.5931,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4670967741935484,
+      "grad_norm": 1.1931136346109332,
+      "learning_rate": 6.431690734886711e-06,
+      "loss": 0.6094,
+      "step": 1991
+    },
+    {
+      "epoch": 0.46733137829912025,
+      "grad_norm": 1.0004037116472506,
+      "learning_rate": 6.427765809237574e-06,
+      "loss": 0.6651,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4675659824046921,
+      "grad_norm": 1.8775423558202449,
+      "learning_rate": 6.423839925455078e-06,
+      "loss": 0.6376,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4678005865102639,
+      "grad_norm": 4.252068076887891,
+      "learning_rate": 6.419913086173769e-06,
+      "loss": 0.608,
+      "step": 1994
+    },
+    {
+      "epoch": 0.46803519061583576,
+      "grad_norm": 1.392209392136907,
+      "learning_rate": 6.415985294028839e-06,
+      "loss": 0.6179,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4682697947214076,
+      "grad_norm": 0.8941148909710063,
+      "learning_rate": 6.412056551656121e-06,
+      "loss": 0.5984,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4685043988269795,
+      "grad_norm": 1.5181958453031448,
+      "learning_rate": 6.40812686169208e-06,
+      "loss": 0.644,
+      "step": 1997
+    },
+    {
+      "epoch": 0.46873900293255133,
+      "grad_norm": 1.5176366895207596,
+      "learning_rate": 6.40419622677382e-06,
+      "loss": 0.6205,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4689736070381232,
+      "grad_norm": 1.0608470980433227,
+      "learning_rate": 6.400264649539083e-06,
+      "loss": 0.621,
+      "step": 1999
+    },
+    {
+      "epoch": 0.46920821114369504,
+      "grad_norm": 1.4891152428412084,
+      "learning_rate": 6.396332132626235e-06,
+      "loss": 0.6087,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46920821114369504,
+      "eval_loss": 0.6271416544914246,
+      "eval_runtime": 34.309,
+      "eval_samples_per_second": 15.914,
+      "eval_steps_per_second": 0.146,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46944281524926684,
+      "grad_norm": 1.323818856110043,
+      "learning_rate": 6.392398678674281e-06,
+      "loss": 0.6231,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4696774193548387,
+      "grad_norm": 1.403156838611236,
+      "learning_rate": 6.388464290322848e-06,
+      "loss": 0.5927,
+      "step": 2002
+    },
+    {
+      "epoch": 0.46991202346041056,
+      "grad_norm": 1.8337101115881043,
+      "learning_rate": 6.384528970212196e-06,
+      "loss": 0.6724,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4701466275659824,
+      "grad_norm": 1.4523158804398866,
+      "learning_rate": 6.380592720983203e-06,
+      "loss": 0.6227,
+      "step": 2004
+    },
+    {
+      "epoch": 0.47038123167155427,
+      "grad_norm": 1.1814001637036708,
+      "learning_rate": 6.376655545277379e-06,
+      "loss": 0.6321,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4706158357771261,
+      "grad_norm": 1.5839664998501093,
+      "learning_rate": 6.3727174457368505e-06,
+      "loss": 0.6386,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4708504398826979,
+      "grad_norm": 1.263564175177129,
+      "learning_rate": 6.368778425004365e-06,
+      "loss": 0.6571,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4710850439882698,
+      "grad_norm": 1.1036086210231406,
+      "learning_rate": 6.364838485723286e-06,
+      "loss": 0.6061,
+      "step": 2008
+    },
+    {
+      "epoch": 0.47131964809384164,
+      "grad_norm": 1.690351485184434,
+      "learning_rate": 6.360897630537599e-06,
+      "loss": 0.6111,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4715542521994135,
+      "grad_norm": 1.6972648310695901,
+      "learning_rate": 6.3569558620918985e-06,
+      "loss": 0.6623,
+      "step": 2010
+    },
+    {
+      "epoch": 0.47178885630498535,
+      "grad_norm": 1.4296033586068853,
+      "learning_rate": 6.353013183031394e-06,
+      "loss": 0.6697,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4720234604105572,
+      "grad_norm": 1.8778101754810188,
+      "learning_rate": 6.349069596001908e-06,
+      "loss": 0.6654,
+      "step": 2012
+    },
+    {
+      "epoch": 0.472258064516129,
+      "grad_norm": 1.1484421913508873,
+      "learning_rate": 6.34512510364987e-06,
+      "loss": 0.6109,
+      "step": 2013
+    },
+    {
+      "epoch": 0.47249266862170086,
+      "grad_norm": 1.2213269093026615,
+      "learning_rate": 6.341179708622315e-06,
+      "loss": 0.6094,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 1.1797986510112424,
+      "learning_rate": 6.337233413566889e-06,
+      "loss": 0.6249,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4729618768328446,
+      "grad_norm": 2.0379696555636713,
+      "learning_rate": 6.3332862211318406e-06,
+      "loss": 0.6219,
+      "step": 2016
+    },
+    {
+      "epoch": 0.47319648093841643,
+      "grad_norm": 0.8989640094613222,
+      "learning_rate": 6.3293381339660155e-06,
+      "loss": 0.6878,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4734310850439883,
+      "grad_norm": 2.6534551632356376,
+      "learning_rate": 6.325389154718865e-06,
+      "loss": 0.6281,
+      "step": 2018
+    },
+    {
+      "epoch": 0.47366568914956014,
+      "grad_norm": 1.0670968163009993,
+      "learning_rate": 6.32143928604044e-06,
+      "loss": 0.6391,
+      "step": 2019
+    },
+    {
+      "epoch": 0.47390029325513194,
+      "grad_norm": 1.2520887385389352,
+      "learning_rate": 6.317488530581384e-06,
+      "loss": 0.5974,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4741348973607038,
+      "grad_norm": 2.1016639103309633,
+      "learning_rate": 6.313536890992935e-06,
+      "loss": 0.6931,
+      "step": 2021
+    },
+    {
+      "epoch": 0.47436950146627566,
+      "grad_norm": 0.8221569405185536,
+      "learning_rate": 6.309584369926932e-06,
+      "loss": 0.6261,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4746041055718475,
+      "grad_norm": 1.0469243110888673,
+      "learning_rate": 6.305630970035796e-06,
+      "loss": 0.5932,
+      "step": 2023
+    },
+    {
+      "epoch": 0.47483870967741937,
+      "grad_norm": 1.2193409312969328,
+      "learning_rate": 6.301676693972547e-06,
+      "loss": 0.6393,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4750733137829912,
+      "grad_norm": 1.0060176975182853,
+      "learning_rate": 6.297721544390782e-06,
+      "loss": 0.5951,
+      "step": 2025
+    },
+    {
+      "epoch": 0.475307917888563,
+      "grad_norm": 1.2805031802136542,
+      "learning_rate": 6.293765523944697e-06,
+      "loss": 0.6822,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4755425219941349,
+      "grad_norm": 1.5525533878292266,
+      "learning_rate": 6.28980863528906e-06,
+      "loss": 0.7123,
+      "step": 2027
+    },
+    {
+      "epoch": 0.47577712609970674,
+      "grad_norm": 1.7537531970959022,
+      "learning_rate": 6.285850881079229e-06,
+      "loss": 0.6235,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4760117302052786,
+      "grad_norm": 2.624656138701152,
+      "learning_rate": 6.281892263971143e-06,
+      "loss": 0.6145,
+      "step": 2029
+    },
+    {
+      "epoch": 0.47624633431085045,
+      "grad_norm": 1.2092601576569733,
+      "learning_rate": 6.277932786621318e-06,
+      "loss": 0.6081,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4764809384164223,
+      "grad_norm": 1.4072005358178026,
+      "learning_rate": 6.273972451686845e-06,
+      "loss": 0.6363,
+      "step": 2031
+    },
+    {
+      "epoch": 0.47671554252199416,
+      "grad_norm": 1.0231130407716105,
+      "learning_rate": 6.270011261825396e-06,
+      "loss": 0.61,
+      "step": 2032
+    },
+    {
+      "epoch": 0.47695014662756596,
+      "grad_norm": 1.4865705604934518,
+      "learning_rate": 6.266049219695211e-06,
+      "loss": 0.6918,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4771847507331378,
+      "grad_norm": 2.200147147581259,
+      "learning_rate": 6.262086327955108e-06,
+      "loss": 0.6011,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4774193548387097,
+      "grad_norm": 1.1387023311455957,
+      "learning_rate": 6.258122589264469e-06,
+      "loss": 0.6083,
+      "step": 2035
+    },
+    {
+      "epoch": 0.47765395894428153,
+      "grad_norm": 1.4672611309047097,
+      "learning_rate": 6.25415800628325e-06,
+      "loss": 0.5993,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4778885630498534,
+      "grad_norm": 1.3735125227353329,
+      "learning_rate": 6.250192581671968e-06,
+      "loss": 0.5535,
+      "step": 2037
+    },
+    {
+      "epoch": 0.47812316715542524,
+      "grad_norm": 1.0229076024186408,
+      "learning_rate": 6.246226318091708e-06,
+      "loss": 0.657,
+      "step": 2038
+    },
+    {
+      "epoch": 0.47835777126099704,
+      "grad_norm": 1.1284116271169373,
+      "learning_rate": 6.24225921820412e-06,
+      "loss": 0.6119,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4785923753665689,
+      "grad_norm": 1.018314898840206,
+      "learning_rate": 6.238291284671409e-06,
+      "loss": 0.6873,
+      "step": 2040
+    },
+    {
+      "epoch": 0.47882697947214076,
+      "grad_norm": 1.2858736025682194,
+      "learning_rate": 6.234322520156347e-06,
+      "loss": 0.5724,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4790615835777126,
+      "grad_norm": 2.88091820997555,
+      "learning_rate": 6.2303529273222585e-06,
+      "loss": 0.6074,
+      "step": 2042
+    },
+    {
+      "epoch": 0.47929618768328447,
+      "grad_norm": 3.0003708776457056,
+      "learning_rate": 6.2263825088330254e-06,
+      "loss": 0.6572,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4795307917888563,
+      "grad_norm": 2.9239657263409526,
+      "learning_rate": 6.222411267353081e-06,
+      "loss": 0.5966,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4797653958944281,
+      "grad_norm": 1.0877362072010932,
+      "learning_rate": 6.218439205547418e-06,
+      "loss": 0.6138,
+      "step": 2045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8093569813862858,
+      "learning_rate": 6.214466326081573e-06,
+      "loss": 0.6149,
+      "step": 2046
+    },
+    {
+      "epoch": 0.48023460410557184,
+      "grad_norm": 1.491766211761286,
+      "learning_rate": 6.210492631621632e-06,
+      "loss": 0.6524,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4804692082111437,
+      "grad_norm": 1.0604179495136983,
+      "learning_rate": 6.206518124834231e-06,
+      "loss": 0.6317,
+      "step": 2048
+    },
+    {
+      "epoch": 0.48070381231671555,
+      "grad_norm": 0.7762572817944767,
+      "learning_rate": 6.202542808386549e-06,
+      "loss": 0.6077,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4809384164222874,
+      "grad_norm": 2.583739623293477,
+      "learning_rate": 6.198566684946308e-06,
+      "loss": 0.6083,
+      "step": 2050
+    },
+    {
+      "epoch": 0.48117302052785926,
+      "grad_norm": 1.5067311705327708,
+      "learning_rate": 6.194589757181772e-06,
+      "loss": 0.6363,
+      "step": 2051
+    },
+    {
+      "epoch": 0.48140762463343106,
+      "grad_norm": 0.9814158342670776,
+      "learning_rate": 6.190612027761748e-06,
+      "loss": 0.6015,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4816422287390029,
+      "grad_norm": 0.9679059478670816,
+      "learning_rate": 6.186633499355576e-06,
+      "loss": 0.5667,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4818768328445748,
+      "grad_norm": 1.2641337429134951,
+      "learning_rate": 6.182654174633132e-06,
+      "loss": 0.6625,
+      "step": 2054
+    },
+    {
+      "epoch": 0.48211143695014663,
+      "grad_norm": 1.9502971263233753,
+      "learning_rate": 6.178674056264831e-06,
+      "loss": 0.6115,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4823460410557185,
+      "grad_norm": 1.4368040195448697,
+      "learning_rate": 6.174693146921618e-06,
+      "loss": 0.6746,
+      "step": 2056
+    },
+    {
+      "epoch": 0.48258064516129034,
+      "grad_norm": 2.524271662293183,
+      "learning_rate": 6.17071144927497e-06,
+      "loss": 0.5983,
+      "step": 2057
+    },
+    {
+      "epoch": 0.48281524926686215,
+      "grad_norm": 1.0766715766159713,
+      "learning_rate": 6.16672896599689e-06,
+      "loss": 0.6108,
+      "step": 2058
+    },
+    {
+      "epoch": 0.483049853372434,
+      "grad_norm": 6.1316689147008745,
+      "learning_rate": 6.162745699759911e-06,
+      "loss": 0.6344,
+      "step": 2059
+    },
+    {
+      "epoch": 0.48328445747800586,
+      "grad_norm": 1.3924286006700812,
+      "learning_rate": 6.158761653237091e-06,
+      "loss": 0.6685,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4835190615835777,
+      "grad_norm": 1.7880709858752692,
+      "learning_rate": 6.15477682910201e-06,
+      "loss": 0.5301,
+      "step": 2061
+    },
+    {
+      "epoch": 0.48375366568914957,
+      "grad_norm": 2.1616777292211107,
+      "learning_rate": 6.150791230028772e-06,
+      "loss": 0.626,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4839882697947214,
+      "grad_norm": 1.494502195180896,
+      "learning_rate": 6.146804858692001e-06,
+      "loss": 0.6205,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4842228739002933,
+      "grad_norm": 1.0405545660119742,
+      "learning_rate": 6.142817717766837e-06,
+      "loss": 0.5933,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4844574780058651,
+      "grad_norm": 4.123818420590978,
+      "learning_rate": 6.138829809928938e-06,
+      "loss": 0.6474,
+      "step": 2065
+    },
+    {
+      "epoch": 0.48469208211143694,
+      "grad_norm": 1.6130108710560063,
+      "learning_rate": 6.1348411378544756e-06,
+      "loss": 0.6335,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4849266862170088,
+      "grad_norm": 3.5519518159682733,
+      "learning_rate": 6.130851704220137e-06,
+      "loss": 0.6165,
+      "step": 2067
+    },
+    {
+      "epoch": 0.48516129032258065,
+      "grad_norm": 1.1605141693731862,
+      "learning_rate": 6.126861511703119e-06,
+      "loss": 0.6823,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4853958944281525,
+      "grad_norm": 1.4032447915593402,
+      "learning_rate": 6.122870562981125e-06,
+      "loss": 0.5898,
+      "step": 2069
+    },
+    {
+      "epoch": 0.48563049853372436,
+      "grad_norm": 2.762947182978864,
+      "learning_rate": 6.118878860732369e-06,
+      "loss": 0.5631,
+      "step": 2070
+    },
+    {
+      "epoch": 0.48586510263929616,
+      "grad_norm": 1.3411722921723264,
+      "learning_rate": 6.1148864076355695e-06,
+      "loss": 0.6272,
+      "step": 2071
+    },
+    {
+      "epoch": 0.486099706744868,
+      "grad_norm": 1.439123735665396,
+      "learning_rate": 6.110893206369951e-06,
+      "loss": 0.5621,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4863343108504399,
+      "grad_norm": 81.63147080931444,
+      "learning_rate": 6.1068992596152355e-06,
+      "loss": 0.619,
+      "step": 2073
+    },
+    {
+      "epoch": 0.48656891495601173,
+      "grad_norm": 1.4230195049501757,
+      "learning_rate": 6.102904570051649e-06,
+      "loss": 0.5931,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4868035190615836,
+      "grad_norm": 2.1737217123008636,
+      "learning_rate": 6.098909140359917e-06,
+      "loss": 0.5828,
+      "step": 2075
+    },
+    {
+      "epoch": 0.48703812316715545,
+      "grad_norm": 2.329390791042317,
+      "learning_rate": 6.094912973221256e-06,
+      "loss": 0.6416,
+      "step": 2076
+    },
+    {
+      "epoch": 0.48727272727272725,
+      "grad_norm": 2.4877429915990135,
+      "learning_rate": 6.090916071317383e-06,
+      "loss": 0.6558,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4875073313782991,
+      "grad_norm": 1.7686598894141479,
+      "learning_rate": 6.086918437330508e-06,
+      "loss": 0.6116,
+      "step": 2078
+    },
+    {
+      "epoch": 0.48774193548387096,
+      "grad_norm": 1.1245664056743825,
+      "learning_rate": 6.082920073943328e-06,
+      "loss": 0.6295,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4879765395894428,
+      "grad_norm": 1.050651182314869,
+      "learning_rate": 6.078920983839032e-06,
+      "loss": 0.6205,
+      "step": 2080
+    },
+    {
+      "epoch": 0.48821114369501467,
+      "grad_norm": 0.9879313298149861,
+      "learning_rate": 6.074921169701296e-06,
+      "loss": 0.5395,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4884457478005865,
+      "grad_norm": 1.681688526777483,
+      "learning_rate": 6.070920634214286e-06,
+      "loss": 0.6156,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4886803519061584,
+      "grad_norm": 1.2333418581645228,
+      "learning_rate": 6.066919380062644e-06,
+      "loss": 0.5937,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4889149560117302,
+      "grad_norm": 1.1003068376854757,
+      "learning_rate": 6.0629174099315e-06,
+      "loss": 0.6386,
+      "step": 2084
+    },
+    {
+      "epoch": 0.48914956011730204,
+      "grad_norm": 1.4249697023558556,
+      "learning_rate": 6.0589147265064655e-06,
+      "loss": 0.6249,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4893841642228739,
+      "grad_norm": 1.2707201307018698,
+      "learning_rate": 6.054911332473627e-06,
+      "loss": 0.6106,
+      "step": 2086
+    },
+    {
+      "epoch": 0.48961876832844575,
+      "grad_norm": 1.1218073590723743,
+      "learning_rate": 6.050907230519547e-06,
+      "loss": 0.6174,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4898533724340176,
+      "grad_norm": 1.218911458915031,
+      "learning_rate": 6.046902423331271e-06,
+      "loss": 0.677,
+      "step": 2088
+    },
+    {
+      "epoch": 0.49008797653958946,
+      "grad_norm": 1.1170171887739733,
+      "learning_rate": 6.0428969135963064e-06,
+      "loss": 0.6809,
+      "step": 2089
+    },
+    {
+      "epoch": 0.49032258064516127,
+      "grad_norm": 1.4478340510895522,
+      "learning_rate": 6.038890704002641e-06,
+      "loss": 0.6629,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4905571847507331,
+      "grad_norm": 2.3220845394586704,
+      "learning_rate": 6.0348837972387295e-06,
+      "loss": 0.7104,
+      "step": 2091
+    },
+    {
+      "epoch": 0.490791788856305,
+      "grad_norm": 1.4967703459157093,
+      "learning_rate": 6.030876195993491e-06,
+      "loss": 0.5655,
+      "step": 2092
+    },
+    {
+      "epoch": 0.49102639296187683,
+      "grad_norm": 3.1962387851332315,
+      "learning_rate": 6.026867902956317e-06,
+      "loss": 0.6309,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4912609970674487,
+      "grad_norm": 4.875947490971553,
+      "learning_rate": 6.022858920817056e-06,
+      "loss": 0.6764,
+      "step": 2094
+    },
+    {
+      "epoch": 0.49149560117302055,
+      "grad_norm": 2.3529662411181818,
+      "learning_rate": 6.018849252266028e-06,
+      "loss": 0.6279,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4917302052785924,
+      "grad_norm": 0.8877451951374036,
+      "learning_rate": 6.014838899994004e-06,
+      "loss": 0.6,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4919648093841642,
+      "grad_norm": 1.0256533418763865,
+      "learning_rate": 6.010827866692219e-06,
+      "loss": 0.6372,
+      "step": 2097
+    },
+    {
+      "epoch": 0.49219941348973606,
+      "grad_norm": 2.645472715766187,
+      "learning_rate": 6.006816155052366e-06,
+      "loss": 0.6368,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4924340175953079,
+      "grad_norm": 1.1803660965278748,
+      "learning_rate": 6.002803767766592e-06,
+      "loss": 0.5749,
+      "step": 2099
+    },
+    {
+      "epoch": 0.49266862170087977,
+      "grad_norm": 2.153341937182534,
+      "learning_rate": 5.9987907075274935e-06,
+      "loss": 0.5923,
+      "step": 2100
+    },
+    {
+      "epoch": 0.49290322580645163,
+      "grad_norm": 1.5279385287985987,
+      "learning_rate": 5.994776977028126e-06,
+      "loss": 0.5865,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4931378299120235,
+      "grad_norm": 1.0401541582219007,
+      "learning_rate": 5.990762578961988e-06,
+      "loss": 0.6482,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4933724340175953,
+      "grad_norm": 1.3610873882619805,
+      "learning_rate": 5.986747516023031e-06,
+      "loss": 0.6234,
+      "step": 2103
+    },
+    {
+      "epoch": 0.49360703812316714,
+      "grad_norm": 1.0965898966482213,
+      "learning_rate": 5.9827317909056474e-06,
+      "loss": 0.6119,
+      "step": 2104
+    },
+    {
+      "epoch": 0.493841642228739,
+      "grad_norm": 0.8575468956558576,
+      "learning_rate": 5.9787154063046805e-06,
+      "loss": 0.5901,
+      "step": 2105
+    },
+    {
+      "epoch": 0.49407624633431085,
+      "grad_norm": 1.7014611619920734,
+      "learning_rate": 5.97469836491541e-06,
+      "loss": 0.6585,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4943108504398827,
+      "grad_norm": 1.4593600436030598,
+      "learning_rate": 5.970680669433561e-06,
+      "loss": 0.6381,
+      "step": 2107
+    },
+    {
+      "epoch": 0.49454545454545457,
+      "grad_norm": 1.2626003058937378,
+      "learning_rate": 5.966662322555295e-06,
+      "loss": 0.6838,
+      "step": 2108
+    },
+    {
+      "epoch": 0.49478005865102637,
+      "grad_norm": 1.5822142751717314,
+      "learning_rate": 5.962643326977211e-06,
+      "loss": 0.64,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4950146627565982,
+      "grad_norm": 1.01398619503039,
+      "learning_rate": 5.958623685396342e-06,
+      "loss": 0.6409,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4952492668621701,
+      "grad_norm": 1.3938547079745711,
+      "learning_rate": 5.95460340051016e-06,
+      "loss": 0.6396,
+      "step": 2111
+    },
+    {
+      "epoch": 0.49548387096774194,
+      "grad_norm": 0.9483622057470202,
+      "learning_rate": 5.950582475016561e-06,
+      "loss": 0.6118,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4957184750733138,
+      "grad_norm": 1.4305872444016032,
+      "learning_rate": 5.946560911613877e-06,
+      "loss": 0.6435,
+      "step": 2113
+    },
+    {
+      "epoch": 0.49595307917888565,
+      "grad_norm": 2.1265470222810574,
+      "learning_rate": 5.9425387130008635e-06,
+      "loss": 0.6321,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4961876832844575,
+      "grad_norm": 1.3248876237644547,
+      "learning_rate": 5.938515881876708e-06,
+      "loss": 0.6395,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4964222873900293,
+      "grad_norm": 1.4061732055086722,
+      "learning_rate": 5.934492420941017e-06,
+      "loss": 0.6115,
+      "step": 2116
+    },
+    {
+      "epoch": 0.49665689149560116,
+      "grad_norm": 2.4947567464829548,
+      "learning_rate": 5.930468332893821e-06,
+      "loss": 0.6783,
+      "step": 2117
+    },
+    {
+      "epoch": 0.496891495601173,
+      "grad_norm": 1.121399487685498,
+      "learning_rate": 5.926443620435572e-06,
+      "loss": 0.6726,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4971260997067449,
+      "grad_norm": 1.5613413013974176,
+      "learning_rate": 5.922418286267143e-06,
+      "loss": 0.6105,
+      "step": 2119
+    },
+    {
+      "epoch": 0.49736070381231673,
+      "grad_norm": 1.0832686155216917,
+      "learning_rate": 5.918392333089819e-06,
+      "loss": 0.6815,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4975953079178886,
+      "grad_norm": 1.0810582259921746,
+      "learning_rate": 5.914365763605308e-06,
+      "loss": 0.6054,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4978299120234604,
+      "grad_norm": 1.199910778135771,
+      "learning_rate": 5.910338580515723e-06,
+      "loss": 0.6353,
+      "step": 2122
+    },
+    {
+      "epoch": 0.49806451612903224,
+      "grad_norm": 1.4493540128067932,
+      "learning_rate": 5.906310786523594e-06,
+      "loss": 0.6516,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4982991202346041,
+      "grad_norm": 1.663545592915166,
+      "learning_rate": 5.9022823843318614e-06,
+      "loss": 0.6498,
+      "step": 2124
+    },
+    {
+      "epoch": 0.49853372434017595,
+      "grad_norm": 2.7499053506077575,
+      "learning_rate": 5.898253376643871e-06,
+      "loss": 0.6302,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4987683284457478,
+      "grad_norm": 2.3568726196225644,
+      "learning_rate": 5.894223766163377e-06,
+      "loss": 0.6074,
+      "step": 2126
+    },
+    {
+      "epoch": 0.49900293255131967,
+      "grad_norm": 1.2555315010985002,
+      "learning_rate": 5.890193555594537e-06,
+      "loss": 0.6101,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4992375366568915,
+      "grad_norm": 1.1208816754811883,
+      "learning_rate": 5.886162747641912e-06,
+      "loss": 0.6378,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4994721407624633,
+      "grad_norm": 0.9039331724175373,
+      "learning_rate": 5.882131345010462e-06,
+      "loss": 0.6231,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4997067448680352,
+      "grad_norm": 1.2402817895612375,
+      "learning_rate": 5.878099350405551e-06,
+      "loss": 0.6197,
+      "step": 2130
+    },
+    {
+      "epoch": 0.49994134897360704,
+      "grad_norm": 5.463467280235245,
+      "learning_rate": 5.874066766532932e-06,
+      "loss": 0.5755,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5001759530791788,
+      "grad_norm": 1.2976140282253599,
+      "learning_rate": 5.870033596098763e-06,
+      "loss": 0.6427,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5004105571847507,
+      "grad_norm": 2.5624849206112397,
+      "learning_rate": 5.8659998418095865e-06,
+      "loss": 0.5447,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5006451612903225,
+      "grad_norm": 1.8390436025804224,
+      "learning_rate": 5.8619655063723455e-06,
+      "loss": 0.6494,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5008797653958944,
+      "grad_norm": 2.1348523946917055,
+      "learning_rate": 5.857930592494366e-06,
+      "loss": 0.5987,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5011143695014663,
+      "grad_norm": 1.3960883090503784,
+      "learning_rate": 5.853895102883366e-06,
+      "loss": 0.6984,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5013489736070381,
+      "grad_norm": 1.5349738414978487,
+      "learning_rate": 5.849859040247447e-06,
+      "loss": 0.596,
+      "step": 2137
+    },
+    {
+      "epoch": 0.50158357771261,
+      "grad_norm": 1.9655076525930564,
+      "learning_rate": 5.845822407295101e-06,
+      "loss": 0.6229,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5018181818181818,
+      "grad_norm": 1.3420007788533819,
+      "learning_rate": 5.841785206735192e-06,
+      "loss": 0.6193,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5020527859237537,
+      "grad_norm": 8.117371270109627,
+      "learning_rate": 5.8377474412769786e-06,
+      "loss": 0.5652,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5022873900293255,
+      "grad_norm": 1.561001686558175,
+      "learning_rate": 5.833709113630086e-06,
+      "loss": 0.6372,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5025219941348974,
+      "grad_norm": 2.092024563321373,
+      "learning_rate": 5.829670226504525e-06,
+      "loss": 0.5863,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5027565982404693,
+      "grad_norm": 10.800969441708352,
+      "learning_rate": 5.825630782610676e-06,
+      "loss": 0.6496,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5029912023460411,
+      "grad_norm": 1.426248979743345,
+      "learning_rate": 5.821590784659298e-06,
+      "loss": 0.6597,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5032258064516129,
+      "grad_norm": 1.2249297694081969,
+      "learning_rate": 5.817550235361519e-06,
+      "loss": 0.607,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5034604105571847,
+      "grad_norm": 1.1137700468619252,
+      "learning_rate": 5.8135091374288374e-06,
+      "loss": 0.6101,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5036950146627566,
+      "grad_norm": 1.4303501400443885,
+      "learning_rate": 5.809467493573122e-06,
+      "loss": 0.6401,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5039296187683284,
+      "grad_norm": 18.40844197854649,
+      "learning_rate": 5.805425306506602e-06,
+      "loss": 0.5962,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5041642228739003,
+      "grad_norm": 1.1199426720492256,
+      "learning_rate": 5.80138257894188e-06,
+      "loss": 0.6741,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5043988269794721,
+      "grad_norm": 1.3560101892234653,
+      "learning_rate": 5.797339313591912e-06,
+      "loss": 0.6349,
+      "step": 2150
+    },
+    {
+      "epoch": 0.504633431085044,
+      "grad_norm": 1.2811267233227819,
+      "learning_rate": 5.7932955131700234e-06,
+      "loss": 0.6083,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5048680351906158,
+      "grad_norm": 1.9114263951195682,
+      "learning_rate": 5.78925118038989e-06,
+      "loss": 0.5744,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5051026392961877,
+      "grad_norm": 1.1503305896877258,
+      "learning_rate": 5.785206317965554e-06,
+      "loss": 0.6615,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5053372434017596,
+      "grad_norm": 1.0972389798714104,
+      "learning_rate": 5.781160928611405e-06,
+      "loss": 0.5918,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5055718475073314,
+      "grad_norm": 2.011327435876117,
+      "learning_rate": 5.777115015042192e-06,
+      "loss": 0.611,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5058064516129033,
+      "grad_norm": 1.1611929902591391,
+      "learning_rate": 5.773068579973012e-06,
+      "loss": 0.6294,
+      "step": 2156
+    },
+    {
+      "epoch": 0.506041055718475,
+      "grad_norm": 1.7643719135611815,
+      "learning_rate": 5.769021626119314e-06,
+      "loss": 0.6624,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5062756598240469,
+      "grad_norm": 20.636142531088044,
+      "learning_rate": 5.764974156196895e-06,
+      "loss": 0.6737,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5065102639296187,
+      "grad_norm": 1.247097777242137,
+      "learning_rate": 5.760926172921897e-06,
+      "loss": 0.6894,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5067448680351906,
+      "grad_norm": 1.8066188779463952,
+      "learning_rate": 5.756877679010808e-06,
+      "loss": 0.6224,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5069794721407624,
+      "grad_norm": 2.330999880939858,
+      "learning_rate": 5.752828677180459e-06,
+      "loss": 0.6269,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5072140762463343,
+      "grad_norm": 1.1600168769235768,
+      "learning_rate": 5.748779170148018e-06,
+      "loss": 0.6612,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5074486803519062,
+      "grad_norm": 2.542947949985644,
+      "learning_rate": 5.744729160630998e-06,
+      "loss": 0.6315,
+      "step": 2163
+    },
+    {
+      "epoch": 0.507683284457478,
+      "grad_norm": 0.9921932303454776,
+      "learning_rate": 5.740678651347244e-06,
+      "loss": 0.5856,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5079178885630499,
+      "grad_norm": 2.9383136063362496,
+      "learning_rate": 5.736627645014941e-06,
+      "loss": 0.6361,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5081524926686217,
+      "grad_norm": 1.5761759179803847,
+      "learning_rate": 5.732576144352603e-06,
+      "loss": 0.6016,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5083870967741936,
+      "grad_norm": 1.4922721065637146,
+      "learning_rate": 5.728524152079078e-06,
+      "loss": 0.6215,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5086217008797654,
+      "grad_norm": 1.0309400868013936,
+      "learning_rate": 5.724471670913545e-06,
+      "loss": 0.6037,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5088563049853373,
+      "grad_norm": 9.738469427385654,
+      "learning_rate": 5.720418703575507e-06,
+      "loss": 0.6541,
+      "step": 2169
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 1.701836346894189,
+      "learning_rate": 5.716365252784799e-06,
+      "loss": 0.6199,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5093255131964809,
+      "grad_norm": 1.1885589396843215,
+      "learning_rate": 5.7123113212615776e-06,
+      "loss": 0.6227,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5095601173020528,
+      "grad_norm": 1.0427266253586323,
+      "learning_rate": 5.70825691172632e-06,
+      "loss": 0.6218,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5097947214076246,
+      "grad_norm": 0.9257854703993581,
+      "learning_rate": 5.704202026899827e-06,
+      "loss": 0.5982,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5100293255131965,
+      "grad_norm": 1.4673358268015058,
+      "learning_rate": 5.700146669503217e-06,
+      "loss": 0.5778,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5102639296187683,
+      "grad_norm": 1.6198375985508418,
+      "learning_rate": 5.696090842257925e-06,
+      "loss": 0.6297,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5104985337243402,
+      "grad_norm": 1.339436711188381,
+      "learning_rate": 5.692034547885705e-06,
+      "loss": 0.6198,
+      "step": 2176
+    },
+    {
+      "epoch": 0.510733137829912,
+      "grad_norm": 1.1579018561803922,
+      "learning_rate": 5.6879777891086195e-06,
+      "loss": 0.6565,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5109677419354839,
+      "grad_norm": 1.5933279101788016,
+      "learning_rate": 5.6839205686490474e-06,
+      "loss": 0.6677,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5112023460410557,
+      "grad_norm": 1.3983475973423523,
+      "learning_rate": 5.6798628892296704e-06,
+      "loss": 0.6398,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5114369501466276,
+      "grad_norm": 1.9172780350428003,
+      "learning_rate": 5.675804753573488e-06,
+      "loss": 0.6646,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5116715542521995,
+      "grad_norm": 1.926769896048311,
+      "learning_rate": 5.6717461644037954e-06,
+      "loss": 0.6587,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5119061583577713,
+      "grad_norm": 1.060879830734715,
+      "learning_rate": 5.667687124444201e-06,
+      "loss": 0.6258,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5121407624633431,
+      "grad_norm": 1.9436882055521738,
+      "learning_rate": 5.663627636418611e-06,
+      "loss": 0.6141,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5123753665689149,
+      "grad_norm": 1.3446154026689008,
+      "learning_rate": 5.659567703051233e-06,
+      "loss": 0.5471,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5126099706744868,
+      "grad_norm": 3.7830776679121305,
+      "learning_rate": 5.655507327066573e-06,
+      "loss": 0.5606,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5128445747800586,
+      "grad_norm": 1.5657620918205266,
+      "learning_rate": 5.651446511189438e-06,
+      "loss": 0.6554,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5130791788856305,
+      "grad_norm": 1.756775187657003,
+      "learning_rate": 5.647385258144924e-06,
+      "loss": 0.6211,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5133137829912023,
+      "grad_norm": 1.1675671400970369,
+      "learning_rate": 5.6433235706584245e-06,
+      "loss": 0.5962,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5135483870967742,
+      "grad_norm": 3.6325308426101017,
+      "learning_rate": 5.639261451455621e-06,
+      "loss": 0.6199,
+      "step": 2189
+    },
+    {
+      "epoch": 0.513782991202346,
+      "grad_norm": 0.9891117588749072,
+      "learning_rate": 5.635198903262491e-06,
+      "loss": 0.6064,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5140175953079179,
+      "grad_norm": 1.1258694136256737,
+      "learning_rate": 5.631135928805294e-06,
+      "loss": 0.587,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5142521994134898,
+      "grad_norm": 1.9338927962022219,
+      "learning_rate": 5.627072530810577e-06,
+      "loss": 0.6229,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5144868035190616,
+      "grad_norm": 0.882848012078404,
+      "learning_rate": 5.6230087120051724e-06,
+      "loss": 0.5634,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5147214076246335,
+      "grad_norm": 1.3356355298231395,
+      "learning_rate": 5.6189444751161945e-06,
+      "loss": 0.6302,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5149560117302053,
+      "grad_norm": 1.9578192368455911,
+      "learning_rate": 5.614879822871039e-06,
+      "loss": 0.6136,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5151906158357771,
+      "grad_norm": 1.4082633999462215,
+      "learning_rate": 5.610814757997377e-06,
+      "loss": 0.5932,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5154252199413489,
+      "grad_norm": 0.8495768077701695,
+      "learning_rate": 5.60674928322316e-06,
+      "loss": 0.6419,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5156598240469208,
+      "grad_norm": 1.633680357289323,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.6364,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5158944281524926,
+      "grad_norm": 0.7862901238288248,
+      "learning_rate": 5.59861711488624e-06,
+      "loss": 0.6226,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 1.8803043183879258,
+      "learning_rate": 5.594550426780804e-06,
+      "loss": 0.6542,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5163636363636364,
+      "grad_norm": 2.771889576398019,
+      "learning_rate": 5.590483339689346e-06,
+      "loss": 0.6825,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5165982404692082,
+      "grad_norm": 1.0003813650951727,
+      "learning_rate": 5.586415856341175e-06,
+      "loss": 0.5556,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5168328445747801,
+      "grad_norm": 1.885413679762788,
+      "learning_rate": 5.582347979465864e-06,
+      "loss": 0.6098,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5170674486803519,
+      "grad_norm": 0.8870621895639946,
+      "learning_rate": 5.578279711793251e-06,
+      "loss": 0.6223,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5173020527859238,
+      "grad_norm": 5.416718963422954,
+      "learning_rate": 5.574211056053434e-06,
+      "loss": 0.6284,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5175366568914956,
+      "grad_norm": 1.1596375934628205,
+      "learning_rate": 5.570142014976773e-06,
+      "loss": 0.6213,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5177712609970675,
+      "grad_norm": 1.8189386028411145,
+      "learning_rate": 5.566072591293892e-06,
+      "loss": 0.616,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5180058651026394,
+      "grad_norm": 0.9619387015495638,
+      "learning_rate": 5.562002787735658e-06,
+      "loss": 0.6506,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5182404692082111,
+      "grad_norm": 0.9584281818523113,
+      "learning_rate": 5.557932607033207e-06,
+      "loss": 0.5707,
+      "step": 2209
+    },
+    {
+      "epoch": 0.518475073313783,
+      "grad_norm": 2.3093845761194434,
+      "learning_rate": 5.553862051917922e-06,
+      "loss": 0.6048,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5187096774193548,
+      "grad_norm": 0.9421914922940519,
+      "learning_rate": 5.549791125121435e-06,
+      "loss": 0.6103,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5189442815249267,
+      "grad_norm": 2.7344370847521886,
+      "learning_rate": 5.545719829375633e-06,
+      "loss": 0.5971,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5191788856304985,
+      "grad_norm": 1.4049621935973984,
+      "learning_rate": 5.541648167412648e-06,
+      "loss": 0.697,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5194134897360704,
+      "grad_norm": 1.8334813413280062,
+      "learning_rate": 5.537576141964854e-06,
+      "loss": 0.6194,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5196480938416422,
+      "grad_norm": 1.2094605468362534,
+      "learning_rate": 5.5335037557648765e-06,
+      "loss": 0.6261,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5198826979472141,
+      "grad_norm": 0.7978972605979516,
+      "learning_rate": 5.5294310115455755e-06,
+      "loss": 0.6063,
+      "step": 2216
+    },
+    {
+      "epoch": 0.520117302052786,
+      "grad_norm": 1.4040737386700064,
+      "learning_rate": 5.525357912040058e-06,
+      "loss": 0.5976,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5203519061583578,
+      "grad_norm": 1.3325233002743604,
+      "learning_rate": 5.521284459981662e-06,
+      "loss": 0.5753,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5205865102639297,
+      "grad_norm": 4.882551507518323,
+      "learning_rate": 5.5172106581039705e-06,
+      "loss": 0.618,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5208211143695015,
+      "grad_norm": 0.7895624670833509,
+      "learning_rate": 5.5131365091407955e-06,
+      "loss": 0.6401,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5210557184750733,
+      "grad_norm": 1.1171434818172954,
+      "learning_rate": 5.5090620158261835e-06,
+      "loss": 0.6545,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5212903225806451,
+      "grad_norm": 1.856141293153569,
+      "learning_rate": 5.504987180894411e-06,
+      "loss": 0.663,
+      "step": 2222
+    },
+    {
+      "epoch": 0.521524926686217,
+      "grad_norm": 1.374983758999393,
+      "learning_rate": 5.500912007079987e-06,
+      "loss": 0.6079,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5217595307917888,
+      "grad_norm": 1.8524416969913131,
+      "learning_rate": 5.496836497117642e-06,
+      "loss": 0.6345,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5219941348973607,
+      "grad_norm": 1.0130769940812943,
+      "learning_rate": 5.492760653742339e-06,
+      "loss": 0.6545,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5222287390029325,
+      "grad_norm": 4.3285463762841365,
+      "learning_rate": 5.4886844796892635e-06,
+      "loss": 0.6231,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5224633431085044,
+      "grad_norm": 2.9862132184586088,
+      "learning_rate": 5.4846079776938144e-06,
+      "loss": 0.6299,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5226979472140763,
+      "grad_norm": 1.893331135968169,
+      "learning_rate": 5.480531150491622e-06,
+      "loss": 0.569,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5229325513196481,
+      "grad_norm": 2.463041300453357,
+      "learning_rate": 5.4764540008185294e-06,
+      "loss": 0.6277,
+      "step": 2229
+    },
+    {
+      "epoch": 0.52316715542522,
+      "grad_norm": 2.316814928454367,
+      "learning_rate": 5.472376531410597e-06,
+      "loss": 0.6186,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5234017595307918,
+      "grad_norm": 3.8991222865824504,
+      "learning_rate": 5.468298745004096e-06,
+      "loss": 0.6329,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5236363636363637,
+      "grad_norm": 1.3242630714170949,
+      "learning_rate": 5.464220644335518e-06,
+      "loss": 0.6249,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5238709677419355,
+      "grad_norm": 0.7227513969002327,
+      "learning_rate": 5.460142232141561e-06,
+      "loss": 0.6156,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5241055718475073,
+      "grad_norm": 1.1764152539151682,
+      "learning_rate": 5.45606351115913e-06,
+      "loss": 0.6911,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5243401759530791,
+      "grad_norm": 0.9231453226180013,
+      "learning_rate": 5.451984484125341e-06,
+      "loss": 0.6387,
+      "step": 2235
+    },
+    {
+      "epoch": 0.524574780058651,
+      "grad_norm": 1.4423818958906918,
+      "learning_rate": 5.447905153777515e-06,
+      "loss": 0.6028,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5248093841642228,
+      "grad_norm": 1.252361802772466,
+      "learning_rate": 5.4438255228531734e-06,
+      "loss": 0.5842,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5250439882697947,
+      "grad_norm": 4.381022046934695,
+      "learning_rate": 5.439745594090042e-06,
+      "loss": 0.6338,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5252785923753666,
+      "grad_norm": 1.3650915934618688,
+      "learning_rate": 5.4356653702260475e-06,
+      "loss": 0.6434,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5255131964809384,
+      "grad_norm": 0.9064496023863303,
+      "learning_rate": 5.431584853999312e-06,
+      "loss": 0.5568,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5257478005865103,
+      "grad_norm": 2.008849151658651,
+      "learning_rate": 5.427504048148153e-06,
+      "loss": 0.6,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5259824046920821,
+      "grad_norm": 2.564532321909153,
+      "learning_rate": 5.423422955411087e-06,
+      "loss": 0.6427,
+      "step": 2242
+    },
+    {
+      "epoch": 0.526217008797654,
+      "grad_norm": 2.5285727151936572,
+      "learning_rate": 5.41934157852682e-06,
+      "loss": 0.5802,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5264516129032258,
+      "grad_norm": 1.0149743248092284,
+      "learning_rate": 5.415259920234247e-06,
+      "loss": 0.6196,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5266862170087977,
+      "grad_norm": 0.970858090329897,
+      "learning_rate": 5.411177983272454e-06,
+      "loss": 0.6638,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5269208211143696,
+      "grad_norm": 1.221113048601327,
+      "learning_rate": 5.407095770380717e-06,
+      "loss": 0.671,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5271554252199413,
+      "grad_norm": 1.5190796071583499,
+      "learning_rate": 5.40301328429849e-06,
+      "loss": 0.6139,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5273900293255132,
+      "grad_norm": 1.7030575936840922,
+      "learning_rate": 5.398930527765416e-06,
+      "loss": 0.6286,
+      "step": 2248
+    },
+    {
+      "epoch": 0.527624633431085,
+      "grad_norm": 1.9951172534871002,
+      "learning_rate": 5.394847503521316e-06,
+      "loss": 0.6786,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5278592375366569,
+      "grad_norm": 1.5364603846571465,
+      "learning_rate": 5.3907642143061974e-06,
+      "loss": 0.6772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5280938416422287,
+      "grad_norm": 1.5308755333089372,
+      "learning_rate": 5.386680662860234e-06,
+      "loss": 0.6152,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5283284457478006,
+      "grad_norm": 1.5489129645290929,
+      "learning_rate": 5.382596851923786e-06,
+      "loss": 0.6646,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5285630498533724,
+      "grad_norm": 1.2992198648344988,
+      "learning_rate": 5.378512784237382e-06,
+      "loss": 0.5925,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5287976539589443,
+      "grad_norm": 1.2879178518454926,
+      "learning_rate": 5.374428462541726e-06,
+      "loss": 0.6355,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5290322580645161,
+      "grad_norm": 2.227818265359871,
+      "learning_rate": 5.37034388957769e-06,
+      "loss": 0.6364,
+      "step": 2255
+    },
+    {
+      "epoch": 0.529266862170088,
+      "grad_norm": 0.9408682021543482,
+      "learning_rate": 5.366259068086316e-06,
+      "loss": 0.5764,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5295014662756599,
+      "grad_norm": 1.5853926177483033,
+      "learning_rate": 5.362174000808813e-06,
+      "loss": 0.6645,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5297360703812317,
+      "grad_norm": 2.7689887231795396,
+      "learning_rate": 5.3580886904865525e-06,
+      "loss": 0.6073,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5299706744868036,
+      "grad_norm": 1.0482328044058897,
+      "learning_rate": 5.354003139861075e-06,
+      "loss": 0.6727,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5302052785923753,
+      "grad_norm": 1.0236839529278232,
+      "learning_rate": 5.349917351674073e-06,
+      "loss": 0.6187,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5304398826979472,
+      "grad_norm": 1.0129900103232927,
+      "learning_rate": 5.345831328667408e-06,
+      "loss": 0.5774,
+      "step": 2261
+    },
+    {
+      "epoch": 0.530674486803519,
+      "grad_norm": 1.0846671484043156,
+      "learning_rate": 5.341745073583092e-06,
+      "loss": 0.6425,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5309090909090909,
+      "grad_norm": 1.9163706435576997,
+      "learning_rate": 5.3376585891633e-06,
+      "loss": 0.5842,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5311436950146627,
+      "grad_norm": 1.5106485398321852,
+      "learning_rate": 5.333571878150349e-06,
+      "loss": 0.6318,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5313782991202346,
+      "grad_norm": 1.4955458517030962,
+      "learning_rate": 5.329484943286721e-06,
+      "loss": 0.6589,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5316129032258065,
+      "grad_norm": 1.5417351005640656,
+      "learning_rate": 5.325397787315042e-06,
+      "loss": 0.6079,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5318475073313783,
+      "grad_norm": 1.1147049544994247,
+      "learning_rate": 5.321310412978087e-06,
+      "loss": 0.6736,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5320821114369502,
+      "grad_norm": 0.9691010115156603,
+      "learning_rate": 5.317222823018775e-06,
+      "loss": 0.6478,
+      "step": 2268
+    },
+    {
+      "epoch": 0.532316715542522,
+      "grad_norm": 1.720804059010316,
+      "learning_rate": 5.313135020180177e-06,
+      "loss": 0.6839,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5325513196480939,
+      "grad_norm": 1.997084001822593,
+      "learning_rate": 5.309047007205498e-06,
+      "loss": 0.647,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5327859237536657,
+      "grad_norm": 3.1830546212101867,
+      "learning_rate": 5.30495878683809e-06,
+      "loss": 0.6638,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5330205278592376,
+      "grad_norm": 1.1523445703582846,
+      "learning_rate": 5.300870361821442e-06,
+      "loss": 0.6396,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5332551319648093,
+      "grad_norm": 1.8802046275179825,
+      "learning_rate": 5.296781734899182e-06,
+      "loss": 0.6272,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5334897360703812,
+      "grad_norm": 1.6473165116582242,
+      "learning_rate": 5.292692908815069e-06,
+      "loss": 0.6731,
+      "step": 2274
+    },
+    {
+      "epoch": 0.533724340175953,
+      "grad_norm": 1.2678460151946527,
+      "learning_rate": 5.288603886313003e-06,
+      "loss": 0.6138,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5339589442815249,
+      "grad_norm": 2.716014175519501,
+      "learning_rate": 5.2845146701370085e-06,
+      "loss": 0.6514,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5341935483870968,
+      "grad_norm": 1.9686345263377154,
+      "learning_rate": 5.280425263031245e-06,
+      "loss": 0.6167,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5344281524926686,
+      "grad_norm": 1.2309264732837615,
+      "learning_rate": 5.276335667739998e-06,
+      "loss": 0.6013,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5346627565982405,
+      "grad_norm": 1.1053218192979433,
+      "learning_rate": 5.272245887007678e-06,
+      "loss": 0.6811,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5348973607038123,
+      "grad_norm": 1.0822971676051554,
+      "learning_rate": 5.268155923578822e-06,
+      "loss": 0.5927,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5351319648093842,
+      "grad_norm": 3.3062833457443386,
+      "learning_rate": 5.264065780198089e-06,
+      "loss": 0.6314,
+      "step": 2281
+    },
+    {
+      "epoch": 0.535366568914956,
+      "grad_norm": 0.9575935721960738,
+      "learning_rate": 5.2599754596102615e-06,
+      "loss": 0.6794,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5356011730205279,
+      "grad_norm": 2.7721463574049956,
+      "learning_rate": 5.255884964560235e-06,
+      "loss": 0.6081,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5358357771260998,
+      "grad_norm": 2.592514843146242,
+      "learning_rate": 5.251794297793027e-06,
+      "loss": 0.6182,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5360703812316715,
+      "grad_norm": 1.3936995820236064,
+      "learning_rate": 5.247703462053767e-06,
+      "loss": 0.6374,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5363049853372434,
+      "grad_norm": 15.214073236526824,
+      "learning_rate": 5.243612460087703e-06,
+      "loss": 0.6459,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5365395894428152,
+      "grad_norm": 1.230089585822257,
+      "learning_rate": 5.239521294640185e-06,
+      "loss": 0.5995,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5367741935483871,
+      "grad_norm": 1.3390750104505194,
+      "learning_rate": 5.235429968456686e-06,
+      "loss": 0.6388,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5370087976539589,
+      "grad_norm": 2.2923778420764647,
+      "learning_rate": 5.231338484282774e-06,
+      "loss": 0.6608,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5372434017595308,
+      "grad_norm": 1.3427084443327424,
+      "learning_rate": 5.227246844864131e-06,
+      "loss": 0.6616,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5374780058651026,
+      "grad_norm": 1.479501580983395,
+      "learning_rate": 5.22315505294654e-06,
+      "loss": 0.6313,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5377126099706745,
+      "grad_norm": 1.0572575657370458,
+      "learning_rate": 5.219063111275889e-06,
+      "loss": 0.5879,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5379472140762463,
+      "grad_norm": 1.739928364512639,
+      "learning_rate": 5.214971022598162e-06,
+      "loss": 0.6682,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5381818181818182,
+      "grad_norm": 1.437674607065896,
+      "learning_rate": 5.210878789659445e-06,
+      "loss": 0.6549,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5384164222873901,
+      "grad_norm": 1.7351018511658245,
+      "learning_rate": 5.206786415205921e-06,
+      "loss": 0.6323,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5386510263929619,
+      "grad_norm": 1.3106284147500318,
+      "learning_rate": 5.202693901983868e-06,
+      "loss": 0.5961,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5388856304985338,
+      "grad_norm": 2.530769324981352,
+      "learning_rate": 5.1986012527396524e-06,
+      "loss": 0.6121,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5391202346041055,
+      "grad_norm": 1.9709138346894095,
+      "learning_rate": 5.194508470219739e-06,
+      "loss": 0.6177,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5393548387096774,
+      "grad_norm": 1.5527505505893144,
+      "learning_rate": 5.190415557170678e-06,
+      "loss": 0.6227,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5395894428152492,
+      "grad_norm": 1.1474346639859412,
+      "learning_rate": 5.186322516339108e-06,
+      "loss": 0.5706,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5398240469208211,
+      "grad_norm": 2.047406907631492,
+      "learning_rate": 5.1822293504717525e-06,
+      "loss": 0.6194,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5400586510263929,
+      "grad_norm": 1.04144786967005,
+      "learning_rate": 5.178136062315421e-06,
+      "loss": 0.6008,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5402932551319648,
+      "grad_norm": 1.3444351715872211,
+      "learning_rate": 5.174042654617001e-06,
+      "loss": 0.6392,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5405278592375367,
+      "grad_norm": 1.4815625258755283,
+      "learning_rate": 5.169949130123465e-06,
+      "loss": 0.641,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5407624633431085,
+      "grad_norm": 1.77251076274484,
+      "learning_rate": 5.165855491581863e-06,
+      "loss": 0.5752,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5409970674486804,
+      "grad_norm": 1.4646850211091804,
+      "learning_rate": 5.161761741739319e-06,
+      "loss": 0.6036,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5412316715542522,
+      "grad_norm": 2.4059610900230135,
+      "learning_rate": 5.157667883343031e-06,
+      "loss": 0.5928,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5414662756598241,
+      "grad_norm": 1.247773585715356,
+      "learning_rate": 5.153573919140275e-06,
+      "loss": 0.6137,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5417008797653959,
+      "grad_norm": 1.3846254026394906,
+      "learning_rate": 5.149479851878393e-06,
+      "loss": 0.6126,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5419354838709678,
+      "grad_norm": 0.8360359792043676,
+      "learning_rate": 5.145385684304799e-06,
+      "loss": 0.6029,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5421700879765395,
+      "grad_norm": 2.6140391672325602,
+      "learning_rate": 5.141291419166974e-06,
+      "loss": 0.6147,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5424046920821114,
+      "grad_norm": 1.3117410475337958,
+      "learning_rate": 5.137197059212463e-06,
+      "loss": 0.5693,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5426392961876833,
+      "grad_norm": 3.6763496348206948,
+      "learning_rate": 5.133102607188875e-06,
+      "loss": 0.6716,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5428739002932551,
+      "grad_norm": 2.718896859240161,
+      "learning_rate": 5.1290080658438814e-06,
+      "loss": 0.5728,
+      "step": 2314
+    },
+    {
+      "epoch": 0.543108504398827,
+      "grad_norm": 1.2701500246992494,
+      "learning_rate": 5.124913437925215e-06,
+      "loss": 0.5884,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5433431085043988,
+      "grad_norm": 1.079743766066829,
+      "learning_rate": 5.120818726180662e-06,
+      "loss": 0.5784,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5435777126099707,
+      "grad_norm": 1.2813886474414964,
+      "learning_rate": 5.116723933358071e-06,
+      "loss": 0.6133,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5438123167155425,
+      "grad_norm": 1.5789949106389531,
+      "learning_rate": 5.112629062205341e-06,
+      "loss": 0.5909,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5440469208211144,
+      "grad_norm": 1.4241228759151066,
+      "learning_rate": 5.108534115470424e-06,
+      "loss": 0.6201,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5442815249266862,
+      "grad_norm": 1.421847608917025,
+      "learning_rate": 5.1044390959013225e-06,
+      "loss": 0.6455,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5445161290322581,
+      "grad_norm": 2.5473408353621596,
+      "learning_rate": 5.100344006246092e-06,
+      "loss": 0.6531,
+      "step": 2321
+    },
+    {
+      "epoch": 0.54475073313783,
+      "grad_norm": 4.869648761811444,
+      "learning_rate": 5.0962488492528276e-06,
+      "loss": 0.6747,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5449853372434018,
+      "grad_norm": 1.044340096197161,
+      "learning_rate": 5.092153627669675e-06,
+      "loss": 0.6818,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5452199413489736,
+      "grad_norm": 1.1285358434215709,
+      "learning_rate": 5.088058344244823e-06,
+      "loss": 0.5337,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 1.0293245804494093,
+      "learning_rate": 5.083963001726501e-06,
+      "loss": 0.6231,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5456891495601173,
+      "grad_norm": 7.794166682885448,
+      "learning_rate": 5.079867602862974e-06,
+      "loss": 0.6135,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5459237536656891,
+      "grad_norm": 1.187300247165658,
+      "learning_rate": 5.075772150402555e-06,
+      "loss": 0.6588,
+      "step": 2327
+    },
+    {
+      "epoch": 0.546158357771261,
+      "grad_norm": 1.1393919289225325,
+      "learning_rate": 5.071676647093581e-06,
+      "loss": 0.6286,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5463929618768328,
+      "grad_norm": 3.7156917470133535,
+      "learning_rate": 5.0675810956844315e-06,
+      "loss": 0.5989,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5466275659824047,
+      "grad_norm": 1.3848509997731784,
+      "learning_rate": 5.0634854989235145e-06,
+      "loss": 0.6482,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5468621700879766,
+      "grad_norm": 1.7479155187976059,
+      "learning_rate": 5.059389859559271e-06,
+      "loss": 0.6185,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5470967741935484,
+      "grad_norm": 2.3408468567460736,
+      "learning_rate": 5.055294180340167e-06,
+      "loss": 0.6286,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5473313782991203,
+      "grad_norm": 1.7450838683077203,
+      "learning_rate": 5.051198464014698e-06,
+      "loss": 0.6586,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5475659824046921,
+      "grad_norm": 1.0214958871229616,
+      "learning_rate": 5.047102713331387e-06,
+      "loss": 0.6728,
+      "step": 2334
+    },
+    {
+      "epoch": 0.547800586510264,
+      "grad_norm": 1.0866483702890142,
+      "learning_rate": 5.04300693103877e-06,
+      "loss": 0.5837,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5480351906158358,
+      "grad_norm": 1.6318482107644605,
+      "learning_rate": 5.038911119885416e-06,
+      "loss": 0.6479,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5482697947214076,
+      "grad_norm": 1.3678399546457385,
+      "learning_rate": 5.034815282619907e-06,
+      "loss": 0.6184,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5485043988269794,
+      "grad_norm": 6.6632809204162,
+      "learning_rate": 5.030719421990845e-06,
+      "loss": 0.6383,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5487390029325513,
+      "grad_norm": 1.020276110373152,
+      "learning_rate": 5.026623540746842e-06,
+      "loss": 0.562,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5489736070381231,
+      "grad_norm": 2.23704785312758,
+      "learning_rate": 5.0225276416365355e-06,
+      "loss": 0.5807,
+      "step": 2340
+    },
+    {
+      "epoch": 0.549208211143695,
+      "grad_norm": 1.3090584423433533,
+      "learning_rate": 5.018431727408563e-06,
+      "loss": 0.5913,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5494428152492669,
+      "grad_norm": 1.1215291210455687,
+      "learning_rate": 5.0143358008115785e-06,
+      "loss": 0.5933,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5496774193548387,
+      "grad_norm": 1.3851313663648712,
+      "learning_rate": 5.010239864594241e-06,
+      "loss": 0.6181,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5499120234604106,
+      "grad_norm": 2.5214427901534107,
+      "learning_rate": 5.00614392150522e-06,
+      "loss": 0.5563,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5501466275659824,
+      "grad_norm": 1.6969333520157845,
+      "learning_rate": 5.002047974293186e-06,
+      "loss": 0.61,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5503812316715543,
+      "grad_norm": 0.9065597378911432,
+      "learning_rate": 4.997952025706816e-06,
+      "loss": 0.6154,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5506158357771261,
+      "grad_norm": 3.403546478196266,
+      "learning_rate": 4.9938560784947815e-06,
+      "loss": 0.6081,
+      "step": 2347
+    },
+    {
+      "epoch": 0.550850439882698,
+      "grad_norm": 1.522859665413107,
+      "learning_rate": 4.98976013540576e-06,
+      "loss": 0.6104,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5510850439882697,
+      "grad_norm": 2.2807731653683274,
+      "learning_rate": 4.985664199188424e-06,
+      "loss": 0.5902,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5513196480938416,
+      "grad_norm": 1.996223618876685,
+      "learning_rate": 4.981568272591439e-06,
+      "loss": 0.6297,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5515542521994135,
+      "grad_norm": 4.8740727246594755,
+      "learning_rate": 4.977472358363466e-06,
+      "loss": 0.648,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5517888563049853,
+      "grad_norm": 1.930770244826502,
+      "learning_rate": 4.9733764592531595e-06,
+      "loss": 0.6538,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5520234604105572,
+      "grad_norm": 1.7525721701911414,
+      "learning_rate": 4.969280578009157e-06,
+      "loss": 0.6159,
+      "step": 2353
+    },
+    {
+      "epoch": 0.552258064516129,
+      "grad_norm": 1.635051903248449,
+      "learning_rate": 4.965184717380094e-06,
+      "loss": 0.623,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5524926686217009,
+      "grad_norm": 1.253187755847933,
+      "learning_rate": 4.961088880114586e-06,
+      "loss": 0.5675,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5527272727272727,
+      "grad_norm": 2.9737604974053533,
+      "learning_rate": 4.956993068961231e-06,
+      "loss": 0.6619,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5529618768328446,
+      "grad_norm": 2.22359482931915,
+      "learning_rate": 4.952897286668616e-06,
+      "loss": 0.6697,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5531964809384164,
+      "grad_norm": 1.7212673210837532,
+      "learning_rate": 4.948801535985303e-06,
+      "loss": 0.6135,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5534310850439883,
+      "grad_norm": 1.175532897757222,
+      "learning_rate": 4.944705819659834e-06,
+      "loss": 0.616,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5536656891495602,
+      "grad_norm": 2.3673754525350335,
+      "learning_rate": 4.940610140440731e-06,
+      "loss": 0.5772,
+      "step": 2360
+    },
+    {
+      "epoch": 0.553900293255132,
+      "grad_norm": 6.049807078263912,
+      "learning_rate": 4.936514501076486e-06,
+      "loss": 0.5753,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5541348973607038,
+      "grad_norm": 1.6170123243691585,
+      "learning_rate": 4.932418904315569e-06,
+      "loss": 0.6588,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5543695014662756,
+      "grad_norm": 1.329215654069329,
+      "learning_rate": 4.928323352906421e-06,
+      "loss": 0.6389,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5546041055718475,
+      "grad_norm": 1.2327922877038138,
+      "learning_rate": 4.924227849597447e-06,
+      "loss": 0.5687,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5548387096774193,
+      "grad_norm": 1.376394351500655,
+      "learning_rate": 4.9201323971370264e-06,
+      "loss": 0.622,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5550733137829912,
+      "grad_norm": 1.1401847022392284,
+      "learning_rate": 4.9160369982735025e-06,
+      "loss": 0.6096,
+      "step": 2366
+    },
+    {
+      "epoch": 0.555307917888563,
+      "grad_norm": 1.6061812093808758,
+      "learning_rate": 4.911941655755179e-06,
+      "loss": 0.6188,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5555425219941349,
+      "grad_norm": 1.320887601449324,
+      "learning_rate": 4.907846372330326e-06,
+      "loss": 0.613,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5557771260997068,
+      "grad_norm": 1.8809095814675294,
+      "learning_rate": 4.903751150747176e-06,
+      "loss": 0.6148,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5560117302052786,
+      "grad_norm": 0.9737004625936243,
+      "learning_rate": 4.89965599375391e-06,
+      "loss": 0.5631,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5562463343108505,
+      "grad_norm": 2.4030696249831283,
+      "learning_rate": 4.895560904098678e-06,
+      "loss": 0.6168,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5564809384164223,
+      "grad_norm": 1.448894327844498,
+      "learning_rate": 4.8914658845295786e-06,
+      "loss": 0.645,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5567155425219942,
+      "grad_norm": 1.636351851164784,
+      "learning_rate": 4.887370937794661e-06,
+      "loss": 0.5676,
+      "step": 2373
+    },
+    {
+      "epoch": 0.556950146627566,
+      "grad_norm": 3.0144010208779095,
+      "learning_rate": 4.8832760666419295e-06,
+      "loss": 0.5601,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5571847507331378,
+      "grad_norm": 1.5360350101568192,
+      "learning_rate": 4.87918127381934e-06,
+      "loss": 0.6082,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5574193548387096,
+      "grad_norm": 1.4622800973510324,
+      "learning_rate": 4.8750865620747865e-06,
+      "loss": 0.5936,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5576539589442815,
+      "grad_norm": 1.3974336295850467,
+      "learning_rate": 4.870991934156119e-06,
+      "loss": 0.5887,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5578885630498533,
+      "grad_norm": 1.1946185311544872,
+      "learning_rate": 4.866897392811127e-06,
+      "loss": 0.6253,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5581231671554252,
+      "grad_norm": 1.1308622762551268,
+      "learning_rate": 4.8628029407875385e-06,
+      "loss": 0.6541,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5583577712609971,
+      "grad_norm": 0.9214548517449351,
+      "learning_rate": 4.858708580833027e-06,
+      "loss": 0.6298,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5585923753665689,
+      "grad_norm": 1.2700773667255392,
+      "learning_rate": 4.854614315695203e-06,
+      "loss": 0.6025,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5588269794721408,
+      "grad_norm": 1.9741834869439465,
+      "learning_rate": 4.850520148121609e-06,
+      "loss": 0.5761,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5590615835777126,
+      "grad_norm": 1.7369595965255913,
+      "learning_rate": 4.846426080859728e-06,
+      "loss": 0.6152,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5592961876832845,
+      "grad_norm": 2.2043774595925543,
+      "learning_rate": 4.84233211665697e-06,
+      "loss": 0.6177,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5595307917888563,
+      "grad_norm": 1.131176231870637,
+      "learning_rate": 4.838238258260684e-06,
+      "loss": 0.6464,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5597653958944282,
+      "grad_norm": 1.2273789428672568,
+      "learning_rate": 4.83414450841814e-06,
+      "loss": 0.6084,
+      "step": 2386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0192662099118455,
+      "learning_rate": 4.830050869876536e-06,
+      "loss": 0.6207,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5602346041055718,
+      "grad_norm": 1.649900195779437,
+      "learning_rate": 4.825957345383e-06,
+      "loss": 0.6458,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5604692082111437,
+      "grad_norm": 1.7812859104244483,
+      "learning_rate": 4.8218639376845825e-06,
+      "loss": 0.6388,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5607038123167155,
+      "grad_norm": 1.7056418402379854,
+      "learning_rate": 4.817770649528249e-06,
+      "loss": 0.5979,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5609384164222874,
+      "grad_norm": 2.759126474933544,
+      "learning_rate": 4.813677483660892e-06,
+      "loss": 0.6573,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5611730205278592,
+      "grad_norm": 1.3796494768679621,
+      "learning_rate": 4.809584442829323e-06,
+      "loss": 0.5753,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5614076246334311,
+      "grad_norm": 2.225836720974062,
+      "learning_rate": 4.805491529780261e-06,
+      "loss": 0.5676,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5616422287390029,
+      "grad_norm": 4.8603603742046,
+      "learning_rate": 4.801398747260348e-06,
+      "loss": 0.576,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5618768328445748,
+      "grad_norm": 2.728295412602604,
+      "learning_rate": 4.797306098016135e-06,
+      "loss": 0.6185,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5621114369501466,
+      "grad_norm": 1.9233726850880661,
+      "learning_rate": 4.79321358479408e-06,
+      "loss": 0.6146,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5623460410557185,
+      "grad_norm": 1.7933146896764474,
+      "learning_rate": 4.789121210340555e-06,
+      "loss": 0.6069,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5625806451612904,
+      "grad_norm": 0.9867527993249323,
+      "learning_rate": 4.78502897740184e-06,
+      "loss": 0.5888,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5628152492668622,
+      "grad_norm": 1.964434092692377,
+      "learning_rate": 4.780936888724113e-06,
+      "loss": 0.6451,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5630498533724341,
+      "grad_norm": 7.884498287566548,
+      "learning_rate": 4.77684494705346e-06,
+      "loss": 0.6011,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5632844574780058,
+      "grad_norm": 2.5780682514079545,
+      "learning_rate": 4.772753155135871e-06,
+      "loss": 0.6058,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5635190615835777,
+      "grad_norm": 2.179264908491652,
+      "learning_rate": 4.768661515717227e-06,
+      "loss": 0.6801,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5637536656891495,
+      "grad_norm": 1.3226549993329324,
+      "learning_rate": 4.764570031543316e-06,
+      "loss": 0.6316,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5639882697947214,
+      "grad_norm": 1.359005961369105,
+      "learning_rate": 4.760478705359816e-06,
+      "loss": 0.6121,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5642228739002932,
+      "grad_norm": 1.6349963956013602,
+      "learning_rate": 4.756387539912299e-06,
+      "loss": 0.6454,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5644574780058651,
+      "grad_norm": 1.8640311979755861,
+      "learning_rate": 4.752296537946236e-06,
+      "loss": 0.6086,
+      "step": 2406
+    },
+    {
+      "epoch": 0.564692082111437,
+      "grad_norm": 2.5519420011751976,
+      "learning_rate": 4.748205702206975e-06,
+      "loss": 0.6204,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5649266862170088,
+      "grad_norm": 4.836370764245565,
+      "learning_rate": 4.744115035439766e-06,
+      "loss": 0.6785,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5651612903225807,
+      "grad_norm": 2.0737177128369084,
+      "learning_rate": 4.740024540389741e-06,
+      "loss": 0.6272,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5653958944281525,
+      "grad_norm": 0.9540299827895479,
+      "learning_rate": 4.735934219801912e-06,
+      "loss": 0.5643,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5656304985337244,
+      "grad_norm": 1.1685301518053146,
+      "learning_rate": 4.73184407642118e-06,
+      "loss": 0.6396,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5658651026392962,
+      "grad_norm": 3.8313120205410023,
+      "learning_rate": 4.727754112992326e-06,
+      "loss": 0.5677,
+      "step": 2412
+    },
+    {
+      "epoch": 0.566099706744868,
+      "grad_norm": 1.2533957183595168,
+      "learning_rate": 4.723664332260004e-06,
+      "loss": 0.6528,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5663343108504398,
+      "grad_norm": 1.9385572074533628,
+      "learning_rate": 4.719574736968755e-06,
+      "loss": 0.5795,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5665689149560117,
+      "grad_norm": 1.4613315996424439,
+      "learning_rate": 4.715485329862992e-06,
+      "loss": 0.7028,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5668035190615835,
+      "grad_norm": 7.705662308862915,
+      "learning_rate": 4.711396113686998e-06,
+      "loss": 0.6981,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5670381231671554,
+      "grad_norm": 1.35730909069504,
+      "learning_rate": 4.707307091184931e-06,
+      "loss": 0.6128,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5672727272727273,
+      "grad_norm": 1.0543319364793236,
+      "learning_rate": 4.7032182651008204e-06,
+      "loss": 0.6413,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5675073313782991,
+      "grad_norm": 1.3648126645029455,
+      "learning_rate": 4.6991296381785585e-06,
+      "loss": 0.6146,
+      "step": 2419
+    },
+    {
+      "epoch": 0.567741935483871,
+      "grad_norm": 1.0596246737273716,
+      "learning_rate": 4.69504121316191e-06,
+      "loss": 0.6238,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5679765395894428,
+      "grad_norm": 2.954131782774614,
+      "learning_rate": 4.690952992794504e-06,
+      "loss": 0.6482,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5682111436950147,
+      "grad_norm": 1.644712519006594,
+      "learning_rate": 4.686864979819825e-06,
+      "loss": 0.6336,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5684457478005865,
+      "grad_norm": 0.9851063697756272,
+      "learning_rate": 4.682777176981225e-06,
+      "loss": 0.595,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5686803519061584,
+      "grad_norm": 1.9024601181916914,
+      "learning_rate": 4.6786895870219155e-06,
+      "loss": 0.6239,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5689149560117303,
+      "grad_norm": 1.9274176199318562,
+      "learning_rate": 4.674602212684959e-06,
+      "loss": 0.626,
+      "step": 2425
+    },
+    {
+      "epoch": 0.569149560117302,
+      "grad_norm": 1.4662092880540012,
+      "learning_rate": 4.670515056713279e-06,
+      "loss": 0.5909,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5693841642228739,
+      "grad_norm": 1.0391261342097122,
+      "learning_rate": 4.666428121849653e-06,
+      "loss": 0.6,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5696187683284457,
+      "grad_norm": 1.191446619321251,
+      "learning_rate": 4.662341410836703e-06,
+      "loss": 0.6285,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5698533724340176,
+      "grad_norm": 1.4999105564002244,
+      "learning_rate": 4.658254926416911e-06,
+      "loss": 0.62,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5700879765395894,
+      "grad_norm": 1.7234983917641036,
+      "learning_rate": 4.654168671332594e-06,
+      "loss": 0.6239,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5703225806451613,
+      "grad_norm": 1.0770283119340396,
+      "learning_rate": 4.6500826483259284e-06,
+      "loss": 0.6065,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5705571847507331,
+      "grad_norm": 1.2128209828917635,
+      "learning_rate": 4.645996860138928e-06,
+      "loss": 0.6046,
+      "step": 2432
+    },
+    {
+      "epoch": 0.570791788856305,
+      "grad_norm": 2.0413579361061887,
+      "learning_rate": 4.641911309513449e-06,
+      "loss": 0.636,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5710263929618768,
+      "grad_norm": 0.8811953036824909,
+      "learning_rate": 4.637825999191189e-06,
+      "loss": 0.6787,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5712609970674487,
+      "grad_norm": 1.426557418556202,
+      "learning_rate": 4.6337409319136865e-06,
+      "loss": 0.6398,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5714956011730206,
+      "grad_norm": 1.380859345024999,
+      "learning_rate": 4.629656110422311e-06,
+      "loss": 0.6174,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5717302052785924,
+      "grad_norm": 1.1169157694598355,
+      "learning_rate": 4.625571537458275e-06,
+      "loss": 0.6291,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5719648093841643,
+      "grad_norm": 1.2417537193839552,
+      "learning_rate": 4.621487215762619e-06,
+      "loss": 0.6316,
+      "step": 2438
+    },
+    {
+      "epoch": 0.572199413489736,
+      "grad_norm": 1.4795045752184326,
+      "learning_rate": 4.617403148076215e-06,
+      "loss": 0.5826,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5724340175953079,
+      "grad_norm": 0.8933352430022446,
+      "learning_rate": 4.613319337139767e-06,
+      "loss": 0.6526,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5726686217008797,
+      "grad_norm": 1.324711975821137,
+      "learning_rate": 4.609235785693805e-06,
+      "loss": 0.5706,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5729032258064516,
+      "grad_norm": 1.1085588036263778,
+      "learning_rate": 4.605152496478685e-06,
+      "loss": 0.6657,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5731378299120234,
+      "grad_norm": 1.0288656736675692,
+      "learning_rate": 4.601069472234584e-06,
+      "loss": 0.5887,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5733724340175953,
+      "grad_norm": 1.1064627616767901,
+      "learning_rate": 4.596986715701511e-06,
+      "loss": 0.6392,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5736070381231672,
+      "grad_norm": 7.134453408230588,
+      "learning_rate": 4.592904229619284e-06,
+      "loss": 0.5801,
+      "step": 2445
+    },
+    {
+      "epoch": 0.573841642228739,
+      "grad_norm": 3.627498858661479,
+      "learning_rate": 4.588822016727546e-06,
+      "loss": 0.6651,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5740762463343109,
+      "grad_norm": 2.127464440263173,
+      "learning_rate": 4.584740079765755e-06,
+      "loss": 0.6006,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5743108504398827,
+      "grad_norm": 2.535380292056328,
+      "learning_rate": 4.580658421473182e-06,
+      "loss": 0.6363,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5745454545454546,
+      "grad_norm": 1.263636915924277,
+      "learning_rate": 4.576577044588912e-06,
+      "loss": 0.5691,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5747800586510264,
+      "grad_norm": 3.8953432680279536,
+      "learning_rate": 4.572495951851848e-06,
+      "loss": 0.6474,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5750146627565983,
+      "grad_norm": 3.449525943312299,
+      "learning_rate": 4.568415146000689e-06,
+      "loss": 0.5945,
+      "step": 2451
+    },
+    {
+      "epoch": 0.57524926686217,
+      "grad_norm": 0.9085189847225632,
+      "learning_rate": 4.564334629773956e-06,
+      "loss": 0.5878,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5754838709677419,
+      "grad_norm": 1.4430234876118253,
+      "learning_rate": 4.560254405909959e-06,
+      "loss": 0.5796,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5757184750733138,
+      "grad_norm": 0.9715140944677086,
+      "learning_rate": 4.556174477146828e-06,
+      "loss": 0.6414,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5759530791788856,
+      "grad_norm": 1.3802544914953896,
+      "learning_rate": 4.552094846222487e-06,
+      "loss": 0.6364,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5761876832844575,
+      "grad_norm": 14.50739306268874,
+      "learning_rate": 4.548015515874661e-06,
+      "loss": 0.6201,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5764222873900293,
+      "grad_norm": 1.3123194340321647,
+      "learning_rate": 4.543936488840871e-06,
+      "loss": 0.6007,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5766568914956012,
+      "grad_norm": 1.5720636883531176,
+      "learning_rate": 4.539857767858442e-06,
+      "loss": 0.6532,
+      "step": 2458
+    },
+    {
+      "epoch": 0.576891495601173,
+      "grad_norm": 6.814471147180956,
+      "learning_rate": 4.5357793556644825e-06,
+      "loss": 0.6058,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5771260997067449,
+      "grad_norm": 2.249498514114243,
+      "learning_rate": 4.531701254995905e-06,
+      "loss": 0.6017,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5773607038123167,
+      "grad_norm": 2.5722080365660474,
+      "learning_rate": 4.527623468589406e-06,
+      "loss": 0.5677,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5775953079178886,
+      "grad_norm": 1.3189916081527315,
+      "learning_rate": 4.523545999181472e-06,
+      "loss": 0.5812,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5778299120234605,
+      "grad_norm": 1.6996214839000159,
+      "learning_rate": 4.519468849508379e-06,
+      "loss": 0.6036,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5780645161290323,
+      "grad_norm": 0.9931281950875175,
+      "learning_rate": 4.515392022306187e-06,
+      "loss": 0.5785,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5782991202346041,
+      "grad_norm": 1.4534131665761378,
+      "learning_rate": 4.511315520310738e-06,
+      "loss": 0.6852,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5785337243401759,
+      "grad_norm": 0.9868659889697355,
+      "learning_rate": 4.50723934625766e-06,
+      "loss": 0.6259,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5787683284457478,
+      "grad_norm": 2.6107327136802203,
+      "learning_rate": 4.503163502882359e-06,
+      "loss": 0.628,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5790029325513196,
+      "grad_norm": 1.0088657547611688,
+      "learning_rate": 4.4990879929200145e-06,
+      "loss": 0.6573,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5792375366568915,
+      "grad_norm": 1.453365422227254,
+      "learning_rate": 4.49501281910559e-06,
+      "loss": 0.6703,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5794721407624633,
+      "grad_norm": 0.8819630378676603,
+      "learning_rate": 4.490937984173818e-06,
+      "loss": 0.594,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5797067448680352,
+      "grad_norm": 4.032052328443642,
+      "learning_rate": 4.486863490859205e-06,
+      "loss": 0.614,
+      "step": 2471
+    },
+    {
+      "epoch": 0.579941348973607,
+      "grad_norm": 1.6205609214614662,
+      "learning_rate": 4.482789341896029e-06,
+      "loss": 0.6446,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5801759530791789,
+      "grad_norm": 1.7741084960213596,
+      "learning_rate": 4.47871554001834e-06,
+      "loss": 0.6264,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5804105571847508,
+      "grad_norm": 3.095282512692249,
+      "learning_rate": 4.474642087959944e-06,
+      "loss": 0.6742,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 1.165975209316844,
+      "learning_rate": 4.470568988454426e-06,
+      "loss": 0.6051,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5808797653958945,
+      "grad_norm": 6.365287398706099,
+      "learning_rate": 4.466496244235126e-06,
+      "loss": 0.5462,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5811143695014662,
+      "grad_norm": 1.0096029586962736,
+      "learning_rate": 4.4624238580351466e-06,
+      "loss": 0.6374,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5813489736070381,
+      "grad_norm": 1.198192719480048,
+      "learning_rate": 4.458351832587354e-06,
+      "loss": 0.6622,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5815835777126099,
+      "grad_norm": 1.669480946734029,
+      "learning_rate": 4.454280170624368e-06,
+      "loss": 0.613,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 1.0721974900794962,
+      "learning_rate": 4.4502088748785646e-06,
+      "loss": 0.5943,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5820527859237536,
+      "grad_norm": 0.7996800998165046,
+      "learning_rate": 4.446137948082081e-06,
+      "loss": 0.638,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5822873900293255,
+      "grad_norm": 1.1849513327296384,
+      "learning_rate": 4.442067392966794e-06,
+      "loss": 0.5961,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5825219941348974,
+      "grad_norm": 2.3522109536517175,
+      "learning_rate": 4.437997212264343e-06,
+      "loss": 0.6678,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5827565982404692,
+      "grad_norm": 1.2178400640964204,
+      "learning_rate": 4.4339274087061116e-06,
+      "loss": 0.6717,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5829912023460411,
+      "grad_norm": 2.640290682903799,
+      "learning_rate": 4.4298579850232274e-06,
+      "loss": 0.6418,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5832258064516129,
+      "grad_norm": 10.345703317890523,
+      "learning_rate": 4.425788943946567e-06,
+      "loss": 0.6121,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5834604105571848,
+      "grad_norm": 6.764934192433235,
+      "learning_rate": 4.421720288206752e-06,
+      "loss": 0.6012,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5836950146627566,
+      "grad_norm": 0.9387854031509927,
+      "learning_rate": 4.4176520205341365e-06,
+      "loss": 0.609,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5839296187683285,
+      "grad_norm": 0.9911978936260616,
+      "learning_rate": 4.413584143658826e-06,
+      "loss": 0.6379,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5841642228739002,
+      "grad_norm": 1.366898656153631,
+      "learning_rate": 4.409516660310655e-06,
+      "loss": 0.6191,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5843988269794721,
+      "grad_norm": 1.0545558254205785,
+      "learning_rate": 4.405449573219198e-06,
+      "loss": 0.6498,
+      "step": 2491
+    },
+    {
+      "epoch": 0.584633431085044,
+      "grad_norm": 3.351501696016705,
+      "learning_rate": 4.401382885113761e-06,
+      "loss": 0.635,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5848680351906158,
+      "grad_norm": 1.1648565999529112,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.5701,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5851026392961877,
+      "grad_norm": 1.1896548676159913,
+      "learning_rate": 4.3932507167768405e-06,
+      "loss": 0.6399,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5853372434017595,
+      "grad_norm": 1.519452926276841,
+      "learning_rate": 4.389185242002622e-06,
+      "loss": 0.6113,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5855718475073314,
+      "grad_norm": 1.4964823361045552,
+      "learning_rate": 4.385120177128963e-06,
+      "loss": 0.612,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5858064516129032,
+      "grad_norm": 1.0272462888998013,
+      "learning_rate": 4.381055524883806e-06,
+      "loss": 0.607,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5860410557184751,
+      "grad_norm": 2.199327762237035,
+      "learning_rate": 4.376991287994829e-06,
+      "loss": 0.6191,
+      "step": 2498
+    },
+    {
+      "epoch": 0.586275659824047,
+      "grad_norm": 1.0706737172415943,
+      "learning_rate": 4.372927469189425e-06,
+      "loss": 0.6234,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5865102639296188,
+      "grad_norm": 1.3477826341717938,
+      "learning_rate": 4.368864071194709e-06,
+      "loss": 0.6277,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5865102639296188,
+      "eval_loss": 0.6218044757843018,
+      "eval_runtime": 38.7474,
+      "eval_samples_per_second": 14.091,
+      "eval_steps_per_second": 0.129,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5867448680351907,
+      "grad_norm": 1.43454169755497,
+      "learning_rate": 4.3648010967375114e-06,
+      "loss": 0.675,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5869794721407625,
+      "grad_norm": 1.6132885857656545,
+      "learning_rate": 4.3607385485443805e-06,
+      "loss": 0.6075,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5872140762463343,
+      "grad_norm": 1.9554429793995742,
+      "learning_rate": 4.356676429341577e-06,
+      "loss": 0.6351,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5874486803519061,
+      "grad_norm": 0.9729149313289474,
+      "learning_rate": 4.352614741855079e-06,
+      "loss": 0.6028,
+      "step": 2504
+    },
+    {
+      "epoch": 0.587683284457478,
+      "grad_norm": 1.1783962682160853,
+      "learning_rate": 4.348553488810563e-06,
+      "loss": 0.6051,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5879178885630498,
+      "grad_norm": 0.8412586023405613,
+      "learning_rate": 4.344492672933427e-06,
+      "loss": 0.623,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5881524926686217,
+      "grad_norm": 8.623322800244942,
+      "learning_rate": 4.340432296948769e-06,
+      "loss": 0.6421,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5883870967741935,
+      "grad_norm": 1.6358070714325266,
+      "learning_rate": 4.336372363581391e-06,
+      "loss": 0.6101,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5886217008797654,
+      "grad_norm": 0.9504336081958826,
+      "learning_rate": 4.3323128755557995e-06,
+      "loss": 0.6325,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5888563049853373,
+      "grad_norm": 4.554846572918646,
+      "learning_rate": 4.328253835596207e-06,
+      "loss": 0.6629,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5890909090909091,
+      "grad_norm": 1.0740787205585829,
+      "learning_rate": 4.324195246426515e-06,
+      "loss": 0.6084,
+      "step": 2511
+    },
+    {
+      "epoch": 0.589325513196481,
+      "grad_norm": 1.8572210056231553,
+      "learning_rate": 4.32013711077033e-06,
+      "loss": 0.6289,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5895601173020528,
+      "grad_norm": 1.1658183387450498,
+      "learning_rate": 4.316079431350956e-06,
+      "loss": 0.5882,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5897947214076247,
+      "grad_norm": 1.2400853161647338,
+      "learning_rate": 4.312022210891381e-06,
+      "loss": 0.6151,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5900293255131965,
+      "grad_norm": 1.0979577190626502,
+      "learning_rate": 4.307965452114296e-06,
+      "loss": 0.5982,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5902639296187683,
+      "grad_norm": 30.268023563133838,
+      "learning_rate": 4.303909157742077e-06,
+      "loss": 0.5786,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5904985337243401,
+      "grad_norm": 3.0804218435920703,
+      "learning_rate": 4.299853330496784e-06,
+      "loss": 0.6121,
+      "step": 2517
+    },
+    {
+      "epoch": 0.590733137829912,
+      "grad_norm": 1.7700600677022102,
+      "learning_rate": 4.295797973100174e-06,
+      "loss": 0.5792,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5909677419354838,
+      "grad_norm": 1.4165185996015008,
+      "learning_rate": 4.291743088273681e-06,
+      "loss": 0.6553,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5912023460410557,
+      "grad_norm": 1.1984550821123752,
+      "learning_rate": 4.287688678738423e-06,
+      "loss": 0.6573,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5914369501466276,
+      "grad_norm": 1.4132918562457883,
+      "learning_rate": 4.283634747215202e-06,
+      "loss": 0.5815,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5916715542521994,
+      "grad_norm": 3.449582328145338,
+      "learning_rate": 4.2795812964244935e-06,
+      "loss": 0.661,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5919061583577713,
+      "grad_norm": 1.2157965402261237,
+      "learning_rate": 4.275528329086457e-06,
+      "loss": 0.6723,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5921407624633431,
+      "grad_norm": 1.1385200395220776,
+      "learning_rate": 4.271475847920924e-06,
+      "loss": 0.6041,
+      "step": 2524
+    },
+    {
+      "epoch": 0.592375366568915,
+      "grad_norm": 1.9621492288963664,
+      "learning_rate": 4.267423855647399e-06,
+      "loss": 0.6661,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5926099706744868,
+      "grad_norm": 1.4433105370089916,
+      "learning_rate": 4.2633723549850595e-06,
+      "loss": 0.6671,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5928445747800587,
+      "grad_norm": 4.715698963848849,
+      "learning_rate": 4.259321348652757e-06,
+      "loss": 0.6528,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5930791788856306,
+      "grad_norm": 1.820896037731156,
+      "learning_rate": 4.2552708393690035e-06,
+      "loss": 0.692,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5933137829912023,
+      "grad_norm": 1.3980967432063924,
+      "learning_rate": 4.251220829851983e-06,
+      "loss": 0.5905,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5935483870967742,
+      "grad_norm": 1.4592183597296384,
+      "learning_rate": 4.247171322819544e-06,
+      "loss": 0.6232,
+      "step": 2530
+    },
+    {
+      "epoch": 0.593782991202346,
+      "grad_norm": 1.6652890905059732,
+      "learning_rate": 4.243122320989194e-06,
+      "loss": 0.5506,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5940175953079179,
+      "grad_norm": 0.890755420015173,
+      "learning_rate": 4.239073827078103e-06,
+      "loss": 0.6309,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5942521994134897,
+      "grad_norm": 2.132100196010778,
+      "learning_rate": 4.2350258438031075e-06,
+      "loss": 0.6081,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5944868035190616,
+      "grad_norm": 1.3237668202449755,
+      "learning_rate": 4.230978373880687e-06,
+      "loss": 0.5812,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5947214076246334,
+      "grad_norm": 1.1462967036815435,
+      "learning_rate": 4.226931420026989e-06,
+      "loss": 0.5801,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5949560117302053,
+      "grad_norm": 3.0768367275227977,
+      "learning_rate": 4.22288498495781e-06,
+      "loss": 0.6285,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5951906158357771,
+      "grad_norm": 0.9785169841442185,
+      "learning_rate": 4.218839071388597e-06,
+      "loss": 0.6169,
+      "step": 2537
+    },
+    {
+      "epoch": 0.595425219941349,
+      "grad_norm": 1.561503407894026,
+      "learning_rate": 4.2147936820344484e-06,
+      "loss": 0.6483,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5956598240469209,
+      "grad_norm": 2.081519039084864,
+      "learning_rate": 4.210748819610112e-06,
+      "loss": 0.6184,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5958944281524927,
+      "grad_norm": 1.4292431629296551,
+      "learning_rate": 4.206704486829978e-06,
+      "loss": 0.6574,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5961290322580645,
+      "grad_norm": 1.744841612567776,
+      "learning_rate": 4.202660686408088e-06,
+      "loss": 0.6214,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5963636363636363,
+      "grad_norm": 1.1919779137620403,
+      "learning_rate": 4.198617421058122e-06,
+      "loss": 0.6328,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5965982404692082,
+      "grad_norm": 4.436466065782175,
+      "learning_rate": 4.194574693493398e-06,
+      "loss": 0.6622,
+      "step": 2543
+    },
+    {
+      "epoch": 0.59683284457478,
+      "grad_norm": 3.042941235668859,
+      "learning_rate": 4.19053250642688e-06,
+      "loss": 0.6053,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5970674486803519,
+      "grad_norm": 0.767431961578569,
+      "learning_rate": 4.186490862571164e-06,
+      "loss": 0.5862,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5973020527859237,
+      "grad_norm": 2.2513791519685866,
+      "learning_rate": 4.182449764638483e-06,
+      "loss": 0.5453,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5975366568914956,
+      "grad_norm": 1.3737399402284083,
+      "learning_rate": 4.178409215340704e-06,
+      "loss": 0.5814,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5977712609970675,
+      "grad_norm": 1.055379340574401,
+      "learning_rate": 4.174369217389326e-06,
+      "loss": 0.613,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5980058651026393,
+      "grad_norm": 1.2827478647327584,
+      "learning_rate": 4.170329773495477e-06,
+      "loss": 0.6532,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5982404692082112,
+      "grad_norm": 1.4526147471149213,
+      "learning_rate": 4.166290886369916e-06,
+      "loss": 0.6089,
+      "step": 2550
+    },
+    {
+      "epoch": 0.598475073313783,
+      "grad_norm": 1.2719172270318275,
+      "learning_rate": 4.162252558723023e-06,
+      "loss": 0.6736,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5987096774193549,
+      "grad_norm": 1.3665798570895367,
+      "learning_rate": 4.158214793264808e-06,
+      "loss": 0.5972,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5989442815249267,
+      "grad_norm": 0.9995262318034842,
+      "learning_rate": 4.154177592704902e-06,
+      "loss": 0.6147,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5991788856304985,
+      "grad_norm": 1.3115689334601,
+      "learning_rate": 4.1501409597525535e-06,
+      "loss": 0.6099,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5994134897360703,
+      "grad_norm": 2.5755184545949765,
+      "learning_rate": 4.146104897116634e-06,
+      "loss": 0.589,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5996480938416422,
+      "grad_norm": 1.188197608601526,
+      "learning_rate": 4.1420694075056365e-06,
+      "loss": 0.6492,
+      "step": 2556
+    },
+    {
+      "epoch": 0.599882697947214,
+      "grad_norm": 1.7042100888826746,
+      "learning_rate": 4.138034493627655e-06,
+      "loss": 0.5418,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6001173020527859,
+      "grad_norm": 0.8867087212357306,
+      "learning_rate": 4.1340001581904135e-06,
+      "loss": 0.626,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6003519061583578,
+      "grad_norm": 2.8658080054154085,
+      "learning_rate": 4.1299664039012384e-06,
+      "loss": 0.5942,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6005865102639296,
+      "grad_norm": 1.2862082960348973,
+      "learning_rate": 4.125933233467069e-06,
+      "loss": 0.6224,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6008211143695015,
+      "grad_norm": 1.1332587290115137,
+      "learning_rate": 4.12190064959445e-06,
+      "loss": 0.589,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6010557184750733,
+      "grad_norm": 1.536860711564503,
+      "learning_rate": 4.1178686549895395e-06,
+      "loss": 0.6411,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6012903225806452,
+      "grad_norm": 4.299445647549985,
+      "learning_rate": 4.113837252358089e-06,
+      "loss": 0.6219,
+      "step": 2563
+    },
+    {
+      "epoch": 0.601524926686217,
+      "grad_norm": 1.138103697486042,
+      "learning_rate": 4.109806444405463e-06,
+      "loss": 0.6009,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6017595307917889,
+      "grad_norm": 0.9390309637600313,
+      "learning_rate": 4.105776233836624e-06,
+      "loss": 0.6466,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6019941348973608,
+      "grad_norm": 1.1217142386900871,
+      "learning_rate": 4.10174662335613e-06,
+      "loss": 0.6242,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6022287390029325,
+      "grad_norm": 19.912152061851284,
+      "learning_rate": 4.097717615668139e-06,
+      "loss": 0.6174,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6024633431085044,
+      "grad_norm": 1.6585648907036556,
+      "learning_rate": 4.093689213476408e-06,
+      "loss": 0.6203,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6026979472140762,
+      "grad_norm": 1.2833892554078743,
+      "learning_rate": 4.089661419484279e-06,
+      "loss": 0.5992,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6029325513196481,
+      "grad_norm": 0.7427174733322811,
+      "learning_rate": 4.085634236394696e-06,
+      "loss": 0.6039,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6031671554252199,
+      "grad_norm": 1.9360940074646746,
+      "learning_rate": 4.0816076669101815e-06,
+      "loss": 0.6306,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6034017595307918,
+      "grad_norm": 1.0696471538576655,
+      "learning_rate": 4.077581713732859e-06,
+      "loss": 0.6405,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6036363636363636,
+      "grad_norm": 1.206946854193182,
+      "learning_rate": 4.073556379564429e-06,
+      "loss": 0.628,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6038709677419355,
+      "grad_norm": 3.042576436469917,
+      "learning_rate": 4.069531667106181e-06,
+      "loss": 0.5874,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6041055718475073,
+      "grad_norm": 2.3226578612064097,
+      "learning_rate": 4.065507579058984e-06,
+      "loss": 0.566,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6043401759530792,
+      "grad_norm": 1.205886716802571,
+      "learning_rate": 4.0614841181232935e-06,
+      "loss": 0.6125,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6045747800586511,
+      "grad_norm": 1.7443724384894912,
+      "learning_rate": 4.057461286999137e-06,
+      "loss": 0.6886,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6048093841642229,
+      "grad_norm": 0.9805606507048945,
+      "learning_rate": 4.053439088386124e-06,
+      "loss": 0.6146,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6050439882697948,
+      "grad_norm": 1.211593514855346,
+      "learning_rate": 4.04941752498344e-06,
+      "loss": 0.5793,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6052785923753665,
+      "grad_norm": 1.4647559318014667,
+      "learning_rate": 4.045396599489841e-06,
+      "loss": 0.6786,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6055131964809384,
+      "grad_norm": 1.40486952361013,
+      "learning_rate": 4.041376314603659e-06,
+      "loss": 0.5964,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6057478005865102,
+      "grad_norm": 0.8290983501422904,
+      "learning_rate": 4.037356673022792e-06,
+      "loss": 0.6164,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6059824046920821,
+      "grad_norm": 1.5157273495726928,
+      "learning_rate": 4.033337677444708e-06,
+      "loss": 0.6199,
+      "step": 2583
+    },
+    {
+      "epoch": 0.606217008797654,
+      "grad_norm": 2.9508326712613395,
+      "learning_rate": 4.02931933056644e-06,
+      "loss": 0.6085,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6064516129032258,
+      "grad_norm": 1.0947176978306004,
+      "learning_rate": 4.025301635084592e-06,
+      "loss": 0.6444,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6066862170087977,
+      "grad_norm": 2.5767801974845717,
+      "learning_rate": 4.021284593695321e-06,
+      "loss": 0.6482,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6069208211143695,
+      "grad_norm": 1.3068368441858098,
+      "learning_rate": 4.017268209094353e-06,
+      "loss": 0.6913,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6071554252199414,
+      "grad_norm": 2.755890994447317,
+      "learning_rate": 4.013252483976972e-06,
+      "loss": 0.6012,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6073900293255132,
+      "grad_norm": 1.2940935368296023,
+      "learning_rate": 4.009237421038013e-06,
+      "loss": 0.6162,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6076246334310851,
+      "grad_norm": 1.5107803665701236,
+      "learning_rate": 4.0052230229718755e-06,
+      "loss": 0.6479,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6078592375366569,
+      "grad_norm": 0.9096356154205831,
+      "learning_rate": 4.001209292472507e-06,
+      "loss": 0.6229,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6080938416422288,
+      "grad_norm": 1.6646603414639927,
+      "learning_rate": 3.997196232233409e-06,
+      "loss": 0.6291,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6083284457478005,
+      "grad_norm": 1.2012478773424764,
+      "learning_rate": 3.9931838449476355e-06,
+      "loss": 0.6158,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6085630498533724,
+      "grad_norm": 3.208605929111589,
+      "learning_rate": 3.989172133307782e-06,
+      "loss": 0.6519,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6087976539589443,
+      "grad_norm": 0.9172720961577118,
+      "learning_rate": 3.985161100005997e-06,
+      "loss": 0.5433,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6090322580645161,
+      "grad_norm": 2.5108512772562976,
+      "learning_rate": 3.981150747733975e-06,
+      "loss": 0.6582,
+      "step": 2596
+    },
+    {
+      "epoch": 0.609266862170088,
+      "grad_norm": 1.74396930458251,
+      "learning_rate": 3.977141079182945e-06,
+      "loss": 0.6168,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6095014662756598,
+      "grad_norm": 1.3166982183851292,
+      "learning_rate": 3.973132097043685e-06,
+      "loss": 0.6899,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6097360703812317,
+      "grad_norm": 1.2046832346059244,
+      "learning_rate": 3.9691238040065105e-06,
+      "loss": 0.6051,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6099706744868035,
+      "grad_norm": 0.9358456079586567,
+      "learning_rate": 3.965116202761271e-06,
+      "loss": 0.625,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6102052785923754,
+      "grad_norm": 1.4680245628396122,
+      "learning_rate": 3.961109295997359e-06,
+      "loss": 0.5869,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6104398826979472,
+      "grad_norm": 2.2864639472629413,
+      "learning_rate": 3.957103086403694e-06,
+      "loss": 0.6568,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6106744868035191,
+      "grad_norm": 1.3100229042611278,
+      "learning_rate": 3.95309757666873e-06,
+      "loss": 0.6953,
+      "step": 2603
+    },
+    {
+      "epoch": 0.610909090909091,
+      "grad_norm": 1.589692790494812,
+      "learning_rate": 3.949092769480453e-06,
+      "loss": 0.5994,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6111436950146627,
+      "grad_norm": 2.239815470993383,
+      "learning_rate": 3.945088667526375e-06,
+      "loss": 0.5607,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6113782991202346,
+      "grad_norm": 1.546349756698268,
+      "learning_rate": 3.941085273493536e-06,
+      "loss": 0.6097,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6116129032258064,
+      "grad_norm": 3.09397743068729,
+      "learning_rate": 3.9370825900685e-06,
+      "loss": 0.7011,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6118475073313783,
+      "grad_norm": 1.585717896104532,
+      "learning_rate": 3.9330806199373595e-06,
+      "loss": 0.5706,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6120821114369501,
+      "grad_norm": 1.400585147618227,
+      "learning_rate": 3.929079365785716e-06,
+      "loss": 0.6331,
+      "step": 2609
+    },
+    {
+      "epoch": 0.612316715542522,
+      "grad_norm": 1.2322264495654516,
+      "learning_rate": 3.925078830298704e-06,
+      "loss": 0.6274,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6125513196480938,
+      "grad_norm": 1.4124526796440844,
+      "learning_rate": 3.92107901616097e-06,
+      "loss": 0.6103,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6127859237536657,
+      "grad_norm": 1.5300209305905332,
+      "learning_rate": 3.917079926056674e-06,
+      "loss": 0.6342,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6130205278592376,
+      "grad_norm": 1.2017607539252948,
+      "learning_rate": 3.913081562669492e-06,
+      "loss": 0.584,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6132551319648094,
+      "grad_norm": 1.171386657241352,
+      "learning_rate": 3.9090839286826185e-06,
+      "loss": 0.611,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6134897360703813,
+      "grad_norm": 1.175493331648172,
+      "learning_rate": 3.9050870267787446e-06,
+      "loss": 0.6258,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6137243401759531,
+      "grad_norm": 0.8381430034876453,
+      "learning_rate": 3.9010908596400865e-06,
+      "loss": 0.6404,
+      "step": 2616
+    },
+    {
+      "epoch": 0.613958944281525,
+      "grad_norm": 1.3056645293670937,
+      "learning_rate": 3.897095429948352e-06,
+      "loss": 0.6062,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6141935483870967,
+      "grad_norm": 0.9232357664815262,
+      "learning_rate": 3.893100740384766e-06,
+      "loss": 0.5713,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6144281524926686,
+      "grad_norm": 1.9520853215963634,
+      "learning_rate": 3.889106793630052e-06,
+      "loss": 0.6338,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6146627565982404,
+      "grad_norm": 1.4988606630842436,
+      "learning_rate": 3.885113592364432e-06,
+      "loss": 0.6179,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6148973607038123,
+      "grad_norm": 1.465478839792279,
+      "learning_rate": 3.881121139267632e-06,
+      "loss": 0.5571,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6151319648093841,
+      "grad_norm": 1.37075089168491,
+      "learning_rate": 3.8771294370188775e-06,
+      "loss": 0.6586,
+      "step": 2622
+    },
+    {
+      "epoch": 0.615366568914956,
+      "grad_norm": 1.1861832507325312,
+      "learning_rate": 3.873138488296883e-06,
+      "loss": 0.5717,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6156011730205279,
+      "grad_norm": 1.0975220488700843,
+      "learning_rate": 3.8691482957798636e-06,
+      "loss": 0.6611,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6158357771260997,
+      "grad_norm": 0.9290249703287269,
+      "learning_rate": 3.865158862145525e-06,
+      "loss": 0.5995,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6160703812316716,
+      "grad_norm": 0.9140054772375495,
+      "learning_rate": 3.861170190071064e-06,
+      "loss": 0.6243,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6163049853372434,
+      "grad_norm": 1.2375783218924745,
+      "learning_rate": 3.8571822822331644e-06,
+      "loss": 0.6148,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6165395894428153,
+      "grad_norm": 5.314375888137883,
+      "learning_rate": 3.853195141308001e-06,
+      "loss": 0.6104,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6167741935483871,
+      "grad_norm": 2.2260423031861105,
+      "learning_rate": 3.8492087699712294e-06,
+      "loss": 0.6345,
+      "step": 2629
+    },
+    {
+      "epoch": 0.617008797653959,
+      "grad_norm": 1.2388179244531388,
+      "learning_rate": 3.8452231708979905e-06,
+      "loss": 0.6356,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6172434017595307,
+      "grad_norm": 1.474997299256861,
+      "learning_rate": 3.8412383467629104e-06,
+      "loss": 0.6138,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6174780058651026,
+      "grad_norm": 1.4157998917734198,
+      "learning_rate": 3.83725430024009e-06,
+      "loss": 0.5826,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6177126099706745,
+      "grad_norm": 1.086677811735564,
+      "learning_rate": 3.833271034003111e-06,
+      "loss": 0.6335,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6179472140762463,
+      "grad_norm": 0.9606402467854953,
+      "learning_rate": 3.8292885507250315e-06,
+      "loss": 0.6142,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6181818181818182,
+      "grad_norm": 1.029850892655021,
+      "learning_rate": 3.8253068530783825e-06,
+      "loss": 0.6191,
+      "step": 2635
+    },
+    {
+      "epoch": 0.61841642228739,
+      "grad_norm": 2.030545771516275,
+      "learning_rate": 3.8213259437351695e-06,
+      "loss": 0.5957,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6186510263929619,
+      "grad_norm": 0.8472838413106262,
+      "learning_rate": 3.817345825366871e-06,
+      "loss": 0.5828,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6188856304985337,
+      "grad_norm": 4.825714667726455,
+      "learning_rate": 3.813366500644426e-06,
+      "loss": 0.5783,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6191202346041056,
+      "grad_norm": 3.2736180582520293,
+      "learning_rate": 3.809387972238255e-06,
+      "loss": 0.6472,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6193548387096774,
+      "grad_norm": 2.0847739365669073,
+      "learning_rate": 3.8054102428182284e-06,
+      "loss": 0.5926,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6195894428152493,
+      "grad_norm": 2.13526093103931,
+      "learning_rate": 3.801433315053693e-06,
+      "loss": 0.6315,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6198240469208212,
+      "grad_norm": 1.25259446518174,
+      "learning_rate": 3.7974571916134533e-06,
+      "loss": 0.6209,
+      "step": 2642
+    },
+    {
+      "epoch": 0.620058651026393,
+      "grad_norm": 1.5747100024152387,
+      "learning_rate": 3.793481875165771e-06,
+      "loss": 0.6398,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6202932551319648,
+      "grad_norm": 0.9601714494423652,
+      "learning_rate": 3.7895073683783683e-06,
+      "loss": 0.6226,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6205278592375366,
+      "grad_norm": 1.506069025950793,
+      "learning_rate": 3.7855336739184303e-06,
+      "loss": 0.611,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6207624633431085,
+      "grad_norm": 1.1899295416433049,
+      "learning_rate": 3.7815607944525826e-06,
+      "loss": 0.5866,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6209970674486803,
+      "grad_norm": 1.8749969036584033,
+      "learning_rate": 3.777588732646919e-06,
+      "loss": 0.596,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6212316715542522,
+      "grad_norm": 1.792002552963097,
+      "learning_rate": 3.7736174911669775e-06,
+      "loss": 0.5952,
+      "step": 2648
+    },
+    {
+      "epoch": 0.621466275659824,
+      "grad_norm": 1.0404897387789034,
+      "learning_rate": 3.7696470726777436e-06,
+      "loss": 0.6029,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6217008797653959,
+      "grad_norm": 1.7960891351545873,
+      "learning_rate": 3.7656774798436545e-06,
+      "loss": 0.6562,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6219354838709678,
+      "grad_norm": 1.6596318518513833,
+      "learning_rate": 3.761708715328593e-06,
+      "loss": 0.611,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6221700879765396,
+      "grad_norm": 0.9896201383882054,
+      "learning_rate": 3.7577407817958815e-06,
+      "loss": 0.6461,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6224046920821115,
+      "grad_norm": 1.739289988297414,
+      "learning_rate": 3.7537736819082926e-06,
+      "loss": 0.6074,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6226392961876833,
+      "grad_norm": 1.1191647740135733,
+      "learning_rate": 3.7498074183280344e-06,
+      "loss": 0.6547,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6228739002932552,
+      "grad_norm": 1.874952600529361,
+      "learning_rate": 3.7458419937167516e-06,
+      "loss": 0.6325,
+      "step": 2655
+    },
+    {
+      "epoch": 0.623108504398827,
+      "grad_norm": 2.979617394432184,
+      "learning_rate": 3.7418774107355318e-06,
+      "loss": 0.6199,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6233431085043988,
+      "grad_norm": 1.1896527781829702,
+      "learning_rate": 3.737913672044894e-06,
+      "loss": 0.6108,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6235777126099706,
+      "grad_norm": 0.8829346383219321,
+      "learning_rate": 3.73395078030479e-06,
+      "loss": 0.5705,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6238123167155425,
+      "grad_norm": 0.7062082275621552,
+      "learning_rate": 3.7299887381746047e-06,
+      "loss": 0.5735,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6240469208211143,
+      "grad_norm": 1.221847445565343,
+      "learning_rate": 3.726027548313157e-06,
+      "loss": 0.597,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6242815249266862,
+      "grad_norm": 2.0902160378726014,
+      "learning_rate": 3.722067213378683e-06,
+      "loss": 0.6836,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6245161290322581,
+      "grad_norm": 1.377813196198677,
+      "learning_rate": 3.718107736028858e-06,
+      "loss": 0.6921,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6247507331378299,
+      "grad_norm": 2.4337369135631692,
+      "learning_rate": 3.7141491189207716e-06,
+      "loss": 0.6248,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6249853372434018,
+      "grad_norm": 1.2767180243430922,
+      "learning_rate": 3.7101913647109415e-06,
+      "loss": 0.6024,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6252199413489736,
+      "grad_norm": 1.82351491486182,
+      "learning_rate": 3.706234476055306e-06,
+      "loss": 0.6211,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6254545454545455,
+      "grad_norm": 1.0719908652535732,
+      "learning_rate": 3.702278455609219e-06,
+      "loss": 0.6546,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6256891495601173,
+      "grad_norm": 1.243280748149877,
+      "learning_rate": 3.6983233060274543e-06,
+      "loss": 0.6106,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6259237536656892,
+      "grad_norm": 2.1666429604577084,
+      "learning_rate": 3.6943690299642055e-06,
+      "loss": 0.6072,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6261583577712609,
+      "grad_norm": 0.917414355692659,
+      "learning_rate": 3.6904156300730697e-06,
+      "loss": 0.6379,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6263929618768328,
+      "grad_norm": 1.0762156591178333,
+      "learning_rate": 3.6864631090070656e-06,
+      "loss": 0.6422,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6266275659824047,
+      "grad_norm": 0.9453096475901798,
+      "learning_rate": 3.6825114694186192e-06,
+      "loss": 0.5849,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6268621700879765,
+      "grad_norm": 1.5752552955005232,
+      "learning_rate": 3.6785607139595625e-06,
+      "loss": 0.6535,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6270967741935484,
+      "grad_norm": 1.6758441060756903,
+      "learning_rate": 3.6746108452811347e-06,
+      "loss": 0.6612,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6273313782991202,
+      "grad_norm": 1.7082489282247946,
+      "learning_rate": 3.670661866033988e-06,
+      "loss": 0.5804,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6275659824046921,
+      "grad_norm": 0.9346939807372437,
+      "learning_rate": 3.666713778868161e-06,
+      "loss": 0.6119,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6278005865102639,
+      "grad_norm": 1.3277551451666765,
+      "learning_rate": 3.662766586433111e-06,
+      "loss": 0.6514,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6280351906158358,
+      "grad_norm": 2.6067439559867163,
+      "learning_rate": 3.6588202913776865e-06,
+      "loss": 0.6681,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6282697947214076,
+      "grad_norm": 3.3067083631174516,
+      "learning_rate": 3.6548748963501324e-06,
+      "loss": 0.5825,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6285043988269795,
+      "grad_norm": 2.4946271815287755,
+      "learning_rate": 3.650930403998093e-06,
+      "loss": 0.6368,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6287390029325514,
+      "grad_norm": 3.530727381232086,
+      "learning_rate": 3.646986816968608e-06,
+      "loss": 0.6473,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6289736070381232,
+      "grad_norm": 1.2456120856690516,
+      "learning_rate": 3.643044137908104e-06,
+      "loss": 0.5802,
+      "step": 2681
+    },
+    {
+      "epoch": 0.629208211143695,
+      "grad_norm": 1.5535352093408097,
+      "learning_rate": 3.6391023694624017e-06,
+      "loss": 0.638,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6294428152492668,
+      "grad_norm": 3.621896711457749,
+      "learning_rate": 3.635161514276715e-06,
+      "loss": 0.5986,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6296774193548387,
+      "grad_norm": 1.2081206262423436,
+      "learning_rate": 3.631221574995637e-06,
+      "loss": 0.6501,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6299120234604105,
+      "grad_norm": 1.6672839582985357,
+      "learning_rate": 3.6272825542631507e-06,
+      "loss": 0.5878,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6301466275659824,
+      "grad_norm": 10.538271916312206,
+      "learning_rate": 3.6233444547226214e-06,
+      "loss": 0.663,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6303812316715542,
+      "grad_norm": 1.0090407633007583,
+      "learning_rate": 3.6194072790167978e-06,
+      "loss": 0.5989,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6306158357771261,
+      "grad_norm": 1.3637371746508158,
+      "learning_rate": 3.615471029787807e-06,
+      "loss": 0.6095,
+      "step": 2688
+    },
+    {
+      "epoch": 0.630850439882698,
+      "grad_norm": 0.8850365791866931,
+      "learning_rate": 3.6115357096771536e-06,
+      "loss": 0.5444,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6310850439882698,
+      "grad_norm": 1.0810936007043943,
+      "learning_rate": 3.6076013213257198e-06,
+      "loss": 0.5888,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6313196480938417,
+      "grad_norm": 1.2337870138836586,
+      "learning_rate": 3.6036678673737668e-06,
+      "loss": 0.5531,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6315542521994135,
+      "grad_norm": 1.0824233995302337,
+      "learning_rate": 3.599735350460919e-06,
+      "loss": 0.6177,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6317888563049854,
+      "grad_norm": 1.223039780177105,
+      "learning_rate": 3.5958037732261806e-06,
+      "loss": 0.5734,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6320234604105572,
+      "grad_norm": 1.2296995133189876,
+      "learning_rate": 3.5918731383079224e-06,
+      "loss": 0.5612,
+      "step": 2694
+    },
+    {
+      "epoch": 0.632258064516129,
+      "grad_norm": 1.06663436020491,
+      "learning_rate": 3.5879434483438814e-06,
+      "loss": 0.6859,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6324926686217008,
+      "grad_norm": 1.1289573518711966,
+      "learning_rate": 3.5840147059711603e-06,
+      "loss": 0.6033,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6327272727272727,
+      "grad_norm": 2.1740518436103193,
+      "learning_rate": 3.5800869138262333e-06,
+      "loss": 0.5861,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6329618768328445,
+      "grad_norm": 1.127612472110948,
+      "learning_rate": 3.5761600745449234e-06,
+      "loss": 0.6525,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6331964809384164,
+      "grad_norm": 1.1748248943049848,
+      "learning_rate": 3.572234190762426e-06,
+      "loss": 0.6545,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6334310850439883,
+      "grad_norm": 1.1353027900528645,
+      "learning_rate": 3.568309265113291e-06,
+      "loss": 0.6244,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6336656891495601,
+      "grad_norm": 1.7665864864207614,
+      "learning_rate": 3.564385300231423e-06,
+      "loss": 0.6311,
+      "step": 2701
+    },
+    {
+      "epoch": 0.633900293255132,
+      "grad_norm": 1.3127731536976572,
+      "learning_rate": 3.5604622987500857e-06,
+      "loss": 0.6428,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6341348973607038,
+      "grad_norm": 0.993273922561666,
+      "learning_rate": 3.5565402633018963e-06,
+      "loss": 0.6029,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6343695014662757,
+      "grad_norm": 0.9025340222281976,
+      "learning_rate": 3.5526191965188173e-06,
+      "loss": 0.6261,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6346041055718475,
+      "grad_norm": 1.4845181882160094,
+      "learning_rate": 3.5486991010321703e-06,
+      "loss": 0.644,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6348387096774194,
+      "grad_norm": 1.4027983004146045,
+      "learning_rate": 3.5447799794726213e-06,
+      "loss": 0.6432,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6350733137829913,
+      "grad_norm": 1.2481329067569296,
+      "learning_rate": 3.5408618344701783e-06,
+      "loss": 0.6242,
+      "step": 2707
+    },
+    {
+      "epoch": 0.635307917888563,
+      "grad_norm": 0.9977286075070906,
+      "learning_rate": 3.536944668654202e-06,
+      "loss": 0.6417,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6355425219941349,
+      "grad_norm": 1.4098047628922843,
+      "learning_rate": 3.5330284846533892e-06,
+      "loss": 0.6009,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6357771260997067,
+      "grad_norm": 3.4440858861484824,
+      "learning_rate": 3.529113285095781e-06,
+      "loss": 0.6783,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6360117302052786,
+      "grad_norm": 1.1546959945770436,
+      "learning_rate": 3.525199072608758e-06,
+      "loss": 0.5936,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6362463343108504,
+      "grad_norm": 1.9316549570873787,
+      "learning_rate": 3.521285849819037e-06,
+      "loss": 0.6582,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6364809384164223,
+      "grad_norm": 1.4124539147742596,
+      "learning_rate": 3.517373619352672e-06,
+      "loss": 0.6443,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6367155425219941,
+      "grad_norm": 1.5754658740131884,
+      "learning_rate": 3.513462383835051e-06,
+      "loss": 0.6192,
+      "step": 2714
+    },
+    {
+      "epoch": 0.636950146627566,
+      "grad_norm": 4.357016819681828,
+      "learning_rate": 3.5095521458908933e-06,
+      "loss": 0.6045,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6371847507331379,
+      "grad_norm": 1.332128586662064,
+      "learning_rate": 3.505642908144248e-06,
+      "loss": 0.5887,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6374193548387097,
+      "grad_norm": 1.1025566333002628,
+      "learning_rate": 3.5017346732184974e-06,
+      "loss": 0.6358,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6376539589442816,
+      "grad_norm": 1.025648018635841,
+      "learning_rate": 3.4978274437363447e-06,
+      "loss": 0.5679,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6378885630498534,
+      "grad_norm": 7.6139405115353656,
+      "learning_rate": 3.4939212223198217e-06,
+      "loss": 0.6501,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6381231671554253,
+      "grad_norm": 1.7807089617714034,
+      "learning_rate": 3.490016011590287e-06,
+      "loss": 0.5845,
+      "step": 2720
+    },
+    {
+      "epoch": 0.638357771260997,
+      "grad_norm": 0.6796377890963405,
+      "learning_rate": 3.486111814168412e-06,
+      "loss": 0.6334,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6385923753665689,
+      "grad_norm": 2.3100830732035416,
+      "learning_rate": 3.4822086326741966e-06,
+      "loss": 0.6216,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6388269794721407,
+      "grad_norm": 1.066199556390954,
+      "learning_rate": 3.478306469726957e-06,
+      "loss": 0.6064,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6390615835777126,
+      "grad_norm": 2.4176759794341476,
+      "learning_rate": 3.4744053279453215e-06,
+      "loss": 0.624,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6392961876832844,
+      "grad_norm": 1.4245791185101886,
+      "learning_rate": 3.470505209947238e-06,
+      "loss": 0.6554,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6395307917888563,
+      "grad_norm": 2.2478579927150144,
+      "learning_rate": 3.4666061183499664e-06,
+      "loss": 0.6072,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6397653958944282,
+      "grad_norm": 42.51034965314345,
+      "learning_rate": 3.4627080557700745e-06,
+      "loss": 0.5753,
+      "step": 2727
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0267583118908696,
+      "learning_rate": 3.458811024823444e-06,
+      "loss": 0.6259,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6402346041055719,
+      "grad_norm": 2.0496655540758906,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.6206,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6404692082111437,
+      "grad_norm": 4.606048010629496,
+      "learning_rate": 3.451020068290024e-06,
+      "loss": 0.6405,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6407038123167156,
+      "grad_norm": 7.1661831612698474,
+      "learning_rate": 3.4471261479315254e-06,
+      "loss": 0.5937,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6409384164222874,
+      "grad_norm": 1.364196386868124,
+      "learning_rate": 3.4432332696628666e-06,
+      "loss": 0.6627,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6411730205278592,
+      "grad_norm": 1.3018487040503812,
+      "learning_rate": 3.439341436096449e-06,
+      "loss": 0.5911,
+      "step": 2733
+    },
+    {
+      "epoch": 0.641407624633431,
+      "grad_norm": 2.5308607910073873,
+      "learning_rate": 3.435450649843972e-06,
+      "loss": 0.6302,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6416422287390029,
+      "grad_norm": 0.9991064818881716,
+      "learning_rate": 3.4315609135164308e-06,
+      "loss": 0.6248,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6418768328445748,
+      "grad_norm": 0.9548112508106517,
+      "learning_rate": 3.4276722297241206e-06,
+      "loss": 0.6133,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6421114369501466,
+      "grad_norm": 0.9423465003312195,
+      "learning_rate": 3.423784601076627e-06,
+      "loss": 0.6046,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6423460410557185,
+      "grad_norm": 1.7320553071311917,
+      "learning_rate": 3.4198980301828256e-06,
+      "loss": 0.6002,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6425806451612903,
+      "grad_norm": 1.4341561810044723,
+      "learning_rate": 3.4160125196508866e-06,
+      "loss": 0.643,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6428152492668622,
+      "grad_norm": 1.04030864388395,
+      "learning_rate": 3.4121280720882664e-06,
+      "loss": 0.6217,
+      "step": 2740
+    },
+    {
+      "epoch": 0.643049853372434,
+      "grad_norm": 1.1039159426036667,
+      "learning_rate": 3.4082446901017074e-06,
+      "loss": 0.64,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6432844574780059,
+      "grad_norm": 1.6120862915442005,
+      "learning_rate": 3.4043623762972373e-06,
+      "loss": 0.6533,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6435190615835777,
+      "grad_norm": 1.2605836948366589,
+      "learning_rate": 3.400481133280171e-06,
+      "loss": 0.6339,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6437536656891496,
+      "grad_norm": 1.2836432076726034,
+      "learning_rate": 3.3966009636550944e-06,
+      "loss": 0.6091,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6439882697947215,
+      "grad_norm": 2.0565431344091625,
+      "learning_rate": 3.392721870025886e-06,
+      "loss": 0.6201,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6442228739002932,
+      "grad_norm": 1.0085068623908457,
+      "learning_rate": 3.388843854995694e-06,
+      "loss": 0.58,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6444574780058651,
+      "grad_norm": 1.1703206393983352,
+      "learning_rate": 3.3849669211669446e-06,
+      "loss": 0.5855,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6446920821114369,
+      "grad_norm": 1.5128832145377682,
+      "learning_rate": 3.3810910711413376e-06,
+      "loss": 0.6566,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6449266862170088,
+      "grad_norm": 1.0285692439770233,
+      "learning_rate": 3.3772163075198512e-06,
+      "loss": 0.5741,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 1.8756765771691108,
+      "learning_rate": 3.373342632902723e-06,
+      "loss": 0.6326,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6453958944281525,
+      "grad_norm": 8.108637468473399,
+      "learning_rate": 3.369470049889473e-06,
+      "loss": 0.5747,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6456304985337243,
+      "grad_norm": 1.5276067869363399,
+      "learning_rate": 3.36559856107888e-06,
+      "loss": 0.6272,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6458651026392962,
+      "grad_norm": 2.293138628477277,
+      "learning_rate": 3.3617281690689895e-06,
+      "loss": 0.6686,
+      "step": 2753
+    },
+    {
+      "epoch": 0.646099706744868,
+      "grad_norm": 0.8556954037490899,
+      "learning_rate": 3.3578588764571144e-06,
+      "loss": 0.6563,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6463343108504399,
+      "grad_norm": 1.614328853113237,
+      "learning_rate": 3.3539906858398285e-06,
+      "loss": 0.6094,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6465689149560118,
+      "grad_norm": 1.047880166499049,
+      "learning_rate": 3.3501235998129624e-06,
+      "loss": 0.5949,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6468035190615836,
+      "grad_norm": 1.226481161791258,
+      "learning_rate": 3.3462576209716125e-06,
+      "loss": 0.6267,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6470381231671555,
+      "grad_norm": 1.3843091369275025,
+      "learning_rate": 3.342392751910123e-06,
+      "loss": 0.6006,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6472727272727272,
+      "grad_norm": 1.2495802125022213,
+      "learning_rate": 3.3385289952221023e-06,
+      "loss": 0.6683,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6475073313782991,
+      "grad_norm": 0.9728160026784236,
+      "learning_rate": 3.3346663535004087e-06,
+      "loss": 0.579,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6477419354838709,
+      "grad_norm": 1.0065823588235314,
+      "learning_rate": 3.3308048293371504e-06,
+      "loss": 0.6237,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6479765395894428,
+      "grad_norm": 1.7313388752219214,
+      "learning_rate": 3.3269444253236877e-06,
+      "loss": 0.6416,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6482111436950146,
+      "grad_norm": 2.9632307552546715,
+      "learning_rate": 3.32308514405063e-06,
+      "loss": 0.6369,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6484457478005865,
+      "grad_norm": 1.5046527093512325,
+      "learning_rate": 3.3192269881078302e-06,
+      "loss": 0.644,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6486803519061584,
+      "grad_norm": 1.7642242622972442,
+      "learning_rate": 3.315369960084388e-06,
+      "loss": 0.5786,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6489149560117302,
+      "grad_norm": 1.3994481817732896,
+      "learning_rate": 3.311514062568649e-06,
+      "loss": 0.6897,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6491495601173021,
+      "grad_norm": 1.017016404276696,
+      "learning_rate": 3.307659298148195e-06,
+      "loss": 0.6394,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6493841642228739,
+      "grad_norm": 1.0081237138282781,
+      "learning_rate": 3.3038056694098485e-06,
+      "loss": 0.6467,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6496187683284458,
+      "grad_norm": 6.285559513489754,
+      "learning_rate": 3.2999531789396746e-06,
+      "loss": 0.6673,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6498533724340176,
+      "grad_norm": 1.9095115053630256,
+      "learning_rate": 3.2961018293229676e-06,
+      "loss": 0.6369,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6500879765395895,
+      "grad_norm": 1.3376844881925287,
+      "learning_rate": 3.2922516231442605e-06,
+      "loss": 0.628,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6503225806451612,
+      "grad_norm": 1.5111841203473888,
+      "learning_rate": 3.2884025629873206e-06,
+      "loss": 0.5885,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6505571847507331,
+      "grad_norm": 1.7208948910426636,
+      "learning_rate": 3.2845546514351384e-06,
+      "loss": 0.6226,
+      "step": 2773
+    },
+    {
+      "epoch": 0.650791788856305,
+      "grad_norm": 16.711657319988067,
+      "learning_rate": 3.280707891069943e-06,
+      "loss": 0.573,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6510263929618768,
+      "grad_norm": 0.895072769673571,
+      "learning_rate": 3.276862284473187e-06,
+      "loss": 0.6127,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6512609970674487,
+      "grad_norm": 1.3120849452846568,
+      "learning_rate": 3.2730178342255458e-06,
+      "loss": 0.594,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6514956011730205,
+      "grad_norm": 1.4245611380007643,
+      "learning_rate": 3.2691745429069234e-06,
+      "loss": 0.6032,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6517302052785924,
+      "grad_norm": 0.9659806790106332,
+      "learning_rate": 3.265332413096444e-06,
+      "loss": 0.513,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6519648093841642,
+      "grad_norm": 2.3951227318079167,
+      "learning_rate": 3.2614914473724506e-06,
+      "loss": 0.6545,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6521994134897361,
+      "grad_norm": 0.9086189715822532,
+      "learning_rate": 3.257651648312512e-06,
+      "loss": 0.6551,
+      "step": 2780
+    },
+    {
+      "epoch": 0.652434017595308,
+      "grad_norm": 1.7055376227017278,
+      "learning_rate": 3.253813018493402e-06,
+      "loss": 0.6354,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6526686217008798,
+      "grad_norm": 1.9206882119330737,
+      "learning_rate": 3.249975560491121e-06,
+      "loss": 0.6246,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6529032258064517,
+      "grad_norm": 1.1316882930673104,
+      "learning_rate": 3.24613927688088e-06,
+      "loss": 0.6016,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6531378299120235,
+      "grad_norm": 1.11577630879089,
+      "learning_rate": 3.2423041702370976e-06,
+      "loss": 0.5841,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6533724340175953,
+      "grad_norm": 1.0826065702956267,
+      "learning_rate": 3.238470243133407e-06,
+      "loss": 0.6303,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6536070381231671,
+      "grad_norm": 3.4741203170593296,
+      "learning_rate": 3.23463749814265e-06,
+      "loss": 0.6435,
+      "step": 2786
+    },
+    {
+      "epoch": 0.653841642228739,
+      "grad_norm": 4.310514045761198,
+      "learning_rate": 3.2308059378368693e-06,
+      "loss": 0.6219,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6540762463343108,
+      "grad_norm": 1.2706772062017948,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.6163,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6543108504398827,
+      "grad_norm": 3.0178359748669195,
+      "learning_rate": 3.2231463815644616e-06,
+      "loss": 0.6427,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 2.2186249513083265,
+      "learning_rate": 3.219318390737944e-06,
+      "loss": 0.576,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6547800586510264,
+      "grad_norm": 4.714384982564151,
+      "learning_rate": 3.2154915948766263e-06,
+      "loss": 0.624,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6550146627565983,
+      "grad_norm": 0.940786148661298,
+      "learning_rate": 3.211665996548564e-06,
+      "loss": 0.6327,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6552492668621701,
+      "grad_norm": 3.598906368613139,
+      "learning_rate": 3.207841598321007e-06,
+      "loss": 0.6062,
+      "step": 2793
+    },
+    {
+      "epoch": 0.655483870967742,
+      "grad_norm": 2.230046756241626,
+      "learning_rate": 3.2040184027604003e-06,
+      "loss": 0.6567,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6557184750733138,
+      "grad_norm": 3.3323044330215237,
+      "learning_rate": 3.2001964124323883e-06,
+      "loss": 0.638,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6559530791788857,
+      "grad_norm": 1.4963999977115132,
+      "learning_rate": 3.1963756299017924e-06,
+      "loss": 0.5887,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6561876832844574,
+      "grad_norm": 1.1389768169399515,
+      "learning_rate": 3.192556057732638e-06,
+      "loss": 0.6353,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6564222873900293,
+      "grad_norm": 0.9912206816074209,
+      "learning_rate": 3.1887376984881313e-06,
+      "loss": 0.6657,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6566568914956011,
+      "grad_norm": 1.3764734320177685,
+      "learning_rate": 3.1849205547306638e-06,
+      "loss": 0.6573,
+      "step": 2799
+    },
+    {
+      "epoch": 0.656891495601173,
+      "grad_norm": 1.7392991946176513,
+      "learning_rate": 3.1811046290218125e-06,
+      "loss": 0.6301,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6571260997067448,
+      "grad_norm": 1.7459654559653102,
+      "learning_rate": 3.177289923922342e-06,
+      "loss": 0.5983,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6573607038123167,
+      "grad_norm": 1.503585318798611,
+      "learning_rate": 3.173476441992187e-06,
+      "loss": 0.5794,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6575953079178886,
+      "grad_norm": 1.7370600512456607,
+      "learning_rate": 3.1696641857904743e-06,
+      "loss": 0.6309,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6578299120234604,
+      "grad_norm": 4.265726896789798,
+      "learning_rate": 3.165853157875495e-06,
+      "loss": 0.6069,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6580645161290323,
+      "grad_norm": 1.220764530339127,
+      "learning_rate": 3.162043360804726e-06,
+      "loss": 0.5676,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6582991202346041,
+      "grad_norm": 0.9774335610640675,
+      "learning_rate": 3.158234797134815e-06,
+      "loss": 0.6017,
+      "step": 2806
+    },
+    {
+      "epoch": 0.658533724340176,
+      "grad_norm": 1.185068209023832,
+      "learning_rate": 3.1544274694215804e-06,
+      "loss": 0.6278,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6587683284457478,
+      "grad_norm": 1.2496617926932865,
+      "learning_rate": 3.1506213802200113e-06,
+      "loss": 0.6298,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6590029325513197,
+      "grad_norm": 0.9430005699055659,
+      "learning_rate": 3.1468165320842712e-06,
+      "loss": 0.6327,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6592375366568914,
+      "grad_norm": 1.4647903486974598,
+      "learning_rate": 3.1430129275676813e-06,
+      "loss": 0.5568,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6594721407624633,
+      "grad_norm": 0.9140100955216531,
+      "learning_rate": 3.139210569222736e-06,
+      "loss": 0.6392,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6597067448680352,
+      "grad_norm": 3.573415723502299,
+      "learning_rate": 3.1354094596010908e-06,
+      "loss": 0.6562,
+      "step": 2812
+    },
+    {
+      "epoch": 0.659941348973607,
+      "grad_norm": 1.3931891053320853,
+      "learning_rate": 3.1316096012535625e-06,
+      "loss": 0.6509,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6601759530791789,
+      "grad_norm": 1.128743321592625,
+      "learning_rate": 3.127810996730128e-06,
+      "loss": 0.6702,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6604105571847507,
+      "grad_norm": 1.255563636699726,
+      "learning_rate": 3.1240136485799254e-06,
+      "loss": 0.5907,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6606451612903226,
+      "grad_norm": 0.9502285457823598,
+      "learning_rate": 3.120217559351247e-06,
+      "loss": 0.6267,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6608797653958944,
+      "grad_norm": 1.3567066440983155,
+      "learning_rate": 3.1164227315915398e-06,
+      "loss": 0.5449,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6611143695014663,
+      "grad_norm": 1.279395443678301,
+      "learning_rate": 3.112629167847409e-06,
+      "loss": 0.6256,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6613489736070381,
+      "grad_norm": 1.3699719026303925,
+      "learning_rate": 3.1088368706646057e-06,
+      "loss": 0.6479,
+      "step": 2819
+    },
+    {
+      "epoch": 0.66158357771261,
+      "grad_norm": 2.792818448408116,
+      "learning_rate": 3.1050458425880335e-06,
+      "loss": 0.5395,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6618181818181819,
+      "grad_norm": 1.5848622001133197,
+      "learning_rate": 3.101256086161747e-06,
+      "loss": 0.6829,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6620527859237537,
+      "grad_norm": 0.9511167105997375,
+      "learning_rate": 3.0974676039289418e-06,
+      "loss": 0.6108,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6622873900293255,
+      "grad_norm": 23.345564151938994,
+      "learning_rate": 3.0936803984319618e-06,
+      "loss": 0.6289,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6625219941348973,
+      "grad_norm": 1.8392029992996093,
+      "learning_rate": 3.089894472212297e-06,
+      "loss": 0.6412,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6627565982404692,
+      "grad_norm": 1.3812046493943864,
+      "learning_rate": 3.0861098278105713e-06,
+      "loss": 0.6064,
+      "step": 2825
+    },
+    {
+      "epoch": 0.662991202346041,
+      "grad_norm": 1.456229608860474,
+      "learning_rate": 3.0823264677665573e-06,
+      "loss": 0.6365,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6632258064516129,
+      "grad_norm": 1.4220174379325206,
+      "learning_rate": 3.078544394619155e-06,
+      "loss": 0.5967,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6634604105571847,
+      "grad_norm": 0.9232152751664393,
+      "learning_rate": 3.0747636109064126e-06,
+      "loss": 0.6265,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6636950146627566,
+      "grad_norm": 1.674191387003085,
+      "learning_rate": 3.0709841191655064e-06,
+      "loss": 0.6271,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6639296187683285,
+      "grad_norm": 1.4379811489781062,
+      "learning_rate": 3.0672059219327453e-06,
+      "loss": 0.6451,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6641642228739003,
+      "grad_norm": 2.3144628414571033,
+      "learning_rate": 3.0634290217435703e-06,
+      "loss": 0.6097,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6643988269794722,
+      "grad_norm": 2.348917719491438,
+      "learning_rate": 3.059653421132557e-06,
+      "loss": 0.5946,
+      "step": 2832
+    },
+    {
+      "epoch": 0.664633431085044,
+      "grad_norm": 1.3169704591372087,
+      "learning_rate": 3.0558791226333974e-06,
+      "loss": 0.5785,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6648680351906159,
+      "grad_norm": 1.109645045604988,
+      "learning_rate": 3.0521061287789215e-06,
+      "loss": 0.6101,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6651026392961877,
+      "grad_norm": 1.332282345179443,
+      "learning_rate": 3.0483344421010797e-06,
+      "loss": 0.6328,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6653372434017595,
+      "grad_norm": 1.5215806892713624,
+      "learning_rate": 3.0445640651309416e-06,
+      "loss": 0.5636,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6655718475073313,
+      "grad_norm": 0.759584834471881,
+      "learning_rate": 3.0407950003987025e-06,
+      "loss": 0.6954,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6658064516129032,
+      "grad_norm": 2.274171222583829,
+      "learning_rate": 3.037027250433676e-06,
+      "loss": 0.6559,
+      "step": 2838
+    },
+    {
+      "epoch": 0.666041055718475,
+      "grad_norm": 1.511166352592508,
+      "learning_rate": 3.0332608177642896e-06,
+      "loss": 0.6084,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6662756598240469,
+      "grad_norm": 1.6393585855755946,
+      "learning_rate": 3.029495704918094e-06,
+      "loss": 0.627,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6665102639296188,
+      "grad_norm": 1.328829377128736,
+      "learning_rate": 3.0257319144217488e-06,
+      "loss": 0.6695,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6667448680351906,
+      "grad_norm": 1.1493865460633572,
+      "learning_rate": 3.021969448801028e-06,
+      "loss": 0.6438,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6669794721407625,
+      "grad_norm": 1.5317020059344715,
+      "learning_rate": 3.0182083105808135e-06,
+      "loss": 0.6045,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6672140762463343,
+      "grad_norm": 1.590973695060672,
+      "learning_rate": 3.0144485022851035e-06,
+      "loss": 0.6542,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6674486803519062,
+      "grad_norm": 1.5991926674214896,
+      "learning_rate": 3.0106900264369956e-06,
+      "loss": 0.6055,
+      "step": 2845
+    },
+    {
+      "epoch": 0.667683284457478,
+      "grad_norm": 1.0388666239291835,
+      "learning_rate": 3.006932885558697e-06,
+      "loss": 0.585,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6679178885630499,
+      "grad_norm": 1.3609644532751701,
+      "learning_rate": 3.0031770821715233e-06,
+      "loss": 0.6355,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6681524926686218,
+      "grad_norm": 10.985182746296877,
+      "learning_rate": 2.9994226187958824e-06,
+      "loss": 0.7029,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6683870967741935,
+      "grad_norm": 3.1268044104649353,
+      "learning_rate": 2.9956694979512936e-06,
+      "loss": 0.6416,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6686217008797654,
+      "grad_norm": 1.1755091839315208,
+      "learning_rate": 2.991917722156367e-06,
+      "loss": 0.6039,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6688563049853372,
+      "grad_norm": 2.2426166787988757,
+      "learning_rate": 2.988167293928815e-06,
+      "loss": 0.637,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6690909090909091,
+      "grad_norm": 1.272917129366812,
+      "learning_rate": 2.984418215785445e-06,
+      "loss": 0.6057,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6693255131964809,
+      "grad_norm": 1.3658970750260513,
+      "learning_rate": 2.9806704902421557e-06,
+      "loss": 0.585,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6695601173020528,
+      "grad_norm": 3.979620433534735,
+      "learning_rate": 2.9769241198139404e-06,
+      "loss": 0.6345,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6697947214076246,
+      "grad_norm": 1.1443883545732938,
+      "learning_rate": 2.9731791070148856e-06,
+      "loss": 0.6229,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6700293255131965,
+      "grad_norm": 1.6220656787609486,
+      "learning_rate": 2.969435454358158e-06,
+      "loss": 0.6288,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6702639296187684,
+      "grad_norm": 1.340768995873484,
+      "learning_rate": 2.965693164356023e-06,
+      "loss": 0.6359,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6704985337243402,
+      "grad_norm": 1.7545770470136055,
+      "learning_rate": 2.961952239519823e-06,
+      "loss": 0.5351,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6707331378299121,
+      "grad_norm": 1.2654033268936953,
+      "learning_rate": 2.9582126823599877e-06,
+      "loss": 0.6208,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6709677419354839,
+      "grad_norm": 1.1272607784509288,
+      "learning_rate": 2.9544744953860265e-06,
+      "loss": 0.6451,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6712023460410557,
+      "grad_norm": 0.9460218742456779,
+      "learning_rate": 2.950737681106537e-06,
+      "loss": 0.5644,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6714369501466275,
+      "grad_norm": 1.8527524541962235,
+      "learning_rate": 2.947002242029182e-06,
+      "loss": 0.6979,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6716715542521994,
+      "grad_norm": 0.818960409812016,
+      "learning_rate": 2.9432681806607145e-06,
+      "loss": 0.633,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6719061583577712,
+      "grad_norm": 6.587958564986356,
+      "learning_rate": 2.9395354995069574e-06,
+      "loss": 0.6928,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6721407624633431,
+      "grad_norm": 1.6530245205627356,
+      "learning_rate": 2.935804201072806e-06,
+      "loss": 0.6065,
+      "step": 2865
+    },
+    {
+      "epoch": 0.672375366568915,
+      "grad_norm": 1.39135578014723,
+      "learning_rate": 2.9320742878622303e-06,
+      "loss": 0.6615,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6726099706744868,
+      "grad_norm": 1.1184181775851425,
+      "learning_rate": 2.9283457623782707e-06,
+      "loss": 0.6423,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6728445747800587,
+      "grad_norm": 1.158912499957988,
+      "learning_rate": 2.9246186271230335e-06,
+      "loss": 0.6295,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6730791788856305,
+      "grad_norm": 1.6279079724630416,
+      "learning_rate": 2.9208928845976947e-06,
+      "loss": 0.5959,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6733137829912024,
+      "grad_norm": 1.705259226439366,
+      "learning_rate": 2.917168537302497e-06,
+      "loss": 0.6354,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6735483870967742,
+      "grad_norm": 1.1830061393698021,
+      "learning_rate": 2.913445587736742e-06,
+      "loss": 0.688,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6737829912023461,
+      "grad_norm": 1.1272705513657213,
+      "learning_rate": 2.9097240383988e-06,
+      "loss": 0.5818,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6740175953079179,
+      "grad_norm": 1.0288641731279962,
+      "learning_rate": 2.9060038917860928e-06,
+      "loss": 0.6086,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6742521994134897,
+      "grad_norm": 1.464748073480879,
+      "learning_rate": 2.902285150395106e-06,
+      "loss": 0.5965,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6744868035190615,
+      "grad_norm": 12.020052613752277,
+      "learning_rate": 2.898567816721389e-06,
+      "loss": 0.6194,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6747214076246334,
+      "grad_norm": 3.604009329849857,
+      "learning_rate": 2.894851893259532e-06,
+      "loss": 0.6338,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6749560117302053,
+      "grad_norm": 1.6719245844127935,
+      "learning_rate": 2.8911373825031897e-06,
+      "loss": 0.6333,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6751906158357771,
+      "grad_norm": 1.0984583224202944,
+      "learning_rate": 2.8874242869450655e-06,
+      "loss": 0.5903,
+      "step": 2878
+    },
+    {
+      "epoch": 0.675425219941349,
+      "grad_norm": 0.9577017499977273,
+      "learning_rate": 2.8837126090769118e-06,
+      "loss": 0.5841,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6756598240469208,
+      "grad_norm": 1.1865292120559294,
+      "learning_rate": 2.8800023513895324e-06,
+      "loss": 0.5805,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6758944281524927,
+      "grad_norm": 2.230851870372906,
+      "learning_rate": 2.876293516372778e-06,
+      "loss": 0.5917,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6761290322580645,
+      "grad_norm": 0.9735266374200137,
+      "learning_rate": 2.872586106515537e-06,
+      "loss": 0.6052,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6763636363636364,
+      "grad_norm": 1.839648421209005,
+      "learning_rate": 2.8688801243057536e-06,
+      "loss": 0.577,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6765982404692082,
+      "grad_norm": 1.2510107958058816,
+      "learning_rate": 2.8651755722304082e-06,
+      "loss": 0.7087,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6768328445747801,
+      "grad_norm": 2.1524909106565837,
+      "learning_rate": 2.8614724527755168e-06,
+      "loss": 0.6245,
+      "step": 2885
+    },
+    {
+      "epoch": 0.677067448680352,
+      "grad_norm": 1.8904527779746534,
+      "learning_rate": 2.857770768426139e-06,
+      "loss": 0.5994,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6773020527859237,
+      "grad_norm": 2.065524797125004,
+      "learning_rate": 2.854070521666376e-06,
+      "loss": 0.6339,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6775366568914956,
+      "grad_norm": 0.9496542794903673,
+      "learning_rate": 2.8503717149793543e-06,
+      "loss": 0.5921,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6777712609970674,
+      "grad_norm": 1.4592674331186697,
+      "learning_rate": 2.846674350847239e-06,
+      "loss": 0.6341,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6780058651026393,
+      "grad_norm": 1.6848807945445805,
+      "learning_rate": 2.8429784317512275e-06,
+      "loss": 0.6479,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6782404692082111,
+      "grad_norm": 2.0942540269295793,
+      "learning_rate": 2.8392839601715482e-06,
+      "loss": 0.5938,
+      "step": 2891
+    },
+    {
+      "epoch": 0.678475073313783,
+      "grad_norm": 1.2271050050156875,
+      "learning_rate": 2.8355909385874546e-06,
+      "loss": 0.546,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6787096774193548,
+      "grad_norm": 4.204770776951658,
+      "learning_rate": 2.831899369477233e-06,
+      "loss": 0.6386,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6789442815249267,
+      "grad_norm": 2.0165045411018014,
+      "learning_rate": 2.8282092553181863e-06,
+      "loss": 0.5919,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6791788856304986,
+      "grad_norm": 1.011918455921512,
+      "learning_rate": 2.8245205985866467e-06,
+      "loss": 0.5626,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6794134897360704,
+      "grad_norm": 0.9938648578959085,
+      "learning_rate": 2.8208334017579735e-06,
+      "loss": 0.6657,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6796480938416423,
+      "grad_norm": 1.1786828064949448,
+      "learning_rate": 2.8171476673065343e-06,
+      "loss": 0.6132,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6798826979472141,
+      "grad_norm": 3.0588265278017173,
+      "learning_rate": 2.8134633977057236e-06,
+      "loss": 0.6338,
+      "step": 2898
+    },
+    {
+      "epoch": 0.680117302052786,
+      "grad_norm": 1.4664012215829267,
+      "learning_rate": 2.8097805954279505e-06,
+      "loss": 0.6034,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6803519061583577,
+      "grad_norm": 0.992836254764791,
+      "learning_rate": 2.8060992629446406e-06,
+      "loss": 0.6164,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6805865102639296,
+      "grad_norm": 1.3595756537761916,
+      "learning_rate": 2.802419402726233e-06,
+      "loss": 0.687,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6808211143695014,
+      "grad_norm": 1.098305893354896,
+      "learning_rate": 2.7987410172421756e-06,
+      "loss": 0.6327,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6810557184750733,
+      "grad_norm": 2.0540377571088806,
+      "learning_rate": 2.7950641089609275e-06,
+      "loss": 0.6314,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6812903225806451,
+      "grad_norm": 1.1741171050623673,
+      "learning_rate": 2.7913886803499657e-06,
+      "loss": 0.5887,
+      "step": 2904
+    },
+    {
+      "epoch": 0.681524926686217,
+      "grad_norm": 0.9085521953652322,
+      "learning_rate": 2.7877147338757605e-06,
+      "loss": 0.6212,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6817595307917889,
+      "grad_norm": 1.0193622872667523,
+      "learning_rate": 2.7840422720037943e-06,
+      "loss": 0.6166,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6819941348973607,
+      "grad_norm": 1.3384204790400827,
+      "learning_rate": 2.7803712971985535e-06,
+      "loss": 0.6386,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6822287390029326,
+      "grad_norm": 1.0633549432371698,
+      "learning_rate": 2.7767018119235263e-06,
+      "loss": 0.6078,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6824633431085044,
+      "grad_norm": 2.2193333192961524,
+      "learning_rate": 2.7730338186411997e-06,
+      "loss": 0.6392,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6826979472140763,
+      "grad_norm": 2.4172875239478633,
+      "learning_rate": 2.769367319813064e-06,
+      "loss": 0.6209,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6829325513196481,
+      "grad_norm": 1.0609735734651438,
+      "learning_rate": 2.7657023178995955e-06,
+      "loss": 0.6235,
+      "step": 2911
+    },
+    {
+      "epoch": 0.68316715542522,
+      "grad_norm": 2.658231085240406,
+      "learning_rate": 2.7620388153602807e-06,
+      "loss": 0.5978,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6834017595307917,
+      "grad_norm": 1.4268951686909808,
+      "learning_rate": 2.7583768146535927e-06,
+      "loss": 0.6674,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6836363636363636,
+      "grad_norm": 0.9821828639937361,
+      "learning_rate": 2.754716318236993e-06,
+      "loss": 0.6747,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6838709677419355,
+      "grad_norm": 1.6885672688129323,
+      "learning_rate": 2.7510573285669382e-06,
+      "loss": 0.6205,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6841055718475073,
+      "grad_norm": 1.5084964667460745,
+      "learning_rate": 2.7473998480988746e-06,
+      "loss": 0.6012,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6843401759530792,
+      "grad_norm": 0.9368814430221031,
+      "learning_rate": 2.7437438792872332e-06,
+      "loss": 0.6208,
+      "step": 2917
+    },
+    {
+      "epoch": 0.684574780058651,
+      "grad_norm": 3.8251928370956545,
+      "learning_rate": 2.7400894245854327e-06,
+      "loss": 0.599,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6848093841642229,
+      "grad_norm": 0.9841376000436076,
+      "learning_rate": 2.7364364864458724e-06,
+      "loss": 0.6451,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6850439882697947,
+      "grad_norm": 1.1442021678808518,
+      "learning_rate": 2.7327850673199384e-06,
+      "loss": 0.6777,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6852785923753666,
+      "grad_norm": 1.7873515601018037,
+      "learning_rate": 2.7291351696579966e-06,
+      "loss": 0.615,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6855131964809384,
+      "grad_norm": 1.1203503910361199,
+      "learning_rate": 2.7254867959093856e-06,
+      "loss": 0.5835,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6857478005865103,
+      "grad_norm": 1.5265436819101847,
+      "learning_rate": 2.721839948522428e-06,
+      "loss": 0.5801,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6859824046920822,
+      "grad_norm": 1.2311245489074993,
+      "learning_rate": 2.7181946299444205e-06,
+      "loss": 0.6397,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6862170087976539,
+      "grad_norm": 0.9843243705694885,
+      "learning_rate": 2.7145508426216345e-06,
+      "loss": 0.6595,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6864516129032258,
+      "grad_norm": 2.3092268531811233,
+      "learning_rate": 2.7109085889993112e-06,
+      "loss": 0.6262,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6866862170087976,
+      "grad_norm": 1.80595676554796,
+      "learning_rate": 2.707267871521666e-06,
+      "loss": 0.6001,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6869208211143695,
+      "grad_norm": 1.1348891790325555,
+      "learning_rate": 2.70362869263188e-06,
+      "loss": 0.615,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6871554252199413,
+      "grad_norm": 2.263449903706541,
+      "learning_rate": 2.699991054772105e-06,
+      "loss": 0.6215,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6873900293255132,
+      "grad_norm": 1.3392135503572136,
+      "learning_rate": 2.6963549603834584e-06,
+      "loss": 0.6426,
+      "step": 2930
+    },
+    {
+      "epoch": 0.687624633431085,
+      "grad_norm": 1.9645893704674489,
+      "learning_rate": 2.692720411906017e-06,
+      "loss": 0.5612,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6878592375366569,
+      "grad_norm": 1.194251282533319,
+      "learning_rate": 2.689087411778823e-06,
+      "loss": 0.5922,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6880938416422288,
+      "grad_norm": 1.7386890174380603,
+      "learning_rate": 2.6854559624398867e-06,
+      "loss": 0.6513,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6883284457478006,
+      "grad_norm": 1.2575489656956949,
+      "learning_rate": 2.681826066326166e-06,
+      "loss": 0.6096,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6885630498533725,
+      "grad_norm": 1.6346961341634954,
+      "learning_rate": 2.6781977258735825e-06,
+      "loss": 0.622,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6887976539589443,
+      "grad_norm": 0.859684406970139,
+      "learning_rate": 2.674570943517014e-06,
+      "loss": 0.6752,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6890322580645162,
+      "grad_norm": 1.0660195697369763,
+      "learning_rate": 2.6709457216902923e-06,
+      "loss": 0.6744,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6892668621700879,
+      "grad_norm": 7.324120026228176,
+      "learning_rate": 2.6673220628262002e-06,
+      "loss": 0.6128,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6895014662756598,
+      "grad_norm": 1.0028191550126604,
+      "learning_rate": 2.6636999693564757e-06,
+      "loss": 0.6286,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6897360703812316,
+      "grad_norm": 6.330141781990575,
+      "learning_rate": 2.6600794437117984e-06,
+      "loss": 0.674,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6899706744868035,
+      "grad_norm": 2.8688801373319834,
+      "learning_rate": 2.656460488321805e-06,
+      "loss": 0.6466,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6902052785923753,
+      "grad_norm": 1.0042272689014873,
+      "learning_rate": 2.6528431056150753e-06,
+      "loss": 0.6179,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6904398826979472,
+      "grad_norm": 0.884456076476593,
+      "learning_rate": 2.649227298019129e-06,
+      "loss": 0.632,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6906744868035191,
+      "grad_norm": 1.1733317012689413,
+      "learning_rate": 2.645613067960435e-06,
+      "loss": 0.6153,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6909090909090909,
+      "grad_norm": 1.2068306093086028,
+      "learning_rate": 2.642000417864401e-06,
+      "loss": 0.5545,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6911436950146628,
+      "grad_norm": 1.0984644873764033,
+      "learning_rate": 2.638389350155373e-06,
+      "loss": 0.5988,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6913782991202346,
+      "grad_norm": 1.5260598428986425,
+      "learning_rate": 2.6347798672566405e-06,
+      "loss": 0.6232,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6916129032258065,
+      "grad_norm": 0.9801830003751519,
+      "learning_rate": 2.631171971590419e-06,
+      "loss": 0.6082,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6918475073313783,
+      "grad_norm": 2.187568539878525,
+      "learning_rate": 2.627565665577872e-06,
+      "loss": 0.648,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6920821114369502,
+      "grad_norm": 1.2116407120973633,
+      "learning_rate": 2.6239609516390886e-06,
+      "loss": 0.5748,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6923167155425219,
+      "grad_norm": 1.058164460232869,
+      "learning_rate": 2.620357832193088e-06,
+      "loss": 0.6096,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6925513196480938,
+      "grad_norm": 1.4310036079844117,
+      "learning_rate": 2.616756309657823e-06,
+      "loss": 0.6508,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6927859237536657,
+      "grad_norm": 0.8324497438810345,
+      "learning_rate": 2.613156386450174e-06,
+      "loss": 0.6373,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6930205278592375,
+      "grad_norm": 1.4335451407998134,
+      "learning_rate": 2.6095580649859488e-06,
+      "loss": 0.5867,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6932551319648094,
+      "grad_norm": 1.8180013742308017,
+      "learning_rate": 2.6059613476798784e-06,
+      "loss": 0.6112,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6934897360703812,
+      "grad_norm": 3.1499798966921913,
+      "learning_rate": 2.6023662369456206e-06,
+      "loss": 0.5904,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6937243401759531,
+      "grad_norm": 2.1569673925272737,
+      "learning_rate": 2.598772735195746e-06,
+      "loss": 0.5609,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6939589442815249,
+      "grad_norm": 1.1310822575355794,
+      "learning_rate": 2.5951808448417603e-06,
+      "loss": 0.6575,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6941935483870968,
+      "grad_norm": 1.1641509734349609,
+      "learning_rate": 2.591590568294078e-06,
+      "loss": 0.5965,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6944281524926686,
+      "grad_norm": 1.062059412976598,
+      "learning_rate": 2.5880019079620294e-06,
+      "loss": 0.6265,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6946627565982405,
+      "grad_norm": 1.3379517850880083,
+      "learning_rate": 2.5844148662538627e-06,
+      "loss": 0.6738,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6948973607038124,
+      "grad_norm": 2.594687775988743,
+      "learning_rate": 2.5808294455767468e-06,
+      "loss": 0.6698,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6951319648093842,
+      "grad_norm": 3.255937518305223,
+      "learning_rate": 2.57724564833675e-06,
+      "loss": 0.5931,
+      "step": 2963
+    },
+    {
+      "epoch": 0.695366568914956,
+      "grad_norm": 2.5251109529175224,
+      "learning_rate": 2.573663476938859e-06,
+      "loss": 0.623,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6956011730205278,
+      "grad_norm": 1.3058098673928853,
+      "learning_rate": 2.57008293378697e-06,
+      "loss": 0.6892,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6958357771260997,
+      "grad_norm": 1.0865533957493687,
+      "learning_rate": 2.5665040212838822e-06,
+      "loss": 0.5507,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6960703812316715,
+      "grad_norm": 0.9426206025996563,
+      "learning_rate": 2.5629267418313065e-06,
+      "loss": 0.608,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6963049853372434,
+      "grad_norm": 2.1931147294957376,
+      "learning_rate": 2.5593510978298487e-06,
+      "loss": 0.6676,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6965395894428152,
+      "grad_norm": 2.713450087881679,
+      "learning_rate": 2.555777091679024e-06,
+      "loss": 0.6456,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6967741935483871,
+      "grad_norm": 1.2620988976737808,
+      "learning_rate": 2.552204725777253e-06,
+      "loss": 0.6396,
+      "step": 2970
+    },
+    {
+      "epoch": 0.697008797653959,
+      "grad_norm": 1.1781007934007224,
+      "learning_rate": 2.5486340025218425e-06,
+      "loss": 0.5823,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6972434017595308,
+      "grad_norm": 5.135887486520184,
+      "learning_rate": 2.5450649243090076e-06,
+      "loss": 0.6487,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6974780058651027,
+      "grad_norm": 2.0839686381113016,
+      "learning_rate": 2.5414974935338555e-06,
+      "loss": 0.6321,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6977126099706745,
+      "grad_norm": 1.3161281018417577,
+      "learning_rate": 2.5379317125903876e-06,
+      "loss": 0.5946,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6979472140762464,
+      "grad_norm": 0.8018410432530535,
+      "learning_rate": 2.5343675838715e-06,
+      "loss": 0.6257,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6981818181818182,
+      "grad_norm": 2.0506368128493393,
+      "learning_rate": 2.53080510976898e-06,
+      "loss": 0.5974,
+      "step": 2976
+    },
+    {
+      "epoch": 0.69841642228739,
+      "grad_norm": 6.629086288287914,
+      "learning_rate": 2.5272442926735e-06,
+      "loss": 0.6418,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6986510263929618,
+      "grad_norm": 1.2259407613073565,
+      "learning_rate": 2.5236851349746242e-06,
+      "loss": 0.5896,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6988856304985337,
+      "grad_norm": 2.028341343535244,
+      "learning_rate": 2.5201276390608087e-06,
+      "loss": 0.6621,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6991202346041056,
+      "grad_norm": 2.619611758559292,
+      "learning_rate": 2.5165718073193823e-06,
+      "loss": 0.6777,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6993548387096774,
+      "grad_norm": 2.5583207750689043,
+      "learning_rate": 2.513017642136566e-06,
+      "loss": 0.6431,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6995894428152493,
+      "grad_norm": 1.3158747470507823,
+      "learning_rate": 2.5094651458974593e-06,
+      "loss": 0.643,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6998240469208211,
+      "grad_norm": 0.9359562415514225,
+      "learning_rate": 2.5059143209860428e-06,
+      "loss": 0.5798,
+      "step": 2983
+    },
+    {
+      "epoch": 0.700058651026393,
+      "grad_norm": 1.0087867885845683,
+      "learning_rate": 2.5023651697851734e-06,
+      "loss": 0.6573,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7002932551319648,
+      "grad_norm": 1.2602193218941862,
+      "learning_rate": 2.4988176946765906e-06,
+      "loss": 0.6273,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7005278592375367,
+      "grad_norm": 1.1527791507490224,
+      "learning_rate": 2.4952718980408973e-06,
+      "loss": 0.7004,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7007624633431085,
+      "grad_norm": 1.8359600994002145,
+      "learning_rate": 2.491727782257584e-06,
+      "loss": 0.6263,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7009970674486804,
+      "grad_norm": 1.0763423476582339,
+      "learning_rate": 2.4881853497050074e-06,
+      "loss": 0.6284,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7012316715542521,
+      "grad_norm": 5.243668554983986,
+      "learning_rate": 2.4846446027603892e-06,
+      "loss": 0.5597,
+      "step": 2989
+    },
+    {
+      "epoch": 0.701466275659824,
+      "grad_norm": 4.7044154282587005,
+      "learning_rate": 2.481105543799827e-06,
+      "loss": 0.6417,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7017008797653959,
+      "grad_norm": 0.8525839737743981,
+      "learning_rate": 2.4775681751982837e-06,
+      "loss": 0.668,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7019354838709677,
+      "grad_norm": 1.6191931203466026,
+      "learning_rate": 2.474032499329588e-06,
+      "loss": 0.6046,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7021700879765396,
+      "grad_norm": 2.6330047781071055,
+      "learning_rate": 2.470498518566433e-06,
+      "loss": 0.5948,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7024046920821114,
+      "grad_norm": 1.1482290058274915,
+      "learning_rate": 2.466966235280369e-06,
+      "loss": 0.6493,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7026392961876833,
+      "grad_norm": 1.9539406976269607,
+      "learning_rate": 2.4634356518418167e-06,
+      "loss": 0.6036,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7028739002932551,
+      "grad_norm": 1.546517172870452,
+      "learning_rate": 2.459906770620051e-06,
+      "loss": 0.5652,
+      "step": 2996
+    },
+    {
+      "epoch": 0.703108504398827,
+      "grad_norm": 0.9728200086431372,
+      "learning_rate": 2.4563795939832014e-06,
+      "loss": 0.5967,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7033431085043989,
+      "grad_norm": 1.1543715824692742,
+      "learning_rate": 2.452854124298257e-06,
+      "loss": 0.5984,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7035777126099707,
+      "grad_norm": 1.5598700116312605,
+      "learning_rate": 2.4493303639310634e-06,
+      "loss": 0.5613,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7038123167155426,
+      "grad_norm": 1.0808904679380773,
+      "learning_rate": 2.445808315246315e-06,
+      "loss": 0.6377,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7038123167155426,
+      "eval_loss": 0.618014931678772,
+      "eval_runtime": 39.8688,
+      "eval_samples_per_second": 13.695,
+      "eval_steps_per_second": 0.125,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7040469208211144,
+      "grad_norm": 1.0083436032813202,
+      "learning_rate": 2.44228798060756e-06,
+      "loss": 0.5637,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7042815249266862,
+      "grad_norm": 1.551336060663135,
+      "learning_rate": 2.438769362377196e-06,
+      "loss": 0.6397,
+      "step": 3002
+    },
+    {
+      "epoch": 0.704516129032258,
+      "grad_norm": 2.0623084373347162,
+      "learning_rate": 2.435252462916467e-06,
+      "loss": 0.5864,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7047507331378299,
+      "grad_norm": 1.460880689822133,
+      "learning_rate": 2.4317372845854663e-06,
+      "loss": 0.6128,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7049853372434017,
+      "grad_norm": 0.7829634791946768,
+      "learning_rate": 2.4282238297431328e-06,
+      "loss": 0.6119,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7052199413489736,
+      "grad_norm": 3.4110934486558393,
+      "learning_rate": 2.424712100747243e-06,
+      "loss": 0.5944,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7054545454545454,
+      "grad_norm": 4.504269536036809,
+      "learning_rate": 2.4212020999544184e-06,
+      "loss": 0.5989,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7056891495601173,
+      "grad_norm": 2.9123036593108784,
+      "learning_rate": 2.4176938297201285e-06,
+      "loss": 0.6156,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7059237536656892,
+      "grad_norm": 1.6082724379510858,
+      "learning_rate": 2.414187292398668e-06,
+      "loss": 0.5904,
+      "step": 3009
+    },
+    {
+      "epoch": 0.706158357771261,
+      "grad_norm": 1.003507769067074,
+      "learning_rate": 2.4106824903431774e-06,
+      "loss": 0.6328,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7063929618768329,
+      "grad_norm": 0.8252337540952676,
+      "learning_rate": 2.40717942590563e-06,
+      "loss": 0.6094,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7066275659824047,
+      "grad_norm": 1.664789198568701,
+      "learning_rate": 2.4036781014368344e-06,
+      "loss": 0.6058,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7068621700879766,
+      "grad_norm": 0.94550532740909,
+      "learning_rate": 2.4001785192864314e-06,
+      "loss": 0.6074,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7070967741935484,
+      "grad_norm": 0.7776157157972251,
+      "learning_rate": 2.3966806818028894e-06,
+      "loss": 0.6772,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7073313782991202,
+      "grad_norm": 1.376934742327379,
+      "learning_rate": 2.3931845913335073e-06,
+      "loss": 0.6232,
+      "step": 3015
+    },
+    {
+      "epoch": 0.707565982404692,
+      "grad_norm": 1.2813169239858986,
+      "learning_rate": 2.3896902502244194e-06,
+      "loss": 0.6091,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7078005865102639,
+      "grad_norm": 1.313302381723937,
+      "learning_rate": 2.3861976608205734e-06,
+      "loss": 0.6414,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7080351906158358,
+      "grad_norm": 2.2660289115882866,
+      "learning_rate": 2.3827068254657493e-06,
+      "loss": 0.6105,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7082697947214076,
+      "grad_norm": 1.2436694682572658,
+      "learning_rate": 2.3792177465025483e-06,
+      "loss": 0.6077,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7085043988269795,
+      "grad_norm": 1.2934200363317827,
+      "learning_rate": 2.375730426272393e-06,
+      "loss": 0.6443,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7087390029325513,
+      "grad_norm": 1.3226873193959154,
+      "learning_rate": 2.3722448671155253e-06,
+      "loss": 0.5967,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7089736070381232,
+      "grad_norm": 1.4469709024801236,
+      "learning_rate": 2.368761071371008e-06,
+      "loss": 0.63,
+      "step": 3022
+    },
+    {
+      "epoch": 0.709208211143695,
+      "grad_norm": 1.691638793475797,
+      "learning_rate": 2.3652790413767125e-06,
+      "loss": 0.5817,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7094428152492669,
+      "grad_norm": 1.283648412545291,
+      "learning_rate": 2.3617987794693358e-06,
+      "loss": 0.6422,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 1.2682867575516177,
+      "learning_rate": 2.3583202879843844e-06,
+      "loss": 0.6562,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7099120234604106,
+      "grad_norm": 1.5039338883580333,
+      "learning_rate": 2.354843569256172e-06,
+      "loss": 0.6024,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7101466275659825,
+      "grad_norm": 1.1283712881073278,
+      "learning_rate": 2.3513686256178276e-06,
+      "loss": 0.62,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7103812316715542,
+      "grad_norm": 1.5069041246396937,
+      "learning_rate": 2.3478954594012884e-06,
+      "loss": 0.6182,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7106158357771261,
+      "grad_norm": 1.3222552024129546,
+      "learning_rate": 2.344424072937298e-06,
+      "loss": 0.6445,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7108504398826979,
+      "grad_norm": 1.649292393317922,
+      "learning_rate": 2.3409544685554064e-06,
+      "loss": 0.6197,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7110850439882698,
+      "grad_norm": 1.1890982073103622,
+      "learning_rate": 2.3374866485839664e-06,
+      "loss": 0.5987,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7113196480938416,
+      "grad_norm": 1.2831891994843283,
+      "learning_rate": 2.3340206153501354e-06,
+      "loss": 0.6385,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7115542521994135,
+      "grad_norm": 1.1886954278865602,
+      "learning_rate": 2.3305563711798696e-06,
+      "loss": 0.5843,
+      "step": 3033
+    },
+    {
+      "epoch": 0.7117888563049853,
+      "grad_norm": 1.1703464338328116,
+      "learning_rate": 2.327093918397928e-06,
+      "loss": 0.6114,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7120234604105572,
+      "grad_norm": 1.3578841915745736,
+      "learning_rate": 2.3236332593278603e-06,
+      "loss": 0.6142,
+      "step": 3035
+    },
+    {
+      "epoch": 0.712258064516129,
+      "grad_norm": 1.211142185636031,
+      "learning_rate": 2.3201743962920202e-06,
+      "loss": 0.6294,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7124926686217009,
+      "grad_norm": 1.084284996182101,
+      "learning_rate": 2.3167173316115528e-06,
+      "loss": 0.6227,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7127272727272728,
+      "grad_norm": 1.0898888210111246,
+      "learning_rate": 2.313262067606396e-06,
+      "loss": 0.557,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7129618768328446,
+      "grad_norm": 1.0564936899031456,
+      "learning_rate": 2.3098086065952808e-06,
+      "loss": 0.5943,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7131964809384165,
+      "grad_norm": 1.9814185949779952,
+      "learning_rate": 2.3063569508957267e-06,
+      "loss": 0.6005,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7134310850439882,
+      "grad_norm": 2.386733138868949,
+      "learning_rate": 2.3029071028240434e-06,
+      "loss": 0.6237,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7136656891495601,
+      "grad_norm": 3.276239044924214,
+      "learning_rate": 2.299459064695328e-06,
+      "loss": 0.6184,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7139002932551319,
+      "grad_norm": 1.5829964122673519,
+      "learning_rate": 2.2960128388234582e-06,
+      "loss": 0.6109,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7141348973607038,
+      "grad_norm": 0.9406077157147652,
+      "learning_rate": 2.292568427521098e-06,
+      "loss": 0.6473,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7143695014662756,
+      "grad_norm": 1.9567918321818636,
+      "learning_rate": 2.289125833099702e-06,
+      "loss": 0.5906,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7146041055718475,
+      "grad_norm": 1.0668767104284897,
+      "learning_rate": 2.2856850578694906e-06,
+      "loss": 0.5668,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7148387096774194,
+      "grad_norm": 1.1161683312619683,
+      "learning_rate": 2.2822461041394745e-06,
+      "loss": 0.6033,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7150733137829912,
+      "grad_norm": 3.070349541073239,
+      "learning_rate": 2.2788089742174374e-06,
+      "loss": 0.5884,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7153079178885631,
+      "grad_norm": 1.18434704318199,
+      "learning_rate": 2.2753736704099418e-06,
+      "loss": 0.5763,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7155425219941349,
+      "grad_norm": 1.4469610247277611,
+      "learning_rate": 2.2719401950223215e-06,
+      "loss": 0.6348,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7157771260997068,
+      "grad_norm": 2.0429440492187587,
+      "learning_rate": 2.268508550358687e-06,
+      "loss": 0.6096,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7160117302052786,
+      "grad_norm": 1.1376126933156385,
+      "learning_rate": 2.2650787387219138e-06,
+      "loss": 0.6232,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7162463343108504,
+      "grad_norm": 1.0273333553927704,
+      "learning_rate": 2.2616507624136564e-06,
+      "loss": 0.6217,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7164809384164222,
+      "grad_norm": 0.9998438699287782,
+      "learning_rate": 2.258224623734333e-06,
+      "loss": 0.6015,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7167155425219941,
+      "grad_norm": 0.9329147141061209,
+      "learning_rate": 2.2548003249831256e-06,
+      "loss": 0.5317,
+      "step": 3055
+    },
+    {
+      "epoch": 0.716950146627566,
+      "grad_norm": 0.8633350498749837,
+      "learning_rate": 2.2513778684579864e-06,
+      "loss": 0.6328,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7171847507331378,
+      "grad_norm": 2.210007930230921,
+      "learning_rate": 2.2479572564556286e-06,
+      "loss": 0.6384,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7174193548387097,
+      "grad_norm": 1.200148315167745,
+      "learning_rate": 2.2445384912715285e-06,
+      "loss": 0.6012,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7176539589442815,
+      "grad_norm": 1.6038881013054826,
+      "learning_rate": 2.241121575199925e-06,
+      "loss": 0.5895,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7178885630498534,
+      "grad_norm": 1.0959333665107625,
+      "learning_rate": 2.2377065105338106e-06,
+      "loss": 0.6262,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7181231671554252,
+      "grad_norm": 0.8947350604959121,
+      "learning_rate": 2.2342932995649395e-06,
+      "loss": 0.6071,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7183577712609971,
+      "grad_norm": 10.184755961967252,
+      "learning_rate": 2.2308819445838258e-06,
+      "loss": 0.6115,
+      "step": 3062
+    },
+    {
+      "epoch": 0.718592375366569,
+      "grad_norm": 0.8963815777111561,
+      "learning_rate": 2.2274724478797284e-06,
+      "loss": 0.6656,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7188269794721408,
+      "grad_norm": 1.6515649916986137,
+      "learning_rate": 2.224064811740666e-06,
+      "loss": 0.6202,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7190615835777127,
+      "grad_norm": 1.046955887754639,
+      "learning_rate": 2.220659038453408e-06,
+      "loss": 0.6358,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7192961876832844,
+      "grad_norm": 6.840520870058273,
+      "learning_rate": 2.217255130303471e-06,
+      "loss": 0.5781,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7195307917888563,
+      "grad_norm": 0.7073585759640614,
+      "learning_rate": 2.2138530895751236e-06,
+      "loss": 0.5675,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7197653958944281,
+      "grad_norm": 0.8842575469228101,
+      "learning_rate": 2.2104529185513807e-06,
+      "loss": 0.6518,
+      "step": 3068
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3747487052793663,
+      "learning_rate": 2.207054619513995e-06,
+      "loss": 0.6211,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7202346041055718,
+      "grad_norm": 4.562252425973848,
+      "learning_rate": 2.2036581947434755e-06,
+      "loss": 0.585,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7204692082111437,
+      "grad_norm": 1.3957950102548555,
+      "learning_rate": 2.2002636465190656e-06,
+      "loss": 0.6095,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7207038123167155,
+      "grad_norm": 1.0275054610921959,
+      "learning_rate": 2.196870977118749e-06,
+      "loss": 0.5958,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7209384164222874,
+      "grad_norm": 2.3443541466146143,
+      "learning_rate": 2.19348018881925e-06,
+      "loss": 0.637,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7211730205278593,
+      "grad_norm": 1.0533846665061215,
+      "learning_rate": 2.1900912838960353e-06,
+      "loss": 0.5426,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7214076246334311,
+      "grad_norm": 1.6562542886649325,
+      "learning_rate": 2.186704264623299e-06,
+      "loss": 0.6025,
+      "step": 3075
+    },
+    {
+      "epoch": 0.721642228739003,
+      "grad_norm": 0.9924826617798908,
+      "learning_rate": 2.183319133273976e-06,
+      "loss": 0.6394,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7218768328445748,
+      "grad_norm": 4.865792980566119,
+      "learning_rate": 2.179935892119732e-06,
+      "loss": 0.6339,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7221114369501467,
+      "grad_norm": 1.0134198127887246,
+      "learning_rate": 2.176554543430965e-06,
+      "loss": 0.5715,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7223460410557184,
+      "grad_norm": 1.4089052269905165,
+      "learning_rate": 2.173175089476803e-06,
+      "loss": 0.5746,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7225806451612903,
+      "grad_norm": 1.008940549715216,
+      "learning_rate": 2.169797532525103e-06,
+      "loss": 0.6386,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7228152492668621,
+      "grad_norm": 0.8327562239505816,
+      "learning_rate": 2.166421874842446e-06,
+      "loss": 0.6307,
+      "step": 3081
+    },
+    {
+      "epoch": 0.723049853372434,
+      "grad_norm": 1.544075456904035,
+      "learning_rate": 2.1630481186941394e-06,
+      "loss": 0.5737,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7232844574780058,
+      "grad_norm": 1.481794493564964,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.6036,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7235190615835777,
+      "grad_norm": 0.8996551378170596,
+      "learning_rate": 2.156306320055443e-06,
+      "loss": 0.5978,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7237536656891496,
+      "grad_norm": 0.755147395183249,
+      "learning_rate": 2.15293828208928e-06,
+      "loss": 0.642,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7239882697947214,
+      "grad_norm": 1.5027580035435841,
+      "learning_rate": 2.149572154705929e-06,
+      "loss": 0.592,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7242228739002933,
+      "grad_norm": 3.2543859736341347,
+      "learning_rate": 2.1462079401643023e-06,
+      "loss": 0.6052,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7244574780058651,
+      "grad_norm": 1.6318345814434447,
+      "learning_rate": 2.1428456407220317e-06,
+      "loss": 0.5663,
+      "step": 3088
+    },
+    {
+      "epoch": 0.724692082111437,
+      "grad_norm": 0.9578170186463885,
+      "learning_rate": 2.1394852586354574e-06,
+      "loss": 0.6225,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7249266862170088,
+      "grad_norm": 0.883183190557881,
+      "learning_rate": 2.1361267961596377e-06,
+      "loss": 0.6094,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7251612903225807,
+      "grad_norm": 1.2821911719857417,
+      "learning_rate": 2.132770255548347e-06,
+      "loss": 0.6445,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7253958944281524,
+      "grad_norm": 1.2509694667357942,
+      "learning_rate": 2.1294156390540597e-06,
+      "loss": 0.6104,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7256304985337243,
+      "grad_norm": 1.0663919817396155,
+      "learning_rate": 2.1260629489279662e-06,
+      "loss": 0.6476,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7258651026392962,
+      "grad_norm": 1.898001441221721,
+      "learning_rate": 2.1227121874199626e-06,
+      "loss": 0.6382,
+      "step": 3094
+    },
+    {
+      "epoch": 0.726099706744868,
+      "grad_norm": 4.713467343013875,
+      "learning_rate": 2.119363356778649e-06,
+      "loss": 0.6383,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7263343108504399,
+      "grad_norm": 0.9249120017017618,
+      "learning_rate": 2.116016459251333e-06,
+      "loss": 0.662,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7265689149560117,
+      "grad_norm": 1.1245590424039145,
+      "learning_rate": 2.112671497084024e-06,
+      "loss": 0.5678,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7268035190615836,
+      "grad_norm": 3.5769692477007946,
+      "learning_rate": 2.109328472521427e-06,
+      "loss": 0.6488,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7270381231671554,
+      "grad_norm": 4.387371038068046,
+      "learning_rate": 2.105987387806956e-06,
+      "loss": 0.5981,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.8345824137163759,
+      "learning_rate": 2.1026482451827185e-06,
+      "loss": 0.5947,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7275073313782991,
+      "grad_norm": 1.2060890079193303,
+      "learning_rate": 2.0993110468895167e-06,
+      "loss": 0.6008,
+      "step": 3101
+    },
+    {
+      "epoch": 0.727741935483871,
+      "grad_norm": 1.8075769290329984,
+      "learning_rate": 2.095975795166851e-06,
+      "loss": 0.5954,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7279765395894429,
+      "grad_norm": 1.1759828909931278,
+      "learning_rate": 2.092642492252915e-06,
+      "loss": 0.598,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7282111436950147,
+      "grad_norm": 1.268712882941752,
+      "learning_rate": 2.0893111403845947e-06,
+      "loss": 0.5805,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7284457478005865,
+      "grad_norm": 1.198108753238446,
+      "learning_rate": 2.085981741797465e-06,
+      "loss": 0.5865,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7286803519061583,
+      "grad_norm": 1.508717280905453,
+      "learning_rate": 2.082654298725793e-06,
+      "loss": 0.6283,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7289149560117302,
+      "grad_norm": 1.6247561744111116,
+      "learning_rate": 2.079328813402531e-06,
+      "loss": 0.6365,
+      "step": 3107
+    },
+    {
+      "epoch": 0.729149560117302,
+      "grad_norm": 1.060366855349647,
+      "learning_rate": 2.0760052880593213e-06,
+      "loss": 0.6289,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7293841642228739,
+      "grad_norm": 1.8964263061112259,
+      "learning_rate": 2.0726837249264824e-06,
+      "loss": 0.5458,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7296187683284457,
+      "grad_norm": 2.249605547258256,
+      "learning_rate": 2.069364126233025e-06,
+      "loss": 0.6055,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7298533724340176,
+      "grad_norm": 1.5909784767012003,
+      "learning_rate": 2.0660464942066376e-06,
+      "loss": 0.6011,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7300879765395895,
+      "grad_norm": 1.5043636812013932,
+      "learning_rate": 2.06273083107369e-06,
+      "loss": 0.5516,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7303225806451613,
+      "grad_norm": 1.4742416464015022,
+      "learning_rate": 2.0594171390592294e-06,
+      "loss": 0.571,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7305571847507332,
+      "grad_norm": 3.8632603210786214,
+      "learning_rate": 2.0561054203869813e-06,
+      "loss": 0.6107,
+      "step": 3114
+    },
+    {
+      "epoch": 0.730791788856305,
+      "grad_norm": 2.841471697964061,
+      "learning_rate": 2.052795677279347e-06,
+      "loss": 0.6255,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7310263929618769,
+      "grad_norm": 1.6206507011890372,
+      "learning_rate": 2.0494879119574017e-06,
+      "loss": 0.5709,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7312609970674486,
+      "grad_norm": 0.9339932762109435,
+      "learning_rate": 2.046182126640895e-06,
+      "loss": 0.6723,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7314956011730205,
+      "grad_norm": 1.8009934526165545,
+      "learning_rate": 2.0428783235482423e-06,
+      "loss": 0.6217,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7317302052785923,
+      "grad_norm": 2.5312349835021952,
+      "learning_rate": 2.039576504896534e-06,
+      "loss": 0.6169,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7319648093841642,
+      "grad_norm": 0.9887613242226666,
+      "learning_rate": 2.036276672901531e-06,
+      "loss": 0.6332,
+      "step": 3120
+    },
+    {
+      "epoch": 0.732199413489736,
+      "grad_norm": 1.678550249720782,
+      "learning_rate": 2.0329788297776525e-06,
+      "loss": 0.6192,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7324340175953079,
+      "grad_norm": 1.1955767796139254,
+      "learning_rate": 2.0296829777379912e-06,
+      "loss": 0.6095,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7326686217008798,
+      "grad_norm": 3.0385761272879916,
+      "learning_rate": 2.0263891189942993e-06,
+      "loss": 0.617,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7329032258064516,
+      "grad_norm": 1.4161970808512983,
+      "learning_rate": 2.023097255756992e-06,
+      "loss": 0.6094,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7331378299120235,
+      "grad_norm": 0.9458671409511216,
+      "learning_rate": 2.0198073902351474e-06,
+      "loss": 0.5817,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7333724340175953,
+      "grad_norm": 3.385699080239693,
+      "learning_rate": 2.0165195246365014e-06,
+      "loss": 0.5997,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7336070381231672,
+      "grad_norm": 2.0482022284730346,
+      "learning_rate": 2.013233661167444e-06,
+      "loss": 0.6438,
+      "step": 3127
+    },
+    {
+      "epoch": 0.733841642228739,
+      "grad_norm": 1.1718917100723913,
+      "learning_rate": 2.0099498020330305e-06,
+      "loss": 0.6445,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7340762463343109,
+      "grad_norm": 1.0495146094318377,
+      "learning_rate": 2.0066679494369664e-06,
+      "loss": 0.6037,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7343108504398826,
+      "grad_norm": 0.9820637177008168,
+      "learning_rate": 2.0033881055816055e-06,
+      "loss": 0.6352,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7345454545454545,
+      "grad_norm": 1.5459291010794618,
+      "learning_rate": 2.000110272667962e-06,
+      "loss": 0.5882,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7347800586510264,
+      "grad_norm": 0.9100604994307435,
+      "learning_rate": 1.996834452895695e-06,
+      "loss": 0.6138,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7350146627565982,
+      "grad_norm": 1.5098249251636204,
+      "learning_rate": 1.993560648463117e-06,
+      "loss": 0.6248,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7352492668621701,
+      "grad_norm": 0.9405517757113439,
+      "learning_rate": 1.9902888615671868e-06,
+      "loss": 0.5962,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7354838709677419,
+      "grad_norm": 2.171708657575388,
+      "learning_rate": 1.987019094403502e-06,
+      "loss": 0.5975,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7357184750733138,
+      "grad_norm": 0.8737450754598194,
+      "learning_rate": 1.983751349166318e-06,
+      "loss": 0.6323,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7359530791788856,
+      "grad_norm": 1.2820614570051516,
+      "learning_rate": 1.980485628048525e-06,
+      "loss": 0.6098,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7361876832844575,
+      "grad_norm": 4.415409656406048,
+      "learning_rate": 1.9772219332416544e-06,
+      "loss": 0.6305,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7364222873900294,
+      "grad_norm": 1.172455834608719,
+      "learning_rate": 1.9739602669358796e-06,
+      "loss": 0.5922,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7366568914956012,
+      "grad_norm": 0.9686245660583657,
+      "learning_rate": 1.970700631320014e-06,
+      "loss": 0.6243,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7368914956011731,
+      "grad_norm": 0.9026122089982698,
+      "learning_rate": 1.967443028581507e-06,
+      "loss": 0.6328,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7371260997067449,
+      "grad_norm": 1.908394707423746,
+      "learning_rate": 1.9641874609064443e-06,
+      "loss": 0.6088,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7373607038123167,
+      "grad_norm": 1.2418215891782105,
+      "learning_rate": 1.960933930479545e-06,
+      "loss": 0.6463,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7375953079178885,
+      "grad_norm": 0.8779404299481255,
+      "learning_rate": 1.957682439484161e-06,
+      "loss": 0.6391,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7378299120234604,
+      "grad_norm": 6.762926930788082,
+      "learning_rate": 1.9544329901022774e-06,
+      "loss": 0.5739,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7380645161290322,
+      "grad_norm": 1.1959431410962325,
+      "learning_rate": 1.951185584514509e-06,
+      "loss": 0.6349,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7382991202346041,
+      "grad_norm": 5.181848344572192,
+      "learning_rate": 1.947940224900095e-06,
+      "loss": 0.6058,
+      "step": 3147
+    },
+    {
+      "epoch": 0.738533724340176,
+      "grad_norm": 1.1024974625091453,
+      "learning_rate": 1.944696913436905e-06,
+      "loss": 0.665,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7387683284457478,
+      "grad_norm": 1.3799331790467135,
+      "learning_rate": 1.941455652301438e-06,
+      "loss": 0.5731,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7390029325513197,
+      "grad_norm": 1.5318353909593199,
+      "learning_rate": 1.9382164436688083e-06,
+      "loss": 0.6505,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7392375366568915,
+      "grad_norm": 1.7414709924900138,
+      "learning_rate": 1.934979289712759e-06,
+      "loss": 0.6059,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7394721407624634,
+      "grad_norm": 2.435808441135469,
+      "learning_rate": 1.9317441926056534e-06,
+      "loss": 0.5639,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7397067448680352,
+      "grad_norm": 1.3984171255855322,
+      "learning_rate": 1.928511154518473e-06,
+      "loss": 0.6034,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7399413489736071,
+      "grad_norm": 1.5570283468737232,
+      "learning_rate": 1.9252801776208214e-06,
+      "loss": 0.6545,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7401759530791789,
+      "grad_norm": 1.7027751874831438,
+      "learning_rate": 1.922051264080912e-06,
+      "loss": 0.6166,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7404105571847507,
+      "grad_norm": 0.892403316284046,
+      "learning_rate": 1.9188244160655767e-06,
+      "loss": 0.6034,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7406451612903225,
+      "grad_norm": 1.080148125578477,
+      "learning_rate": 1.9155996357402682e-06,
+      "loss": 0.5889,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7408797653958944,
+      "grad_norm": 1.0278013041046992,
+      "learning_rate": 1.912376925269041e-06,
+      "loss": 0.6308,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7411143695014663,
+      "grad_norm": 1.474904404158769,
+      "learning_rate": 1.9091562868145665e-06,
+      "loss": 0.5791,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7413489736070381,
+      "grad_norm": 0.8000659816307549,
+      "learning_rate": 1.9059377225381237e-06,
+      "loss": 0.604,
+      "step": 3160
+    },
+    {
+      "epoch": 0.74158357771261,
+      "grad_norm": 1.0811011568216184,
+      "learning_rate": 1.9027212345996004e-06,
+      "loss": 0.5715,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7418181818181818,
+      "grad_norm": 1.351120093416797,
+      "learning_rate": 1.8995068251574917e-06,
+      "loss": 0.6177,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7420527859237537,
+      "grad_norm": 2.1420836870860547,
+      "learning_rate": 1.8962944963688984e-06,
+      "loss": 0.6011,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7422873900293255,
+      "grad_norm": 0.9624553925494452,
+      "learning_rate": 1.8930842503895203e-06,
+      "loss": 0.583,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7425219941348974,
+      "grad_norm": 1.2035760958104174,
+      "learning_rate": 1.889876089373663e-06,
+      "loss": 0.6134,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7427565982404692,
+      "grad_norm": 11.10356395089785,
+      "learning_rate": 1.8866700154742373e-06,
+      "loss": 0.5597,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7429912023460411,
+      "grad_norm": 1.3005055557435543,
+      "learning_rate": 1.8834660308427455e-06,
+      "loss": 0.587,
+      "step": 3167
+    },
+    {
+      "epoch": 0.743225806451613,
+      "grad_norm": 0.8253032565548221,
+      "learning_rate": 1.8802641376292913e-06,
+      "loss": 0.5997,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7434604105571847,
+      "grad_norm": 1.1465569201298969,
+      "learning_rate": 1.8770643379825753e-06,
+      "loss": 0.5903,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7436950146627566,
+      "grad_norm": 1.6163983338351435,
+      "learning_rate": 1.873866634049894e-06,
+      "loss": 0.5688,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7439296187683284,
+      "grad_norm": 1.8421419088577964,
+      "learning_rate": 1.8706710279771344e-06,
+      "loss": 0.6198,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7441642228739003,
+      "grad_norm": 1.1628522233856817,
+      "learning_rate": 1.867477521908781e-06,
+      "loss": 0.6059,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7443988269794721,
+      "grad_norm": 1.1566810176453983,
+      "learning_rate": 1.8642861179878996e-06,
+      "loss": 0.6138,
+      "step": 3173
+    },
+    {
+      "epoch": 0.744633431085044,
+      "grad_norm": 1.7677356551185617,
+      "learning_rate": 1.8610968183561568e-06,
+      "loss": 0.6072,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7448680351906158,
+      "grad_norm": 3.209839176190197,
+      "learning_rate": 1.8579096251538014e-06,
+      "loss": 0.5988,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7451026392961877,
+      "grad_norm": 2.7197831295626513,
+      "learning_rate": 1.8547245405196656e-06,
+      "loss": 0.6332,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7453372434017596,
+      "grad_norm": 1.1208861821326568,
+      "learning_rate": 1.8515415665911718e-06,
+      "loss": 0.6438,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7455718475073314,
+      "grad_norm": 1.1706103880893302,
+      "learning_rate": 1.8483607055043234e-06,
+      "loss": 0.6166,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7458064516129033,
+      "grad_norm": 1.047508343781435,
+      "learning_rate": 1.845181959393707e-06,
+      "loss": 0.658,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7460410557184751,
+      "grad_norm": 1.4765537772475474,
+      "learning_rate": 1.8420053303924907e-06,
+      "loss": 0.6553,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7462756598240469,
+      "grad_norm": 1.4285242167071128,
+      "learning_rate": 1.8388308206324163e-06,
+      "loss": 0.628,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7465102639296187,
+      "grad_norm": 1.5413870290351732,
+      "learning_rate": 1.835658432243812e-06,
+      "loss": 0.6168,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7467448680351906,
+      "grad_norm": 1.162157085894512,
+      "learning_rate": 1.832488167355579e-06,
+      "loss": 0.6126,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7469794721407624,
+      "grad_norm": 1.870748574842299,
+      "learning_rate": 1.829320028095189e-06,
+      "loss": 0.6164,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7472140762463343,
+      "grad_norm": 3.5789738134555473,
+      "learning_rate": 1.8261540165886922e-06,
+      "loss": 0.6272,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7474486803519061,
+      "grad_norm": 1.2668968034268808,
+      "learning_rate": 1.8229901349607094e-06,
+      "loss": 0.5763,
+      "step": 3186
+    },
+    {
+      "epoch": 0.747683284457478,
+      "grad_norm": 1.439955611019166,
+      "learning_rate": 1.8198283853344328e-06,
+      "loss": 0.626,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7479178885630499,
+      "grad_norm": 4.236895960995736,
+      "learning_rate": 1.8166687698316237e-06,
+      "loss": 0.6326,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7481524926686217,
+      "grad_norm": 0.9687410433379223,
+      "learning_rate": 1.8135112905726105e-06,
+      "loss": 0.6125,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7483870967741936,
+      "grad_norm": 1.4870595340848511,
+      "learning_rate": 1.810355949676289e-06,
+      "loss": 0.6338,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7486217008797654,
+      "grad_norm": 1.1158536244934172,
+      "learning_rate": 1.8072027492601185e-06,
+      "loss": 0.595,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7488563049853373,
+      "grad_norm": 1.2743991860418764,
+      "learning_rate": 1.804051691440125e-06,
+      "loss": 0.592,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7490909090909091,
+      "grad_norm": 0.9837471161778617,
+      "learning_rate": 1.8009027783308914e-06,
+      "loss": 0.6272,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7493255131964809,
+      "grad_norm": 1.2492660965151676,
+      "learning_rate": 1.7977560120455645e-06,
+      "loss": 0.5991,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7495601173020527,
+      "grad_norm": 1.1936748068108065,
+      "learning_rate": 1.794611394695855e-06,
+      "loss": 0.6599,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7497947214076246,
+      "grad_norm": 1.6810118197692079,
+      "learning_rate": 1.7914689283920217e-06,
+      "loss": 0.6005,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7500293255131965,
+      "grad_norm": 1.5887747861032746,
+      "learning_rate": 1.7883286152428876e-06,
+      "loss": 0.5649,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7502639296187683,
+      "grad_norm": 1.8736086111526098,
+      "learning_rate": 1.7851904573558276e-06,
+      "loss": 0.6145,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7504985337243402,
+      "grad_norm": 1.2735683110887916,
+      "learning_rate": 1.7820544568367714e-06,
+      "loss": 0.5978,
+      "step": 3199
+    },
+    {
+      "epoch": 0.750733137829912,
+      "grad_norm": 1.2712134023945538,
+      "learning_rate": 1.7789206157902023e-06,
+      "loss": 0.5872,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7509677419354839,
+      "grad_norm": 1.2131521657913562,
+      "learning_rate": 1.7757889363191484e-06,
+      "loss": 0.6117,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7512023460410557,
+      "grad_norm": 1.4220997200621723,
+      "learning_rate": 1.7726594205251917e-06,
+      "loss": 0.5951,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7514369501466276,
+      "grad_norm": 1.4570245317132258,
+      "learning_rate": 1.7695320705084678e-06,
+      "loss": 0.6275,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7516715542521994,
+      "grad_norm": 3.1706261340018926,
+      "learning_rate": 1.7664068883676478e-06,
+      "loss": 0.5926,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7519061583577713,
+      "grad_norm": 1.6096835212893639,
+      "learning_rate": 1.7632838761999539e-06,
+      "loss": 0.6522,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7521407624633432,
+      "grad_norm": 2.065351607750116,
+      "learning_rate": 1.760163036101153e-06,
+      "loss": 0.5881,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7523753665689149,
+      "grad_norm": 1.1357256647355625,
+      "learning_rate": 1.7570443701655514e-06,
+      "loss": 0.6865,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7526099706744868,
+      "grad_norm": 1.777097383952137,
+      "learning_rate": 1.7539278804859993e-06,
+      "loss": 0.5925,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7528445747800586,
+      "grad_norm": 2.4940549327241506,
+      "learning_rate": 1.7508135691538852e-06,
+      "loss": 0.5852,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7530791788856305,
+      "grad_norm": 1.4418911966589327,
+      "learning_rate": 1.7477014382591317e-06,
+      "loss": 0.6765,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7533137829912023,
+      "grad_norm": 1.6796218309635522,
+      "learning_rate": 1.7445914898902072e-06,
+      "loss": 0.596,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7535483870967742,
+      "grad_norm": 2.4018444163431614,
+      "learning_rate": 1.7414837261341095e-06,
+      "loss": 0.6741,
+      "step": 3212
+    },
+    {
+      "epoch": 0.753782991202346,
+      "grad_norm": 1.9184834069018,
+      "learning_rate": 1.7383781490763678e-06,
+      "loss": 0.6114,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7540175953079179,
+      "grad_norm": 2.32381471595833,
+      "learning_rate": 1.7352747608010495e-06,
+      "loss": 0.6452,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7542521994134898,
+      "grad_norm": 1.4461303924946431,
+      "learning_rate": 1.732173563390751e-06,
+      "loss": 0.6317,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7544868035190616,
+      "grad_norm": 1.882361661059761,
+      "learning_rate": 1.7290745589265979e-06,
+      "loss": 0.6106,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7547214076246335,
+      "grad_norm": 1.237694940304862,
+      "learning_rate": 1.7259777494882452e-06,
+      "loss": 0.5929,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7549560117302053,
+      "grad_norm": 2.42434472479123,
+      "learning_rate": 1.722883137153874e-06,
+      "loss": 0.5957,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7551906158357772,
+      "grad_norm": 1.168549637164787,
+      "learning_rate": 1.7197907240001915e-06,
+      "loss": 0.5911,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7554252199413489,
+      "grad_norm": 1.4714442871452822,
+      "learning_rate": 1.716700512102429e-06,
+      "loss": 0.5573,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7556598240469208,
+      "grad_norm": 1.2759502742647681,
+      "learning_rate": 1.7136125035343416e-06,
+      "loss": 0.5863,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7558944281524926,
+      "grad_norm": 1.105407342353222,
+      "learning_rate": 1.7105267003682019e-06,
+      "loss": 0.6292,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7561290322580645,
+      "grad_norm": 2.1463339906352097,
+      "learning_rate": 1.7074431046748075e-06,
+      "loss": 0.6516,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7563636363636363,
+      "grad_norm": 1.3566818248828167,
+      "learning_rate": 1.7043617185234718e-06,
+      "loss": 0.6561,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7565982404692082,
+      "grad_norm": 1.391925597739797,
+      "learning_rate": 1.701282543982027e-06,
+      "loss": 0.6495,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7568328445747801,
+      "grad_norm": 1.052403619890732,
+      "learning_rate": 1.6982055831168187e-06,
+      "loss": 0.6365,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7570674486803519,
+      "grad_norm": 0.9718241393674357,
+      "learning_rate": 1.6951308379927102e-06,
+      "loss": 0.6318,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7573020527859238,
+      "grad_norm": 1.5025567820685952,
+      "learning_rate": 1.6920583106730749e-06,
+      "loss": 0.6091,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7575366568914956,
+      "grad_norm": 1.082148010090495,
+      "learning_rate": 1.688988003219802e-06,
+      "loss": 0.641,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7577712609970675,
+      "grad_norm": 1.183491718195806,
+      "learning_rate": 1.685919917693284e-06,
+      "loss": 0.6996,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7580058651026393,
+      "grad_norm": 1.0837438815945954,
+      "learning_rate": 1.682854056152427e-06,
+      "loss": 0.6328,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7582404692082112,
+      "grad_norm": 1.8852225998302086,
+      "learning_rate": 1.6797904206546495e-06,
+      "loss": 0.6047,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7584750733137829,
+      "grad_norm": 3.0390157560807944,
+      "learning_rate": 1.676729013255865e-06,
+      "loss": 0.5883,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7587096774193548,
+      "grad_norm": 1.4808623864273756,
+      "learning_rate": 1.6736698360105003e-06,
+      "loss": 0.582,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7589442815249267,
+      "grad_norm": 1.3712063641076089,
+      "learning_rate": 1.6706128909714819e-06,
+      "loss": 0.6305,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7591788856304985,
+      "grad_norm": 1.4806061227165328,
+      "learning_rate": 1.6675581801902403e-06,
+      "loss": 0.5892,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7594134897360704,
+      "grad_norm": 0.8762509533048858,
+      "learning_rate": 1.6645057057167059e-06,
+      "loss": 0.616,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7596480938416422,
+      "grad_norm": 1.1895038329000147,
+      "learning_rate": 1.6614554695993085e-06,
+      "loss": 0.6203,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7598826979472141,
+      "grad_norm": 2.3442459561327187,
+      "learning_rate": 1.658407473884972e-06,
+      "loss": 0.6605,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7601173020527859,
+      "grad_norm": 2.0882630174993557,
+      "learning_rate": 1.6553617206191247e-06,
+      "loss": 0.6289,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7603519061583578,
+      "grad_norm": 2.046616949155505,
+      "learning_rate": 1.6523182118456855e-06,
+      "loss": 0.6084,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7605865102639296,
+      "grad_norm": 1.489896729297481,
+      "learning_rate": 1.6492769496070642e-06,
+      "loss": 0.6511,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7608211143695015,
+      "grad_norm": 0.9596870948622265,
+      "learning_rate": 1.6462379359441682e-06,
+      "loss": 0.5915,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7610557184750734,
+      "grad_norm": 0.8713245464286309,
+      "learning_rate": 1.6432011728963936e-06,
+      "loss": 0.6004,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7612903225806451,
+      "grad_norm": 0.9299015133859897,
+      "learning_rate": 1.640166662501626e-06,
+      "loss": 0.5891,
+      "step": 3245
+    },
+    {
+      "epoch": 0.761524926686217,
+      "grad_norm": 2.007429978690424,
+      "learning_rate": 1.6371344067962413e-06,
+      "loss": 0.5883,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7617595307917888,
+      "grad_norm": 0.7109673104179589,
+      "learning_rate": 1.6341044078150964e-06,
+      "loss": 0.6513,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7619941348973607,
+      "grad_norm": 2.522006058325781,
+      "learning_rate": 1.631076667591543e-06,
+      "loss": 0.5992,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7622287390029325,
+      "grad_norm": 1.1163809240656386,
+      "learning_rate": 1.6280511881574119e-06,
+      "loss": 0.6283,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7624633431085044,
+      "grad_norm": 1.186669010120512,
+      "learning_rate": 1.625027971543014e-06,
+      "loss": 0.5698,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7626979472140762,
+      "grad_norm": 1.3475417859065295,
+      "learning_rate": 1.622007019777146e-06,
+      "loss": 0.5819,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7629325513196481,
+      "grad_norm": 1.3745759835709592,
+      "learning_rate": 1.6189883348870837e-06,
+      "loss": 0.6516,
+      "step": 3252
+    },
+    {
+      "epoch": 0.76316715542522,
+      "grad_norm": 0.9443505990754515,
+      "learning_rate": 1.615971918898581e-06,
+      "loss": 0.5694,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7634017595307918,
+      "grad_norm": 0.7468947245991708,
+      "learning_rate": 1.6129577738358704e-06,
+      "loss": 0.6225,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7636363636363637,
+      "grad_norm": 1.9303938491739363,
+      "learning_rate": 1.6099459017216606e-06,
+      "loss": 0.6212,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7638709677419355,
+      "grad_norm": 3.6901644762067445,
+      "learning_rate": 1.606936304577129e-06,
+      "loss": 0.6265,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7641055718475074,
+      "grad_norm": 2.807876702342237,
+      "learning_rate": 1.6039289844219369e-06,
+      "loss": 0.5838,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7643401759530791,
+      "grad_norm": 9.607632025709526,
+      "learning_rate": 1.6009239432742114e-06,
+      "loss": 0.6052,
+      "step": 3258
+    },
+    {
+      "epoch": 0.764574780058651,
+      "grad_norm": 2.4273439455654806,
+      "learning_rate": 1.5979211831505475e-06,
+      "loss": 0.6034,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7648093841642228,
+      "grad_norm": 2.565759391994739,
+      "learning_rate": 1.5949207060660138e-06,
+      "loss": 0.6173,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7650439882697947,
+      "grad_norm": 1.960429815561841,
+      "learning_rate": 1.5919225140341493e-06,
+      "loss": 0.6673,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7652785923753666,
+      "grad_norm": 0.725331766460454,
+      "learning_rate": 1.5889266090669525e-06,
+      "loss": 0.653,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7655131964809384,
+      "grad_norm": 2.4796172938947345,
+      "learning_rate": 1.5859329931748918e-06,
+      "loss": 0.6185,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7657478005865103,
+      "grad_norm": 1.4892279586421546,
+      "learning_rate": 1.582941668366898e-06,
+      "loss": 0.6473,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7659824046920821,
+      "grad_norm": 2.410209132399163,
+      "learning_rate": 1.579952636650366e-06,
+      "loss": 0.6977,
+      "step": 3265
+    },
+    {
+      "epoch": 0.766217008797654,
+      "grad_norm": 4.559312836778912,
+      "learning_rate": 1.5769659000311489e-06,
+      "loss": 0.6302,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7664516129032258,
+      "grad_norm": 1.389797322576221,
+      "learning_rate": 1.5739814605135645e-06,
+      "loss": 0.5953,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7666862170087977,
+      "grad_norm": 3.8276136266417145,
+      "learning_rate": 1.5709993201003827e-06,
+      "loss": 0.6052,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7669208211143695,
+      "grad_norm": 0.8715327110089265,
+      "learning_rate": 1.5680194807928328e-06,
+      "loss": 0.5987,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7671554252199414,
+      "grad_norm": 2.118969297046594,
+      "learning_rate": 1.5650419445906062e-06,
+      "loss": 0.5737,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7673900293255131,
+      "grad_norm": 3.9105429891345294,
+      "learning_rate": 1.5620667134918389e-06,
+      "loss": 0.6153,
+      "step": 3271
+    },
+    {
+      "epoch": 0.767624633431085,
+      "grad_norm": 1.7709376221271529,
+      "learning_rate": 1.5590937894931263e-06,
+      "loss": 0.6535,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7678592375366569,
+      "grad_norm": 1.6513279442972413,
+      "learning_rate": 1.5561231745895127e-06,
+      "loss": 0.6863,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7680938416422287,
+      "grad_norm": 1.1747444830281086,
+      "learning_rate": 1.5531548707744958e-06,
+      "loss": 0.6106,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7683284457478006,
+      "grad_norm": 1.035057859924388,
+      "learning_rate": 1.5501888800400205e-06,
+      "loss": 0.6253,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7685630498533724,
+      "grad_norm": 1.3552243673443785,
+      "learning_rate": 1.5472252043764775e-06,
+      "loss": 0.5905,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7687976539589443,
+      "grad_norm": 2.0780799228519258,
+      "learning_rate": 1.544263845772705e-06,
+      "loss": 0.6417,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7690322580645161,
+      "grad_norm": 1.105916728228006,
+      "learning_rate": 1.541304806215993e-06,
+      "loss": 0.6249,
+      "step": 3278
+    },
+    {
+      "epoch": 0.769266862170088,
+      "grad_norm": 1.7097551252747039,
+      "learning_rate": 1.5383480876920636e-06,
+      "loss": 0.5781,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7695014662756599,
+      "grad_norm": 1.759568915164601,
+      "learning_rate": 1.5353936921850898e-06,
+      "loss": 0.6555,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7697360703812317,
+      "grad_norm": 1.8502303481125513,
+      "learning_rate": 1.532441621677682e-06,
+      "loss": 0.6013,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7699706744868036,
+      "grad_norm": 1.5160790920259053,
+      "learning_rate": 1.5294918781508927e-06,
+      "loss": 0.5995,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7702052785923754,
+      "grad_norm": 1.2596422018253846,
+      "learning_rate": 1.5265444635842109e-06,
+      "loss": 0.6085,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7704398826979472,
+      "grad_norm": 1.2229870670497078,
+      "learning_rate": 1.5235993799555648e-06,
+      "loss": 0.6861,
+      "step": 3284
+    },
+    {
+      "epoch": 0.770674486803519,
+      "grad_norm": 2.460475240227371,
+      "learning_rate": 1.5206566292413123e-06,
+      "loss": 0.5767,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7709090909090909,
+      "grad_norm": 1.4910074561987794,
+      "learning_rate": 1.5177162134162554e-06,
+      "loss": 0.5958,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7711436950146627,
+      "grad_norm": 1.411344643686666,
+      "learning_rate": 1.5147781344536244e-06,
+      "loss": 0.6117,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7713782991202346,
+      "grad_norm": 1.0327102545913112,
+      "learning_rate": 1.511842394325077e-06,
+      "loss": 0.6305,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7716129032258064,
+      "grad_norm": 6.381707169698441,
+      "learning_rate": 1.5089089950007085e-06,
+      "loss": 0.6463,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7718475073313783,
+      "grad_norm": 1.1841337846414857,
+      "learning_rate": 1.5059779384490398e-06,
+      "loss": 0.6157,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7720821114369502,
+      "grad_norm": 0.984782978062202,
+      "learning_rate": 1.5030492266370206e-06,
+      "loss": 0.5964,
+      "step": 3291
+    },
+    {
+      "epoch": 0.772316715542522,
+      "grad_norm": 2.110256966634448,
+      "learning_rate": 1.5001228615300263e-06,
+      "loss": 0.6114,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7725513196480939,
+      "grad_norm": 1.8412856984855048,
+      "learning_rate": 1.497198845091858e-06,
+      "loss": 0.6028,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7727859237536657,
+      "grad_norm": 1.0422277408556229,
+      "learning_rate": 1.494277179284741e-06,
+      "loss": 0.6307,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7730205278592376,
+      "grad_norm": 3.1943671941253546,
+      "learning_rate": 1.491357866069324e-06,
+      "loss": 0.6182,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7732551319648094,
+      "grad_norm": 1.3588517103236077,
+      "learning_rate": 1.488440907404673e-06,
+      "loss": 0.6184,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7734897360703812,
+      "grad_norm": 1.028187402162887,
+      "learning_rate": 1.4855263052482783e-06,
+      "loss": 0.626,
+      "step": 3297
+    },
+    {
+      "epoch": 0.773724340175953,
+      "grad_norm": 4.933856366207103,
+      "learning_rate": 1.482614061556047e-06,
+      "loss": 0.5675,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7739589442815249,
+      "grad_norm": 1.287848599817723,
+      "learning_rate": 1.4797041782823035e-06,
+      "loss": 0.6397,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 1.5406686761677237,
+      "learning_rate": 1.4767966573797886e-06,
+      "loss": 0.6583,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7744281524926686,
+      "grad_norm": 1.8461556875770175,
+      "learning_rate": 1.4738915007996574e-06,
+      "loss": 0.6391,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7746627565982405,
+      "grad_norm": 1.3167080216353213,
+      "learning_rate": 1.470988710491479e-06,
+      "loss": 0.6444,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7748973607038123,
+      "grad_norm": 1.5707265556655843,
+      "learning_rate": 1.4680882884032333e-06,
+      "loss": 0.6416,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7751319648093842,
+      "grad_norm": 0.8937236046784137,
+      "learning_rate": 1.465190236481314e-06,
+      "loss": 0.6134,
+      "step": 3304
+    },
+    {
+      "epoch": 0.775366568914956,
+      "grad_norm": 2.3669902915170695,
+      "learning_rate": 1.4622945566705188e-06,
+      "loss": 0.6163,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7756011730205279,
+      "grad_norm": 1.6036349563529806,
+      "learning_rate": 1.4594012509140566e-06,
+      "loss": 0.577,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7758357771260997,
+      "grad_norm": 1.2137070662818732,
+      "learning_rate": 1.456510321153548e-06,
+      "loss": 0.6103,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7760703812316716,
+      "grad_norm": 1.0992178316279348,
+      "learning_rate": 1.4536217693290094e-06,
+      "loss": 0.6119,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7763049853372433,
+      "grad_norm": 0.7671658076737455,
+      "learning_rate": 1.450735597378869e-06,
+      "loss": 0.679,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7765395894428152,
+      "grad_norm": 5.060920076699434,
+      "learning_rate": 1.447851807239954e-06,
+      "loss": 0.6404,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7767741935483871,
+      "grad_norm": 1.260822906990352,
+      "learning_rate": 1.4449704008474958e-06,
+      "loss": 0.6321,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7770087976539589,
+      "grad_norm": 1.6548344036130198,
+      "learning_rate": 1.4420913801351238e-06,
+      "loss": 0.576,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7772434017595308,
+      "grad_norm": 2.153777299956326,
+      "learning_rate": 1.4392147470348706e-06,
+      "loss": 0.6446,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7774780058651026,
+      "grad_norm": 1.823554274049185,
+      "learning_rate": 1.4363405034771578e-06,
+      "loss": 0.6322,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7777126099706745,
+      "grad_norm": 1.5847096202332276,
+      "learning_rate": 1.4334686513908142e-06,
+      "loss": 0.5185,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7779472140762463,
+      "grad_norm": 3.054059340863301,
+      "learning_rate": 1.4305991927030583e-06,
+      "loss": 0.6468,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7781818181818182,
+      "grad_norm": 3.064431629624091,
+      "learning_rate": 1.4277321293395002e-06,
+      "loss": 0.6333,
+      "step": 3317
+    },
+    {
+      "epoch": 0.77841642228739,
+      "grad_norm": 1.2313262936700589,
+      "learning_rate": 1.424867463224147e-06,
+      "loss": 0.635,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7786510263929619,
+      "grad_norm": 1.9063299113047427,
+      "learning_rate": 1.4220051962793952e-06,
+      "loss": 0.6024,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7788856304985338,
+      "grad_norm": 1.9564053300211146,
+      "learning_rate": 1.4191453304260305e-06,
+      "loss": 0.6241,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7791202346041056,
+      "grad_norm": 8.484623799086485,
+      "learning_rate": 1.4162878675832313e-06,
+      "loss": 0.6786,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7793548387096774,
+      "grad_norm": 3.435443621340251,
+      "learning_rate": 1.4134328096685546e-06,
+      "loss": 0.6067,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7795894428152492,
+      "grad_norm": 1.2425308672133275,
+      "learning_rate": 1.4105801585979546e-06,
+      "loss": 0.6332,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7798240469208211,
+      "grad_norm": 1.1949842893128475,
+      "learning_rate": 1.407729916285764e-06,
+      "loss": 0.5866,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7800586510263929,
+      "grad_norm": 2.9335507153002567,
+      "learning_rate": 1.4048820846446976e-06,
+      "loss": 0.5974,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7802932551319648,
+      "grad_norm": 2.6593441141100866,
+      "learning_rate": 1.4020366655858565e-06,
+      "loss": 0.6377,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7805278592375366,
+      "grad_norm": 2.245946615688388,
+      "learning_rate": 1.3991936610187208e-06,
+      "loss": 0.6443,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7807624633431085,
+      "grad_norm": 2.714745053001885,
+      "learning_rate": 1.396353072851151e-06,
+      "loss": 0.6029,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7809970674486804,
+      "grad_norm": 2.998846477985479,
+      "learning_rate": 1.3935149029893852e-06,
+      "loss": 0.6465,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7812316715542522,
+      "grad_norm": 1.791661294773031,
+      "learning_rate": 1.3906791533380398e-06,
+      "loss": 0.6322,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7814662756598241,
+      "grad_norm": 2.223055666219806,
+      "learning_rate": 1.387845825800106e-06,
+      "loss": 0.6332,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7817008797653959,
+      "grad_norm": 1.040131659888429,
+      "learning_rate": 1.38501492227695e-06,
+      "loss": 0.6324,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7819354838709678,
+      "grad_norm": 1.9881791871211159,
+      "learning_rate": 1.3821864446683126e-06,
+      "loss": 0.6359,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7821700879765396,
+      "grad_norm": 1.6620906650821152,
+      "learning_rate": 1.3793603948723027e-06,
+      "loss": 0.5877,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7824046920821114,
+      "grad_norm": 1.7234082345660948,
+      "learning_rate": 1.3765367747854031e-06,
+      "loss": 0.6428,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7826392961876832,
+      "grad_norm": 1.3111511248431982,
+      "learning_rate": 1.3737155863024692e-06,
+      "loss": 0.5781,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7828739002932551,
+      "grad_norm": 0.8901252471828437,
+      "learning_rate": 1.3708968313167165e-06,
+      "loss": 0.5707,
+      "step": 3337
+    },
+    {
+      "epoch": 0.783108504398827,
+      "grad_norm": 2.656244122080412,
+      "learning_rate": 1.3680805117197343e-06,
+      "loss": 0.6339,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7833431085043988,
+      "grad_norm": 0.9263622289797381,
+      "learning_rate": 1.3652666294014744e-06,
+      "loss": 0.5491,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7835777126099707,
+      "grad_norm": 0.9188490814641627,
+      "learning_rate": 1.362455186250254e-06,
+      "loss": 0.6536,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7838123167155425,
+      "grad_norm": 1.1099594772663544,
+      "learning_rate": 1.3596461841527553e-06,
+      "loss": 0.6266,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7840469208211144,
+      "grad_norm": 2.1744789612847817,
+      "learning_rate": 1.356839624994017e-06,
+      "loss": 0.6167,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7842815249266862,
+      "grad_norm": 1.2617211581651355,
+      "learning_rate": 1.3540355106574415e-06,
+      "loss": 0.6239,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7845161290322581,
+      "grad_norm": 1.9742936170049474,
+      "learning_rate": 1.3512338430247962e-06,
+      "loss": 0.5839,
+      "step": 3344
+    },
+    {
+      "epoch": 0.78475073313783,
+      "grad_norm": 1.0826767865312767,
+      "learning_rate": 1.348434623976196e-06,
+      "loss": 0.6618,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7849853372434018,
+      "grad_norm": 2.185132353567675,
+      "learning_rate": 1.3456378553901195e-06,
+      "loss": 0.5943,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7852199413489737,
+      "grad_norm": 1.048066889707301,
+      "learning_rate": 1.3428435391433992e-06,
+      "loss": 0.6473,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7854545454545454,
+      "grad_norm": 1.231194490593031,
+      "learning_rate": 1.340051677111222e-06,
+      "loss": 0.5767,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7856891495601173,
+      "grad_norm": 1.4452062806085455,
+      "learning_rate": 1.3372622711671274e-06,
+      "loss": 0.6548,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7859237536656891,
+      "grad_norm": 1.1827767162128813,
+      "learning_rate": 1.3344753231830094e-06,
+      "loss": 0.6323,
+      "step": 3350
+    },
+    {
+      "epoch": 0.786158357771261,
+      "grad_norm": 0.9452392352350582,
+      "learning_rate": 1.3316908350291047e-06,
+      "loss": 0.6468,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7863929618768328,
+      "grad_norm": 2.4247209122080378,
+      "learning_rate": 1.3289088085740097e-06,
+      "loss": 0.5795,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7866275659824047,
+      "grad_norm": 1.2327284477556488,
+      "learning_rate": 1.3261292456846648e-06,
+      "loss": 0.6387,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7868621700879765,
+      "grad_norm": 0.950831340782894,
+      "learning_rate": 1.323352148226351e-06,
+      "loss": 0.5497,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7870967741935484,
+      "grad_norm": 1.0809008969675262,
+      "learning_rate": 1.320577518062704e-06,
+      "loss": 0.6244,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7873313782991203,
+      "grad_norm": 1.1115781736671435,
+      "learning_rate": 1.317805357055698e-06,
+      "loss": 0.5876,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7875659824046921,
+      "grad_norm": 1.3060102063465862,
+      "learning_rate": 1.3150356670656533e-06,
+      "loss": 0.6285,
+      "step": 3357
+    },
+    {
+      "epoch": 0.787800586510264,
+      "grad_norm": 2.1245940015772193,
+      "learning_rate": 1.31226844995123e-06,
+      "loss": 0.6092,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7880351906158358,
+      "grad_norm": 1.8045950580074606,
+      "learning_rate": 1.3095037075694312e-06,
+      "loss": 0.6315,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7882697947214077,
+      "grad_norm": 1.0368790165036073,
+      "learning_rate": 1.3067414417755931e-06,
+      "loss": 0.6301,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7885043988269794,
+      "grad_norm": 12.983901457301942,
+      "learning_rate": 1.3039816544233986e-06,
+      "loss": 0.5728,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7887390029325513,
+      "grad_norm": 1.354932499189322,
+      "learning_rate": 1.3012243473648633e-06,
+      "loss": 0.6179,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7889736070381231,
+      "grad_norm": 3.908539668487798,
+      "learning_rate": 1.2984695224503351e-06,
+      "loss": 0.618,
+      "step": 3363
+    },
+    {
+      "epoch": 0.789208211143695,
+      "grad_norm": 0.860678478738657,
+      "learning_rate": 1.2957171815285014e-06,
+      "loss": 0.6253,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7894428152492668,
+      "grad_norm": 1.5337010548258105,
+      "learning_rate": 1.2929673264463798e-06,
+      "loss": 0.6361,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7896774193548387,
+      "grad_norm": 1.8146665153796193,
+      "learning_rate": 1.2902199590493209e-06,
+      "loss": 0.5863,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7899120234604106,
+      "grad_norm": 1.1848692429594583,
+      "learning_rate": 1.2874750811810062e-06,
+      "loss": 0.5762,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7901466275659824,
+      "grad_norm": 1.0084090285521348,
+      "learning_rate": 1.2847326946834427e-06,
+      "loss": 0.5974,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7903812316715543,
+      "grad_norm": 1.8962677897193514,
+      "learning_rate": 1.2819928013969717e-06,
+      "loss": 0.645,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7906158357771261,
+      "grad_norm": 2.5131656674707563,
+      "learning_rate": 1.279255403160259e-06,
+      "loss": 0.6142,
+      "step": 3370
+    },
+    {
+      "epoch": 0.790850439882698,
+      "grad_norm": 1.1443527817593289,
+      "learning_rate": 1.2765205018102928e-06,
+      "loss": 0.6114,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7910850439882698,
+      "grad_norm": 2.162940758826294,
+      "learning_rate": 1.2737880991823888e-06,
+      "loss": 0.6757,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7913196480938416,
+      "grad_norm": 2.711688830845445,
+      "learning_rate": 1.2710581971101855e-06,
+      "loss": 0.6145,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7915542521994134,
+      "grad_norm": 1.7964302931137748,
+      "learning_rate": 1.2683307974256436e-06,
+      "loss": 0.6805,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7917888563049853,
+      "grad_norm": 1.47766141071686,
+      "learning_rate": 1.2656059019590445e-06,
+      "loss": 0.6007,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7920234604105572,
+      "grad_norm": 1.13335750370577,
+      "learning_rate": 1.2628835125389877e-06,
+      "loss": 0.6222,
+      "step": 3376
+    },
+    {
+      "epoch": 0.792258064516129,
+      "grad_norm": 1.947783932586144,
+      "learning_rate": 1.2601636309923925e-06,
+      "loss": 0.5923,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7924926686217009,
+      "grad_norm": 1.1599492959540594,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.6269,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7927272727272727,
+      "grad_norm": 1.2256969852581998,
+      "learning_rate": 1.2547313988188464e-06,
+      "loss": 0.6487,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7929618768328446,
+      "grad_norm": 1.1128353455031734,
+      "learning_rate": 1.2520190518373126e-06,
+      "loss": 0.6368,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7931964809384164,
+      "grad_norm": 1.3646467602570638,
+      "learning_rate": 1.2493092200200713e-06,
+      "loss": 0.5822,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7934310850439883,
+      "grad_norm": 1.9955223075484694,
+      "learning_rate": 1.24660190518562e-06,
+      "loss": 0.5729,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7936656891495601,
+      "grad_norm": 1.0226382679391863,
+      "learning_rate": 1.2438971091507552e-06,
+      "loss": 0.5821,
+      "step": 3383
+    },
+    {
+      "epoch": 0.793900293255132,
+      "grad_norm": 0.8494424787549543,
+      "learning_rate": 1.2411948337305922e-06,
+      "loss": 0.5865,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7941348973607039,
+      "grad_norm": 1.443145038082585,
+      "learning_rate": 1.238495080738551e-06,
+      "loss": 0.6101,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7943695014662756,
+      "grad_norm": 2.338883443189615,
+      "learning_rate": 1.23579785198636e-06,
+      "loss": 0.6457,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7946041055718475,
+      "grad_norm": 0.9019626554116807,
+      "learning_rate": 1.233103149284055e-06,
+      "loss": 0.6284,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7948387096774193,
+      "grad_norm": 1.958804857967841,
+      "learning_rate": 1.2304109744399717e-06,
+      "loss": 0.6538,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7950733137829912,
+      "grad_norm": 0.733177226538545,
+      "learning_rate": 1.2277213292607526e-06,
+      "loss": 0.6726,
+      "step": 3389
+    },
+    {
+      "epoch": 0.795307917888563,
+      "grad_norm": 1.2349663035512626,
+      "learning_rate": 1.2250342155513473e-06,
+      "loss": 0.6444,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7955425219941349,
+      "grad_norm": 1.1684590968148054,
+      "learning_rate": 1.2223496351149977e-06,
+      "loss": 0.5985,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7957771260997067,
+      "grad_norm": 0.9992866474664069,
+      "learning_rate": 1.2196675897532512e-06,
+      "loss": 0.6353,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7960117302052786,
+      "grad_norm": 1.2238212730894584,
+      "learning_rate": 1.216988081265953e-06,
+      "loss": 0.5725,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7962463343108505,
+      "grad_norm": 1.8022939523151673,
+      "learning_rate": 1.2143111114512457e-06,
+      "loss": 0.7001,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7964809384164223,
+      "grad_norm": 0.9958332458784978,
+      "learning_rate": 1.211636682105568e-06,
+      "loss": 0.6095,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7967155425219942,
+      "grad_norm": 1.2532084596987885,
+      "learning_rate": 1.2089647950236556e-06,
+      "loss": 0.572,
+      "step": 3396
+    },
+    {
+      "epoch": 0.796950146627566,
+      "grad_norm": 2.4333055534970245,
+      "learning_rate": 1.2062954519985321e-06,
+      "loss": 0.5893,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7971847507331379,
+      "grad_norm": 5.836441593581866,
+      "learning_rate": 1.203628654821523e-06,
+      "loss": 0.5637,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7974193548387096,
+      "grad_norm": 1.2065652498080184,
+      "learning_rate": 1.2009644052822407e-06,
+      "loss": 0.573,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7976539589442815,
+      "grad_norm": 11.18946386473759,
+      "learning_rate": 1.1983027051685841e-06,
+      "loss": 0.6198,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7978885630498533,
+      "grad_norm": 1.2357894570784649,
+      "learning_rate": 1.1956435562667479e-06,
+      "loss": 0.6572,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7981231671554252,
+      "grad_norm": 1.8245800883873866,
+      "learning_rate": 1.1929869603612104e-06,
+      "loss": 0.6301,
+      "step": 3402
+    },
+    {
+      "epoch": 0.798357771260997,
+      "grad_norm": 0.778615321641996,
+      "learning_rate": 1.1903329192347397e-06,
+      "loss": 0.5784,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7985923753665689,
+      "grad_norm": 0.9277145250923433,
+      "learning_rate": 1.1876814346683868e-06,
+      "loss": 0.6406,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7988269794721408,
+      "grad_norm": 1.15173702577741,
+      "learning_rate": 1.1850325084414882e-06,
+      "loss": 0.5859,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7990615835777126,
+      "grad_norm": 2.1054307208618273,
+      "learning_rate": 1.1823861423316647e-06,
+      "loss": 0.6402,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7992961876832845,
+      "grad_norm": 2.9105516733225145,
+      "learning_rate": 1.1797423381148165e-06,
+      "loss": 0.6764,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7995307917888563,
+      "grad_norm": 1.2208509199172028,
+      "learning_rate": 1.1771010975651287e-06,
+      "loss": 0.607,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7997653958944282,
+      "grad_norm": 3.9541231802179593,
+      "learning_rate": 1.17446242245506e-06,
+      "loss": 0.6052,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2953208369938987,
+      "learning_rate": 1.1718263145553522e-06,
+      "loss": 0.634,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8002346041055719,
+      "grad_norm": 4.350314084605691,
+      "learning_rate": 1.169192775635023e-06,
+      "loss": 0.5889,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8004692082111436,
+      "grad_norm": 1.2319447970218724,
+      "learning_rate": 1.1665618074613655e-06,
+      "loss": 0.6296,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8007038123167155,
+      "grad_norm": 3.177908540022959,
+      "learning_rate": 1.1639334117999495e-06,
+      "loss": 0.5964,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8009384164222874,
+      "grad_norm": 3.1498970263816672,
+      "learning_rate": 1.1613075904146159e-06,
+      "loss": 0.611,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8011730205278592,
+      "grad_norm": 2.3301323900155784,
+      "learning_rate": 1.1586843450674807e-06,
+      "loss": 0.5801,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8014076246334311,
+      "grad_norm": 2.028950348561728,
+      "learning_rate": 1.15606367751893e-06,
+      "loss": 0.6185,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8016422287390029,
+      "grad_norm": 5.647118137567237,
+      "learning_rate": 1.1534455895276176e-06,
+      "loss": 0.6362,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8018768328445748,
+      "grad_norm": 1.8848009142357725,
+      "learning_rate": 1.1508300828504682e-06,
+      "loss": 0.6238,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8021114369501466,
+      "grad_norm": 1.6414677876576313,
+      "learning_rate": 1.1482171592426793e-06,
+      "loss": 0.5847,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8023460410557185,
+      "grad_norm": 1.5514532825097258,
+      "learning_rate": 1.145606820457706e-06,
+      "loss": 0.6086,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8025806451612904,
+      "grad_norm": 1.2375321438653892,
+      "learning_rate": 1.1429990682472736e-06,
+      "loss": 0.5699,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8028152492668622,
+      "grad_norm": 2.131660061599189,
+      "learning_rate": 1.140393904361372e-06,
+      "loss": 0.6265,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8030498533724341,
+      "grad_norm": 1.2230117810356314,
+      "learning_rate": 1.137791330548253e-06,
+      "loss": 0.6236,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8032844574780059,
+      "grad_norm": 3.017257196280345,
+      "learning_rate": 1.1351913485544302e-06,
+      "loss": 0.5798,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8035190615835777,
+      "grad_norm": 1.0228870218795352,
+      "learning_rate": 1.13259396012468e-06,
+      "loss": 0.6636,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8037536656891495,
+      "grad_norm": 0.9908655019649436,
+      "learning_rate": 1.1299991670020316e-06,
+      "loss": 0.6472,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8039882697947214,
+      "grad_norm": 1.2559780700329926,
+      "learning_rate": 1.1274069709277817e-06,
+      "loss": 0.6223,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8042228739002932,
+      "grad_norm": 2.5058246431430704,
+      "learning_rate": 1.1248173736414807e-06,
+      "loss": 0.6046,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8044574780058651,
+      "grad_norm": 1.4441822987176873,
+      "learning_rate": 1.1222303768809317e-06,
+      "loss": 0.5347,
+      "step": 3429
+    },
+    {
+      "epoch": 0.804692082111437,
+      "grad_norm": 0.9891278592157156,
+      "learning_rate": 1.1196459823821954e-06,
+      "loss": 0.6416,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8049266862170088,
+      "grad_norm": 3.709255185597513,
+      "learning_rate": 1.117064191879587e-06,
+      "loss": 0.6121,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8051612903225807,
+      "grad_norm": 1.0709487276616474,
+      "learning_rate": 1.1144850071056729e-06,
+      "loss": 0.5815,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8053958944281525,
+      "grad_norm": 1.6016675274815613,
+      "learning_rate": 1.111908429791273e-06,
+      "loss": 0.6087,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8056304985337244,
+      "grad_norm": 2.5992323255194822,
+      "learning_rate": 1.1093344616654512e-06,
+      "loss": 0.6105,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8058651026392962,
+      "grad_norm": 1.2465891584533022,
+      "learning_rate": 1.1067631044555287e-06,
+      "loss": 0.6321,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8060997067448681,
+      "grad_norm": 1.0656886684813924,
+      "learning_rate": 1.104194359887072e-06,
+      "loss": 0.6944,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8063343108504398,
+      "grad_norm": 0.8576471430758463,
+      "learning_rate": 1.1016282296838887e-06,
+      "loss": 0.5715,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8065689149560117,
+      "grad_norm": 3.7196268162761603,
+      "learning_rate": 1.099064715568039e-06,
+      "loss": 0.64,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8068035190615835,
+      "grad_norm": 0.9884368668341875,
+      "learning_rate": 1.0965038192598248e-06,
+      "loss": 0.5934,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8070381231671554,
+      "grad_norm": 0.917770902955031,
+      "learning_rate": 1.0939455424777912e-06,
+      "loss": 0.6006,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8072727272727273,
+      "grad_norm": 1.085614535387496,
+      "learning_rate": 1.091389886938725e-06,
+      "loss": 0.6211,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8075073313782991,
+      "grad_norm": 1.0339829162989158,
+      "learning_rate": 1.0888368543576566e-06,
+      "loss": 0.5927,
+      "step": 3442
+    },
+    {
+      "epoch": 0.807741935483871,
+      "grad_norm": 0.978303960663678,
+      "learning_rate": 1.08628644644785e-06,
+      "loss": 0.6304,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8079765395894428,
+      "grad_norm": 0.9835665862573097,
+      "learning_rate": 1.0837386649208165e-06,
+      "loss": 0.622,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8082111436950147,
+      "grad_norm": 1.105736761674846,
+      "learning_rate": 1.0811935114862992e-06,
+      "loss": 0.6267,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8084457478005865,
+      "grad_norm": 1.3358721873164068,
+      "learning_rate": 1.0786509878522773e-06,
+      "loss": 0.608,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8086803519061584,
+      "grad_norm": 1.4031151785314153,
+      "learning_rate": 1.0761110957249654e-06,
+      "loss": 0.6285,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8089149560117302,
+      "grad_norm": 1.2800841677325638,
+      "learning_rate": 1.0735738368088188e-06,
+      "loss": 0.6121,
+      "step": 3448
+    },
+    {
+      "epoch": 0.8091495601173021,
+      "grad_norm": 2.996863614109977,
+      "learning_rate": 1.0710392128065161e-06,
+      "loss": 0.6352,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8093841642228738,
+      "grad_norm": 1.2062631636769132,
+      "learning_rate": 1.0685072254189726e-06,
+      "loss": 0.587,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8096187683284457,
+      "grad_norm": 1.2920479404932967,
+      "learning_rate": 1.0659778763453343e-06,
+      "loss": 0.6107,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8098533724340176,
+      "grad_norm": 1.8474260191248044,
+      "learning_rate": 1.0634511672829756e-06,
+      "loss": 0.589,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8100879765395894,
+      "grad_norm": 0.9470352861129286,
+      "learning_rate": 1.0609270999275e-06,
+      "loss": 0.5796,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8103225806451613,
+      "grad_norm": 8.115248280785327,
+      "learning_rate": 1.058405675972739e-06,
+      "loss": 0.5878,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8105571847507331,
+      "grad_norm": 1.1784816455084388,
+      "learning_rate": 1.055886897110745e-06,
+      "loss": 0.5983,
+      "step": 3455
+    },
+    {
+      "epoch": 0.810791788856305,
+      "grad_norm": 0.9124461933172028,
+      "learning_rate": 1.0533707650318043e-06,
+      "loss": 0.5784,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8110263929618768,
+      "grad_norm": 1.049205448579507,
+      "learning_rate": 1.0508572814244206e-06,
+      "loss": 0.5951,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8112609970674487,
+      "grad_norm": 1.431606379787145,
+      "learning_rate": 1.0483464479753208e-06,
+      "loss": 0.6573,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8114956011730206,
+      "grad_norm": 0.9809468144743023,
+      "learning_rate": 1.045838266369455e-06,
+      "loss": 0.6089,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8117302052785924,
+      "grad_norm": 3.727388887782009,
+      "learning_rate": 1.0433327382899927e-06,
+      "loss": 0.5923,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8119648093841643,
+      "grad_norm": 0.8896991320966477,
+      "learning_rate": 1.040829865418324e-06,
+      "loss": 0.5877,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8121994134897361,
+      "grad_norm": 1.018395009550909,
+      "learning_rate": 1.0383296494340567e-06,
+      "loss": 0.587,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8124340175953079,
+      "grad_norm": 1.287468978398467,
+      "learning_rate": 1.0358320920150133e-06,
+      "loss": 0.5934,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8126686217008797,
+      "grad_norm": 2.7926171161451867,
+      "learning_rate": 1.0333371948372344e-06,
+      "loss": 0.5986,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8129032258064516,
+      "grad_norm": 1.8736280494825253,
+      "learning_rate": 1.0308449595749781e-06,
+      "loss": 0.613,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8131378299120234,
+      "grad_norm": 0.9349203319508304,
+      "learning_rate": 1.0283553879007097e-06,
+      "loss": 0.6019,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8133724340175953,
+      "grad_norm": 2.1523798510918186,
+      "learning_rate": 1.0258684814851116e-06,
+      "loss": 0.5917,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8136070381231671,
+      "grad_norm": 0.9878462442135967,
+      "learning_rate": 1.0233842419970773e-06,
+      "loss": 0.6228,
+      "step": 3468
+    },
+    {
+      "epoch": 0.813841642228739,
+      "grad_norm": 2.3537190534706305,
+      "learning_rate": 1.0209026711037095e-06,
+      "loss": 0.5871,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8140762463343109,
+      "grad_norm": 1.29018941978892,
+      "learning_rate": 1.0184237704703198e-06,
+      "loss": 0.6462,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8143108504398827,
+      "grad_norm": 2.022691422348032,
+      "learning_rate": 1.0159475417604302e-06,
+      "loss": 0.5954,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8145454545454546,
+      "grad_norm": 0.8903482850727282,
+      "learning_rate": 1.0134739866357645e-06,
+      "loss": 0.5839,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8147800586510264,
+      "grad_norm": 4.840718158812474,
+      "learning_rate": 1.0110031067562592e-06,
+      "loss": 0.6187,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8150146627565983,
+      "grad_norm": 2.0406564338103657,
+      "learning_rate": 1.0085349037800517e-06,
+      "loss": 0.6272,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8152492668621701,
+      "grad_norm": 3.409140244776667,
+      "learning_rate": 1.0060693793634817e-06,
+      "loss": 0.6034,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8154838709677419,
+      "grad_norm": 7.243806883878034,
+      "learning_rate": 1.0036065351610941e-06,
+      "loss": 0.6321,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8157184750733137,
+      "grad_norm": 7.195637294241193,
+      "learning_rate": 1.0011463728256338e-06,
+      "loss": 0.5766,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8159530791788856,
+      "grad_norm": 1.3114893035862916,
+      "learning_rate": 9.986888940080468e-07,
+      "loss": 0.6126,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8161876832844575,
+      "grad_norm": 2.4355311049465334,
+      "learning_rate": 9.962341003574777e-07,
+      "loss": 0.6483,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8164222873900293,
+      "grad_norm": 1.6477126802661715,
+      "learning_rate": 9.937819935212689e-07,
+      "loss": 0.5685,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8166568914956012,
+      "grad_norm": 0.9376230126087468,
+      "learning_rate": 9.913325751449604e-07,
+      "loss": 0.6642,
+      "step": 3481
+    },
+    {
+      "epoch": 0.816891495601173,
+      "grad_norm": 3.068464537116145,
+      "learning_rate": 9.88885846872289e-07,
+      "loss": 0.6068,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8171260997067449,
+      "grad_norm": 1.012110598195739,
+      "learning_rate": 9.86441810345183e-07,
+      "loss": 0.6291,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8173607038123167,
+      "grad_norm": 4.890928410756853,
+      "learning_rate": 9.840004672037662e-07,
+      "loss": 0.6299,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8175953079178886,
+      "grad_norm": 0.9490324308707202,
+      "learning_rate": 9.815618190863569e-07,
+      "loss": 0.6318,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8178299120234604,
+      "grad_norm": 1.0164596010633509,
+      "learning_rate": 9.791258676294624e-07,
+      "loss": 0.674,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8180645161290323,
+      "grad_norm": 1.5936458082809284,
+      "learning_rate": 9.7669261446778e-07,
+      "loss": 0.7048,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8182991202346042,
+      "grad_norm": 1.3760063676507426,
+      "learning_rate": 9.742620612341992e-07,
+      "loss": 0.592,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8185337243401759,
+      "grad_norm": 1.2236214272978867,
+      "learning_rate": 9.718342095597938e-07,
+      "loss": 0.6422,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8187683284457478,
+      "grad_norm": 0.9393637249751821,
+      "learning_rate": 9.694090610738282e-07,
+      "loss": 0.5863,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8190029325513196,
+      "grad_norm": 3.0544605828552607,
+      "learning_rate": 9.669866174037512e-07,
+      "loss": 0.6415,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8192375366568915,
+      "grad_norm": 0.9127177445804946,
+      "learning_rate": 9.645668801751945e-07,
+      "loss": 0.5778,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8194721407624633,
+      "grad_norm": 4.315266733744367,
+      "learning_rate": 9.621498510119753e-07,
+      "loss": 0.656,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8197067448680352,
+      "grad_norm": 1.6398331279048695,
+      "learning_rate": 9.597355315360972e-07,
+      "loss": 0.6588,
+      "step": 3494
+    },
+    {
+      "epoch": 0.819941348973607,
+      "grad_norm": 1.4045317259022378,
+      "learning_rate": 9.57323923367739e-07,
+      "loss": 0.5761,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8201759530791789,
+      "grad_norm": 0.7886031222397599,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.5902,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8204105571847508,
+      "grad_norm": 1.5585839263058723,
+      "learning_rate": 9.52508847425212e-07,
+      "loss": 0.5764,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8206451612903226,
+      "grad_norm": 1.0925902592568077,
+      "learning_rate": 9.501053828823054e-07,
+      "loss": 0.6187,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8208797653958945,
+      "grad_norm": 2.38334965345538,
+      "learning_rate": 9.4770463610944e-07,
+      "loss": 0.6244,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8211143695014663,
+      "grad_norm": 1.0805049509677784,
+      "learning_rate": 9.453066087176916e-07,
+      "loss": 0.6368,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8211143695014663,
+      "eval_loss": 0.6160818338394165,
+      "eval_runtime": 35.8884,
+      "eval_samples_per_second": 15.214,
+      "eval_steps_per_second": 0.139,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4262,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.006925861060608e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}