diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30887 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 4395,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011376564277588168,
+      "grad_norm": 0.9736748337745667,
+      "learning_rate": 0.0009997724687144482,
+      "loss": 2.1831,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022753128555176336,
+      "grad_norm": NaN,
+      "learning_rate": 0.0009997724687144482,
+      "loss": 3.1748,
+      "step": 2
+    },
+    {
+      "epoch": 0.0034129692832764505,
+      "grad_norm": NaN,
+      "learning_rate": 0.0009997724687144482,
+      "loss": 4.6431,
+      "step": 3
+    },
+    {
+      "epoch": 0.004550625711035267,
+      "grad_norm": 1.4586130380630493,
+      "learning_rate": 0.0009995449374288964,
+      "loss": 3.0014,
+      "step": 4
+    },
+    {
+      "epoch": 0.005688282138794084,
+      "grad_norm": 1.1064860820770264,
+      "learning_rate": 0.0009993174061433449,
+      "loss": 3.0825,
+      "step": 5
+    },
+    {
+      "epoch": 0.006825938566552901,
+      "grad_norm": 1.7668497562408447,
+      "learning_rate": 0.000999089874857793,
+      "loss": 4.1777,
+      "step": 6
+    },
+    {
+      "epoch": 0.007963594994311717,
+      "grad_norm": 1.0811728239059448,
+      "learning_rate": 0.000998862343572241,
+      "loss": 2.9395,
+      "step": 7
+    },
+    {
+      "epoch": 0.009101251422070534,
+      "grad_norm": 1.3404239416122437,
+      "learning_rate": 0.0009986348122866895,
+      "loss": 3.0641,
+      "step": 8
+    },
+    {
+      "epoch": 0.010238907849829351,
+      "grad_norm": 1.7664979696273804,
+      "learning_rate": 0.0009984072810011377,
+      "loss": 4.9861,
+      "step": 9
+    },
+    {
+      "epoch": 0.011376564277588168,
+      "grad_norm": 1.3834271430969238,
+      "learning_rate": 0.000998179749715586,
+      "loss": 3.3572,
+      "step": 10
+    },
+    {
+      "epoch": 0.012514220705346985,
+      "grad_norm": 1.6286646127700806,
+      "learning_rate": 0.0009979522184300341,
+      "loss": 4.866,
+      "step": 11
+    },
+    {
+      "epoch": 0.013651877133105802,
+      "grad_norm": 0.9623754620552063,
+      "learning_rate": 0.0009977246871444823,
+      "loss": 2.6571,
+      "step": 12
+    },
+    {
+      "epoch": 0.01478953356086462,
+      "grad_norm": 1.4946273565292358,
+      "learning_rate": 0.0009974971558589305,
+      "loss": 3.2579,
+      "step": 13
+    },
+    {
+      "epoch": 0.015927189988623434,
+      "grad_norm": 1.5644868612289429,
+      "learning_rate": 0.000997269624573379,
+      "loss": 4.3634,
+      "step": 14
+    },
+    {
+      "epoch": 0.017064846416382253,
+      "grad_norm": 1.1862064599990845,
+      "learning_rate": 0.0009970420932878272,
+      "loss": 2.5753,
+      "step": 15
+    },
+    {
+      "epoch": 0.01820250284414107,
+      "grad_norm": 1.8198087215423584,
+      "learning_rate": 0.0009968145620022754,
+      "loss": 2.7812,
+      "step": 16
+    },
+    {
+      "epoch": 0.019340159271899887,
+      "grad_norm": 1.6003659963607788,
+      "learning_rate": 0.0009965870307167236,
+      "loss": 3.8034,
+      "step": 17
+    },
+    {
+      "epoch": 0.020477815699658702,
+      "grad_norm": 0.6375657916069031,
+      "learning_rate": 0.0009963594994311718,
+      "loss": 1.8391,
+      "step": 18
+    },
+    {
+      "epoch": 0.02161547212741752,
+      "grad_norm": 1.5585576295852661,
+      "learning_rate": 0.00099613196814562,
+      "loss": 3.863,
+      "step": 19
+    },
+    {
+      "epoch": 0.022753128555176336,
+      "grad_norm": 1.4096543788909912,
+      "learning_rate": 0.0009959044368600682,
+      "loss": 2.7089,
+      "step": 20
+    },
+    {
+      "epoch": 0.023890784982935155,
+      "grad_norm": 1.2902367115020752,
+      "learning_rate": 0.0009956769055745164,
+      "loss": 3.508,
+      "step": 21
+    },
+    {
+      "epoch": 0.02502844141069397,
+      "grad_norm": 1.1524360179901123,
+      "learning_rate": 0.0009954493742889647,
+      "loss": 3.5223,
+      "step": 22
+    },
+    {
+      "epoch": 0.026166097838452786,
+      "grad_norm": 1.2636401653289795,
+      "learning_rate": 0.000995221843003413,
+      "loss": 2.9982,
+      "step": 23
+    },
+    {
+      "epoch": 0.027303754266211604,
+      "grad_norm": 1.3827288150787354,
+      "learning_rate": 0.0009949943117178613,
+      "loss": 4.1671,
+      "step": 24
+    },
+    {
+      "epoch": 0.02844141069397042,
+      "grad_norm": 1.1216269731521606,
+      "learning_rate": 0.0009947667804323095,
+      "loss": 2.0255,
+      "step": 25
+    },
+    {
+      "epoch": 0.02957906712172924,
+      "grad_norm": 1.6343213319778442,
+      "learning_rate": 0.0009945392491467577,
+      "loss": 3.2769,
+      "step": 26
+    },
+    {
+      "epoch": 0.030716723549488054,
+      "grad_norm": 1.988418698310852,
+      "learning_rate": 0.000994311717861206,
+      "loss": 4.1173,
+      "step": 27
+    },
+    {
+      "epoch": 0.03185437997724687,
+      "grad_norm": 1.2403652667999268,
+      "learning_rate": 0.0009940841865756541,
+      "loss": 2.7539,
+      "step": 28
+    },
+    {
+      "epoch": 0.03299203640500569,
+      "grad_norm": 1.1057771444320679,
+      "learning_rate": 0.0009938566552901023,
+      "loss": 3.5235,
+      "step": 29
+    },
+    {
+      "epoch": 0.034129692832764506,
+      "grad_norm": 1.1171739101409912,
+      "learning_rate": 0.0009936291240045505,
+      "loss": 2.8045,
+      "step": 30
+    },
+    {
+      "epoch": 0.03526734926052332,
+      "grad_norm": 1.4165009260177612,
+      "learning_rate": 0.000993401592718999,
+      "loss": 4.1957,
+      "step": 31
+    },
+    {
+      "epoch": 0.03640500568828214,
+      "grad_norm": 1.4927332401275635,
+      "learning_rate": 0.0009931740614334472,
+      "loss": 3.5043,
+      "step": 32
+    },
+    {
+      "epoch": 0.03754266211604096,
+      "grad_norm": 1.4980137348175049,
+      "learning_rate": 0.0009929465301478954,
+      "loss": 4.4631,
+      "step": 33
+    },
+    {
+      "epoch": 0.038680318543799774,
+      "grad_norm": 1.375657320022583,
+      "learning_rate": 0.0009927189988623436,
+      "loss": 3.3355,
+      "step": 34
+    },
+    {
+      "epoch": 0.03981797497155859,
+      "grad_norm": 1.5445199012756348,
+      "learning_rate": 0.0009924914675767918,
+      "loss": 4.7957,
+      "step": 35
+    },
+    {
+      "epoch": 0.040955631399317405,
+      "grad_norm": 1.644087314605713,
+      "learning_rate": 0.00099226393629124,
+      "loss": 3.2537,
+      "step": 36
+    },
+    {
+      "epoch": 0.04209328782707622,
+      "grad_norm": 1.307708978652954,
+      "learning_rate": 0.0009920364050056882,
+      "loss": 3.6423,
+      "step": 37
+    },
+    {
+      "epoch": 0.04323094425483504,
+      "grad_norm": 1.1098673343658447,
+      "learning_rate": 0.0009918088737201364,
+      "loss": 2.7579,
+      "step": 38
+    },
+    {
+      "epoch": 0.04436860068259386,
+      "grad_norm": 1.436892032623291,
+      "learning_rate": 0.0009915813424345847,
+      "loss": 2.6732,
+      "step": 39
+    },
+    {
+      "epoch": 0.04550625711035267,
+      "grad_norm": 0.9877327680587769,
+      "learning_rate": 0.000991353811149033,
+      "loss": 2.4214,
+      "step": 40
+    },
+    {
+      "epoch": 0.04664391353811149,
+      "grad_norm": 1.6885147094726562,
+      "learning_rate": 0.0009911262798634813,
+      "loss": 3.0437,
+      "step": 41
+    },
+    {
+      "epoch": 0.04778156996587031,
+      "grad_norm": 1.2590283155441284,
+      "learning_rate": 0.0009908987485779295,
+      "loss": 3.1304,
+      "step": 42
+    },
+    {
+      "epoch": 0.048919226393629126,
+      "grad_norm": 1.3581849336624146,
+      "learning_rate": 0.0009906712172923777,
+      "loss": 3.5613,
+      "step": 43
+    },
+    {
+      "epoch": 0.05005688282138794,
+      "grad_norm": 1.5315910577774048,
+      "learning_rate": 0.000990443686006826,
+      "loss": 5.662,
+      "step": 44
+    },
+    {
+      "epoch": 0.051194539249146756,
+      "grad_norm": 0.9183916449546814,
+      "learning_rate": 0.0009902161547212743,
+      "loss": 2.6213,
+      "step": 45
+    },
+    {
+      "epoch": 0.05233219567690557,
+      "grad_norm": 1.1212267875671387,
+      "learning_rate": 0.0009899886234357223,
+      "loss": 2.613,
+      "step": 46
+    },
+    {
+      "epoch": 0.053469852104664393,
+      "grad_norm": 0.9141287803649902,
+      "learning_rate": 0.0009897610921501705,
+      "loss": 2.2044,
+      "step": 47
+    },
+    {
+      "epoch": 0.05460750853242321,
+      "grad_norm": 1.1506439447402954,
+      "learning_rate": 0.000989533560864619,
+      "loss": 3.5604,
+      "step": 48
+    },
+    {
+      "epoch": 0.055745164960182024,
+      "grad_norm": 0.9535030722618103,
+      "learning_rate": 0.0009893060295790672,
+      "loss": 2.5689,
+      "step": 49
+    },
+    {
+      "epoch": 0.05688282138794084,
+      "grad_norm": 0.8492175936698914,
+      "learning_rate": 0.0009890784982935154,
+      "loss": 2.4686,
+      "step": 50
+    },
+    {
+      "epoch": 0.05802047781569966,
+      "grad_norm": 1.176944613456726,
+      "learning_rate": 0.0009888509670079636,
+      "loss": 2.9579,
+      "step": 51
+    },
+    {
+      "epoch": 0.05915813424345848,
+      "grad_norm": 1.0339919328689575,
+      "learning_rate": 0.0009886234357224118,
+      "loss": 2.6745,
+      "step": 52
+    },
+    {
+      "epoch": 0.06029579067121729,
+      "grad_norm": 0.9942947626113892,
+      "learning_rate": 0.00098839590443686,
+      "loss": 1.9818,
+      "step": 53
+    },
+    {
+      "epoch": 0.06143344709897611,
+      "grad_norm": 0.8406325578689575,
+      "learning_rate": 0.0009881683731513085,
+      "loss": 1.4172,
+      "step": 54
+    },
+    {
+      "epoch": 0.06257110352673492,
+      "grad_norm": 1.1649565696716309,
+      "learning_rate": 0.0009879408418657567,
+      "loss": 3.7562,
+      "step": 55
+    },
+    {
+      "epoch": 0.06370875995449374,
+      "grad_norm": 1.0086359977722168,
+      "learning_rate": 0.0009877133105802047,
+      "loss": 1.6131,
+      "step": 56
+    },
+    {
+      "epoch": 0.06484641638225255,
+      "grad_norm": 1.4940587282180786,
+      "learning_rate": 0.000987485779294653,
+      "loss": 2.0438,
+      "step": 57
+    },
+    {
+      "epoch": 0.06598407281001138,
+      "grad_norm": 1.365527868270874,
+      "learning_rate": 0.0009872582480091013,
+      "loss": 3.1455,
+      "step": 58
+    },
+    {
+      "epoch": 0.0671217292377702,
+      "grad_norm": 1.9875714778900146,
+      "learning_rate": 0.0009870307167235495,
+      "loss": 4.7546,
+      "step": 59
+    },
+    {
+      "epoch": 0.06825938566552901,
+      "grad_norm": 1.2569215297698975,
+      "learning_rate": 0.0009868031854379977,
+      "loss": 2.9149,
+      "step": 60
+    },
+    {
+      "epoch": 0.06939704209328783,
+      "grad_norm": 1.7797069549560547,
+      "learning_rate": 0.000986575654152446,
+      "loss": 4.0154,
+      "step": 61
+    },
+    {
+      "epoch": 0.07053469852104664,
+      "grad_norm": 1.348248839378357,
+      "learning_rate": 0.0009863481228668941,
+      "loss": 4.3475,
+      "step": 62
+    },
+    {
+      "epoch": 0.07167235494880546,
+      "grad_norm": 1.334631323814392,
+      "learning_rate": 0.0009861205915813426,
+      "loss": 4.094,
+      "step": 63
+    },
+    {
+      "epoch": 0.07281001137656427,
+      "grad_norm": 1.1636334657669067,
+      "learning_rate": 0.0009858930602957908,
+      "loss": 3.4968,
+      "step": 64
+    },
+    {
+      "epoch": 0.07394766780432309,
+      "grad_norm": 1.045125126838684,
+      "learning_rate": 0.000985665529010239,
+      "loss": 2.4535,
+      "step": 65
+    },
+    {
+      "epoch": 0.07508532423208192,
+      "grad_norm": 0.9353369474411011,
+      "learning_rate": 0.0009854379977246872,
+      "loss": 2.0262,
+      "step": 66
+    },
+    {
+      "epoch": 0.07622298065984073,
+      "grad_norm": 1.0836305618286133,
+      "learning_rate": 0.0009852104664391354,
+      "loss": 3.03,
+      "step": 67
+    },
+    {
+      "epoch": 0.07736063708759955,
+      "grad_norm": 0.8448026776313782,
+      "learning_rate": 0.0009849829351535836,
+      "loss": 2.0041,
+      "step": 68
+    },
+    {
+      "epoch": 0.07849829351535836,
+      "grad_norm": 1.4605052471160889,
+      "learning_rate": 0.0009847554038680318,
+      "loss": 3.9341,
+      "step": 69
+    },
+    {
+      "epoch": 0.07963594994311718,
+      "grad_norm": 1.040484070777893,
+      "learning_rate": 0.00098452787258248,
+      "loss": 3.1491,
+      "step": 70
+    },
+    {
+      "epoch": 0.080773606370876,
+      "grad_norm": 1.9364869594573975,
+      "learning_rate": 0.0009843003412969285,
+      "loss": 4.5523,
+      "step": 71
+    },
+    {
+      "epoch": 0.08191126279863481,
+      "grad_norm": 1.1543667316436768,
+      "learning_rate": 0.0009840728100113767,
+      "loss": 3.0408,
+      "step": 72
+    },
+    {
+      "epoch": 0.08304891922639362,
+      "grad_norm": 1.410697340965271,
+      "learning_rate": 0.0009838452787258249,
+      "loss": 3.5993,
+      "step": 73
+    },
+    {
+      "epoch": 0.08418657565415244,
+      "grad_norm": 1.4086410999298096,
+      "learning_rate": 0.000983617747440273,
+      "loss": 3.7115,
+      "step": 74
+    },
+    {
+      "epoch": 0.08532423208191127,
+      "grad_norm": 1.3084092140197754,
+      "learning_rate": 0.0009833902161547213,
+      "loss": 2.0049,
+      "step": 75
+    },
+    {
+      "epoch": 0.08646188850967008,
+      "grad_norm": 1.598218321800232,
+      "learning_rate": 0.0009831626848691695,
+      "loss": 3.0724,
+      "step": 76
+    },
+    {
+      "epoch": 0.0875995449374289,
+      "grad_norm": 0.9294804334640503,
+      "learning_rate": 0.0009829351535836177,
+      "loss": 2.953,
+      "step": 77
+    },
+    {
+      "epoch": 0.08873720136518772,
+      "grad_norm": 0.882921576499939,
+      "learning_rate": 0.000982707622298066,
+      "loss": 1.8918,
+      "step": 78
+    },
+    {
+      "epoch": 0.08987485779294653,
+      "grad_norm": 0.8168361186981201,
+      "learning_rate": 0.0009824800910125141,
+      "loss": 1.9626,
+      "step": 79
+    },
+    {
+      "epoch": 0.09101251422070535,
+      "grad_norm": 1.6279023885726929,
+      "learning_rate": 0.0009822525597269626,
+      "loss": 4.2579,
+      "step": 80
+    },
+    {
+      "epoch": 0.09215017064846416,
+      "grad_norm": 1.3332301378250122,
+      "learning_rate": 0.0009820250284414108,
+      "loss": 2.8298,
+      "step": 81
+    },
+    {
+      "epoch": 0.09328782707622298,
+      "grad_norm": 1.2498598098754883,
+      "learning_rate": 0.000981797497155859,
+      "loss": 2.9971,
+      "step": 82
+    },
+    {
+      "epoch": 0.09442548350398179,
+      "grad_norm": 0.8949465751647949,
+      "learning_rate": 0.0009815699658703072,
+      "loss": 1.7321,
+      "step": 83
+    },
+    {
+      "epoch": 0.09556313993174062,
+      "grad_norm": 0.8322932124137878,
+      "learning_rate": 0.0009813424345847554,
+      "loss": 1.9678,
+      "step": 84
+    },
+    {
+      "epoch": 0.09670079635949944,
+      "grad_norm": 1.2313084602355957,
+      "learning_rate": 0.0009811149032992036,
+      "loss": 2.7896,
+      "step": 85
+    },
+    {
+      "epoch": 0.09783845278725825,
+      "grad_norm": 0.749366283416748,
+      "learning_rate": 0.0009808873720136518,
+      "loss": 1.7082,
+      "step": 86
+    },
+    {
+      "epoch": 0.09897610921501707,
+      "grad_norm": 1.3737764358520508,
+      "learning_rate": 0.0009806598407281,
+      "loss": 4.0502,
+      "step": 87
+    },
+    {
+      "epoch": 0.10011376564277588,
+      "grad_norm": 1.0343174934387207,
+      "learning_rate": 0.0009804323094425485,
+      "loss": 4.1111,
+      "step": 88
+    },
+    {
+      "epoch": 0.1012514220705347,
+      "grad_norm": 1.1936379671096802,
+      "learning_rate": 0.0009802047781569967,
+      "loss": 3.3507,
+      "step": 89
+    },
+    {
+      "epoch": 0.10238907849829351,
+      "grad_norm": 1.1276384592056274,
+      "learning_rate": 0.0009799772468714449,
+      "loss": 2.6818,
+      "step": 90
+    },
+    {
+      "epoch": 0.10352673492605233,
+      "grad_norm": 2.1285316944122314,
+      "learning_rate": 0.000979749715585893,
+      "loss": 5.7983,
+      "step": 91
+    },
+    {
+      "epoch": 0.10466439135381114,
+      "grad_norm": 1.4918886423110962,
+      "learning_rate": 0.0009795221843003413,
+      "loss": 2.4772,
+      "step": 92
+    },
+    {
+      "epoch": 0.10580204778156997,
+      "grad_norm": 1.229989767074585,
+      "learning_rate": 0.0009792946530147895,
+      "loss": 3.1748,
+      "step": 93
+    },
+    {
+      "epoch": 0.10693970420932879,
+      "grad_norm": 1.2130271196365356,
+      "learning_rate": 0.000979067121729238,
+      "loss": 1.964,
+      "step": 94
+    },
+    {
+      "epoch": 0.1080773606370876,
+      "grad_norm": 1.5376529693603516,
+      "learning_rate": 0.000978839590443686,
+      "loss": 3.2146,
+      "step": 95
+    },
+    {
+      "epoch": 0.10921501706484642,
+      "grad_norm": 1.0450708866119385,
+      "learning_rate": 0.0009786120591581341,
+      "loss": 2.9779,
+      "step": 96
+    },
+    {
+      "epoch": 0.11035267349260523,
+      "grad_norm": 0.8169401288032532,
+      "learning_rate": 0.0009783845278725826,
+      "loss": 2.9477,
+      "step": 97
+    },
+    {
+      "epoch": 0.11149032992036405,
+      "grad_norm": 1.361669898033142,
+      "learning_rate": 0.0009781569965870308,
+      "loss": 2.842,
+      "step": 98
+    },
+    {
+      "epoch": 0.11262798634812286,
+      "grad_norm": 0.8985223174095154,
+      "learning_rate": 0.000977929465301479,
+      "loss": 2.3365,
+      "step": 99
+    },
+    {
+      "epoch": 0.11376564277588168,
+      "grad_norm": 0.9749724864959717,
+      "learning_rate": 0.0009777019340159272,
+      "loss": 1.5105,
+      "step": 100
+    },
+    {
+      "epoch": 0.1149032992036405,
+      "grad_norm": 1.0484577417373657,
+      "learning_rate": 0.0009774744027303754,
+      "loss": 1.8364,
+      "step": 101
+    },
+    {
+      "epoch": 0.11604095563139932,
+      "grad_norm": 0.9952252507209778,
+      "learning_rate": 0.0009772468714448236,
+      "loss": 2.8899,
+      "step": 102
+    },
+    {
+      "epoch": 0.11717861205915814,
+      "grad_norm": 1.3469035625457764,
+      "learning_rate": 0.000977019340159272,
+      "loss": 3.7179,
+      "step": 103
+    },
+    {
+      "epoch": 0.11831626848691695,
+      "grad_norm": 0.9715613722801208,
+      "learning_rate": 0.0009767918088737202,
+      "loss": 1.8968,
+      "step": 104
+    },
+    {
+      "epoch": 0.11945392491467577,
+      "grad_norm": 0.7201526165008545,
+      "learning_rate": 0.0009765642775881682,
+      "loss": 1.6332,
+      "step": 105
+    },
+    {
+      "epoch": 0.12059158134243458,
+      "grad_norm": 0.9688479900360107,
+      "learning_rate": 0.0009763367463026166,
+      "loss": 2.8186,
+      "step": 106
+    },
+    {
+      "epoch": 0.1217292377701934,
+      "grad_norm": 1.2450870275497437,
+      "learning_rate": 0.0009761092150170649,
+      "loss": 3.0069,
+      "step": 107
+    },
+    {
+      "epoch": 0.12286689419795221,
+      "grad_norm": 1.3213499784469604,
+      "learning_rate": 0.0009758816837315131,
+      "loss": 3.8223,
+      "step": 108
+    },
+    {
+      "epoch": 0.12400455062571103,
+      "grad_norm": 1.4514771699905396,
+      "learning_rate": 0.0009756541524459613,
+      "loss": 4.2365,
+      "step": 109
+    },
+    {
+      "epoch": 0.12514220705346984,
+      "grad_norm": 0.8881714344024658,
+      "learning_rate": 0.0009754266211604096,
+      "loss": 2.1791,
+      "step": 110
+    },
+    {
+      "epoch": 0.12627986348122866,
+      "grad_norm": 1.0673104524612427,
+      "learning_rate": 0.0009751990898748578,
+      "loss": 2.073,
+      "step": 111
+    },
+    {
+      "epoch": 0.12741751990898748,
+      "grad_norm": 1.6061453819274902,
+      "learning_rate": 0.000974971558589306,
+      "loss": 3.0393,
+      "step": 112
+    },
+    {
+      "epoch": 0.1285551763367463,
+      "grad_norm": 0.9326847791671753,
+      "learning_rate": 0.0009747440273037544,
+      "loss": 1.549,
+      "step": 113
+    },
+    {
+      "epoch": 0.1296928327645051,
+      "grad_norm": 0.8536593914031982,
+      "learning_rate": 0.0009745164960182025,
+      "loss": 1.7561,
+      "step": 114
+    },
+    {
+      "epoch": 0.13083048919226395,
+      "grad_norm": 1.1350816488265991,
+      "learning_rate": 0.0009742889647326507,
+      "loss": 4.0099,
+      "step": 115
+    },
+    {
+      "epoch": 0.13196814562002276,
+      "grad_norm": 1.5794169902801514,
+      "learning_rate": 0.000974061433447099,
+      "loss": 4.4439,
+      "step": 116
+    },
+    {
+      "epoch": 0.13310580204778158,
+      "grad_norm": 1.0470244884490967,
+      "learning_rate": 0.0009738339021615472,
+      "loss": 2.459,
+      "step": 117
+    },
+    {
+      "epoch": 0.1342434584755404,
+      "grad_norm": 0.9671889543533325,
+      "learning_rate": 0.0009736063708759955,
+      "loss": 2.7996,
+      "step": 118
+    },
+    {
+      "epoch": 0.1353811149032992,
+      "grad_norm": 1.0588785409927368,
+      "learning_rate": 0.0009733788395904437,
+      "loss": 2.4111,
+      "step": 119
+    },
+    {
+      "epoch": 0.13651877133105803,
+      "grad_norm": 1.3822710514068604,
+      "learning_rate": 0.0009731513083048919,
+      "loss": 3.7897,
+      "step": 120
+    },
+    {
+      "epoch": 0.13765642775881684,
+      "grad_norm": 1.1157410144805908,
+      "learning_rate": 0.0009729237770193402,
+      "loss": 2.8744,
+      "step": 121
+    },
+    {
+      "epoch": 0.13879408418657566,
+      "grad_norm": 1.1391868591308594,
+      "learning_rate": 0.0009726962457337885,
+      "loss": 3.3561,
+      "step": 122
+    },
+    {
+      "epoch": 0.13993174061433447,
+      "grad_norm": 0.7412275671958923,
+      "learning_rate": 0.0009724687144482367,
+      "loss": 2.0783,
+      "step": 123
+    },
+    {
+      "epoch": 0.1410693970420933,
+      "grad_norm": 1.821694254875183,
+      "learning_rate": 0.0009722411831626849,
+      "loss": 4.1648,
+      "step": 124
+    },
+    {
+      "epoch": 0.1422070534698521,
+      "grad_norm": 0.9905460476875305,
+      "learning_rate": 0.0009720136518771331,
+      "loss": 2.3263,
+      "step": 125
+    },
+    {
+      "epoch": 0.14334470989761092,
+      "grad_norm": NaN,
+      "learning_rate": 0.0009720136518771331,
+      "loss": 3.7223,
+      "step": 126
+    },
+    {
+      "epoch": 0.14448236632536973,
+      "grad_norm": 0.9033792614936829,
+      "learning_rate": 0.0009717861205915813,
+      "loss": 2.0929,
+      "step": 127
+    },
+    {
+      "epoch": 0.14562002275312855,
+      "grad_norm": 0.9408984184265137,
+      "learning_rate": 0.0009715585893060296,
+      "loss": 2.4161,
+      "step": 128
+    },
+    {
+      "epoch": 0.14675767918088736,
+      "grad_norm": 0.9367227554321289,
+      "learning_rate": 0.0009713310580204778,
+      "loss": 2.1832,
+      "step": 129
+    },
+    {
+      "epoch": 0.14789533560864618,
+      "grad_norm": 1.7745789289474487,
+      "learning_rate": 0.000971103526734926,
+      "loss": 3.0164,
+      "step": 130
+    },
+    {
+      "epoch": 0.149032992036405,
+      "grad_norm": 1.1147022247314453,
+      "learning_rate": 0.0009708759954493744,
+      "loss": 2.4758,
+      "step": 131
+    },
+    {
+      "epoch": 0.15017064846416384,
+      "grad_norm": 1.2068356275558472,
+      "learning_rate": 0.0009706484641638226,
+      "loss": 2.1346,
+      "step": 132
+    },
+    {
+      "epoch": 0.15130830489192265,
+      "grad_norm": 1.1658344268798828,
+      "learning_rate": 0.0009704209328782708,
+      "loss": 2.3267,
+      "step": 133
+    },
+    {
+      "epoch": 0.15244596131968147,
+      "grad_norm": 1.0015108585357666,
+      "learning_rate": 0.0009701934015927191,
+      "loss": 2.6728,
+      "step": 134
+    },
+    {
+      "epoch": 0.15358361774744028,
+      "grad_norm": 0.8626338243484497,
+      "learning_rate": 0.0009699658703071672,
+      "loss": 2.2007,
+      "step": 135
+    },
+    {
+      "epoch": 0.1547212741751991,
+      "grad_norm": 1.3096460103988647,
+      "learning_rate": 0.0009697383390216154,
+      "loss": 2.7757,
+      "step": 136
+    },
+    {
+      "epoch": 0.1558589306029579,
+      "grad_norm": 0.5792064070701599,
+      "learning_rate": 0.0009695108077360637,
+      "loss": 1.4733,
+      "step": 137
+    },
+    {
+      "epoch": 0.15699658703071673,
+      "grad_norm": 1.103218913078308,
+      "learning_rate": 0.0009692832764505119,
+      "loss": 3.3553,
+      "step": 138
+    },
+    {
+      "epoch": 0.15813424345847554,
+      "grad_norm": 1.1538619995117188,
+      "learning_rate": 0.0009690557451649603,
+      "loss": 2.6128,
+      "step": 139
+    },
+    {
+      "epoch": 0.15927189988623436,
+      "grad_norm": 0.8585355877876282,
+      "learning_rate": 0.0009688282138794085,
+      "loss": 1.995,
+      "step": 140
+    },
+    {
+      "epoch": 0.16040955631399317,
+      "grad_norm": 0.9618576765060425,
+      "learning_rate": 0.0009686006825938567,
+      "loss": 2.7825,
+      "step": 141
+    },
+    {
+      "epoch": 0.161547212741752,
+      "grad_norm": 1.1649035215377808,
+      "learning_rate": 0.000968373151308305,
+      "loss": 3.4463,
+      "step": 142
+    },
+    {
+      "epoch": 0.1626848691695108,
+      "grad_norm": 0.7993972301483154,
+      "learning_rate": 0.0009681456200227532,
+      "loss": 1.343,
+      "step": 143
+    },
+    {
+      "epoch": 0.16382252559726962,
+      "grad_norm": 1.05124032497406,
+      "learning_rate": 0.0009679180887372013,
+      "loss": 3.3284,
+      "step": 144
+    },
+    {
+      "epoch": 0.16496018202502843,
+      "grad_norm": 1.1311984062194824,
+      "learning_rate": 0.0009676905574516496,
+      "loss": 2.7092,
+      "step": 145
+    },
+    {
+      "epoch": 0.16609783845278725,
+      "grad_norm": 1.1213219165802002,
+      "learning_rate": 0.0009674630261660978,
+      "loss": 2.905,
+      "step": 146
+    },
+    {
+      "epoch": 0.16723549488054607,
+      "grad_norm": 1.0415153503417969,
+      "learning_rate": 0.000967235494880546,
+      "loss": 2.5957,
+      "step": 147
+    },
+    {
+      "epoch": 0.16837315130830488,
+      "grad_norm": 1.1598129272460938,
+      "learning_rate": 0.0009670079635949944,
+      "loss": 2.9467,
+      "step": 148
+    },
+    {
+      "epoch": 0.1695108077360637,
+      "grad_norm": 1.0593596696853638,
+      "learning_rate": 0.0009667804323094426,
+      "loss": 3.2002,
+      "step": 149
+    },
+    {
+      "epoch": 0.17064846416382254,
+      "grad_norm": 1.003199815750122,
+      "learning_rate": 0.0009665529010238908,
+      "loss": 2.107,
+      "step": 150
+    },
+    {
+      "epoch": 0.17178612059158135,
+      "grad_norm": 1.135500431060791,
+      "learning_rate": 0.0009663253697383391,
+      "loss": 2.8572,
+      "step": 151
+    },
+    {
+      "epoch": 0.17292377701934017,
+      "grad_norm": 3.164407730102539,
+      "learning_rate": 0.0009660978384527873,
+      "loss": 2.5331,
+      "step": 152
+    },
+    {
+      "epoch": 0.17406143344709898,
+      "grad_norm": 1.1366970539093018,
+      "learning_rate": 0.0009658703071672355,
+      "loss": 2.8843,
+      "step": 153
+    },
+    {
+      "epoch": 0.1751990898748578,
+      "grad_norm": 1.3237171173095703,
+      "learning_rate": 0.0009656427758816837,
+      "loss": 3.0479,
+      "step": 154
+    },
+    {
+      "epoch": 0.17633674630261661,
+      "grad_norm": 0.7948922514915466,
+      "learning_rate": 0.0009654152445961319,
+      "loss": 2.3427,
+      "step": 155
+    },
+    {
+      "epoch": 0.17747440273037543,
+      "grad_norm": 1.0152616500854492,
+      "learning_rate": 0.0009651877133105801,
+      "loss": 2.1661,
+      "step": 156
+    },
+    {
+      "epoch": 0.17861205915813425,
+      "grad_norm": 0.6189094185829163,
+      "learning_rate": 0.0009649601820250285,
+      "loss": 1.7452,
+      "step": 157
+    },
+    {
+      "epoch": 0.17974971558589306,
+      "grad_norm": 1.024207353591919,
+      "learning_rate": 0.0009647326507394767,
+      "loss": 3.1021,
+      "step": 158
+    },
+    {
+      "epoch": 0.18088737201365188,
+      "grad_norm": 0.9709880352020264,
+      "learning_rate": 0.000964505119453925,
+      "loss": 2.5928,
+      "step": 159
+    },
+    {
+      "epoch": 0.1820250284414107,
+      "grad_norm": 0.9320815801620483,
+      "learning_rate": 0.0009642775881683732,
+      "loss": 2.5435,
+      "step": 160
+    },
+    {
+      "epoch": 0.1831626848691695,
+      "grad_norm": 1.1968194246292114,
+      "learning_rate": 0.0009640500568828214,
+      "loss": 2.7312,
+      "step": 161
+    },
+    {
+      "epoch": 0.18430034129692832,
+      "grad_norm": 0.6706811189651489,
+      "learning_rate": 0.0009638225255972697,
+      "loss": 1.6675,
+      "step": 162
+    },
+    {
+      "epoch": 0.18543799772468714,
+      "grad_norm": 0.9634714722633362,
+      "learning_rate": 0.0009635949943117179,
+      "loss": 1.9415,
+      "step": 163
+    },
+    {
+      "epoch": 0.18657565415244595,
+      "grad_norm": 1.2071101665496826,
+      "learning_rate": 0.000963367463026166,
+      "loss": 2.3851,
+      "step": 164
+    },
+    {
+      "epoch": 0.18771331058020477,
+      "grad_norm": 1.1103309392929077,
+      "learning_rate": 0.0009631399317406144,
+      "loss": 2.4283,
+      "step": 165
+    },
+    {
+      "epoch": 0.18885096700796358,
+      "grad_norm": 0.7859079241752625,
+      "learning_rate": 0.0009629124004550626,
+      "loss": 1.6424,
+      "step": 166
+    },
+    {
+      "epoch": 0.1899886234357224,
+      "grad_norm": 1.008284568786621,
+      "learning_rate": 0.0009626848691695108,
+      "loss": 2.4152,
+      "step": 167
+    },
+    {
+      "epoch": 0.19112627986348124,
+      "grad_norm": 1.1314949989318848,
+      "learning_rate": 0.0009624573378839591,
+      "loss": 3.8599,
+      "step": 168
+    },
+    {
+      "epoch": 0.19226393629124006,
+      "grad_norm": 1.0891621112823486,
+      "learning_rate": 0.0009622298065984073,
+      "loss": 2.8165,
+      "step": 169
+    },
+    {
+      "epoch": 0.19340159271899887,
+      "grad_norm": 1.5934785604476929,
+      "learning_rate": 0.0009620022753128555,
+      "loss": 2.2033,
+      "step": 170
+    },
+    {
+      "epoch": 0.1945392491467577,
+      "grad_norm": 0.9060195684432983,
+      "learning_rate": 0.0009617747440273038,
+      "loss": 1.9928,
+      "step": 171
+    },
+    {
+      "epoch": 0.1956769055745165,
+      "grad_norm": 0.8390949964523315,
+      "learning_rate": 0.000961547212741752,
+      "loss": 2.1069,
+      "step": 172
+    },
+    {
+      "epoch": 0.19681456200227532,
+      "grad_norm": 1.1905990839004517,
+      "learning_rate": 0.0009613196814562003,
+      "loss": 2.5079,
+      "step": 173
+    },
+    {
+      "epoch": 0.19795221843003413,
+      "grad_norm": 1.3425909280776978,
+      "learning_rate": 0.0009610921501706485,
+      "loss": 3.3445,
+      "step": 174
+    },
+    {
+      "epoch": 0.19908987485779295,
+      "grad_norm": 0.9901446104049683,
+      "learning_rate": 0.0009608646188850967,
+      "loss": 3.0084,
+      "step": 175
+    },
+    {
+      "epoch": 0.20022753128555176,
+      "grad_norm": 1.7754806280136108,
+      "learning_rate": 0.0009606370875995449,
+      "loss": 2.8587,
+      "step": 176
+    },
+    {
+      "epoch": 0.20136518771331058,
+      "grad_norm": 1.0438529253005981,
+      "learning_rate": 0.0009604095563139932,
+      "loss": 2.5869,
+      "step": 177
+    },
+    {
+      "epoch": 0.2025028441410694,
+      "grad_norm": 1.0804096460342407,
+      "learning_rate": 0.0009601820250284414,
+      "loss": 3.065,
+      "step": 178
+    },
+    {
+      "epoch": 0.2036405005688282,
+      "grad_norm": 1.2588906288146973,
+      "learning_rate": 0.0009599544937428897,
+      "loss": 1.9953,
+      "step": 179
+    },
+    {
+      "epoch": 0.20477815699658702,
+      "grad_norm": 1.1029516458511353,
+      "learning_rate": 0.0009597269624573379,
+      "loss": 2.2572,
+      "step": 180
+    },
+    {
+      "epoch": 0.20591581342434584,
+      "grad_norm": 1.1141575574874878,
+      "learning_rate": 0.0009594994311717862,
+      "loss": 2.3295,
+      "step": 181
+    },
+    {
+      "epoch": 0.20705346985210465,
+      "grad_norm": 1.2331255674362183,
+      "learning_rate": 0.0009592718998862345,
+      "loss": 2.8308,
+      "step": 182
+    },
+    {
+      "epoch": 0.20819112627986347,
+      "grad_norm": 1.3520818948745728,
+      "learning_rate": 0.0009590443686006826,
+      "loss": 3.12,
+      "step": 183
+    },
+    {
+      "epoch": 0.20932878270762229,
+      "grad_norm": 1.2728475332260132,
+      "learning_rate": 0.0009588168373151308,
+      "loss": 2.8458,
+      "step": 184
+    },
+    {
+      "epoch": 0.21046643913538113,
+      "grad_norm": 0.779961884021759,
+      "learning_rate": 0.0009585893060295791,
+      "loss": 1.7038,
+      "step": 185
+    },
+    {
+      "epoch": 0.21160409556313994,
+      "grad_norm": 0.7108963131904602,
+      "learning_rate": 0.0009583617747440273,
+      "loss": 1.3587,
+      "step": 186
+    },
+    {
+      "epoch": 0.21274175199089876,
+      "grad_norm": 0.7882399559020996,
+      "learning_rate": 0.0009581342434584755,
+      "loss": 1.7968,
+      "step": 187
+    },
+    {
+      "epoch": 0.21387940841865757,
+      "grad_norm": 1.3022332191467285,
+      "learning_rate": 0.0009579067121729238,
+      "loss": 3.4653,
+      "step": 188
+    },
+    {
+      "epoch": 0.2150170648464164,
+      "grad_norm": 0.9861379265785217,
+      "learning_rate": 0.000957679180887372,
+      "loss": 2.5134,
+      "step": 189
+    },
+    {
+      "epoch": 0.2161547212741752,
+      "grad_norm": 1.3816951513290405,
+      "learning_rate": 0.0009574516496018203,
+      "loss": 1.7213,
+      "step": 190
+    },
+    {
+      "epoch": 0.21729237770193402,
+      "grad_norm": 0.7945815920829773,
+      "learning_rate": 0.0009572241183162686,
+      "loss": 1.533,
+      "step": 191
+    },
+    {
+      "epoch": 0.21843003412969283,
+      "grad_norm": 0.8966066837310791,
+      "learning_rate": 0.0009569965870307168,
+      "loss": 1.9149,
+      "step": 192
+    },
+    {
+      "epoch": 0.21956769055745165,
+      "grad_norm": 1.2404614686965942,
+      "learning_rate": 0.0009567690557451649,
+      "loss": 2.8121,
+      "step": 193
+    },
+    {
+      "epoch": 0.22070534698521047,
+      "grad_norm": 1.0486642122268677,
+      "learning_rate": 0.0009565415244596132,
+      "loss": 2.5131,
+      "step": 194
+    },
+    {
+      "epoch": 0.22184300341296928,
+      "grad_norm": 0.9867016673088074,
+      "learning_rate": 0.0009563139931740614,
+      "loss": 2.9784,
+      "step": 195
+    },
+    {
+      "epoch": 0.2229806598407281,
+      "grad_norm": 1.2857524156570435,
+      "learning_rate": 0.0009560864618885096,
+      "loss": 2.7056,
+      "step": 196
+    },
+    {
+      "epoch": 0.2241183162684869,
+      "grad_norm": 0.8611189126968384,
+      "learning_rate": 0.0009558589306029579,
+      "loss": 1.6405,
+      "step": 197
+    },
+    {
+      "epoch": 0.22525597269624573,
+      "grad_norm": 1.1097594499588013,
+      "learning_rate": 0.0009556313993174062,
+      "loss": 2.3999,
+      "step": 198
+    },
+    {
+      "epoch": 0.22639362912400454,
+      "grad_norm": 1.3522695302963257,
+      "learning_rate": 0.0009554038680318545,
+      "loss": 2.4307,
+      "step": 199
+    },
+    {
+      "epoch": 0.22753128555176336,
+      "grad_norm": 0.7342440485954285,
+      "learning_rate": 0.0009551763367463027,
+      "loss": 1.4805,
+      "step": 200
+    },
+    {
+      "epoch": 0.22866894197952217,
+      "grad_norm": 1.099187970161438,
+      "learning_rate": 0.0009549488054607509,
+      "loss": 2.4419,
+      "step": 201
+    },
+    {
+      "epoch": 0.229806598407281,
+      "grad_norm": 0.7036212682723999,
+      "learning_rate": 0.0009547212741751992,
+      "loss": 1.5371,
+      "step": 202
+    },
+    {
+      "epoch": 0.23094425483503983,
+      "grad_norm": 0.7153437733650208,
+      "learning_rate": 0.0009544937428896473,
+      "loss": 2.0429,
+      "step": 203
+    },
+    {
+      "epoch": 0.23208191126279865,
+      "grad_norm": 0.9110562801361084,
+      "learning_rate": 0.0009542662116040955,
+      "loss": 2.2197,
+      "step": 204
+    },
+    {
+      "epoch": 0.23321956769055746,
+      "grad_norm": 0.9033525586128235,
+      "learning_rate": 0.0009540386803185438,
+      "loss": 2.0181,
+      "step": 205
+    },
+    {
+      "epoch": 0.23435722411831628,
+      "grad_norm": 1.1249139308929443,
+      "learning_rate": 0.000953811149032992,
+      "loss": 2.6982,
+      "step": 206
+    },
+    {
+      "epoch": 0.2354948805460751,
+      "grad_norm": 2.2283267974853516,
+      "learning_rate": 0.0009535836177474403,
+      "loss": 3.7574,
+      "step": 207
+    },
+    {
+      "epoch": 0.2366325369738339,
+      "grad_norm": 1.0475785732269287,
+      "learning_rate": 0.0009533560864618886,
+      "loss": 2.3081,
+      "step": 208
+    },
+    {
+      "epoch": 0.23777019340159272,
+      "grad_norm": 0.9977370500564575,
+      "learning_rate": 0.0009531285551763368,
+      "loss": 2.5508,
+      "step": 209
+    },
+    {
+      "epoch": 0.23890784982935154,
+      "grad_norm": 1.2504466772079468,
+      "learning_rate": 0.000952901023890785,
+      "loss": 2.9425,
+      "step": 210
+    },
+    {
+      "epoch": 0.24004550625711035,
+      "grad_norm": 1.4991356134414673,
+      "learning_rate": 0.0009526734926052333,
+      "loss": 2.6766,
+      "step": 211
+    },
+    {
+      "epoch": 0.24118316268486917,
+      "grad_norm": 0.952013373374939,
+      "learning_rate": 0.0009524459613196815,
+      "loss": 1.3622,
+      "step": 212
+    },
+    {
+      "epoch": 0.24232081911262798,
+      "grad_norm": 0.9106293320655823,
+      "learning_rate": 0.0009522184300341296,
+      "loss": 2.2534,
+      "step": 213
+    },
+    {
+      "epoch": 0.2434584755403868,
+      "grad_norm": 1.6993825435638428,
+      "learning_rate": 0.0009519908987485779,
+      "loss": 3.5046,
+      "step": 214
+    },
+    {
+      "epoch": 0.2445961319681456,
+      "grad_norm": 1.4656329154968262,
+      "learning_rate": 0.0009517633674630262,
+      "loss": 3.9916,
+      "step": 215
+    },
+    {
+      "epoch": 0.24573378839590443,
+      "grad_norm": 0.9613212943077087,
+      "learning_rate": 0.0009515358361774744,
+      "loss": 2.7701,
+      "step": 216
+    },
+    {
+      "epoch": 0.24687144482366324,
+      "grad_norm": 1.357133388519287,
+      "learning_rate": 0.0009513083048919227,
+      "loss": 3.3809,
+      "step": 217
+    },
+    {
+      "epoch": 0.24800910125142206,
+      "grad_norm": 1.2995182275772095,
+      "learning_rate": 0.0009510807736063709,
+      "loss": 2.889,
+      "step": 218
+    },
+    {
+      "epoch": 0.24914675767918087,
+      "grad_norm": 1.4708192348480225,
+      "learning_rate": 0.0009508532423208191,
+      "loss": 2.6981,
+      "step": 219
+    },
+    {
+      "epoch": 0.2502844141069397,
+      "grad_norm": 1.0964555740356445,
+      "learning_rate": 0.0009506257110352674,
+      "loss": 3.448,
+      "step": 220
+    },
+    {
+      "epoch": 0.2514220705346985,
+      "grad_norm": 0.9647961854934692,
+      "learning_rate": 0.0009503981797497156,
+      "loss": 1.367,
+      "step": 221
+    },
+    {
+      "epoch": 0.2525597269624573,
+      "grad_norm": 1.0223067998886108,
+      "learning_rate": 0.0009501706484641638,
+      "loss": 2.653,
+      "step": 222
+    },
+    {
+      "epoch": 0.25369738339021614,
+      "grad_norm": 1.2166616916656494,
+      "learning_rate": 0.000949943117178612,
+      "loss": 3.8313,
+      "step": 223
+    },
+    {
+      "epoch": 0.25483503981797495,
+      "grad_norm": 1.2261899709701538,
+      "learning_rate": 0.0009497155858930603,
+      "loss": 3.3107,
+      "step": 224
+    },
+    {
+      "epoch": 0.25597269624573377,
+      "grad_norm": 1.2194271087646484,
+      "learning_rate": 0.0009494880546075086,
+      "loss": 2.7319,
+      "step": 225
+    },
+    {
+      "epoch": 0.2571103526734926,
+      "grad_norm": 1.196639895439148,
+      "learning_rate": 0.0009492605233219568,
+      "loss": 3.1096,
+      "step": 226
+    },
+    {
+      "epoch": 0.2582480091012514,
+      "grad_norm": 0.9778220057487488,
+      "learning_rate": 0.000949032992036405,
+      "loss": 2.5783,
+      "step": 227
+    },
+    {
+      "epoch": 0.2593856655290102,
+      "grad_norm": 1.0489749908447266,
+      "learning_rate": 0.0009488054607508533,
+      "loss": 2.1872,
+      "step": 228
+    },
+    {
+      "epoch": 0.2605233219567691,
+      "grad_norm": 2.25301194190979,
+      "learning_rate": 0.0009485779294653015,
+      "loss": 2.4844,
+      "step": 229
+    },
+    {
+      "epoch": 0.2616609783845279,
+      "grad_norm": 1.065746784210205,
+      "learning_rate": 0.0009483503981797497,
+      "loss": 3.7665,
+      "step": 230
+    },
+    {
+      "epoch": 0.2627986348122867,
+      "grad_norm": 1.1636378765106201,
+      "learning_rate": 0.0009481228668941981,
+      "loss": 2.3635,
+      "step": 231
+    },
+    {
+      "epoch": 0.26393629124004553,
+      "grad_norm": 0.915281355381012,
+      "learning_rate": 0.0009478953356086462,
+      "loss": 2.5689,
+      "step": 232
+    },
+    {
+      "epoch": 0.26507394766780434,
+      "grad_norm": 1.088327407836914,
+      "learning_rate": 0.0009476678043230944,
+      "loss": 2.3029,
+      "step": 233
+    },
+    {
+      "epoch": 0.26621160409556316,
+      "grad_norm": 0.8353843092918396,
+      "learning_rate": 0.0009474402730375427,
+      "loss": 2.1205,
+      "step": 234
+    },
+    {
+      "epoch": 0.267349260523322,
+      "grad_norm": 0.8425211906433105,
+      "learning_rate": 0.0009472127417519909,
+      "loss": 1.6575,
+      "step": 235
+    },
+    {
+      "epoch": 0.2684869169510808,
+      "grad_norm": 1.309173822402954,
+      "learning_rate": 0.0009469852104664391,
+      "loss": 3.7625,
+      "step": 236
+    },
+    {
+      "epoch": 0.2696245733788396,
+      "grad_norm": 1.0323517322540283,
+      "learning_rate": 0.0009467576791808874,
+      "loss": 2.7422,
+      "step": 237
+    },
+    {
+      "epoch": 0.2707622298065984,
+      "grad_norm": 1.0105620622634888,
+      "learning_rate": 0.0009465301478953356,
+      "loss": 2.0942,
+      "step": 238
+    },
+    {
+      "epoch": 0.27189988623435724,
+      "grad_norm": 1.2887792587280273,
+      "learning_rate": 0.0009463026166097838,
+      "loss": 3.3921,
+      "step": 239
+    },
+    {
+      "epoch": 0.27303754266211605,
+      "grad_norm": 0.8635759949684143,
+      "learning_rate": 0.0009460750853242322,
+      "loss": 1.6201,
+      "step": 240
+    },
+    {
+      "epoch": 0.27417519908987487,
+      "grad_norm": 1.0267283916473389,
+      "learning_rate": 0.0009458475540386804,
+      "loss": 2.6196,
+      "step": 241
+    },
+    {
+      "epoch": 0.2753128555176337,
+      "grad_norm": 1.3228164911270142,
+      "learning_rate": 0.0009456200227531286,
+      "loss": 2.9269,
+      "step": 242
+    },
+    {
+      "epoch": 0.2764505119453925,
+      "grad_norm": 1.1229983568191528,
+      "learning_rate": 0.0009453924914675768,
+      "loss": 2.0913,
+      "step": 243
+    },
+    {
+      "epoch": 0.2775881683731513,
+      "grad_norm": 1.4177569150924683,
+      "learning_rate": 0.000945164960182025,
+      "loss": 2.1312,
+      "step": 244
+    },
+    {
+      "epoch": 0.2787258248009101,
+      "grad_norm": 0.8409637212753296,
+      "learning_rate": 0.0009449374288964733,
+      "loss": 2.4469,
+      "step": 245
+    },
+    {
+      "epoch": 0.27986348122866894,
+      "grad_norm": 0.9386504292488098,
+      "learning_rate": 0.0009447098976109215,
+      "loss": 1.9967,
+      "step": 246
+    },
+    {
+      "epoch": 0.28100113765642776,
+      "grad_norm": 1.1676746606826782,
+      "learning_rate": 0.0009444823663253697,
+      "loss": 2.1986,
+      "step": 247
+    },
+    {
+      "epoch": 0.2821387940841866,
+      "grad_norm": 0.7538396716117859,
+      "learning_rate": 0.0009442548350398181,
+      "loss": 1.3639,
+      "step": 248
+    },
+    {
+      "epoch": 0.2832764505119454,
+      "grad_norm": 0.9919435381889343,
+      "learning_rate": 0.0009440273037542663,
+      "loss": 2.5573,
+      "step": 249
+    },
+    {
+      "epoch": 0.2844141069397042,
+      "grad_norm": 0.7212353944778442,
+      "learning_rate": 0.0009437997724687145,
+      "loss": 2.7537,
+      "step": 250
+    },
+    {
+      "epoch": 0.285551763367463,
+      "grad_norm": 1.0949044227600098,
+      "learning_rate": 0.0009435722411831627,
+      "loss": 1.8613,
+      "step": 251
+    },
+    {
+      "epoch": 0.28668941979522183,
+      "grad_norm": 1.1615073680877686,
+      "learning_rate": 0.0009433447098976109,
+      "loss": 2.3952,
+      "step": 252
+    },
+    {
+      "epoch": 0.28782707622298065,
+      "grad_norm": 1.1122605800628662,
+      "learning_rate": 0.0009431171786120591,
+      "loss": 2.5904,
+      "step": 253
+    },
+    {
+      "epoch": 0.28896473265073946,
+      "grad_norm": 1.0439378023147583,
+      "learning_rate": 0.0009428896473265074,
+      "loss": 2.2653,
+      "step": 254
+    },
+    {
+      "epoch": 0.2901023890784983,
+      "grad_norm": 2.413464069366455,
+      "learning_rate": 0.0009426621160409556,
+      "loss": 4.6422,
+      "step": 255
+    },
+    {
+      "epoch": 0.2912400455062571,
+      "grad_norm": 0.9542593955993652,
+      "learning_rate": 0.0009424345847554038,
+      "loss": 2.3552,
+      "step": 256
+    },
+    {
+      "epoch": 0.2923777019340159,
+      "grad_norm": 1.229703426361084,
+      "learning_rate": 0.0009422070534698522,
+      "loss": 2.993,
+      "step": 257
+    },
+    {
+      "epoch": 0.2935153583617747,
+      "grad_norm": 1.0281733274459839,
+      "learning_rate": 0.0009419795221843004,
+      "loss": 2.3383,
+      "step": 258
+    },
+    {
+      "epoch": 0.29465301478953354,
+      "grad_norm": 0.8514539003372192,
+      "learning_rate": 0.0009417519908987486,
+      "loss": 1.9407,
+      "step": 259
+    },
+    {
+      "epoch": 0.29579067121729236,
+      "grad_norm": 0.6966246962547302,
+      "learning_rate": 0.0009415244596131969,
+      "loss": 1.6663,
+      "step": 260
+    },
+    {
+      "epoch": 0.29692832764505117,
+      "grad_norm": 1.4110140800476074,
+      "learning_rate": 0.000941296928327645,
+      "loss": 4.4839,
+      "step": 261
+    },
+    {
+      "epoch": 0.29806598407281,
+      "grad_norm": 1.094144582748413,
+      "learning_rate": 0.0009410693970420933,
+      "loss": 2.2603,
+      "step": 262
+    },
+    {
+      "epoch": 0.2992036405005688,
+      "grad_norm": 1.0117661952972412,
+      "learning_rate": 0.0009408418657565415,
+      "loss": 1.9057,
+      "step": 263
+    },
+    {
+      "epoch": 0.3003412969283277,
+      "grad_norm": 1.1968410015106201,
+      "learning_rate": 0.0009406143344709897,
+      "loss": 2.5717,
+      "step": 264
+    },
+    {
+      "epoch": 0.3014789533560865,
+      "grad_norm": 1.324204444885254,
+      "learning_rate": 0.0009403868031854381,
+      "loss": 2.4079,
+      "step": 265
+    },
+    {
+      "epoch": 0.3026166097838453,
+      "grad_norm": 1.3995249271392822,
+      "learning_rate": 0.0009401592718998863,
+      "loss": 3.3401,
+      "step": 266
+    },
+    {
+      "epoch": 0.3037542662116041,
+      "grad_norm": 1.0412198305130005,
+      "learning_rate": 0.0009399317406143345,
+      "loss": 3.0768,
+      "step": 267
+    },
+    {
+      "epoch": 0.30489192263936293,
+      "grad_norm": 0.6889500617980957,
+      "learning_rate": 0.0009397042093287828,
+      "loss": 1.8718,
+      "step": 268
+    },
+    {
+      "epoch": 0.30602957906712175,
+      "grad_norm": 1.1345181465148926,
+      "learning_rate": 0.000939476678043231,
+      "loss": 3.4683,
+      "step": 269
+    },
+    {
+      "epoch": 0.30716723549488056,
+      "grad_norm": 0.8793913722038269,
+      "learning_rate": 0.0009392491467576792,
+      "loss": 2.2562,
+      "step": 270
+    },
+    {
+      "epoch": 0.3083048919226394,
+      "grad_norm": 1.0166738033294678,
+      "learning_rate": 0.0009390216154721274,
+      "loss": 1.328,
+      "step": 271
+    },
+    {
+      "epoch": 0.3094425483503982,
+      "grad_norm": 0.817034900188446,
+      "learning_rate": 0.0009387940841865756,
+      "loss": 2.4001,
+      "step": 272
+    },
+    {
+      "epoch": 0.310580204778157,
+      "grad_norm": 0.9647735357284546,
+      "learning_rate": 0.0009385665529010238,
+      "loss": 2.2161,
+      "step": 273
+    },
+    {
+      "epoch": 0.3117178612059158,
+      "grad_norm": 0.9152204394340515,
+      "learning_rate": 0.0009383390216154722,
+      "loss": 2.5515,
+      "step": 274
+    },
+    {
+      "epoch": 0.31285551763367464,
+      "grad_norm": 1.4312481880187988,
+      "learning_rate": 0.0009381114903299204,
+      "loss": 4.4949,
+      "step": 275
+    },
+    {
+      "epoch": 0.31399317406143346,
+      "grad_norm": 0.9626976847648621,
+      "learning_rate": 0.0009378839590443686,
+      "loss": 1.8928,
+      "step": 276
+    },
+    {
+      "epoch": 0.31513083048919227,
+      "grad_norm": 2.2159972190856934,
+      "learning_rate": 0.0009376564277588169,
+      "loss": 4.2681,
+      "step": 277
+    },
+    {
+      "epoch": 0.3162684869169511,
+      "grad_norm": 1.171950101852417,
+      "learning_rate": 0.0009374288964732651,
+      "loss": 2.1355,
+      "step": 278
+    },
+    {
+      "epoch": 0.3174061433447099,
+      "grad_norm": 0.6437592506408691,
+      "learning_rate": 0.0009372013651877133,
+      "loss": 1.5357,
+      "step": 279
+    },
+    {
+      "epoch": 0.3185437997724687,
+      "grad_norm": 1.0927355289459229,
+      "learning_rate": 0.0009369738339021616,
+      "loss": 2.1487,
+      "step": 280
+    },
+    {
+      "epoch": 0.31968145620022753,
+      "grad_norm": 1.1274104118347168,
+      "learning_rate": 0.0009367463026166097,
+      "loss": 2.6737,
+      "step": 281
+    },
+    {
+      "epoch": 0.32081911262798635,
+      "grad_norm": 0.9172748327255249,
+      "learning_rate": 0.000936518771331058,
+      "loss": 1.8367,
+      "step": 282
+    },
+    {
+      "epoch": 0.32195676905574516,
+      "grad_norm": 1.2308052778244019,
+      "learning_rate": 0.0009362912400455063,
+      "loss": 2.4757,
+      "step": 283
+    },
+    {
+      "epoch": 0.323094425483504,
+      "grad_norm": 0.8640644550323486,
+      "learning_rate": 0.0009360637087599545,
+      "loss": 2.0472,
+      "step": 284
+    },
+    {
+      "epoch": 0.3242320819112628,
+      "grad_norm": 1.4058482646942139,
+      "learning_rate": 0.0009358361774744028,
+      "loss": 2.8178,
+      "step": 285
+    },
+    {
+      "epoch": 0.3253697383390216,
+      "grad_norm": 1.1872564554214478,
+      "learning_rate": 0.000935608646188851,
+      "loss": 2.3008,
+      "step": 286
+    },
+    {
+      "epoch": 0.3265073947667804,
+      "grad_norm": 0.7326186299324036,
+      "learning_rate": 0.0009353811149032992,
+      "loss": 1.173,
+      "step": 287
+    },
+    {
+      "epoch": 0.32764505119453924,
+      "grad_norm": 1.229687213897705,
+      "learning_rate": 0.0009351535836177475,
+      "loss": 1.8621,
+      "step": 288
+    },
+    {
+      "epoch": 0.32878270762229805,
+      "grad_norm": 1.0105575323104858,
+      "learning_rate": 0.0009349260523321957,
+      "loss": 2.6121,
+      "step": 289
+    },
+    {
+      "epoch": 0.32992036405005687,
+      "grad_norm": 1.0747400522232056,
+      "learning_rate": 0.0009346985210466438,
+      "loss": 2.3038,
+      "step": 290
+    },
+    {
+      "epoch": 0.3310580204778157,
+      "grad_norm": 1.3244582414627075,
+      "learning_rate": 0.0009344709897610922,
+      "loss": 1.9476,
+      "step": 291
+    },
+    {
+      "epoch": 0.3321956769055745,
+      "grad_norm": 0.8411991000175476,
+      "learning_rate": 0.0009342434584755404,
+      "loss": 1.8068,
+      "step": 292
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.9534935355186462,
+      "learning_rate": 0.0009340159271899886,
+      "loss": 2.0956,
+      "step": 293
+    },
+    {
+      "epoch": 0.33447098976109213,
+      "grad_norm": 1.0585649013519287,
+      "learning_rate": 0.0009337883959044369,
+      "loss": 2.7111,
+      "step": 294
+    },
+    {
+      "epoch": 0.33560864618885095,
+      "grad_norm": 0.7823183536529541,
+      "learning_rate": 0.0009335608646188851,
+      "loss": 1.9446,
+      "step": 295
+    },
+    {
+      "epoch": 0.33674630261660976,
+      "grad_norm": 0.8460807204246521,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 2.6191,
+      "step": 296
+    },
+    {
+      "epoch": 0.3378839590443686,
+      "grad_norm": 1.1409828662872314,
+      "learning_rate": 0.0009331058020477816,
+      "loss": 3.1742,
+      "step": 297
+    },
+    {
+      "epoch": 0.3390216154721274,
+      "grad_norm": 1.1377147436141968,
+      "learning_rate": 0.0009328782707622299,
+      "loss": 2.7242,
+      "step": 298
+    },
+    {
+      "epoch": 0.34015927189988626,
+      "grad_norm": 1.4707083702087402,
+      "learning_rate": 0.0009326507394766781,
+      "loss": 3.7897,
+      "step": 299
+    },
+    {
+      "epoch": 0.3412969283276451,
+      "grad_norm": 0.964156448841095,
+      "learning_rate": 0.0009324232081911263,
+      "loss": 2.0425,
+      "step": 300
+    },
+    {
+      "epoch": 0.3424345847554039,
+      "grad_norm": 1.323971152305603,
+      "learning_rate": 0.0009321956769055745,
+      "loss": 3.1034,
+      "step": 301
+    },
+    {
+      "epoch": 0.3435722411831627,
+      "grad_norm": 1.3859295845031738,
+      "learning_rate": 0.0009319681456200227,
+      "loss": 2.9186,
+      "step": 302
+    },
+    {
+      "epoch": 0.3447098976109215,
+      "grad_norm": 0.8150478601455688,
+      "learning_rate": 0.000931740614334471,
+      "loss": 2.3752,
+      "step": 303
+    },
+    {
+      "epoch": 0.34584755403868034,
+      "grad_norm": 0.8035518527030945,
+      "learning_rate": 0.0009315130830489192,
+      "loss": 2.252,
+      "step": 304
+    },
+    {
+      "epoch": 0.34698521046643915,
+      "grad_norm": 1.2278333902359009,
+      "learning_rate": 0.0009312855517633675,
+      "loss": 2.3787,
+      "step": 305
+    },
+    {
+      "epoch": 0.34812286689419797,
+      "grad_norm": 1.116999626159668,
+      "learning_rate": 0.0009310580204778157,
+      "loss": 1.4018,
+      "step": 306
+    },
+    {
+      "epoch": 0.3492605233219568,
+      "grad_norm": 1.0040169954299927,
+      "learning_rate": 0.000930830489192264,
+      "loss": 2.0511,
+      "step": 307
+    },
+    {
+      "epoch": 0.3503981797497156,
+      "grad_norm": 0.901432991027832,
+      "learning_rate": 0.0009306029579067123,
+      "loss": 1.8721,
+      "step": 308
+    },
+    {
+      "epoch": 0.3515358361774744,
+      "grad_norm": 1.451587438583374,
+      "learning_rate": 0.0009303754266211605,
+      "loss": 3.0643,
+      "step": 309
+    },
+    {
+      "epoch": 0.35267349260523323,
+      "grad_norm": 1.2055668830871582,
+      "learning_rate": 0.0009301478953356086,
+      "loss": 3.2057,
+      "step": 310
+    },
+    {
+      "epoch": 0.35381114903299204,
+      "grad_norm": 0.8358264565467834,
+      "learning_rate": 0.0009299203640500569,
+      "loss": 2.0797,
+      "step": 311
+    },
+    {
+      "epoch": 0.35494880546075086,
+      "grad_norm": 0.7835274338722229,
+      "learning_rate": 0.0009296928327645051,
+      "loss": 1.2359,
+      "step": 312
+    },
+    {
+      "epoch": 0.3560864618885097,
+      "grad_norm": 0.9940115809440613,
+      "learning_rate": 0.0009294653014789533,
+      "loss": 2.2001,
+      "step": 313
+    },
+    {
+      "epoch": 0.3572241183162685,
+      "grad_norm": 1.147113561630249,
+      "learning_rate": 0.0009292377701934016,
+      "loss": 2.0408,
+      "step": 314
+    },
+    {
+      "epoch": 0.3583617747440273,
+      "grad_norm": 0.4681651294231415,
+      "learning_rate": 0.0009290102389078499,
+      "loss": 0.7199,
+      "step": 315
+    },
+    {
+      "epoch": 0.3594994311717861,
+      "grad_norm": 1.2197333574295044,
+      "learning_rate": 0.0009287827076222981,
+      "loss": 3.6619,
+      "step": 316
+    },
+    {
+      "epoch": 0.36063708759954494,
+      "grad_norm": 0.9935864806175232,
+      "learning_rate": 0.0009285551763367464,
+      "loss": 2.0533,
+      "step": 317
+    },
+    {
+      "epoch": 0.36177474402730375,
+      "grad_norm": 0.9229695200920105,
+      "learning_rate": 0.0009283276450511946,
+      "loss": 1.4724,
+      "step": 318
+    },
+    {
+      "epoch": 0.36291240045506257,
+      "grad_norm": 1.3133660554885864,
+      "learning_rate": 0.0009281001137656428,
+      "loss": 2.9495,
+      "step": 319
+    },
+    {
+      "epoch": 0.3640500568828214,
+      "grad_norm": 0.9246838092803955,
+      "learning_rate": 0.000927872582480091,
+      "loss": 1.6493,
+      "step": 320
+    },
+    {
+      "epoch": 0.3651877133105802,
+      "grad_norm": 1.4214354753494263,
+      "learning_rate": 0.0009276450511945392,
+      "loss": 2.2151,
+      "step": 321
+    },
+    {
+      "epoch": 0.366325369738339,
+      "grad_norm": 1.167389988899231,
+      "learning_rate": 0.0009274175199089874,
+      "loss": 2.1775,
+      "step": 322
+    },
+    {
+      "epoch": 0.36746302616609783,
+      "grad_norm": 1.3376412391662598,
+      "learning_rate": 0.0009271899886234357,
+      "loss": 3.2754,
+      "step": 323
+    },
+    {
+      "epoch": 0.36860068259385664,
+      "grad_norm": 0.9739036560058594,
+      "learning_rate": 0.000926962457337884,
+      "loss": 2.151,
+      "step": 324
+    },
+    {
+      "epoch": 0.36973833902161546,
+      "grad_norm": 0.9827941656112671,
+      "learning_rate": 0.0009267349260523323,
+      "loss": 3.0028,
+      "step": 325
+    },
+    {
+      "epoch": 0.3708759954493743,
+      "grad_norm": 0.811549723148346,
+      "learning_rate": 0.0009265073947667805,
+      "loss": 2.1614,
+      "step": 326
+    },
+    {
+      "epoch": 0.3720136518771331,
+      "grad_norm": 0.9825149178504944,
+      "learning_rate": 0.0009262798634812287,
+      "loss": 1.7451,
+      "step": 327
+    },
+    {
+      "epoch": 0.3731513083048919,
+      "grad_norm": 1.237622857093811,
+      "learning_rate": 0.000926052332195677,
+      "loss": 3.3102,
+      "step": 328
+    },
+    {
+      "epoch": 0.3742889647326507,
+      "grad_norm": 0.9079214334487915,
+      "learning_rate": 0.0009258248009101251,
+      "loss": 2.0276,
+      "step": 329
+    },
+    {
+      "epoch": 0.37542662116040953,
+      "grad_norm": 0.7722564935684204,
+      "learning_rate": 0.0009255972696245733,
+      "loss": 2.1105,
+      "step": 330
+    },
+    {
+      "epoch": 0.37656427758816835,
+      "grad_norm": 0.9066492319107056,
+      "learning_rate": 0.0009253697383390216,
+      "loss": 3.4384,
+      "step": 331
+    },
+    {
+      "epoch": 0.37770193401592717,
+      "grad_norm": 0.8562538623809814,
+      "learning_rate": 0.0009251422070534699,
+      "loss": 2.9233,
+      "step": 332
+    },
+    {
+      "epoch": 0.378839590443686,
+      "grad_norm": 1.1646368503570557,
+      "learning_rate": 0.0009249146757679181,
+      "loss": 2.551,
+      "step": 333
+    },
+    {
+      "epoch": 0.3799772468714448,
+      "grad_norm": 0.9063814878463745,
+      "learning_rate": 0.0009246871444823664,
+      "loss": 2.3003,
+      "step": 334
+    },
+    {
+      "epoch": 0.38111490329920367,
+      "grad_norm": 0.993689775466919,
+      "learning_rate": 0.0009244596131968146,
+      "loss": 3.3941,
+      "step": 335
+    },
+    {
+      "epoch": 0.3822525597269625,
+      "grad_norm": 1.2754982709884644,
+      "learning_rate": 0.0009242320819112628,
+      "loss": 2.5285,
+      "step": 336
+    },
+    {
+      "epoch": 0.3833902161547213,
+      "grad_norm": 1.0740411281585693,
+      "learning_rate": 0.0009240045506257111,
+      "loss": 1.4935,
+      "step": 337
+    },
+    {
+      "epoch": 0.3845278725824801,
+      "grad_norm": 1.089651107788086,
+      "learning_rate": 0.0009237770193401593,
+      "loss": 2.136,
+      "step": 338
+    },
+    {
+      "epoch": 0.3856655290102389,
+      "grad_norm": 0.8579985499382019,
+      "learning_rate": 0.0009235494880546074,
+      "loss": 1.9851,
+      "step": 339
+    },
+    {
+      "epoch": 0.38680318543799774,
+      "grad_norm": 1.0594276189804077,
+      "learning_rate": 0.0009233219567690558,
+      "loss": 2.244,
+      "step": 340
+    },
+    {
+      "epoch": 0.38794084186575656,
+      "grad_norm": 0.9024211168289185,
+      "learning_rate": 0.000923094425483504,
+      "loss": 2.5911,
+      "step": 341
+    },
+    {
+      "epoch": 0.3890784982935154,
+      "grad_norm": 1.2810311317443848,
+      "learning_rate": 0.0009228668941979522,
+      "loss": 3.0762,
+      "step": 342
+    },
+    {
+      "epoch": 0.3902161547212742,
+      "grad_norm": 0.9511512517929077,
+      "learning_rate": 0.0009226393629124005,
+      "loss": 2.871,
+      "step": 343
+    },
+    {
+      "epoch": 0.391353811149033,
+      "grad_norm": 1.0560070276260376,
+      "learning_rate": 0.0009224118316268487,
+      "loss": 1.4818,
+      "step": 344
+    },
+    {
+      "epoch": 0.3924914675767918,
+      "grad_norm": 0.7268106341362,
+      "learning_rate": 0.000922184300341297,
+      "loss": 0.9631,
+      "step": 345
+    },
+    {
+      "epoch": 0.39362912400455063,
+      "grad_norm": 1.2521812915802002,
+      "learning_rate": 0.0009219567690557452,
+      "loss": 2.0506,
+      "step": 346
+    },
+    {
+      "epoch": 0.39476678043230945,
+      "grad_norm": 0.8288787603378296,
+      "learning_rate": 0.0009217292377701934,
+      "loss": 1.7319,
+      "step": 347
+    },
+    {
+      "epoch": 0.39590443686006827,
+      "grad_norm": 1.5332680940628052,
+      "learning_rate": 0.0009215017064846418,
+      "loss": 1.8226,
+      "step": 348
+    },
+    {
+      "epoch": 0.3970420932878271,
+      "grad_norm": 0.918202817440033,
+      "learning_rate": 0.0009212741751990899,
+      "loss": 1.9544,
+      "step": 349
+    },
+    {
+      "epoch": 0.3981797497155859,
+      "grad_norm": 1.5337828397750854,
+      "learning_rate": 0.0009210466439135381,
+      "loss": 1.7653,
+      "step": 350
+    },
+    {
+      "epoch": 0.3993174061433447,
+      "grad_norm": 0.8244525194168091,
+      "learning_rate": 0.0009208191126279864,
+      "loss": 1.0721,
+      "step": 351
+    },
+    {
+      "epoch": 0.4004550625711035,
+      "grad_norm": 1.727785587310791,
+      "learning_rate": 0.0009205915813424346,
+      "loss": 3.8503,
+      "step": 352
+    },
+    {
+      "epoch": 0.40159271899886234,
+      "grad_norm": 0.8593099117279053,
+      "learning_rate": 0.0009203640500568828,
+      "loss": 1.5773,
+      "step": 353
+    },
+    {
+      "epoch": 0.40273037542662116,
+      "grad_norm": 0.929568886756897,
+      "learning_rate": 0.0009201365187713311,
+      "loss": 2.0465,
+      "step": 354
+    },
+    {
+      "epoch": 0.40386803185437997,
+      "grad_norm": 0.8598810434341431,
+      "learning_rate": 0.0009199089874857793,
+      "loss": 2.3678,
+      "step": 355
+    },
+    {
+      "epoch": 0.4050056882821388,
+      "grad_norm": 1.19975745677948,
+      "learning_rate": 0.0009196814562002275,
+      "loss": 1.9968,
+      "step": 356
+    },
+    {
+      "epoch": 0.4061433447098976,
+      "grad_norm": 0.8773411512374878,
+      "learning_rate": 0.0009194539249146759,
+      "loss": 2.3507,
+      "step": 357
+    },
+    {
+      "epoch": 0.4072810011376564,
+      "grad_norm": NaN,
+      "learning_rate": 0.0009194539249146759,
+      "loss": 3.3585,
+      "step": 358
+    },
+    {
+      "epoch": 0.40841865756541523,
+      "grad_norm": 1.068331241607666,
+      "learning_rate": 0.000919226393629124,
+      "loss": 2.2658,
+      "step": 359
+    },
+    {
+      "epoch": 0.40955631399317405,
+      "grad_norm": 0.8418660163879395,
+      "learning_rate": 0.0009189988623435722,
+      "loss": 1.1336,
+      "step": 360
+    },
+    {
+      "epoch": 0.41069397042093286,
+      "grad_norm": 1.0328933000564575,
+      "learning_rate": 0.0009187713310580205,
+      "loss": 2.308,
+      "step": 361
+    },
+    {
+      "epoch": 0.4118316268486917,
+      "grad_norm": 0.7511950135231018,
+      "learning_rate": 0.0009185437997724687,
+      "loss": 2.0096,
+      "step": 362
+    },
+    {
+      "epoch": 0.4129692832764505,
+      "grad_norm": 1.0027238130569458,
+      "learning_rate": 0.0009183162684869169,
+      "loss": 2.4037,
+      "step": 363
+    },
+    {
+      "epoch": 0.4141069397042093,
+      "grad_norm": 0.7929302453994751,
+      "learning_rate": 0.0009180887372013652,
+      "loss": 1.7147,
+      "step": 364
+    },
+    {
+      "epoch": 0.4152445961319681,
+      "grad_norm": 0.9058247804641724,
+      "learning_rate": 0.0009178612059158134,
+      "loss": 1.2271,
+      "step": 365
+    },
+    {
+      "epoch": 0.41638225255972694,
+      "grad_norm": 1.107584834098816,
+      "learning_rate": 0.0009176336746302618,
+      "loss": 1.976,
+      "step": 366
+    },
+    {
+      "epoch": 0.41751990898748575,
+      "grad_norm": 0.9150335788726807,
+      "learning_rate": 0.00091740614334471,
+      "loss": 1.5669,
+      "step": 367
+    },
+    {
+      "epoch": 0.41865756541524457,
+      "grad_norm": 1.2605892419815063,
+      "learning_rate": 0.0009171786120591582,
+      "loss": 3.2092,
+      "step": 368
+    },
+    {
+      "epoch": 0.4197952218430034,
+      "grad_norm": 1.8941680192947388,
+      "learning_rate": 0.0009169510807736064,
+      "loss": 2.2166,
+      "step": 369
+    },
+    {
+      "epoch": 0.42093287827076226,
+      "grad_norm": 1.0579288005828857,
+      "learning_rate": 0.0009167235494880546,
+      "loss": 2.2364,
+      "step": 370
+    },
+    {
+      "epoch": 0.42207053469852107,
+      "grad_norm": 0.8153421878814697,
+      "learning_rate": 0.0009164960182025028,
+      "loss": 2.2531,
+      "step": 371
+    },
+    {
+      "epoch": 0.4232081911262799,
+      "grad_norm": 0.9382014274597168,
+      "learning_rate": 0.0009162684869169511,
+      "loss": 1.7155,
+      "step": 372
+    },
+    {
+      "epoch": 0.4243458475540387,
+      "grad_norm": 0.8863131403923035,
+      "learning_rate": 0.0009160409556313993,
+      "loss": 2.0033,
+      "step": 373
+    },
+    {
+      "epoch": 0.4254835039817975,
+      "grad_norm": 1.0132858753204346,
+      "learning_rate": 0.0009158134243458475,
+      "loss": 2.7736,
+      "step": 374
+    },
+    {
+      "epoch": 0.42662116040955633,
+      "grad_norm": 1.237510085105896,
+      "learning_rate": 0.0009155858930602959,
+      "loss": 2.3289,
+      "step": 375
+    },
+    {
+      "epoch": 0.42775881683731515,
+      "grad_norm": 1.4978190660476685,
+      "learning_rate": 0.0009153583617747441,
+      "loss": 3.452,
+      "step": 376
+    },
+    {
+      "epoch": 0.42889647326507396,
+      "grad_norm": 0.9788629412651062,
+      "learning_rate": 0.0009151308304891923,
+      "loss": 2.536,
+      "step": 377
+    },
+    {
+      "epoch": 0.4300341296928328,
+      "grad_norm": 1.0312438011169434,
+      "learning_rate": 0.0009149032992036406,
+      "loss": 2.1322,
+      "step": 378
+    },
+    {
+      "epoch": 0.4311717861205916,
+      "grad_norm": 0.7808533906936646,
+      "learning_rate": 0.0009146757679180887,
+      "loss": 1.7681,
+      "step": 379
+    },
+    {
+      "epoch": 0.4323094425483504,
+      "grad_norm": 1.5941463708877563,
+      "learning_rate": 0.0009144482366325369,
+      "loss": 3.6207,
+      "step": 380
+    },
+    {
+      "epoch": 0.4334470989761092,
+      "grad_norm": 0.9317173361778259,
+      "learning_rate": 0.0009142207053469852,
+      "loss": 1.5295,
+      "step": 381
+    },
+    {
+      "epoch": 0.43458475540386804,
+      "grad_norm": 1.0690711736679077,
+      "learning_rate": 0.0009139931740614334,
+      "loss": 2.6297,
+      "step": 382
+    },
+    {
+      "epoch": 0.43572241183162685,
+      "grad_norm": 0.9129312634468079,
+      "learning_rate": 0.0009137656427758817,
+      "loss": 1.3539,
+      "step": 383
+    },
+    {
+      "epoch": 0.43686006825938567,
+      "grad_norm": 1.0281989574432373,
+      "learning_rate": 0.00091353811149033,
+      "loss": 2.2572,
+      "step": 384
+    },
+    {
+      "epoch": 0.4379977246871445,
+      "grad_norm": 0.9232156872749329,
+      "learning_rate": 0.0009133105802047782,
+      "loss": 2.7774,
+      "step": 385
+    },
+    {
+      "epoch": 0.4391353811149033,
+      "grad_norm": 0.8758882880210876,
+      "learning_rate": 0.0009130830489192265,
+      "loss": 1.8537,
+      "step": 386
+    },
+    {
+      "epoch": 0.4402730375426621,
+      "grad_norm": 0.6057271361351013,
+      "learning_rate": 0.0009128555176336747,
+      "loss": 1.1669,
+      "step": 387
+    },
+    {
+      "epoch": 0.44141069397042093,
+      "grad_norm": 0.764013409614563,
+      "learning_rate": 0.0009126279863481229,
+      "loss": 1.5494,
+      "step": 388
+    },
+    {
+      "epoch": 0.44254835039817975,
+      "grad_norm": 0.7935448288917542,
+      "learning_rate": 0.0009124004550625711,
+      "loss": 1.5411,
+      "step": 389
+    },
+    {
+      "epoch": 0.44368600682593856,
+      "grad_norm": 0.9698971509933472,
+      "learning_rate": 0.0009121729237770193,
+      "loss": 2.1653,
+      "step": 390
+    },
+    {
+      "epoch": 0.4448236632536974,
+      "grad_norm": 0.9462388753890991,
+      "learning_rate": 0.0009119453924914675,
+      "loss": 3.2917,
+      "step": 391
+    },
+    {
+      "epoch": 0.4459613196814562,
+      "grad_norm": 1.1189157962799072,
+      "learning_rate": 0.0009117178612059159,
+      "loss": 1.8733,
+      "step": 392
+    },
+    {
+      "epoch": 0.447098976109215,
+      "grad_norm": 1.251373529434204,
+      "learning_rate": 0.0009114903299203641,
+      "loss": 2.239,
+      "step": 393
+    },
+    {
+      "epoch": 0.4482366325369738,
+      "grad_norm": 1.385899305343628,
+      "learning_rate": 0.0009112627986348123,
+      "loss": 3.8287,
+      "step": 394
+    },
+    {
+      "epoch": 0.44937428896473264,
+      "grad_norm": 0.8532758951187134,
+      "learning_rate": 0.0009110352673492606,
+      "loss": 1.5912,
+      "step": 395
+    },
+    {
+      "epoch": 0.45051194539249145,
+      "grad_norm": 0.9151597023010254,
+      "learning_rate": 0.0009108077360637088,
+      "loss": 1.841,
+      "step": 396
+    },
+    {
+      "epoch": 0.45164960182025027,
+      "grad_norm": 1.054795265197754,
+      "learning_rate": 0.000910580204778157,
+      "loss": 2.7137,
+      "step": 397
+    },
+    {
+      "epoch": 0.4527872582480091,
+      "grad_norm": 0.7954731583595276,
+      "learning_rate": 0.0009103526734926052,
+      "loss": 1.7309,
+      "step": 398
+    },
+    {
+      "epoch": 0.4539249146757679,
+      "grad_norm": 1.1954761743545532,
+      "learning_rate": 0.0009101251422070534,
+      "loss": 3.0152,
+      "step": 399
+    },
+    {
+      "epoch": 0.4550625711035267,
+      "grad_norm": 1.0547521114349365,
+      "learning_rate": 0.0009098976109215017,
+      "loss": 3.1303,
+      "step": 400
+    },
+    {
+      "epoch": 0.45620022753128553,
+      "grad_norm": 1.2710171937942505,
+      "learning_rate": 0.00090967007963595,
+      "loss": 3.1468,
+      "step": 401
+    },
+    {
+      "epoch": 0.45733788395904434,
+      "grad_norm": 0.8694390058517456,
+      "learning_rate": 0.0009094425483503982,
+      "loss": 1.815,
+      "step": 402
+    },
+    {
+      "epoch": 0.45847554038680316,
+      "grad_norm": 0.9537279605865479,
+      "learning_rate": 0.0009092150170648464,
+      "loss": 1.9914,
+      "step": 403
+    },
+    {
+      "epoch": 0.459613196814562,
+      "grad_norm": 0.9804282188415527,
+      "learning_rate": 0.0009089874857792947,
+      "loss": 2.4996,
+      "step": 404
+    },
+    {
+      "epoch": 0.46075085324232085,
+      "grad_norm": 1.2874373197555542,
+      "learning_rate": 0.0009087599544937429,
+      "loss": 2.2957,
+      "step": 405
+    },
+    {
+      "epoch": 0.46188850967007966,
+      "grad_norm": 0.8987186551094055,
+      "learning_rate": 0.0009085324232081912,
+      "loss": 2.2726,
+      "step": 406
+    },
+    {
+      "epoch": 0.4630261660978385,
+      "grad_norm": 1.0590325593948364,
+      "learning_rate": 0.0009083048919226394,
+      "loss": 2.5264,
+      "step": 407
+    },
+    {
+      "epoch": 0.4641638225255973,
+      "grad_norm": 1.0790257453918457,
+      "learning_rate": 0.0009080773606370875,
+      "loss": 2.7717,
+      "step": 408
+    },
+    {
+      "epoch": 0.4653014789533561,
+      "grad_norm": 1.1287912130355835,
+      "learning_rate": 0.0009078498293515359,
+      "loss": 1.9665,
+      "step": 409
+    },
+    {
+      "epoch": 0.4664391353811149,
+      "grad_norm": 0.8903608918190002,
+      "learning_rate": 0.0009076222980659841,
+      "loss": 3.0162,
+      "step": 410
+    },
+    {
+      "epoch": 0.46757679180887374,
+      "grad_norm": 0.7123409509658813,
+      "learning_rate": 0.0009073947667804323,
+      "loss": 1.5362,
+      "step": 411
+    },
+    {
+      "epoch": 0.46871444823663255,
+      "grad_norm": 0.9984010457992554,
+      "learning_rate": 0.0009071672354948806,
+      "loss": 2.6421,
+      "step": 412
+    },
+    {
+      "epoch": 0.46985210466439137,
+      "grad_norm": 0.9573709964752197,
+      "learning_rate": 0.0009069397042093288,
+      "loss": 3.0399,
+      "step": 413
+    },
+    {
+      "epoch": 0.4709897610921502,
+      "grad_norm": 1.0717376470565796,
+      "learning_rate": 0.000906712172923777,
+      "loss": 2.5299,
+      "step": 414
+    },
+    {
+      "epoch": 0.472127417519909,
+      "grad_norm": 1.3975024223327637,
+      "learning_rate": 0.0009064846416382253,
+      "loss": 2.6824,
+      "step": 415
+    },
+    {
+      "epoch": 0.4732650739476678,
+      "grad_norm": 0.9038397073745728,
+      "learning_rate": 0.0009062571103526736,
+      "loss": 2.0442,
+      "step": 416
+    },
+    {
+      "epoch": 0.47440273037542663,
+      "grad_norm": 1.1075880527496338,
+      "learning_rate": 0.0009060295790671218,
+      "loss": 2.1013,
+      "step": 417
+    },
+    {
+      "epoch": 0.47554038680318544,
+      "grad_norm": 0.8074356913566589,
+      "learning_rate": 0.00090580204778157,
+      "loss": 2.9299,
+      "step": 418
+    },
+    {
+      "epoch": 0.47667804323094426,
+      "grad_norm": 0.9493515491485596,
+      "learning_rate": 0.0009055745164960182,
+      "loss": 2.0034,
+      "step": 419
+    },
+    {
+      "epoch": 0.4778156996587031,
+      "grad_norm": 1.3625547885894775,
+      "learning_rate": 0.0009053469852104664,
+      "loss": 2.4718,
+      "step": 420
+    },
+    {
+      "epoch": 0.4789533560864619,
+      "grad_norm": 1.2692298889160156,
+      "learning_rate": 0.0009051194539249147,
+      "loss": 2.676,
+      "step": 421
+    },
+    {
+      "epoch": 0.4800910125142207,
+      "grad_norm": 1.1094417572021484,
+      "learning_rate": 0.0009048919226393629,
+      "loss": 1.5969,
+      "step": 422
+    },
+    {
+      "epoch": 0.4812286689419795,
+      "grad_norm": 0.8246015310287476,
+      "learning_rate": 0.0009046643913538111,
+      "loss": 1.9902,
+      "step": 423
+    },
+    {
+      "epoch": 0.48236632536973834,
+      "grad_norm": 1.2059593200683594,
+      "learning_rate": 0.0009044368600682594,
+      "loss": 2.2454,
+      "step": 424
+    },
+    {
+      "epoch": 0.48350398179749715,
+      "grad_norm": 1.2407282590866089,
+      "learning_rate": 0.0009042093287827077,
+      "loss": 1.9269,
+      "step": 425
+    },
+    {
+      "epoch": 0.48464163822525597,
+      "grad_norm": 1.407323956489563,
+      "learning_rate": 0.0009039817974971559,
+      "loss": 3.2682,
+      "step": 426
+    },
+    {
+      "epoch": 0.4857792946530148,
+      "grad_norm": 1.0163967609405518,
+      "learning_rate": 0.0009037542662116041,
+      "loss": 2.3068,
+      "step": 427
+    },
+    {
+      "epoch": 0.4869169510807736,
+      "grad_norm": 0.9030061364173889,
+      "learning_rate": 0.0009035267349260523,
+      "loss": 1.8157,
+      "step": 428
+    },
+    {
+      "epoch": 0.4880546075085324,
+      "grad_norm": 1.6989500522613525,
+      "learning_rate": 0.0009032992036405006,
+      "loss": 3.7374,
+      "step": 429
+    },
+    {
+      "epoch": 0.4891922639362912,
+      "grad_norm": 1.1923679113388062,
+      "learning_rate": 0.0009030716723549488,
+      "loss": 2.565,
+      "step": 430
+    },
+    {
+      "epoch": 0.49032992036405004,
+      "grad_norm": 0.7808157205581665,
+      "learning_rate": 0.000902844141069397,
+      "loss": 2.05,
+      "step": 431
+    },
+    {
+      "epoch": 0.49146757679180886,
+      "grad_norm": 1.2099822759628296,
+      "learning_rate": 0.0009026166097838453,
+      "loss": 2.8765,
+      "step": 432
+    },
+    {
+      "epoch": 0.4926052332195677,
+      "grad_norm": 0.628551185131073,
+      "learning_rate": 0.0009023890784982936,
+      "loss": 1.3558,
+      "step": 433
+    },
+    {
+      "epoch": 0.4937428896473265,
+      "grad_norm": 1.0964410305023193,
+      "learning_rate": 0.0009021615472127418,
+      "loss": 3.4868,
+      "step": 434
+    },
+    {
+      "epoch": 0.4948805460750853,
+      "grad_norm": 1.1500060558319092,
+      "learning_rate": 0.0009019340159271901,
+      "loss": 2.6184,
+      "step": 435
+    },
+    {
+      "epoch": 0.4960182025028441,
+      "grad_norm": 0.9226061701774597,
+      "learning_rate": 0.0009017064846416383,
+      "loss": 2.2053,
+      "step": 436
+    },
+    {
+      "epoch": 0.49715585893060293,
+      "grad_norm": 0.9873738288879395,
+      "learning_rate": 0.0009014789533560864,
+      "loss": 2.4479,
+      "step": 437
+    },
+    {
+      "epoch": 0.49829351535836175,
+      "grad_norm": 0.6207978129386902,
+      "learning_rate": 0.0009012514220705347,
+      "loss": 1.1146,
+      "step": 438
+    },
+    {
+      "epoch": 0.49943117178612056,
+      "grad_norm": 0.7853182554244995,
+      "learning_rate": 0.0009010238907849829,
+      "loss": 2.4183,
+      "step": 439
+    },
+    {
+      "epoch": 0.5005688282138794,
+      "grad_norm": 1.1462852954864502,
+      "learning_rate": 0.0009007963594994311,
+      "loss": 2.3492,
+      "step": 440
+    },
+    {
+      "epoch": 0.5017064846416383,
+      "grad_norm": 0.7885423302650452,
+      "learning_rate": 0.0009005688282138795,
+      "loss": 1.2548,
+      "step": 441
+    },
+    {
+      "epoch": 0.502844141069397,
+      "grad_norm": 1.071851134300232,
+      "learning_rate": 0.0009003412969283277,
+      "loss": 2.1613,
+      "step": 442
+    },
+    {
+      "epoch": 0.5039817974971559,
+      "grad_norm": 0.934515655040741,
+      "learning_rate": 0.0009001137656427759,
+      "loss": 2.1865,
+      "step": 443
+    },
+    {
+      "epoch": 0.5051194539249146,
+      "grad_norm": 0.8959553241729736,
+      "learning_rate": 0.0008998862343572242,
+      "loss": 1.3645,
+      "step": 444
+    },
+    {
+      "epoch": 0.5062571103526735,
+      "grad_norm": 0.6154829263687134,
+      "learning_rate": 0.0008996587030716724,
+      "loss": 0.9074,
+      "step": 445
+    },
+    {
+      "epoch": 0.5073947667804323,
+      "grad_norm": 0.8270671367645264,
+      "learning_rate": 0.0008994311717861206,
+      "loss": 2.3498,
+      "step": 446
+    },
+    {
+      "epoch": 0.5085324232081911,
+      "grad_norm": 1.0123733282089233,
+      "learning_rate": 0.0008992036405005688,
+      "loss": 2.0715,
+      "step": 447
+    },
+    {
+      "epoch": 0.5096700796359499,
+      "grad_norm": 1.2498971223831177,
+      "learning_rate": 0.000898976109215017,
+      "loss": 3.0001,
+      "step": 448
+    },
+    {
+      "epoch": 0.5108077360637088,
+      "grad_norm": 1.0939654111862183,
+      "learning_rate": 0.0008987485779294653,
+      "loss": 1.7905,
+      "step": 449
+    },
+    {
+      "epoch": 0.5119453924914675,
+      "grad_norm": 1.0476197004318237,
+      "learning_rate": 0.0008985210466439136,
+      "loss": 2.68,
+      "step": 450
+    },
+    {
+      "epoch": 0.5130830489192264,
+      "grad_norm": 1.1398872137069702,
+      "learning_rate": 0.0008982935153583618,
+      "loss": 3.1602,
+      "step": 451
+    },
+    {
+      "epoch": 0.5142207053469852,
+      "grad_norm": 1.092517614364624,
+      "learning_rate": 0.0008980659840728101,
+      "loss": 1.772,
+      "step": 452
+    },
+    {
+      "epoch": 0.515358361774744,
+      "grad_norm": 1.1467770338058472,
+      "learning_rate": 0.0008978384527872583,
+      "loss": 3.0966,
+      "step": 453
+    },
+    {
+      "epoch": 0.5164960182025028,
+      "grad_norm": 0.9609680771827698,
+      "learning_rate": 0.0008976109215017065,
+      "loss": 2.1103,
+      "step": 454
+    },
+    {
+      "epoch": 0.5176336746302617,
+      "grad_norm": 1.0481035709381104,
+      "learning_rate": 0.0008973833902161548,
+      "loss": 1.9659,
+      "step": 455
+    },
+    {
+      "epoch": 0.5187713310580204,
+      "grad_norm": 0.8882219791412354,
+      "learning_rate": 0.000897155858930603,
+      "loss": 1.6778,
+      "step": 456
+    },
+    {
+      "epoch": 0.5199089874857793,
+      "grad_norm": 0.8644529581069946,
+      "learning_rate": 0.0008969283276450511,
+      "loss": 2.0891,
+      "step": 457
+    },
+    {
+      "epoch": 0.5210466439135382,
+      "grad_norm": 0.9775826930999756,
+      "learning_rate": 0.0008967007963594995,
+      "loss": 2.6407,
+      "step": 458
+    },
+    {
+      "epoch": 0.5221843003412969,
+      "grad_norm": 1.4607579708099365,
+      "learning_rate": 0.0008964732650739477,
+      "loss": 2.8334,
+      "step": 459
+    },
+    {
+      "epoch": 0.5233219567690558,
+      "grad_norm": 1.4416364431381226,
+      "learning_rate": 0.0008962457337883959,
+      "loss": 2.8912,
+      "step": 460
+    },
+    {
+      "epoch": 0.5244596131968146,
+      "grad_norm": 0.7647553086280823,
+      "learning_rate": 0.0008960182025028442,
+      "loss": 2.1203,
+      "step": 461
+    },
+    {
+      "epoch": 0.5255972696245734,
+      "grad_norm": 0.9002684950828552,
+      "learning_rate": 0.0008957906712172924,
+      "loss": 1.9081,
+      "step": 462
+    },
+    {
+      "epoch": 0.5267349260523322,
+      "grad_norm": 0.6581302285194397,
+      "learning_rate": 0.0008955631399317406,
+      "loss": 2.001,
+      "step": 463
+    },
+    {
+      "epoch": 0.5278725824800911,
+      "grad_norm": 1.0612741708755493,
+      "learning_rate": 0.0008953356086461889,
+      "loss": 1.9066,
+      "step": 464
+    },
+    {
+      "epoch": 0.5290102389078498,
+      "grad_norm": 1.4963552951812744,
+      "learning_rate": 0.0008951080773606371,
+      "loss": 2.1809,
+      "step": 465
+    },
+    {
+      "epoch": 0.5301478953356087,
+      "grad_norm": 1.1061559915542603,
+      "learning_rate": 0.0008948805460750852,
+      "loss": 3.068,
+      "step": 466
+    },
+    {
+      "epoch": 0.5312855517633674,
+      "grad_norm": 0.9316163659095764,
+      "learning_rate": 0.0008946530147895336,
+      "loss": 1.9232,
+      "step": 467
+    },
+    {
+      "epoch": 0.5324232081911263,
+      "grad_norm": 0.8861683011054993,
+      "learning_rate": 0.0008944254835039818,
+      "loss": 2.1368,
+      "step": 468
+    },
+    {
+      "epoch": 0.5335608646188851,
+      "grad_norm": 0.835658848285675,
+      "learning_rate": 0.0008941979522184301,
+      "loss": 1.8731,
+      "step": 469
+    },
+    {
+      "epoch": 0.534698521046644,
+      "grad_norm": 1.1577521562576294,
+      "learning_rate": 0.0008939704209328783,
+      "loss": 3.0726,
+      "step": 470
+    },
+    {
+      "epoch": 0.5358361774744027,
+      "grad_norm": 1.3069539070129395,
+      "learning_rate": 0.0008937428896473265,
+      "loss": 2.8827,
+      "step": 471
+    },
+    {
+      "epoch": 0.5369738339021616,
+      "grad_norm": 0.884005606174469,
+      "learning_rate": 0.0008935153583617748,
+      "loss": 2.2167,
+      "step": 472
+    },
+    {
+      "epoch": 0.5381114903299203,
+      "grad_norm": 1.0972729921340942,
+      "learning_rate": 0.000893287827076223,
+      "loss": 2.8096,
+      "step": 473
+    },
+    {
+      "epoch": 0.5392491467576792,
+      "grad_norm": 0.6620914340019226,
+      "learning_rate": 0.0008930602957906712,
+      "loss": 1.8139,
+      "step": 474
+    },
+    {
+      "epoch": 0.540386803185438,
+      "grad_norm": 1.0320155620574951,
+      "learning_rate": 0.0008928327645051196,
+      "loss": 2.7753,
+      "step": 475
+    },
+    {
+      "epoch": 0.5415244596131968,
+      "grad_norm": 1.195923924446106,
+      "learning_rate": 0.0008926052332195677,
+      "loss": 2.8982,
+      "step": 476
+    },
+    {
+      "epoch": 0.5426621160409556,
+      "grad_norm": 0.6206729412078857,
+      "learning_rate": 0.0008923777019340159,
+      "loss": 1.6572,
+      "step": 477
+    },
+    {
+      "epoch": 0.5437997724687145,
+      "grad_norm": 0.867581844329834,
+      "learning_rate": 0.0008921501706484642,
+      "loss": 1.3072,
+      "step": 478
+    },
+    {
+      "epoch": 0.5449374288964732,
+      "grad_norm": 1.066758632659912,
+      "learning_rate": 0.0008919226393629124,
+      "loss": 2.2283,
+      "step": 479
+    },
+    {
+      "epoch": 0.5460750853242321,
+      "grad_norm": 0.9295621514320374,
+      "learning_rate": 0.0008916951080773606,
+      "loss": 1.8157,
+      "step": 480
+    },
+    {
+      "epoch": 0.5472127417519909,
+      "grad_norm": 1.0595327615737915,
+      "learning_rate": 0.0008914675767918089,
+      "loss": 2.486,
+      "step": 481
+    },
+    {
+      "epoch": 0.5483503981797497,
+      "grad_norm": 1.1606998443603516,
+      "learning_rate": 0.0008912400455062571,
+      "loss": 1.9862,
+      "step": 482
+    },
+    {
+      "epoch": 0.5494880546075085,
+      "grad_norm": 1.3078361749649048,
+      "learning_rate": 0.0008910125142207054,
+      "loss": 2.824,
+      "step": 483
+    },
+    {
+      "epoch": 0.5506257110352674,
+      "grad_norm": 1.1636848449707031,
+      "learning_rate": 0.0008907849829351537,
+      "loss": 2.9846,
+      "step": 484
+    },
+    {
+      "epoch": 0.5517633674630261,
+      "grad_norm": 1.6126782894134521,
+      "learning_rate": 0.0008905574516496019,
+      "loss": 3.7502,
+      "step": 485
+    },
+    {
+      "epoch": 0.552901023890785,
+      "grad_norm": 0.9118810296058655,
+      "learning_rate": 0.00089032992036405,
+      "loss": 2.7482,
+      "step": 486
+    },
+    {
+      "epoch": 0.5540386803185438,
+      "grad_norm": 0.9121948480606079,
+      "learning_rate": 0.0008901023890784983,
+      "loss": 1.8244,
+      "step": 487
+    },
+    {
+      "epoch": 0.5551763367463026,
+      "grad_norm": 1.749747633934021,
+      "learning_rate": 0.0008898748577929465,
+      "loss": 3.4315,
+      "step": 488
+    },
+    {
+      "epoch": 0.5563139931740614,
+      "grad_norm": 0.9559519290924072,
+      "learning_rate": 0.0008896473265073947,
+      "loss": 1.2761,
+      "step": 489
+    },
+    {
+      "epoch": 0.5574516496018203,
+      "grad_norm": 1.3748880624771118,
+      "learning_rate": 0.000889419795221843,
+      "loss": 2.4184,
+      "step": 490
+    },
+    {
+      "epoch": 0.558589306029579,
+      "grad_norm": 0.9325023889541626,
+      "learning_rate": 0.0008891922639362912,
+      "loss": 1.6256,
+      "step": 491
+    },
+    {
+      "epoch": 0.5597269624573379,
+      "grad_norm": 1.2054123878479004,
+      "learning_rate": 0.0008889647326507396,
+      "loss": 2.4543,
+      "step": 492
+    },
+    {
+      "epoch": 0.5608646188850968,
+      "grad_norm": 1.1576976776123047,
+      "learning_rate": 0.0008887372013651878,
+      "loss": 1.7206,
+      "step": 493
+    },
+    {
+      "epoch": 0.5620022753128555,
+      "grad_norm": 0.8814551830291748,
+      "learning_rate": 0.000888509670079636,
+      "loss": 1.8812,
+      "step": 494
+    },
+    {
+      "epoch": 0.5631399317406144,
+      "grad_norm": 0.8528147339820862,
+      "learning_rate": 0.0008882821387940843,
+      "loss": 2.411,
+      "step": 495
+    },
+    {
+      "epoch": 0.5642775881683731,
+      "grad_norm": 1.0148645639419556,
+      "learning_rate": 0.0008880546075085324,
+      "loss": 2.1765,
+      "step": 496
+    },
+    {
+      "epoch": 0.565415244596132,
+      "grad_norm": 0.7151885032653809,
+      "learning_rate": 0.0008878270762229806,
+      "loss": 1.2428,
+      "step": 497
+    },
+    {
+      "epoch": 0.5665529010238908,
+      "grad_norm": 1.7494755983352661,
+      "learning_rate": 0.0008875995449374289,
+      "loss": 4.1417,
+      "step": 498
+    },
+    {
+      "epoch": 0.5676905574516496,
+      "grad_norm": 0.8421460390090942,
+      "learning_rate": 0.0008873720136518771,
+      "loss": 1.7912,
+      "step": 499
+    },
+    {
+      "epoch": 0.5688282138794084,
+      "grad_norm": 0.9811123013496399,
+      "learning_rate": 0.0008871444823663254,
+      "loss": 2.1765,
+      "step": 500
+    },
+    {
+      "epoch": 0.5699658703071673,
+      "grad_norm": 0.9422712922096252,
+      "learning_rate": 0.0008869169510807737,
+      "loss": 1.7157,
+      "step": 501
+    },
+    {
+      "epoch": 0.571103526734926,
+      "grad_norm": 0.8885846138000488,
+      "learning_rate": 0.0008866894197952219,
+      "loss": 2.2024,
+      "step": 502
+    },
+    {
+      "epoch": 0.5722411831626849,
+      "grad_norm": 1.6120095252990723,
+      "learning_rate": 0.0008864618885096701,
+      "loss": 3.8523,
+      "step": 503
+    },
+    {
+      "epoch": 0.5733788395904437,
+      "grad_norm": 1.2259505987167358,
+      "learning_rate": 0.0008862343572241184,
+      "loss": 2.3551,
+      "step": 504
+    },
+    {
+      "epoch": 0.5745164960182025,
+      "grad_norm": 0.9865310788154602,
+      "learning_rate": 0.0008860068259385665,
+      "loss": 1.0719,
+      "step": 505
+    },
+    {
+      "epoch": 0.5756541524459613,
+      "grad_norm": 0.945303201675415,
+      "learning_rate": 0.0008857792946530147,
+      "loss": 1.8286,
+      "step": 506
+    },
+    {
+      "epoch": 0.5767918088737202,
+      "grad_norm": 1.099501371383667,
+      "learning_rate": 0.000885551763367463,
+      "loss": 2.4123,
+      "step": 507
+    },
+    {
+      "epoch": 0.5779294653014789,
+      "grad_norm": 1.3092284202575684,
+      "learning_rate": 0.0008853242320819112,
+      "loss": 2.8365,
+      "step": 508
+    },
+    {
+      "epoch": 0.5790671217292378,
+      "grad_norm": 0.9153149127960205,
+      "learning_rate": 0.0008850967007963595,
+      "loss": 1.4962,
+      "step": 509
+    },
+    {
+      "epoch": 0.5802047781569966,
+      "grad_norm": 1.0560232400894165,
+      "learning_rate": 0.0008848691695108078,
+      "loss": 2.0615,
+      "step": 510
+    },
+    {
+      "epoch": 0.5813424345847554,
+      "grad_norm": 0.6506679058074951,
+      "learning_rate": 0.000884641638225256,
+      "loss": 1.7931,
+      "step": 511
+    },
+    {
+      "epoch": 0.5824800910125142,
+      "grad_norm": 0.6726229786872864,
+      "learning_rate": 0.0008844141069397043,
+      "loss": 1.5145,
+      "step": 512
+    },
+    {
+      "epoch": 0.5836177474402731,
+      "grad_norm": 0.8806062340736389,
+      "learning_rate": 0.0008841865756541525,
+      "loss": 1.3302,
+      "step": 513
+    },
+    {
+      "epoch": 0.5847554038680318,
+      "grad_norm": 0.8398321270942688,
+      "learning_rate": 0.0008839590443686007,
+      "loss": 1.0688,
+      "step": 514
+    },
+    {
+      "epoch": 0.5858930602957907,
+      "grad_norm": 1.0214747190475464,
+      "learning_rate": 0.0008837315130830489,
+      "loss": 2.7142,
+      "step": 515
+    },
+    {
+      "epoch": 0.5870307167235495,
+      "grad_norm": 1.1030226945877075,
+      "learning_rate": 0.0008835039817974971,
+      "loss": 1.7378,
+      "step": 516
+    },
+    {
+      "epoch": 0.5881683731513083,
+      "grad_norm": 0.9792185425758362,
+      "learning_rate": 0.0008832764505119454,
+      "loss": 1.9917,
+      "step": 517
+    },
+    {
+      "epoch": 0.5893060295790671,
+      "grad_norm": 0.7293819785118103,
+      "learning_rate": 0.0008830489192263937,
+      "loss": 1.7711,
+      "step": 518
+    },
+    {
+      "epoch": 0.590443686006826,
+      "grad_norm": 0.9090821743011475,
+      "learning_rate": 0.0008828213879408419,
+      "loss": 2.3188,
+      "step": 519
+    },
+    {
+      "epoch": 0.5915813424345847,
+      "grad_norm": 0.9316034317016602,
+      "learning_rate": 0.0008825938566552901,
+      "loss": 1.9596,
+      "step": 520
+    },
+    {
+      "epoch": 0.5927189988623436,
+      "grad_norm": 1.022985577583313,
+      "learning_rate": 0.0008823663253697384,
+      "loss": 1.7448,
+      "step": 521
+    },
+    {
+      "epoch": 0.5938566552901023,
+      "grad_norm": 1.1398788690567017,
+      "learning_rate": 0.0008821387940841866,
+      "loss": 2.932,
+      "step": 522
+    },
+    {
+      "epoch": 0.5949943117178612,
+      "grad_norm": 1.1560657024383545,
+      "learning_rate": 0.0008819112627986348,
+      "loss": 2.2647,
+      "step": 523
+    },
+    {
+      "epoch": 0.59613196814562,
+      "grad_norm": 1.1563608646392822,
+      "learning_rate": 0.0008816837315130831,
+      "loss": 2.6389,
+      "step": 524
+    },
+    {
+      "epoch": 0.5972696245733788,
+      "grad_norm": 0.6971971988677979,
+      "learning_rate": 0.0008814562002275312,
+      "loss": 1.6904,
+      "step": 525
+    },
+    {
+      "epoch": 0.5984072810011376,
+      "grad_norm": 1.1338340044021606,
+      "learning_rate": 0.0008812286689419795,
+      "loss": 2.7721,
+      "step": 526
+    },
+    {
+      "epoch": 0.5995449374288965,
+      "grad_norm": 0.9443331360816956,
+      "learning_rate": 0.0008810011376564278,
+      "loss": 1.673,
+      "step": 527
+    },
+    {
+      "epoch": 0.6006825938566553,
+      "grad_norm": 0.7245666980743408,
+      "learning_rate": 0.000880773606370876,
+      "loss": 1.716,
+      "step": 528
+    },
+    {
+      "epoch": 0.6018202502844141,
+      "grad_norm": 0.9884563684463501,
+      "learning_rate": 0.0008805460750853242,
+      "loss": 3.0318,
+      "step": 529
+    },
+    {
+      "epoch": 0.602957906712173,
+      "grad_norm": 1.2986050844192505,
+      "learning_rate": 0.0008803185437997725,
+      "loss": 2.7565,
+      "step": 530
+    },
+    {
+      "epoch": 0.6040955631399317,
+      "grad_norm": 0.9133585691452026,
+      "learning_rate": 0.0008800910125142207,
+      "loss": 2.1352,
+      "step": 531
+    },
+    {
+      "epoch": 0.6052332195676906,
+      "grad_norm": 0.7802772521972656,
+      "learning_rate": 0.000879863481228669,
+      "loss": 1.487,
+      "step": 532
+    },
+    {
+      "epoch": 0.6063708759954494,
+      "grad_norm": 1.488508701324463,
+      "learning_rate": 0.0008796359499431173,
+      "loss": 3.6104,
+      "step": 533
+    },
+    {
+      "epoch": 0.6075085324232082,
+      "grad_norm": 1.4523831605911255,
+      "learning_rate": 0.0008794084186575654,
+      "loss": 4.1854,
+      "step": 534
+    },
+    {
+      "epoch": 0.608646188850967,
+      "grad_norm": 0.9844598174095154,
+      "learning_rate": 0.0008791808873720137,
+      "loss": 2.2759,
+      "step": 535
+    },
+    {
+      "epoch": 0.6097838452787259,
+      "grad_norm": 1.0784645080566406,
+      "learning_rate": 0.0008789533560864619,
+      "loss": 2.6678,
+      "step": 536
+    },
+    {
+      "epoch": 0.6109215017064846,
+      "grad_norm": 1.2045968770980835,
+      "learning_rate": 0.0008787258248009101,
+      "loss": 2.4359,
+      "step": 537
+    },
+    {
+      "epoch": 0.6120591581342435,
+      "grad_norm": 1.1759347915649414,
+      "learning_rate": 0.0008784982935153584,
+      "loss": 2.6307,
+      "step": 538
+    },
+    {
+      "epoch": 0.6131968145620023,
+      "grad_norm": 1.0141427516937256,
+      "learning_rate": 0.0008782707622298066,
+      "loss": 2.5533,
+      "step": 539
+    },
+    {
+      "epoch": 0.6143344709897611,
+      "grad_norm": 0.986733615398407,
+      "learning_rate": 0.0008780432309442548,
+      "loss": 1.1977,
+      "step": 540
+    },
+    {
+      "epoch": 0.6154721274175199,
+      "grad_norm": 1.7550513744354248,
+      "learning_rate": 0.0008778156996587031,
+      "loss": 1.977,
+      "step": 541
+    },
+    {
+      "epoch": 0.6166097838452788,
+      "grad_norm": 1.0284448862075806,
+      "learning_rate": 0.0008775881683731514,
+      "loss": 2.2885,
+      "step": 542
+    },
+    {
+      "epoch": 0.6177474402730375,
+      "grad_norm": 0.8970025181770325,
+      "learning_rate": 0.0008773606370875996,
+      "loss": 1.519,
+      "step": 543
+    },
+    {
+      "epoch": 0.6188850967007964,
+      "grad_norm": 1.1675552129745483,
+      "learning_rate": 0.0008771331058020478,
+      "loss": 2.5995,
+      "step": 544
+    },
+    {
+      "epoch": 0.6200227531285551,
+      "grad_norm": 0.7909214496612549,
+      "learning_rate": 0.000876905574516496,
+      "loss": 2.4494,
+      "step": 545
+    },
+    {
+      "epoch": 0.621160409556314,
+      "grad_norm": 1.1816651821136475,
+      "learning_rate": 0.0008766780432309442,
+      "loss": 2.0798,
+      "step": 546
+    },
+    {
+      "epoch": 0.6222980659840728,
+      "grad_norm": 1.2361869812011719,
+      "learning_rate": 0.0008764505119453925,
+      "loss": 2.1153,
+      "step": 547
+    },
+    {
+      "epoch": 0.6234357224118316,
+      "grad_norm": 1.498818278312683,
+      "learning_rate": 0.0008762229806598407,
+      "loss": 3.6819,
+      "step": 548
+    },
+    {
+      "epoch": 0.6245733788395904,
+      "grad_norm": 1.2410242557525635,
+      "learning_rate": 0.0008759954493742889,
+      "loss": 2.3829,
+      "step": 549
+    },
+    {
+      "epoch": 0.6257110352673493,
+      "grad_norm": 1.1705360412597656,
+      "learning_rate": 0.0008757679180887373,
+      "loss": 1.718,
+      "step": 550
+    },
+    {
+      "epoch": 0.626848691695108,
+      "grad_norm": 0.9694558382034302,
+      "learning_rate": 0.0008755403868031855,
+      "loss": 2.1928,
+      "step": 551
+    },
+    {
+      "epoch": 0.6279863481228669,
+      "grad_norm": 0.6821228265762329,
+      "learning_rate": 0.0008753128555176338,
+      "loss": 1.406,
+      "step": 552
+    },
+    {
+      "epoch": 0.6291240045506257,
+      "grad_norm": 0.8572499752044678,
+      "learning_rate": 0.000875085324232082,
+      "loss": 1.6783,
+      "step": 553
+    },
+    {
+      "epoch": 0.6302616609783845,
+      "grad_norm": 0.8171662092208862,
+      "learning_rate": 0.0008748577929465301,
+      "loss": 2.4399,
+      "step": 554
+    },
+    {
+      "epoch": 0.6313993174061433,
+      "grad_norm": 1.320075273513794,
+      "learning_rate": 0.0008746302616609784,
+      "loss": 1.4817,
+      "step": 555
+    },
+    {
+      "epoch": 0.6325369738339022,
+      "grad_norm": 0.7951433658599854,
+      "learning_rate": 0.0008744027303754266,
+      "loss": 1.7655,
+      "step": 556
+    },
+    {
+      "epoch": 0.6336746302616609,
+      "grad_norm": 0.7013391256332397,
+      "learning_rate": 0.0008741751990898748,
+      "loss": 1.3957,
+      "step": 557
+    },
+    {
+      "epoch": 0.6348122866894198,
+      "grad_norm": 0.8427788019180298,
+      "learning_rate": 0.0008739476678043232,
+      "loss": 1.5612,
+      "step": 558
+    },
+    {
+      "epoch": 0.6359499431171786,
+      "grad_norm": 1.0773825645446777,
+      "learning_rate": 0.0008737201365187714,
+      "loss": 1.8739,
+      "step": 559
+    },
+    {
+      "epoch": 0.6370875995449374,
+      "grad_norm": 0.7293964624404907,
+      "learning_rate": 0.0008734926052332196,
+      "loss": 1.1344,
+      "step": 560
+    },
+    {
+      "epoch": 0.6382252559726962,
+      "grad_norm": 0.8789951205253601,
+      "learning_rate": 0.0008732650739476679,
+      "loss": 2.1382,
+      "step": 561
+    },
+    {
+      "epoch": 0.6393629124004551,
+      "grad_norm": 1.86644446849823,
+      "learning_rate": 0.0008730375426621161,
+      "loss": 4.1822,
+      "step": 562
+    },
+    {
+      "epoch": 0.6405005688282139,
+      "grad_norm": 1.1930569410324097,
+      "learning_rate": 0.0008728100113765643,
+      "loss": 2.7447,
+      "step": 563
+    },
+    {
+      "epoch": 0.6416382252559727,
+      "grad_norm": 0.9682884812355042,
+      "learning_rate": 0.0008725824800910125,
+      "loss": 2.0626,
+      "step": 564
+    },
+    {
+      "epoch": 0.6427758816837316,
+      "grad_norm": 0.7793911695480347,
+      "learning_rate": 0.0008723549488054607,
+      "loss": 1.643,
+      "step": 565
+    },
+    {
+      "epoch": 0.6439135381114903,
+      "grad_norm": 1.0574642419815063,
+      "learning_rate": 0.0008721274175199089,
+      "loss": 2.1083,
+      "step": 566
+    },
+    {
+      "epoch": 0.6450511945392492,
+      "grad_norm": 0.8339124917984009,
+      "learning_rate": 0.0008718998862343573,
+      "loss": 1.7013,
+      "step": 567
+    },
+    {
+      "epoch": 0.646188850967008,
+      "grad_norm": 0.7775769829750061,
+      "learning_rate": 0.0008716723549488055,
+      "loss": 1.3428,
+      "step": 568
+    },
+    {
+      "epoch": 0.6473265073947668,
+      "grad_norm": 1.089309573173523,
+      "learning_rate": 0.0008714448236632537,
+      "loss": 2.1378,
+      "step": 569
+    },
+    {
+      "epoch": 0.6484641638225256,
+      "grad_norm": 1.043135166168213,
+      "learning_rate": 0.000871217292377702,
+      "loss": 3.3488,
+      "step": 570
+    },
+    {
+      "epoch": 0.6496018202502845,
+      "grad_norm": 0.9596251249313354,
+      "learning_rate": 0.0008709897610921502,
+      "loss": 2.3173,
+      "step": 571
+    },
+    {
+      "epoch": 0.6507394766780432,
+      "grad_norm": 0.8774250745773315,
+      "learning_rate": 0.0008707622298065985,
+      "loss": 2.1654,
+      "step": 572
+    },
+    {
+      "epoch": 0.6518771331058021,
+      "grad_norm": 0.7768117189407349,
+      "learning_rate": 0.0008705346985210466,
+      "loss": 1.56,
+      "step": 573
+    },
+    {
+      "epoch": 0.6530147895335608,
+      "grad_norm": 0.9584445953369141,
+      "learning_rate": 0.0008703071672354948,
+      "loss": 1.8768,
+      "step": 574
+    },
+    {
+      "epoch": 0.6541524459613197,
+      "grad_norm": 0.7490730285644531,
+      "learning_rate": 0.0008700796359499432,
+      "loss": 1.5192,
+      "step": 575
+    },
+    {
+      "epoch": 0.6552901023890785,
+      "grad_norm": 0.779798686504364,
+      "learning_rate": 0.0008698521046643914,
+      "loss": 1.6678,
+      "step": 576
+    },
+    {
+      "epoch": 0.6564277588168373,
+      "grad_norm": 1.2081419229507446,
+      "learning_rate": 0.0008696245733788396,
+      "loss": 2.5276,
+      "step": 577
+    },
+    {
+      "epoch": 0.6575654152445961,
+      "grad_norm": 1.5270092487335205,
+      "learning_rate": 0.0008693970420932879,
+      "loss": 3.8846,
+      "step": 578
+    },
+    {
+      "epoch": 0.658703071672355,
+      "grad_norm": 0.9504697918891907,
+      "learning_rate": 0.0008691695108077361,
+      "loss": 1.8834,
+      "step": 579
+    },
+    {
+      "epoch": 0.6598407281001137,
+      "grad_norm": 0.8998041749000549,
+      "learning_rate": 0.0008689419795221843,
+      "loss": 2.0045,
+      "step": 580
+    },
+    {
+      "epoch": 0.6609783845278726,
+      "grad_norm": 0.8134379386901855,
+      "learning_rate": 0.0008687144482366326,
+      "loss": 0.9556,
+      "step": 581
+    },
+    {
+      "epoch": 0.6621160409556314,
+      "grad_norm": 0.7257466316223145,
+      "learning_rate": 0.0008684869169510808,
+      "loss": 1.2292,
+      "step": 582
+    },
+    {
+      "epoch": 0.6632536973833902,
+      "grad_norm": 1.3256492614746094,
+      "learning_rate": 0.0008682593856655289,
+      "loss": 3.8102,
+      "step": 583
+    },
+    {
+      "epoch": 0.664391353811149,
+      "grad_norm": 0.7932450771331787,
+      "learning_rate": 0.0008680318543799773,
+      "loss": 2.2484,
+      "step": 584
+    },
+    {
+      "epoch": 0.6655290102389079,
+      "grad_norm": 1.0340895652770996,
+      "learning_rate": 0.0008678043230944255,
+      "loss": 2.1459,
+      "step": 585
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.0674877166748047,
+      "learning_rate": 0.0008675767918088737,
+      "loss": 2.7635,
+      "step": 586
+    },
+    {
+      "epoch": 0.6678043230944255,
+      "grad_norm": 1.0885896682739258,
+      "learning_rate": 0.000867349260523322,
+      "loss": 2.525,
+      "step": 587
+    },
+    {
+      "epoch": 0.6689419795221843,
+      "grad_norm": 0.6129403710365295,
+      "learning_rate": 0.0008671217292377702,
+      "loss": 1.225,
+      "step": 588
+    },
+    {
+      "epoch": 0.6700796359499431,
+      "grad_norm": 0.9944592118263245,
+      "learning_rate": 0.0008668941979522184,
+      "loss": 1.9925,
+      "step": 589
+    },
+    {
+      "epoch": 0.6712172923777019,
+      "grad_norm": 1.1834667921066284,
+      "learning_rate": 0.0008666666666666667,
+      "loss": 2.3584,
+      "step": 590
+    },
+    {
+      "epoch": 0.6723549488054608,
+      "grad_norm": 1.3296507596969604,
+      "learning_rate": 0.0008664391353811149,
+      "loss": 2.7873,
+      "step": 591
+    },
+    {
+      "epoch": 0.6734926052332195,
+      "grad_norm": 0.7346708178520203,
+      "learning_rate": 0.0008662116040955633,
+      "loss": 1.5522,
+      "step": 592
+    },
+    {
+      "epoch": 0.6746302616609784,
+      "grad_norm": 1.405709147453308,
+      "learning_rate": 0.0008659840728100114,
+      "loss": 3.4668,
+      "step": 593
+    },
+    {
+      "epoch": 0.6757679180887372,
+      "grad_norm": 0.925556480884552,
+      "learning_rate": 0.0008657565415244596,
+      "loss": 2.5614,
+      "step": 594
+    },
+    {
+      "epoch": 0.676905574516496,
+      "grad_norm": 0.6485567688941956,
+      "learning_rate": 0.0008655290102389079,
+      "loss": 1.3796,
+      "step": 595
+    },
+    {
+      "epoch": 0.6780432309442548,
+      "grad_norm": 1.2317935228347778,
+      "learning_rate": 0.0008653014789533561,
+      "loss": 2.7619,
+      "step": 596
+    },
+    {
+      "epoch": 0.6791808873720137,
+      "grad_norm": 1.8749306201934814,
+      "learning_rate": 0.0008650739476678043,
+      "loss": 2.0424,
+      "step": 597
+    },
+    {
+      "epoch": 0.6803185437997725,
+      "grad_norm": 1.0405480861663818,
+      "learning_rate": 0.0008648464163822526,
+      "loss": 1.6496,
+      "step": 598
+    },
+    {
+      "epoch": 0.6814562002275313,
+      "grad_norm": 0.7675598859786987,
+      "learning_rate": 0.0008646188850967008,
+      "loss": 1.5025,
+      "step": 599
+    },
+    {
+      "epoch": 0.6825938566552902,
+      "grad_norm": 0.9507393836975098,
+      "learning_rate": 0.000864391353811149,
+      "loss": 2.1572,
+      "step": 600
+    },
+    {
+      "epoch": 0.6837315130830489,
+      "grad_norm": 0.7208441495895386,
+      "learning_rate": 0.0008641638225255974,
+      "loss": 1.8724,
+      "step": 601
+    },
+    {
+      "epoch": 0.6848691695108078,
+      "grad_norm": 1.0471227169036865,
+      "learning_rate": 0.0008639362912400455,
+      "loss": 1.5183,
+      "step": 602
+    },
+    {
+      "epoch": 0.6860068259385665,
+      "grad_norm": 0.7617486119270325,
+      "learning_rate": 0.0008637087599544937,
+      "loss": 1.9664,
+      "step": 603
+    },
+    {
+      "epoch": 0.6871444823663254,
+      "grad_norm": 1.7303751707077026,
+      "learning_rate": 0.000863481228668942,
+      "loss": 3.4134,
+      "step": 604
+    },
+    {
+      "epoch": 0.6882821387940842,
+      "grad_norm": 1.1979167461395264,
+      "learning_rate": 0.0008632536973833902,
+      "loss": 2.1287,
+      "step": 605
+    },
+    {
+      "epoch": 0.689419795221843,
+      "grad_norm": 0.8797517418861389,
+      "learning_rate": 0.0008630261660978384,
+      "loss": 1.4359,
+      "step": 606
+    },
+    {
+      "epoch": 0.6905574516496018,
+      "grad_norm": 1.3012560606002808,
+      "learning_rate": 0.0008627986348122867,
+      "loss": 1.9451,
+      "step": 607
+    },
+    {
+      "epoch": 0.6916951080773607,
+      "grad_norm": 0.8899447917938232,
+      "learning_rate": 0.000862571103526735,
+      "loss": 1.9974,
+      "step": 608
+    },
+    {
+      "epoch": 0.6928327645051194,
+      "grad_norm": 1.449118733406067,
+      "learning_rate": 0.0008623435722411832,
+      "loss": 3.0255,
+      "step": 609
+    },
+    {
+      "epoch": 0.6939704209328783,
+      "grad_norm": 1.4355418682098389,
+      "learning_rate": 0.0008621160409556315,
+      "loss": 3.6942,
+      "step": 610
+    },
+    {
+      "epoch": 0.6951080773606371,
+      "grad_norm": 0.9907065033912659,
+      "learning_rate": 0.0008618885096700797,
+      "loss": 1.8645,
+      "step": 611
+    },
+    {
+      "epoch": 0.6962457337883959,
+      "grad_norm": 0.8287332057952881,
+      "learning_rate": 0.0008616609783845278,
+      "loss": 1.909,
+      "step": 612
+    },
+    {
+      "epoch": 0.6973833902161547,
+      "grad_norm": 1.207824468612671,
+      "learning_rate": 0.0008614334470989761,
+      "loss": 2.1948,
+      "step": 613
+    },
+    {
+      "epoch": 0.6985210466439136,
+      "grad_norm": 0.6670779585838318,
+      "learning_rate": 0.0008612059158134243,
+      "loss": 1.2597,
+      "step": 614
+    },
+    {
+      "epoch": 0.6996587030716723,
+      "grad_norm": 0.6023557186126709,
+      "learning_rate": 0.0008609783845278726,
+      "loss": 1.2117,
+      "step": 615
+    },
+    {
+      "epoch": 0.7007963594994312,
+      "grad_norm": 1.5673305988311768,
+      "learning_rate": 0.0008607508532423208,
+      "loss": 3.1698,
+      "step": 616
+    },
+    {
+      "epoch": 0.70193401592719,
+      "grad_norm": 0.9000980257987976,
+      "learning_rate": 0.000860523321956769,
+      "loss": 1.9187,
+      "step": 617
+    },
+    {
+      "epoch": 0.7030716723549488,
+      "grad_norm": 0.9901409149169922,
+      "learning_rate": 0.0008602957906712174,
+      "loss": 2.3486,
+      "step": 618
+    },
+    {
+      "epoch": 0.7042093287827076,
+      "grad_norm": 0.7944636940956116,
+      "learning_rate": 0.0008600682593856656,
+      "loss": 2.0677,
+      "step": 619
+    },
+    {
+      "epoch": 0.7053469852104665,
+      "grad_norm": 0.9612780213356018,
+      "learning_rate": 0.0008598407281001138,
+      "loss": 1.693,
+      "step": 620
+    },
+    {
+      "epoch": 0.7064846416382252,
+      "grad_norm": 1.2384001016616821,
+      "learning_rate": 0.0008596131968145621,
+      "loss": 2.9955,
+      "step": 621
+    },
+    {
+      "epoch": 0.7076222980659841,
+      "grad_norm": 1.5061485767364502,
+      "learning_rate": 0.0008593856655290102,
+      "loss": 1.945,
+      "step": 622
+    },
+    {
+      "epoch": 0.7087599544937428,
+      "grad_norm": 1.0285781621932983,
+      "learning_rate": 0.0008591581342434584,
+      "loss": 2.5533,
+      "step": 623
+    },
+    {
+      "epoch": 0.7098976109215017,
+      "grad_norm": 0.783734917640686,
+      "learning_rate": 0.0008589306029579067,
+      "loss": 1.6898,
+      "step": 624
+    },
+    {
+      "epoch": 0.7110352673492605,
+      "grad_norm": 0.7737388610839844,
+      "learning_rate": 0.000858703071672355,
+      "loss": 2.4584,
+      "step": 625
+    },
+    {
+      "epoch": 0.7121729237770194,
+      "grad_norm": 0.6664589047431946,
+      "learning_rate": 0.0008584755403868032,
+      "loss": 1.7401,
+      "step": 626
+    },
+    {
+      "epoch": 0.7133105802047781,
+      "grad_norm": 0.7452929019927979,
+      "learning_rate": 0.0008582480091012515,
+      "loss": 1.3864,
+      "step": 627
+    },
+    {
+      "epoch": 0.714448236632537,
+      "grad_norm": 0.8343956470489502,
+      "learning_rate": 0.0008580204778156997,
+      "loss": 1.4364,
+      "step": 628
+    },
+    {
+      "epoch": 0.7155858930602957,
+      "grad_norm": 0.8422945737838745,
+      "learning_rate": 0.0008577929465301479,
+      "loss": 2.3053,
+      "step": 629
+    },
+    {
+      "epoch": 0.7167235494880546,
+      "grad_norm": 1.2495871782302856,
+      "learning_rate": 0.0008575654152445962,
+      "loss": 3.8026,
+      "step": 630
+    },
+    {
+      "epoch": 0.7178612059158134,
+      "grad_norm": 1.0053642988204956,
+      "learning_rate": 0.0008573378839590444,
+      "loss": 2.2322,
+      "step": 631
+    },
+    {
+      "epoch": 0.7189988623435722,
+      "grad_norm": 1.342475414276123,
+      "learning_rate": 0.0008571103526734925,
+      "loss": 2.2204,
+      "step": 632
+    },
+    {
+      "epoch": 0.7201365187713311,
+      "grad_norm": 1.1478005647659302,
+      "learning_rate": 0.0008568828213879408,
+      "loss": 2.0024,
+      "step": 633
+    },
+    {
+      "epoch": 0.7212741751990899,
+      "grad_norm": 1.3100638389587402,
+      "learning_rate": 0.0008566552901023891,
+      "loss": 1.7865,
+      "step": 634
+    },
+    {
+      "epoch": 0.7224118316268487,
+      "grad_norm": 0.7397316694259644,
+      "learning_rate": 0.0008564277588168374,
+      "loss": 0.9055,
+      "step": 635
+    },
+    {
+      "epoch": 0.7235494880546075,
+      "grad_norm": 1.5261900424957275,
+      "learning_rate": 0.0008562002275312856,
+      "loss": 2.518,
+      "step": 636
+    },
+    {
+      "epoch": 0.7246871444823664,
+      "grad_norm": 1.1781413555145264,
+      "learning_rate": 0.0008559726962457338,
+      "loss": 2.4799,
+      "step": 637
+    },
+    {
+      "epoch": 0.7258248009101251,
+      "grad_norm": 0.9748155474662781,
+      "learning_rate": 0.0008557451649601821,
+      "loss": 1.9048,
+      "step": 638
+    },
+    {
+      "epoch": 0.726962457337884,
+      "grad_norm": 0.9338977336883545,
+      "learning_rate": 0.0008555176336746303,
+      "loss": 2.0054,
+      "step": 639
+    },
+    {
+      "epoch": 0.7281001137656428,
+      "grad_norm": 1.5872938632965088,
+      "learning_rate": 0.0008552901023890785,
+      "loss": 2.7386,
+      "step": 640
+    },
+    {
+      "epoch": 0.7292377701934016,
+      "grad_norm": 0.899350643157959,
+      "learning_rate": 0.0008550625711035267,
+      "loss": 1.694,
+      "step": 641
+    },
+    {
+      "epoch": 0.7303754266211604,
+      "grad_norm": 1.3371248245239258,
+      "learning_rate": 0.000854835039817975,
+      "loss": 2.4103,
+      "step": 642
+    },
+    {
+      "epoch": 0.7315130830489193,
+      "grad_norm": 0.9004554152488708,
+      "learning_rate": 0.0008546075085324232,
+      "loss": 1.7984,
+      "step": 643
+    },
+    {
+      "epoch": 0.732650739476678,
+      "grad_norm": 1.229750394821167,
+      "learning_rate": 0.0008543799772468715,
+      "loss": 2.8663,
+      "step": 644
+    },
+    {
+      "epoch": 0.7337883959044369,
+      "grad_norm": 0.8719218373298645,
+      "learning_rate": 0.0008541524459613197,
+      "loss": 1.5535,
+      "step": 645
+    },
+    {
+      "epoch": 0.7349260523321957,
+      "grad_norm": 1.2239844799041748,
+      "learning_rate": 0.0008539249146757679,
+      "loss": 2.5728,
+      "step": 646
+    },
+    {
+      "epoch": 0.7360637087599545,
+      "grad_norm": 0.9655494093894958,
+      "learning_rate": 0.0008536973833902162,
+      "loss": 1.7532,
+      "step": 647
+    },
+    {
+      "epoch": 0.7372013651877133,
+      "grad_norm": 0.8480125665664673,
+      "learning_rate": 0.0008534698521046644,
+      "loss": 1.3947,
+      "step": 648
+    },
+    {
+      "epoch": 0.7383390216154722,
+      "grad_norm": 1.0782033205032349,
+      "learning_rate": 0.0008532423208191126,
+      "loss": 2.2726,
+      "step": 649
+    },
+    {
+      "epoch": 0.7394766780432309,
+      "grad_norm": 0.8517910838127136,
+      "learning_rate": 0.000853014789533561,
+      "loss": 1.7146,
+      "step": 650
+    },
+    {
+      "epoch": 0.7406143344709898,
+      "grad_norm": 0.6702635288238525,
+      "learning_rate": 0.0008527872582480091,
+      "loss": 1.4064,
+      "step": 651
+    },
+    {
+      "epoch": 0.7417519908987485,
+      "grad_norm": 0.807049572467804,
+      "learning_rate": 0.0008525597269624573,
+      "loss": 1.5396,
+      "step": 652
+    },
+    {
+      "epoch": 0.7428896473265074,
+      "grad_norm": 1.3272011280059814,
+      "learning_rate": 0.0008523321956769056,
+      "loss": 2.6018,
+      "step": 653
+    },
+    {
+      "epoch": 0.7440273037542662,
+      "grad_norm": 0.8511213064193726,
+      "learning_rate": 0.0008521046643913538,
+      "loss": 2.0945,
+      "step": 654
+    },
+    {
+      "epoch": 0.745164960182025,
+      "grad_norm": 1.0640743970870972,
+      "learning_rate": 0.0008518771331058021,
+      "loss": 3.3749,
+      "step": 655
+    },
+    {
+      "epoch": 0.7463026166097838,
+      "grad_norm": 0.9631751179695129,
+      "learning_rate": 0.0008516496018202503,
+      "loss": 2.4991,
+      "step": 656
+    },
+    {
+      "epoch": 0.7474402730375427,
+      "grad_norm": 1.0921177864074707,
+      "learning_rate": 0.0008514220705346985,
+      "loss": 1.9951,
+      "step": 657
+    },
+    {
+      "epoch": 0.7485779294653014,
+      "grad_norm": 1.032777190208435,
+      "learning_rate": 0.0008511945392491469,
+      "loss": 1.9648,
+      "step": 658
+    },
+    {
+      "epoch": 0.7497155858930603,
+      "grad_norm": 0.7291549444198608,
+      "learning_rate": 0.0008509670079635951,
+      "loss": 2.0568,
+      "step": 659
+    },
+    {
+      "epoch": 0.7508532423208191,
+      "grad_norm": 1.0137559175491333,
+      "learning_rate": 0.0008507394766780433,
+      "loss": 2.5676,
+      "step": 660
+    },
+    {
+      "epoch": 0.7519908987485779,
+      "grad_norm": 0.9185921549797058,
+      "learning_rate": 0.0008505119453924915,
+      "loss": 1.6855,
+      "step": 661
+    },
+    {
+      "epoch": 0.7531285551763367,
+      "grad_norm": 1.0048507452011108,
+      "learning_rate": 0.0008502844141069397,
+      "loss": 1.5001,
+      "step": 662
+    },
+    {
+      "epoch": 0.7542662116040956,
+      "grad_norm": 1.2506744861602783,
+      "learning_rate": 0.0008500568828213879,
+      "loss": 2.4675,
+      "step": 663
+    },
+    {
+      "epoch": 0.7554038680318543,
+      "grad_norm": 1.7534509897232056,
+      "learning_rate": 0.0008498293515358362,
+      "loss": 4.192,
+      "step": 664
+    },
+    {
+      "epoch": 0.7565415244596132,
+      "grad_norm": 1.0792453289031982,
+      "learning_rate": 0.0008496018202502844,
+      "loss": 2.2308,
+      "step": 665
+    },
+    {
+      "epoch": 0.757679180887372,
+      "grad_norm": 0.9935582280158997,
+      "learning_rate": 0.0008493742889647326,
+      "loss": 1.499,
+      "step": 666
+    },
+    {
+      "epoch": 0.7588168373151308,
+      "grad_norm": 0.9015896916389465,
+      "learning_rate": 0.000849146757679181,
+      "loss": 1.6236,
+      "step": 667
+    },
+    {
+      "epoch": 0.7599544937428896,
+      "grad_norm": 1.0141130685806274,
+      "learning_rate": 0.0008489192263936292,
+      "loss": 1.9243,
+      "step": 668
+    },
+    {
+      "epoch": 0.7610921501706485,
+      "grad_norm": 0.5651462078094482,
+      "learning_rate": 0.0008486916951080774,
+      "loss": 0.778,
+      "step": 669
+    },
+    {
+      "epoch": 0.7622298065984073,
+      "grad_norm": 1.0191128253936768,
+      "learning_rate": 0.0008484641638225257,
+      "loss": 2.0027,
+      "step": 670
+    },
+    {
+      "epoch": 0.7633674630261661,
+      "grad_norm": 1.0855798721313477,
+      "learning_rate": 0.0008482366325369738,
+      "loss": 2.4567,
+      "step": 671
+    },
+    {
+      "epoch": 0.764505119453925,
+      "grad_norm": 1.7190563678741455,
+      "learning_rate": 0.000848009101251422,
+      "loss": 2.7192,
+      "step": 672
+    },
+    {
+      "epoch": 0.7656427758816837,
+      "grad_norm": 1.0108932256698608,
+      "learning_rate": 0.0008477815699658703,
+      "loss": 1.2952,
+      "step": 673
+    },
+    {
+      "epoch": 0.7667804323094426,
+      "grad_norm": 2.3063066005706787,
+      "learning_rate": 0.0008475540386803185,
+      "loss": 1.726,
+      "step": 674
+    },
+    {
+      "epoch": 0.7679180887372014,
+      "grad_norm": 0.5806533098220825,
+      "learning_rate": 0.0008473265073947669,
+      "loss": 1.3231,
+      "step": 675
+    },
+    {
+      "epoch": 0.7690557451649602,
+      "grad_norm": 0.9485315680503845,
+      "learning_rate": 0.0008470989761092151,
+      "loss": 1.7927,
+      "step": 676
+    },
+    {
+      "epoch": 0.770193401592719,
+      "grad_norm": 0.7798807621002197,
+      "learning_rate": 0.0008468714448236633,
+      "loss": 1.4712,
+      "step": 677
+    },
+    {
+      "epoch": 0.7713310580204779,
+      "grad_norm": 1.0551048517227173,
+      "learning_rate": 0.0008466439135381116,
+      "loss": 1.9981,
+      "step": 678
+    },
+    {
+      "epoch": 0.7724687144482366,
+      "grad_norm": 1.0884678363800049,
+      "learning_rate": 0.0008464163822525598,
+      "loss": 2.0978,
+      "step": 679
+    },
+    {
+      "epoch": 0.7736063708759955,
+      "grad_norm": 0.9386458992958069,
+      "learning_rate": 0.0008461888509670079,
+      "loss": 2.5401,
+      "step": 680
+    },
+    {
+      "epoch": 0.7747440273037542,
+      "grad_norm": 1.1224387884140015,
+      "learning_rate": 0.0008459613196814562,
+      "loss": 2.5177,
+      "step": 681
+    },
+    {
+      "epoch": 0.7758816837315131,
+      "grad_norm": 0.7325891852378845,
+      "learning_rate": 0.0008457337883959044,
+      "loss": 1.9845,
+      "step": 682
+    },
+    {
+      "epoch": 0.7770193401592719,
+      "grad_norm": 0.8216614723205566,
+      "learning_rate": 0.0008455062571103526,
+      "loss": 1.7352,
+      "step": 683
+    },
+    {
+      "epoch": 0.7781569965870307,
+      "grad_norm": 0.8514799475669861,
+      "learning_rate": 0.000845278725824801,
+      "loss": 1.6026,
+      "step": 684
+    },
+    {
+      "epoch": 0.7792946530147895,
+      "grad_norm": 1.0461024045944214,
+      "learning_rate": 0.0008450511945392492,
+      "loss": 2.7851,
+      "step": 685
+    },
+    {
+      "epoch": 0.7804323094425484,
+      "grad_norm": 1.0738078355789185,
+      "learning_rate": 0.0008448236632536974,
+      "loss": 2.2409,
+      "step": 686
+    },
+    {
+      "epoch": 0.7815699658703071,
+      "grad_norm": 1.5895496606826782,
+      "learning_rate": 0.0008445961319681457,
+      "loss": 2.6157,
+      "step": 687
+    },
+    {
+      "epoch": 0.782707622298066,
+      "grad_norm": 1.095568299293518,
+      "learning_rate": 0.0008443686006825939,
+      "loss": 3.4401,
+      "step": 688
+    },
+    {
+      "epoch": 0.7838452787258248,
+      "grad_norm": 1.2721914052963257,
+      "learning_rate": 0.0008441410693970421,
+      "loss": 3.7557,
+      "step": 689
+    },
+    {
+      "epoch": 0.7849829351535836,
+      "grad_norm": 0.8100789785385132,
+      "learning_rate": 0.0008439135381114903,
+      "loss": 1.8537,
+      "step": 690
+    },
+    {
+      "epoch": 0.7861205915813424,
+      "grad_norm": 0.8364444375038147,
+      "learning_rate": 0.0008436860068259385,
+      "loss": 1.9646,
+      "step": 691
+    },
+    {
+      "epoch": 0.7872582480091013,
+      "grad_norm": 0.8454108834266663,
+      "learning_rate": 0.0008434584755403867,
+      "loss": 1.3103,
+      "step": 692
+    },
+    {
+      "epoch": 0.78839590443686,
+      "grad_norm": 1.0426613092422485,
+      "learning_rate": 0.0008432309442548351,
+      "loss": 2.1801,
+      "step": 693
+    },
+    {
+      "epoch": 0.7895335608646189,
+      "grad_norm": 0.6206464171409607,
+      "learning_rate": 0.0008430034129692833,
+      "loss": 0.6764,
+      "step": 694
+    },
+    {
+      "epoch": 0.7906712172923777,
+      "grad_norm": 1.0742131471633911,
+      "learning_rate": 0.0008427758816837315,
+      "loss": 2.0438,
+      "step": 695
+    },
+    {
+      "epoch": 0.7918088737201365,
+      "grad_norm": 1.2915891408920288,
+      "learning_rate": 0.0008425483503981798,
+      "loss": 2.3515,
+      "step": 696
+    },
+    {
+      "epoch": 0.7929465301478953,
+      "grad_norm": 1.545964002609253,
+      "learning_rate": 0.000842320819112628,
+      "loss": 3.3463,
+      "step": 697
+    },
+    {
+      "epoch": 0.7940841865756542,
+      "grad_norm": 0.8047581911087036,
+      "learning_rate": 0.0008420932878270763,
+      "loss": 2.392,
+      "step": 698
+    },
+    {
+      "epoch": 0.7952218430034129,
+      "grad_norm": 1.0554293394088745,
+      "learning_rate": 0.0008418657565415245,
+      "loss": 1.8506,
+      "step": 699
+    },
+    {
+      "epoch": 0.7963594994311718,
+      "grad_norm": 1.518147587776184,
+      "learning_rate": 0.0008416382252559726,
+      "loss": 4.3646,
+      "step": 700
+    },
+    {
+      "epoch": 0.7974971558589306,
+      "grad_norm": 0.942899227142334,
+      "learning_rate": 0.000841410693970421,
+      "loss": 1.8301,
+      "step": 701
+    },
+    {
+      "epoch": 0.7986348122866894,
+      "grad_norm": 1.1088321208953857,
+      "learning_rate": 0.0008411831626848692,
+      "loss": 3.0843,
+      "step": 702
+    },
+    {
+      "epoch": 0.7997724687144482,
+      "grad_norm": 1.0423277616500854,
+      "learning_rate": 0.0008409556313993174,
+      "loss": 2.6175,
+      "step": 703
+    },
+    {
+      "epoch": 0.800910125142207,
+      "grad_norm": 0.7414242625236511,
+      "learning_rate": 0.0008407281001137657,
+      "loss": 1.399,
+      "step": 704
+    },
+    {
+      "epoch": 0.8020477815699659,
+      "grad_norm": 1.0195127725601196,
+      "learning_rate": 0.0008405005688282139,
+      "loss": 2.1314,
+      "step": 705
+    },
+    {
+      "epoch": 0.8031854379977247,
+      "grad_norm": 1.4709278345108032,
+      "learning_rate": 0.0008402730375426621,
+      "loss": 3.2922,
+      "step": 706
+    },
+    {
+      "epoch": 0.8043230944254836,
+      "grad_norm": 1.656906247138977,
+      "learning_rate": 0.0008400455062571104,
+      "loss": 3.844,
+      "step": 707
+    },
+    {
+      "epoch": 0.8054607508532423,
+      "grad_norm": 0.985885500907898,
+      "learning_rate": 0.0008398179749715586,
+      "loss": 2.9415,
+      "step": 708
+    },
+    {
+      "epoch": 0.8065984072810012,
+      "grad_norm": 1.5020736455917358,
+      "learning_rate": 0.0008395904436860067,
+      "loss": 2.183,
+      "step": 709
+    },
+    {
+      "epoch": 0.8077360637087599,
+      "grad_norm": 0.7547074556350708,
+      "learning_rate": 0.0008393629124004551,
+      "loss": 1.8464,
+      "step": 710
+    },
+    {
+      "epoch": 0.8088737201365188,
+      "grad_norm": 1.1752091646194458,
+      "learning_rate": 0.0008391353811149033,
+      "loss": 2.5145,
+      "step": 711
+    },
+    {
+      "epoch": 0.8100113765642776,
+      "grad_norm": 0.7590292096138,
+      "learning_rate": 0.0008389078498293515,
+      "loss": 1.4564,
+      "step": 712
+    },
+    {
+      "epoch": 0.8111490329920364,
+      "grad_norm": 0.8058563470840454,
+      "learning_rate": 0.0008386803185437998,
+      "loss": 1.7101,
+      "step": 713
+    },
+    {
+      "epoch": 0.8122866894197952,
+      "grad_norm": 0.971361517906189,
+      "learning_rate": 0.000838452787258248,
+      "loss": 1.8707,
+      "step": 714
+    },
+    {
+      "epoch": 0.8134243458475541,
+      "grad_norm": 1.3266187906265259,
+      "learning_rate": 0.0008382252559726962,
+      "loss": 2.8622,
+      "step": 715
+    },
+    {
+      "epoch": 0.8145620022753128,
+      "grad_norm": 1.2026985883712769,
+      "learning_rate": 0.0008379977246871445,
+      "loss": 2.386,
+      "step": 716
+    },
+    {
+      "epoch": 0.8156996587030717,
+      "grad_norm": 1.041385531425476,
+      "learning_rate": 0.0008377701934015928,
+      "loss": 2.1941,
+      "step": 717
+    },
+    {
+      "epoch": 0.8168373151308305,
+      "grad_norm": 0.9578999280929565,
+      "learning_rate": 0.0008375426621160411,
+      "loss": 2.4792,
+      "step": 718
+    },
+    {
+      "epoch": 0.8179749715585893,
+      "grad_norm": 1.569138765335083,
+      "learning_rate": 0.0008373151308304892,
+      "loss": 2.5056,
+      "step": 719
+    },
+    {
+      "epoch": 0.8191126279863481,
+      "grad_norm": 1.2005671262741089,
+      "learning_rate": 0.0008370875995449374,
+      "loss": 2.8101,
+      "step": 720
+    },
+    {
+      "epoch": 0.820250284414107,
+      "grad_norm": 0.8075481057167053,
+      "learning_rate": 0.0008368600682593857,
+      "loss": 1.1596,
+      "step": 721
+    },
+    {
+      "epoch": 0.8213879408418657,
+      "grad_norm": 1.131332278251648,
+      "learning_rate": 0.0008366325369738339,
+      "loss": 2.2506,
+      "step": 722
+    },
+    {
+      "epoch": 0.8225255972696246,
+      "grad_norm": 1.0563104152679443,
+      "learning_rate": 0.0008364050056882821,
+      "loss": 1.9322,
+      "step": 723
+    },
+    {
+      "epoch": 0.8236632536973834,
+      "grad_norm": 0.7609984874725342,
+      "learning_rate": 0.0008361774744027304,
+      "loss": 1.8229,
+      "step": 724
+    },
+    {
+      "epoch": 0.8248009101251422,
+      "grad_norm": 0.8261270523071289,
+      "learning_rate": 0.0008359499431171786,
+      "loss": 2.785,
+      "step": 725
+    },
+    {
+      "epoch": 0.825938566552901,
+      "grad_norm": 0.5597876310348511,
+      "learning_rate": 0.0008357224118316269,
+      "loss": 1.0098,
+      "step": 726
+    },
+    {
+      "epoch": 0.8270762229806599,
+      "grad_norm": 1.219977617263794,
+      "learning_rate": 0.0008354948805460752,
+      "loss": 3.3632,
+      "step": 727
+    },
+    {
+      "epoch": 0.8282138794084186,
+      "grad_norm": 1.1048645973205566,
+      "learning_rate": 0.0008352673492605234,
+      "loss": 3.235,
+      "step": 728
+    },
+    {
+      "epoch": 0.8293515358361775,
+      "grad_norm": 1.0460500717163086,
+      "learning_rate": 0.0008350398179749715,
+      "loss": 1.7857,
+      "step": 729
+    },
+    {
+      "epoch": 0.8304891922639362,
+      "grad_norm": 1.144655466079712,
+      "learning_rate": 0.0008348122866894198,
+      "loss": 2.25,
+      "step": 730
+    },
+    {
+      "epoch": 0.8316268486916951,
+      "grad_norm": 0.9807350039482117,
+      "learning_rate": 0.000834584755403868,
+      "loss": 2.4989,
+      "step": 731
+    },
+    {
+      "epoch": 0.8327645051194539,
+      "grad_norm": 0.9099501371383667,
+      "learning_rate": 0.0008343572241183162,
+      "loss": 2.3824,
+      "step": 732
+    },
+    {
+      "epoch": 0.8339021615472128,
+      "grad_norm": 0.6631197333335876,
+      "learning_rate": 0.0008341296928327645,
+      "loss": 1.6735,
+      "step": 733
+    },
+    {
+      "epoch": 0.8350398179749715,
+      "grad_norm": 0.6548817157745361,
+      "learning_rate": 0.0008339021615472128,
+      "loss": 1.5283,
+      "step": 734
+    },
+    {
+      "epoch": 0.8361774744027304,
+      "grad_norm": 0.6837170720100403,
+      "learning_rate": 0.000833674630261661,
+      "loss": 1.3024,
+      "step": 735
+    },
+    {
+      "epoch": 0.8373151308304891,
+      "grad_norm": 0.8492507934570312,
+      "learning_rate": 0.0008334470989761093,
+      "loss": 1.6662,
+      "step": 736
+    },
+    {
+      "epoch": 0.838452787258248,
+      "grad_norm": 1.2175747156143188,
+      "learning_rate": 0.0008332195676905575,
+      "loss": 2.804,
+      "step": 737
+    },
+    {
+      "epoch": 0.8395904436860068,
+      "grad_norm": 1.0982182025909424,
+      "learning_rate": 0.0008329920364050058,
+      "loss": 2.1759,
+      "step": 738
+    },
+    {
+      "epoch": 0.8407281001137656,
+      "grad_norm": 1.106663465499878,
+      "learning_rate": 0.0008327645051194539,
+      "loss": 2.0419,
+      "step": 739
+    },
+    {
+      "epoch": 0.8418657565415245,
+      "grad_norm": 0.7106451988220215,
+      "learning_rate": 0.0008325369738339021,
+      "loss": 1.2006,
+      "step": 740
+    },
+    {
+      "epoch": 0.8430034129692833,
+      "grad_norm": 1.1868878602981567,
+      "learning_rate": 0.0008323094425483504,
+      "loss": 2.0938,
+      "step": 741
+    },
+    {
+      "epoch": 0.8441410693970421,
+      "grad_norm": 1.006433367729187,
+      "learning_rate": 0.0008320819112627986,
+      "loss": 1.7796,
+      "step": 742
+    },
+    {
+      "epoch": 0.8452787258248009,
+      "grad_norm": 1.0175524950027466,
+      "learning_rate": 0.0008318543799772469,
+      "loss": 2.5526,
+      "step": 743
+    },
+    {
+      "epoch": 0.8464163822525598,
+      "grad_norm": 0.9404505491256714,
+      "learning_rate": 0.0008316268486916952,
+      "loss": 1.9948,
+      "step": 744
+    },
+    {
+      "epoch": 0.8475540386803185,
+      "grad_norm": 0.9086321592330933,
+      "learning_rate": 0.0008313993174061434,
+      "loss": 1.9277,
+      "step": 745
+    },
+    {
+      "epoch": 0.8486916951080774,
+      "grad_norm": 0.7922766208648682,
+      "learning_rate": 0.0008311717861205916,
+      "loss": 2.3355,
+      "step": 746
+    },
+    {
+      "epoch": 0.8498293515358362,
+      "grad_norm": 0.8049002289772034,
+      "learning_rate": 0.0008309442548350399,
+      "loss": 1.7969,
+      "step": 747
+    },
+    {
+      "epoch": 0.850967007963595,
+      "grad_norm": 0.8303267955780029,
+      "learning_rate": 0.000830716723549488,
+      "loss": 1.4973,
+      "step": 748
+    },
+    {
+      "epoch": 0.8521046643913538,
+      "grad_norm": 0.7163656949996948,
+      "learning_rate": 0.0008304891922639362,
+      "loss": 1.7677,
+      "step": 749
+    },
+    {
+      "epoch": 0.8532423208191127,
+      "grad_norm": 1.022519588470459,
+      "learning_rate": 0.0008302616609783845,
+      "loss": 1.6188,
+      "step": 750
+    },
+    {
+      "epoch": 0.8543799772468714,
+      "grad_norm": 1.2985116243362427,
+      "learning_rate": 0.0008300341296928328,
+      "loss": 2.3702,
+      "step": 751
+    },
+    {
+      "epoch": 0.8555176336746303,
+      "grad_norm": 0.7469916939735413,
+      "learning_rate": 0.000829806598407281,
+      "loss": 1.1039,
+      "step": 752
+    },
+    {
+      "epoch": 0.856655290102389,
+      "grad_norm": 0.8801395297050476,
+      "learning_rate": 0.0008295790671217293,
+      "loss": 1.615,
+      "step": 753
+    },
+    {
+      "epoch": 0.8577929465301479,
+      "grad_norm": 0.6937797665596008,
+      "learning_rate": 0.0008293515358361775,
+      "loss": 1.366,
+      "step": 754
+    },
+    {
+      "epoch": 0.8589306029579067,
+      "grad_norm": 0.92084801197052,
+      "learning_rate": 0.0008291240045506257,
+      "loss": 1.1217,
+      "step": 755
+    },
+    {
+      "epoch": 0.8600682593856656,
+      "grad_norm": 1.1955760717391968,
+      "learning_rate": 0.000828896473265074,
+      "loss": 2.3874,
+      "step": 756
+    },
+    {
+      "epoch": 0.8612059158134243,
+      "grad_norm": 1.270263433456421,
+      "learning_rate": 0.0008286689419795222,
+      "loss": 0.7639,
+      "step": 757
+    },
+    {
+      "epoch": 0.8623435722411832,
+      "grad_norm": 1.0862644910812378,
+      "learning_rate": 0.0008284414106939704,
+      "loss": 1.666,
+      "step": 758
+    },
+    {
+      "epoch": 0.863481228668942,
+      "grad_norm": 1.0658801794052124,
+      "learning_rate": 0.0008282138794084187,
+      "loss": 3.0639,
+      "step": 759
+    },
+    {
+      "epoch": 0.8646188850967008,
+      "grad_norm": 0.7104091644287109,
+      "learning_rate": 0.0008279863481228669,
+      "loss": 1.7136,
+      "step": 760
+    },
+    {
+      "epoch": 0.8657565415244596,
+      "grad_norm": 1.0300426483154297,
+      "learning_rate": 0.0008277588168373152,
+      "loss": 1.9242,
+      "step": 761
+    },
+    {
+      "epoch": 0.8668941979522184,
+      "grad_norm": 0.9637190699577332,
+      "learning_rate": 0.0008275312855517634,
+      "loss": 2.0804,
+      "step": 762
+    },
+    {
+      "epoch": 0.8680318543799772,
+      "grad_norm": 0.9905063509941101,
+      "learning_rate": 0.0008273037542662116,
+      "loss": 1.7953,
+      "step": 763
+    },
+    {
+      "epoch": 0.8691695108077361,
+      "grad_norm": 1.0103565454483032,
+      "learning_rate": 0.0008270762229806599,
+      "loss": 2.3104,
+      "step": 764
+    },
+    {
+      "epoch": 0.8703071672354948,
+      "grad_norm": 0.6847909688949585,
+      "learning_rate": 0.0008268486916951081,
+      "loss": 0.9037,
+      "step": 765
+    },
+    {
+      "epoch": 0.8714448236632537,
+      "grad_norm": 0.958070695400238,
+      "learning_rate": 0.0008266211604095563,
+      "loss": 1.3769,
+      "step": 766
+    },
+    {
+      "epoch": 0.8725824800910125,
+      "grad_norm": 1.081380009651184,
+      "learning_rate": 0.0008263936291240047,
+      "loss": 2.6554,
+      "step": 767
+    },
+    {
+      "epoch": 0.8737201365187713,
+      "grad_norm": 1.003764271736145,
+      "learning_rate": 0.0008261660978384528,
+      "loss": 2.3287,
+      "step": 768
+    },
+    {
+      "epoch": 0.8748577929465301,
+      "grad_norm": 0.9018321633338928,
+      "learning_rate": 0.000825938566552901,
+      "loss": 1.1058,
+      "step": 769
+    },
+    {
+      "epoch": 0.875995449374289,
+      "grad_norm": 1.702228307723999,
+      "learning_rate": 0.0008257110352673493,
+      "loss": 2.4671,
+      "step": 770
+    },
+    {
+      "epoch": 0.8771331058020477,
+      "grad_norm": 0.7024636268615723,
+      "learning_rate": 0.0008254835039817975,
+      "loss": 1.3952,
+      "step": 771
+    },
+    {
+      "epoch": 0.8782707622298066,
+      "grad_norm": 0.8887207508087158,
+      "learning_rate": 0.0008252559726962457,
+      "loss": 1.8992,
+      "step": 772
+    },
+    {
+      "epoch": 0.8794084186575654,
+      "grad_norm": 0.8344920873641968,
+      "learning_rate": 0.000825028441410694,
+      "loss": 1.003,
+      "step": 773
+    },
+    {
+      "epoch": 0.8805460750853242,
+      "grad_norm": 1.174970030784607,
+      "learning_rate": 0.0008248009101251422,
+      "loss": 2.1063,
+      "step": 774
+    },
+    {
+      "epoch": 0.8816837315130831,
+      "grad_norm": 1.4693725109100342,
+      "learning_rate": 0.0008245733788395904,
+      "loss": 2.349,
+      "step": 775
+    },
+    {
+      "epoch": 0.8828213879408419,
+      "grad_norm": 0.8909391164779663,
+      "learning_rate": 0.0008243458475540388,
+      "loss": 2.3083,
+      "step": 776
+    },
+    {
+      "epoch": 0.8839590443686007,
+      "grad_norm": 1.178287148475647,
+      "learning_rate": 0.000824118316268487,
+      "loss": 2.2453,
+      "step": 777
+    },
+    {
+      "epoch": 0.8850967007963595,
+      "grad_norm": 1.1147403717041016,
+      "learning_rate": 0.0008238907849829351,
+      "loss": 3.2198,
+      "step": 778
+    },
+    {
+      "epoch": 0.8862343572241184,
+      "grad_norm": 0.8934231996536255,
+      "learning_rate": 0.0008236632536973834,
+      "loss": 2.1325,
+      "step": 779
+    },
+    {
+      "epoch": 0.8873720136518771,
+      "grad_norm": 0.7500196695327759,
+      "learning_rate": 0.0008234357224118316,
+      "loss": 1.062,
+      "step": 780
+    },
+    {
+      "epoch": 0.888509670079636,
+      "grad_norm": 1.1905717849731445,
+      "learning_rate": 0.0008232081911262799,
+      "loss": 2.6013,
+      "step": 781
+    },
+    {
+      "epoch": 0.8896473265073948,
+      "grad_norm": 0.7128300070762634,
+      "learning_rate": 0.0008229806598407281,
+      "loss": 1.4425,
+      "step": 782
+    },
+    {
+      "epoch": 0.8907849829351536,
+      "grad_norm": 0.9810397028923035,
+      "learning_rate": 0.0008227531285551763,
+      "loss": 2.2742,
+      "step": 783
+    },
+    {
+      "epoch": 0.8919226393629124,
+      "grad_norm": 1.1497716903686523,
+      "learning_rate": 0.0008225255972696247,
+      "loss": 2.6226,
+      "step": 784
+    },
+    {
+      "epoch": 0.8930602957906713,
+      "grad_norm": 1.3112133741378784,
+      "learning_rate": 0.0008222980659840729,
+      "loss": 1.8846,
+      "step": 785
+    },
+    {
+      "epoch": 0.89419795221843,
+      "grad_norm": 1.030337929725647,
+      "learning_rate": 0.0008220705346985211,
+      "loss": 2.6641,
+      "step": 786
+    },
+    {
+      "epoch": 0.8953356086461889,
+      "grad_norm": 1.1316418647766113,
+      "learning_rate": 0.0008218430034129693,
+      "loss": 1.8058,
+      "step": 787
+    },
+    {
+      "epoch": 0.8964732650739476,
+      "grad_norm": 1.315293788909912,
+      "learning_rate": 0.0008216154721274175,
+      "loss": 1.6608,
+      "step": 788
+    },
+    {
+      "epoch": 0.8976109215017065,
+      "grad_norm": 1.0237807035446167,
+      "learning_rate": 0.0008213879408418657,
+      "loss": 2.9869,
+      "step": 789
+    },
+    {
+      "epoch": 0.8987485779294653,
+      "grad_norm": 1.0724749565124512,
+      "learning_rate": 0.000821160409556314,
+      "loss": 1.7659,
+      "step": 790
+    },
+    {
+      "epoch": 0.8998862343572241,
+      "grad_norm": 0.8942409753799438,
+      "learning_rate": 0.0008209328782707622,
+      "loss": 1.5329,
+      "step": 791
+    },
+    {
+      "epoch": 0.9010238907849829,
+      "grad_norm": 1.5801607370376587,
+      "learning_rate": 0.0008207053469852104,
+      "loss": 1.8717,
+      "step": 792
+    },
+    {
+      "epoch": 0.9021615472127418,
+      "grad_norm": 0.8906893730163574,
+      "learning_rate": 0.0008204778156996588,
+      "loss": 1.5824,
+      "step": 793
+    },
+    {
+      "epoch": 0.9032992036405005,
+      "grad_norm": 0.6598983407020569,
+      "learning_rate": 0.000820250284414107,
+      "loss": 1.4168,
+      "step": 794
+    },
+    {
+      "epoch": 0.9044368600682594,
+      "grad_norm": 0.7640515565872192,
+      "learning_rate": 0.0008200227531285552,
+      "loss": 1.7068,
+      "step": 795
+    },
+    {
+      "epoch": 0.9055745164960182,
+      "grad_norm": 1.3205150365829468,
+      "learning_rate": 0.0008197952218430035,
+      "loss": 2.5249,
+      "step": 796
+    },
+    {
+      "epoch": 0.906712172923777,
+      "grad_norm": 1.2995262145996094,
+      "learning_rate": 0.0008195676905574516,
+      "loss": 2.9522,
+      "step": 797
+    },
+    {
+      "epoch": 0.9078498293515358,
+      "grad_norm": 0.8465941548347473,
+      "learning_rate": 0.0008193401592718998,
+      "loss": 2.0508,
+      "step": 798
+    },
+    {
+      "epoch": 0.9089874857792947,
+      "grad_norm": 1.2404921054840088,
+      "learning_rate": 0.0008191126279863481,
+      "loss": 2.4488,
+      "step": 799
+    },
+    {
+      "epoch": 0.9101251422070534,
+      "grad_norm": 1.00394868850708,
+      "learning_rate": 0.0008188850967007963,
+      "loss": 1.665,
+      "step": 800
+    },
+    {
+      "epoch": 0.9112627986348123,
+      "grad_norm": 1.0288304090499878,
+      "learning_rate": 0.0008186575654152447,
+      "loss": 2.2679,
+      "step": 801
+    },
+    {
+      "epoch": 0.9124004550625711,
+      "grad_norm": 1.0812597274780273,
+      "learning_rate": 0.0008184300341296929,
+      "loss": 2.092,
+      "step": 802
+    },
+    {
+      "epoch": 0.9135381114903299,
+      "grad_norm": 0.9165686964988708,
+      "learning_rate": 0.0008182025028441411,
+      "loss": 1.6572,
+      "step": 803
+    },
+    {
+      "epoch": 0.9146757679180887,
+      "grad_norm": 1.254876971244812,
+      "learning_rate": 0.0008179749715585894,
+      "loss": 2.2742,
+      "step": 804
+    },
+    {
+      "epoch": 0.9158134243458476,
+      "grad_norm": 0.8111211657524109,
+      "learning_rate": 0.0008177474402730376,
+      "loss": 1.4599,
+      "step": 805
+    },
+    {
+      "epoch": 0.9169510807736063,
+      "grad_norm": 1.2796216011047363,
+      "learning_rate": 0.0008175199089874858,
+      "loss": 2.8677,
+      "step": 806
+    },
+    {
+      "epoch": 0.9180887372013652,
+      "grad_norm": 0.5638169050216675,
+      "learning_rate": 0.000817292377701934,
+      "loss": 1.0711,
+      "step": 807
+    },
+    {
+      "epoch": 0.919226393629124,
+      "grad_norm": 1.0181145668029785,
+      "learning_rate": 0.0008170648464163822,
+      "loss": 2.546,
+      "step": 808
+    },
+    {
+      "epoch": 0.9203640500568828,
+      "grad_norm": 1.0140022039413452,
+      "learning_rate": 0.0008168373151308304,
+      "loss": 1.6735,
+      "step": 809
+    },
+    {
+      "epoch": 0.9215017064846417,
+      "grad_norm": 0.6265180706977844,
+      "learning_rate": 0.0008166097838452788,
+      "loss": 1.581,
+      "step": 810
+    },
+    {
+      "epoch": 0.9226393629124005,
+      "grad_norm": 0.9092225432395935,
+      "learning_rate": 0.000816382252559727,
+      "loss": 2.0735,
+      "step": 811
+    },
+    {
+      "epoch": 0.9237770193401593,
+      "grad_norm": 1.3408722877502441,
+      "learning_rate": 0.0008161547212741752,
+      "loss": 3.3588,
+      "step": 812
+    },
+    {
+      "epoch": 0.9249146757679181,
+      "grad_norm": 1.0309351682662964,
+      "learning_rate": 0.0008159271899886235,
+      "loss": 2.2884,
+      "step": 813
+    },
+    {
+      "epoch": 0.926052332195677,
+      "grad_norm": 1.130393147468567,
+      "learning_rate": 0.0008156996587030717,
+      "loss": 1.994,
+      "step": 814
+    },
+    {
+      "epoch": 0.9271899886234357,
+      "grad_norm": 0.9755035042762756,
+      "learning_rate": 0.0008154721274175199,
+      "loss": 2.4471,
+      "step": 815
+    },
+    {
+      "epoch": 0.9283276450511946,
+      "grad_norm": 0.7888014912605286,
+      "learning_rate": 0.0008152445961319681,
+      "loss": 1.7634,
+      "step": 816
+    },
+    {
+      "epoch": 0.9294653014789533,
+      "grad_norm": 1.2425562143325806,
+      "learning_rate": 0.0008150170648464163,
+      "loss": 2.6725,
+      "step": 817
+    },
+    {
+      "epoch": 0.9306029579067122,
+      "grad_norm": 1.112234354019165,
+      "learning_rate": 0.0008147895335608646,
+      "loss": 2.0245,
+      "step": 818
+    },
+    {
+      "epoch": 0.931740614334471,
+      "grad_norm": 1.8099498748779297,
+      "learning_rate": 0.0008145620022753129,
+      "loss": 3.6121,
+      "step": 819
+    },
+    {
+      "epoch": 0.9328782707622298,
+      "grad_norm": 0.594555139541626,
+      "learning_rate": 0.0008143344709897611,
+      "loss": 1.1207,
+      "step": 820
+    },
+    {
+      "epoch": 0.9340159271899886,
+      "grad_norm": 1.2875663042068481,
+      "learning_rate": 0.0008141069397042094,
+      "loss": 2.6515,
+      "step": 821
+    },
+    {
+      "epoch": 0.9351535836177475,
+      "grad_norm": 1.2231221199035645,
+      "learning_rate": 0.0008138794084186576,
+      "loss": 2.1126,
+      "step": 822
+    },
+    {
+      "epoch": 0.9362912400455062,
+      "grad_norm": 0.6779431700706482,
+      "learning_rate": 0.0008136518771331058,
+      "loss": 1.0446,
+      "step": 823
+    },
+    {
+      "epoch": 0.9374288964732651,
+      "grad_norm": 0.8786545991897583,
+      "learning_rate": 0.0008134243458475541,
+      "loss": 1.8822,
+      "step": 824
+    },
+    {
+      "epoch": 0.9385665529010239,
+      "grad_norm": 0.7241141200065613,
+      "learning_rate": 0.0008131968145620023,
+      "loss": 1.3241,
+      "step": 825
+    },
+    {
+      "epoch": 0.9397042093287827,
+      "grad_norm": 1.032473087310791,
+      "learning_rate": 0.0008129692832764504,
+      "loss": 2.3931,
+      "step": 826
+    },
+    {
+      "epoch": 0.9408418657565415,
+      "grad_norm": 0.8265206217765808,
+      "learning_rate": 0.0008127417519908988,
+      "loss": 1.842,
+      "step": 827
+    },
+    {
+      "epoch": 0.9419795221843004,
+      "grad_norm": 0.9826211929321289,
+      "learning_rate": 0.000812514220705347,
+      "loss": 2.6482,
+      "step": 828
+    },
+    {
+      "epoch": 0.9431171786120591,
+      "grad_norm": 1.065434455871582,
+      "learning_rate": 0.0008122866894197952,
+      "loss": 2.2127,
+      "step": 829
+    },
+    {
+      "epoch": 0.944254835039818,
+      "grad_norm": 0.6911349892616272,
+      "learning_rate": 0.0008120591581342435,
+      "loss": 1.7043,
+      "step": 830
+    },
+    {
+      "epoch": 0.9453924914675768,
+      "grad_norm": 1.031029224395752,
+      "learning_rate": 0.0008118316268486917,
+      "loss": 1.971,
+      "step": 831
+    },
+    {
+      "epoch": 0.9465301478953356,
+      "grad_norm": 0.9776269197463989,
+      "learning_rate": 0.0008116040955631399,
+      "loss": 1.7573,
+      "step": 832
+    },
+    {
+      "epoch": 0.9476678043230944,
+      "grad_norm": 1.006766438484192,
+      "learning_rate": 0.0008113765642775882,
+      "loss": 2.1867,
+      "step": 833
+    },
+    {
+      "epoch": 0.9488054607508533,
+      "grad_norm": 1.4474374055862427,
+      "learning_rate": 0.0008111490329920365,
+      "loss": 1.8203,
+      "step": 834
+    },
+    {
+      "epoch": 0.949943117178612,
+      "grad_norm": 1.0589663982391357,
+      "learning_rate": 0.0008109215017064847,
+      "loss": 2.376,
+      "step": 835
+    },
+    {
+      "epoch": 0.9510807736063709,
+      "grad_norm": 0.8817835450172424,
+      "learning_rate": 0.0008106939704209329,
+      "loss": 2.1685,
+      "step": 836
+    },
+    {
+      "epoch": 0.9522184300341296,
+      "grad_norm": 1.0257823467254639,
+      "learning_rate": 0.0008104664391353811,
+      "loss": 2.7541,
+      "step": 837
+    },
+    {
+      "epoch": 0.9533560864618885,
+      "grad_norm": 0.6072942018508911,
+      "learning_rate": 0.0008102389078498293,
+      "loss": 0.9657,
+      "step": 838
+    },
+    {
+      "epoch": 0.9544937428896473,
+      "grad_norm": 0.6976664066314697,
+      "learning_rate": 0.0008100113765642776,
+      "loss": 1.7921,
+      "step": 839
+    },
+    {
+      "epoch": 0.9556313993174061,
+      "grad_norm": 0.7202094197273254,
+      "learning_rate": 0.0008097838452787258,
+      "loss": 1.6851,
+      "step": 840
+    },
+    {
+      "epoch": 0.9567690557451649,
+      "grad_norm": 0.7134101390838623,
+      "learning_rate": 0.0008095563139931741,
+      "loss": 1.2245,
+      "step": 841
+    },
+    {
+      "epoch": 0.9579067121729238,
+      "grad_norm": 0.9187147617340088,
+      "learning_rate": 0.0008093287827076223,
+      "loss": 1.9654,
+      "step": 842
+    },
+    {
+      "epoch": 0.9590443686006825,
+      "grad_norm": 1.6910814046859741,
+      "learning_rate": 0.0008091012514220706,
+      "loss": 3.1648,
+      "step": 843
+    },
+    {
+      "epoch": 0.9601820250284414,
+      "grad_norm": 1.1383386850357056,
+      "learning_rate": 0.0008088737201365189,
+      "loss": 1.5833,
+      "step": 844
+    },
+    {
+      "epoch": 0.9613196814562003,
+      "grad_norm": 1.1125869750976562,
+      "learning_rate": 0.0008086461888509671,
+      "loss": 2.4418,
+      "step": 845
+    },
+    {
+      "epoch": 0.962457337883959,
+      "grad_norm": 1.096062183380127,
+      "learning_rate": 0.0008084186575654152,
+      "loss": 1.5872,
+      "step": 846
+    },
+    {
+      "epoch": 0.9635949943117179,
+      "grad_norm": 1.1159650087356567,
+      "learning_rate": 0.0008081911262798635,
+      "loss": 0.9669,
+      "step": 847
+    },
+    {
+      "epoch": 0.9647326507394767,
+      "grad_norm": 0.8109821081161499,
+      "learning_rate": 0.0008079635949943117,
+      "loss": 2.3282,
+      "step": 848
+    },
+    {
+      "epoch": 0.9658703071672355,
+      "grad_norm": 0.6839150786399841,
+      "learning_rate": 0.0008077360637087599,
+      "loss": 1.5686,
+      "step": 849
+    },
+    {
+      "epoch": 0.9670079635949943,
+      "grad_norm": 1.0192281007766724,
+      "learning_rate": 0.0008075085324232082,
+      "loss": 2.3769,
+      "step": 850
+    },
+    {
+      "epoch": 0.9681456200227532,
+      "grad_norm": 1.0800739526748657,
+      "learning_rate": 0.0008072810011376565,
+      "loss": 1.9699,
+      "step": 851
+    },
+    {
+      "epoch": 0.9692832764505119,
+      "grad_norm": 1.4787782430648804,
+      "learning_rate": 0.0008070534698521047,
+      "loss": 2.1332,
+      "step": 852
+    },
+    {
+      "epoch": 0.9704209328782708,
+      "grad_norm": 0.8719042539596558,
+      "learning_rate": 0.000806825938566553,
+      "loss": 1.9816,
+      "step": 853
+    },
+    {
+      "epoch": 0.9715585893060296,
+      "grad_norm": 0.4718307554721832,
+      "learning_rate": 0.0008065984072810012,
+      "loss": 0.907,
+      "step": 854
+    },
+    {
+      "epoch": 0.9726962457337884,
+      "grad_norm": 0.7292342185974121,
+      "learning_rate": 0.0008063708759954493,
+      "loss": 1.3498,
+      "step": 855
+    },
+    {
+      "epoch": 0.9738339021615472,
+      "grad_norm": 1.0442382097244263,
+      "learning_rate": 0.0008061433447098976,
+      "loss": 0.8467,
+      "step": 856
+    },
+    {
+      "epoch": 0.9749715585893061,
+      "grad_norm": 1.4007940292358398,
+      "learning_rate": 0.0008059158134243458,
+      "loss": 2.5427,
+      "step": 857
+    },
+    {
+      "epoch": 0.9761092150170648,
+      "grad_norm": 0.6884713172912598,
+      "learning_rate": 0.000805688282138794,
+      "loss": 1.4121,
+      "step": 858
+    },
+    {
+      "epoch": 0.9772468714448237,
+      "grad_norm": 0.8210100531578064,
+      "learning_rate": 0.0008054607508532424,
+      "loss": 1.4311,
+      "step": 859
+    },
+    {
+      "epoch": 0.9783845278725825,
+      "grad_norm": 1.017969012260437,
+      "learning_rate": 0.0008052332195676906,
+      "loss": 2.1577,
+      "step": 860
+    },
+    {
+      "epoch": 0.9795221843003413,
+      "grad_norm": 0.6463725566864014,
+      "learning_rate": 0.0008050056882821389,
+      "loss": 1.1913,
+      "step": 861
+    },
+    {
+      "epoch": 0.9806598407281001,
+      "grad_norm": 0.609213650226593,
+      "learning_rate": 0.0008047781569965871,
+      "loss": 1.4609,
+      "step": 862
+    },
+    {
+      "epoch": 0.981797497155859,
+      "grad_norm": 1.2324273586273193,
+      "learning_rate": 0.0008045506257110353,
+      "loss": 1.5182,
+      "step": 863
+    },
+    {
+      "epoch": 0.9829351535836177,
+      "grad_norm": 1.6558321714401245,
+      "learning_rate": 0.0008043230944254836,
+      "loss": 2.2892,
+      "step": 864
+    },
+    {
+      "epoch": 0.9840728100113766,
+      "grad_norm": 1.1968799829483032,
+      "learning_rate": 0.0008040955631399317,
+      "loss": 2.6779,
+      "step": 865
+    },
+    {
+      "epoch": 0.9852104664391353,
+      "grad_norm": 0.6827085614204407,
+      "learning_rate": 0.0008038680318543799,
+      "loss": 1.8453,
+      "step": 866
+    },
+    {
+      "epoch": 0.9863481228668942,
+      "grad_norm": 0.70023113489151,
+      "learning_rate": 0.0008036405005688282,
+      "loss": 1.5166,
+      "step": 867
+    },
+    {
+      "epoch": 0.987485779294653,
+      "grad_norm": 1.1971696615219116,
+      "learning_rate": 0.0008034129692832765,
+      "loss": 3.0315,
+      "step": 868
+    },
+    {
+      "epoch": 0.9886234357224118,
+      "grad_norm": 1.1876801252365112,
+      "learning_rate": 0.0008031854379977247,
+      "loss": 2.263,
+      "step": 869
+    },
+    {
+      "epoch": 0.9897610921501706,
+      "grad_norm": 1.1676205396652222,
+      "learning_rate": 0.000802957906712173,
+      "loss": 1.9647,
+      "step": 870
+    },
+    {
+      "epoch": 0.9908987485779295,
+      "grad_norm": 1.2230088710784912,
+      "learning_rate": 0.0008027303754266212,
+      "loss": 3.6692,
+      "step": 871
+    },
+    {
+      "epoch": 0.9920364050056882,
+      "grad_norm": 0.9202735424041748,
+      "learning_rate": 0.0008025028441410694,
+      "loss": 1.9178,
+      "step": 872
+    },
+    {
+      "epoch": 0.9931740614334471,
+      "grad_norm": 0.685429036617279,
+      "learning_rate": 0.0008022753128555177,
+      "loss": 1.6819,
+      "step": 873
+    },
+    {
+      "epoch": 0.9943117178612059,
+      "grad_norm": 0.9149478077888489,
+      "learning_rate": 0.0008020477815699659,
+      "loss": 1.7829,
+      "step": 874
+    },
+    {
+      "epoch": 0.9954493742889647,
+      "grad_norm": 0.9612428545951843,
+      "learning_rate": 0.000801820250284414,
+      "loss": 2.3288,
+      "step": 875
+    },
+    {
+      "epoch": 0.9965870307167235,
+      "grad_norm": 1.1229710578918457,
+      "learning_rate": 0.0008015927189988624,
+      "loss": 1.8854,
+      "step": 876
+    },
+    {
+      "epoch": 0.9977246871444824,
+      "grad_norm": 1.3799995183944702,
+      "learning_rate": 0.0008013651877133106,
+      "loss": 3.6562,
+      "step": 877
+    },
+    {
+      "epoch": 0.9988623435722411,
+      "grad_norm": 0.7298029661178589,
+      "learning_rate": 0.0008011376564277588,
+      "loss": 1.7173,
+      "step": 878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1531485319137573,
+      "learning_rate": 0.0008009101251422071,
+      "loss": 2.2511,
+      "step": 879
+    },
+    {
+      "epoch": 1.0,
+      "eval_f1": 0.8883,
+      "eval_gen_len": 49.5182,
+      "eval_loss": 1.9363936185836792,
+      "eval_precision": 0.8875,
+      "eval_recall": 0.8892,
+      "eval_rouge1": 0.4398,
+      "eval_rouge2": 0.1855,
+      "eval_rougeL": 0.3668,
+      "eval_rougeLsum": 0.411,
+      "eval_runtime": 28.6515,
+      "eval_samples_per_second": 3.839,
+      "eval_steps_per_second": 0.489,
+      "step": 879
+    },
+    {
+      "epoch": 1.0011376564277588,
+      "grad_norm": 0.6251806616783142,
+      "learning_rate": 0.0008006825938566553,
+      "loss": 1.0413,
+      "step": 880
+    },
+    {
+      "epoch": 1.0022753128555177,
+      "grad_norm": 0.9199168086051941,
+      "learning_rate": 0.0008004550625711036,
+      "loss": 1.691,
+      "step": 881
+    },
+    {
+      "epoch": 1.0034129692832765,
+      "grad_norm": 1.4455974102020264,
+      "learning_rate": 0.0008002275312855518,
+      "loss": 4.0842,
+      "step": 882
+    },
+    {
+      "epoch": 1.0045506257110353,
+      "grad_norm": 1.1115312576293945,
+      "learning_rate": 0.0008,
+      "loss": 2.4961,
+      "step": 883
+    },
+    {
+      "epoch": 1.005688282138794,
+      "grad_norm": 0.952612042427063,
+      "learning_rate": 0.0007997724687144482,
+      "loss": 1.8239,
+      "step": 884
+    },
+    {
+      "epoch": 1.006825938566553,
+      "grad_norm": 0.9800905585289001,
+      "learning_rate": 0.0007995449374288965,
+      "loss": 1.981,
+      "step": 885
+    },
+    {
+      "epoch": 1.0079635949943118,
+      "grad_norm": 1.0975840091705322,
+      "learning_rate": 0.0007993174061433447,
+      "loss": 2.4143,
+      "step": 886
+    },
+    {
+      "epoch": 1.0091012514220705,
+      "grad_norm": 0.936336874961853,
+      "learning_rate": 0.000799089874857793,
+      "loss": 1.6641,
+      "step": 887
+    },
+    {
+      "epoch": 1.0102389078498293,
+      "grad_norm": 0.5330931544303894,
+      "learning_rate": 0.0007988623435722412,
+      "loss": 1.2285,
+      "step": 888
+    },
+    {
+      "epoch": 1.0113765642775883,
+      "grad_norm": 1.2912349700927734,
+      "learning_rate": 0.0007986348122866894,
+      "loss": 1.594,
+      "step": 889
+    },
+    {
+      "epoch": 1.012514220705347,
+      "grad_norm": 0.8078871369361877,
+      "learning_rate": 0.0007984072810011377,
+      "loss": 1.0581,
+      "step": 890
+    },
+    {
+      "epoch": 1.0136518771331058,
+      "grad_norm": 1.071419358253479,
+      "learning_rate": 0.0007981797497155859,
+      "loss": 1.7577,
+      "step": 891
+    },
+    {
+      "epoch": 1.0147895335608645,
+      "grad_norm": 1.0178149938583374,
+      "learning_rate": 0.0007979522184300341,
+      "loss": 2.4778,
+      "step": 892
+    },
+    {
+      "epoch": 1.0159271899886235,
+      "grad_norm": 0.9748836159706116,
+      "learning_rate": 0.0007977246871444825,
+      "loss": 1.8737,
+      "step": 893
+    },
+    {
+      "epoch": 1.0170648464163823,
+      "grad_norm": 0.7067264914512634,
+      "learning_rate": 0.0007974971558589306,
+      "loss": 1.3954,
+      "step": 894
+    },
+    {
+      "epoch": 1.018202502844141,
+      "grad_norm": 1.2109535932540894,
+      "learning_rate": 0.0007972696245733788,
+      "loss": 2.3141,
+      "step": 895
+    },
+    {
+      "epoch": 1.0193401592718998,
+      "grad_norm": 0.8363299369812012,
+      "learning_rate": 0.0007970420932878271,
+      "loss": 2.0917,
+      "step": 896
+    },
+    {
+      "epoch": 1.0204778156996588,
+      "grad_norm": 0.7036440968513489,
+      "learning_rate": 0.0007968145620022753,
+      "loss": 1.5766,
+      "step": 897
+    },
+    {
+      "epoch": 1.0216154721274175,
+      "grad_norm": 0.6679667830467224,
+      "learning_rate": 0.0007965870307167235,
+      "loss": 1.8018,
+      "step": 898
+    },
+    {
+      "epoch": 1.0227531285551763,
+      "grad_norm": 1.0804336071014404,
+      "learning_rate": 0.0007963594994311718,
+      "loss": 2.0411,
+      "step": 899
+    },
+    {
+      "epoch": 1.023890784982935,
+      "grad_norm": 0.9264572262763977,
+      "learning_rate": 0.00079613196814562,
+      "loss": 2.1497,
+      "step": 900
+    },
+    {
+      "epoch": 1.025028441410694,
+      "grad_norm": 1.2212860584259033,
+      "learning_rate": 0.0007959044368600682,
+      "loss": 2.7216,
+      "step": 901
+    },
+    {
+      "epoch": 1.0261660978384528,
+      "grad_norm": 1.1044024229049683,
+      "learning_rate": 0.0007956769055745166,
+      "loss": 2.3266,
+      "step": 902
+    },
+    {
+      "epoch": 1.0273037542662116,
+      "grad_norm": 0.7855750918388367,
+      "learning_rate": 0.0007954493742889648,
+      "loss": 1.4662,
+      "step": 903
+    },
+    {
+      "epoch": 1.0284414106939703,
+      "grad_norm": 1.0114825963974,
+      "learning_rate": 0.000795221843003413,
+      "loss": 2.0779,
+      "step": 904
+    },
+    {
+      "epoch": 1.0295790671217293,
+      "grad_norm": 0.9704268574714661,
+      "learning_rate": 0.0007949943117178612,
+      "loss": 1.6038,
+      "step": 905
+    },
+    {
+      "epoch": 1.030716723549488,
+      "grad_norm": 1.155097246170044,
+      "learning_rate": 0.0007947667804323094,
+      "loss": 2.13,
+      "step": 906
+    },
+    {
+      "epoch": 1.0318543799772468,
+      "grad_norm": 1.3842484951019287,
+      "learning_rate": 0.0007945392491467577,
+      "loss": 3.5606,
+      "step": 907
+    },
+    {
+      "epoch": 1.0329920364050056,
+      "grad_norm": 1.134261131286621,
+      "learning_rate": 0.0007943117178612059,
+      "loss": 2.9482,
+      "step": 908
+    },
+    {
+      "epoch": 1.0341296928327646,
+      "grad_norm": 1.0498160123825073,
+      "learning_rate": 0.0007940841865756541,
+      "loss": 2.3777,
+      "step": 909
+    },
+    {
+      "epoch": 1.0352673492605233,
+      "grad_norm": 0.7719895243644714,
+      "learning_rate": 0.0007938566552901025,
+      "loss": 2.2765,
+      "step": 910
+    },
+    {
+      "epoch": 1.036405005688282,
+      "grad_norm": 1.402180790901184,
+      "learning_rate": 0.0007936291240045507,
+      "loss": 3.0459,
+      "step": 911
+    },
+    {
+      "epoch": 1.0375426621160408,
+      "grad_norm": 0.8468472361564636,
+      "learning_rate": 0.0007934015927189989,
+      "loss": 0.922,
+      "step": 912
+    },
+    {
+      "epoch": 1.0386803185437998,
+      "grad_norm": 0.8631690740585327,
+      "learning_rate": 0.0007931740614334472,
+      "loss": 1.2322,
+      "step": 913
+    },
+    {
+      "epoch": 1.0398179749715586,
+      "grad_norm": 1.3747214078903198,
+      "learning_rate": 0.0007929465301478953,
+      "loss": 2.485,
+      "step": 914
+    },
+    {
+      "epoch": 1.0409556313993173,
+      "grad_norm": 0.7905521988868713,
+      "learning_rate": 0.0007927189988623435,
+      "loss": 2.1488,
+      "step": 915
+    },
+    {
+      "epoch": 1.0420932878270763,
+      "grad_norm": 1.0372867584228516,
+      "learning_rate": 0.0007924914675767918,
+      "loss": 2.3263,
+      "step": 916
+    },
+    {
+      "epoch": 1.043230944254835,
+      "grad_norm": 0.964745819568634,
+      "learning_rate": 0.00079226393629124,
+      "loss": 1.4245,
+      "step": 917
+    },
+    {
+      "epoch": 1.0443686006825939,
+      "grad_norm": 1.2599427700042725,
+      "learning_rate": 0.0007920364050056883,
+      "loss": 2.8828,
+      "step": 918
+    },
+    {
+      "epoch": 1.0455062571103526,
+      "grad_norm": 0.5311685800552368,
+      "learning_rate": 0.0007918088737201366,
+      "loss": 1.2969,
+      "step": 919
+    },
+    {
+      "epoch": 1.0466439135381116,
+      "grad_norm": 1.0544791221618652,
+      "learning_rate": 0.0007915813424345848,
+      "loss": 2.3014,
+      "step": 920
+    },
+    {
+      "epoch": 1.0477815699658704,
+      "grad_norm": 0.7876049280166626,
+      "learning_rate": 0.000791353811149033,
+      "loss": 2.2945,
+      "step": 921
+    },
+    {
+      "epoch": 1.0489192263936291,
+      "grad_norm": 0.7309691309928894,
+      "learning_rate": 0.0007911262798634813,
+      "loss": 1.2384,
+      "step": 922
+    },
+    {
+      "epoch": 1.0500568828213879,
+      "grad_norm": 0.8504576086997986,
+      "learning_rate": 0.0007908987485779294,
+      "loss": 1.719,
+      "step": 923
+    },
+    {
+      "epoch": 1.0511945392491469,
+      "grad_norm": 1.7143735885620117,
+      "learning_rate": 0.0007906712172923777,
+      "loss": 3.9275,
+      "step": 924
+    },
+    {
+      "epoch": 1.0523321956769056,
+      "grad_norm": 1.8625571727752686,
+      "learning_rate": 0.0007904436860068259,
+      "loss": 4.4724,
+      "step": 925
+    },
+    {
+      "epoch": 1.0534698521046644,
+      "grad_norm": 1.0929653644561768,
+      "learning_rate": 0.0007902161547212741,
+      "loss": 1.9623,
+      "step": 926
+    },
+    {
+      "epoch": 1.0546075085324231,
+      "grad_norm": 1.0899792909622192,
+      "learning_rate": 0.0007899886234357225,
+      "loss": 1.6986,
+      "step": 927
+    },
+    {
+      "epoch": 1.0557451649601821,
+      "grad_norm": 1.1779440641403198,
+      "learning_rate": 0.0007897610921501707,
+      "loss": 1.905,
+      "step": 928
+    },
+    {
+      "epoch": 1.0568828213879409,
+      "grad_norm": 0.6737589240074158,
+      "learning_rate": 0.0007895335608646189,
+      "loss": 1.734,
+      "step": 929
+    },
+    {
+      "epoch": 1.0580204778156996,
+      "grad_norm": 1.0531206130981445,
+      "learning_rate": 0.0007893060295790672,
+      "loss": 1.4693,
+      "step": 930
+    },
+    {
+      "epoch": 1.0591581342434584,
+      "grad_norm": 0.5907047390937805,
+      "learning_rate": 0.0007890784982935154,
+      "loss": 1.4901,
+      "step": 931
+    },
+    {
+      "epoch": 1.0602957906712174,
+      "grad_norm": 0.8621008396148682,
+      "learning_rate": 0.0007888509670079636,
+      "loss": 1.5857,
+      "step": 932
+    },
+    {
+      "epoch": 1.0614334470989761,
+      "grad_norm": 0.9032453298568726,
+      "learning_rate": 0.0007886234357224118,
+      "loss": 1.8026,
+      "step": 933
+    },
+    {
+      "epoch": 1.062571103526735,
+      "grad_norm": 1.4888174533843994,
+      "learning_rate": 0.00078839590443686,
+      "loss": 1.8748,
+      "step": 934
+    },
+    {
+      "epoch": 1.0637087599544937,
+      "grad_norm": 0.6122904419898987,
+      "learning_rate": 0.0007881683731513083,
+      "loss": 0.7945,
+      "step": 935
+    },
+    {
+      "epoch": 1.0648464163822526,
+      "grad_norm": 1.013519048690796,
+      "learning_rate": 0.0007879408418657566,
+      "loss": 1.9398,
+      "step": 936
+    },
+    {
+      "epoch": 1.0659840728100114,
+      "grad_norm": 0.7138906717300415,
+      "learning_rate": 0.0007877133105802048,
+      "loss": 1.6637,
+      "step": 937
+    },
+    {
+      "epoch": 1.0671217292377702,
+      "grad_norm": 1.3058874607086182,
+      "learning_rate": 0.000787485779294653,
+      "loss": 3.3544,
+      "step": 938
+    },
+    {
+      "epoch": 1.068259385665529,
+      "grad_norm": 0.9075451493263245,
+      "learning_rate": 0.0007872582480091013,
+      "loss": 1.7222,
+      "step": 939
+    },
+    {
+      "epoch": 1.069397042093288,
+      "grad_norm": 0.9286370873451233,
+      "learning_rate": 0.0007870307167235495,
+      "loss": 2.0651,
+      "step": 940
+    },
+    {
+      "epoch": 1.0705346985210467,
+      "grad_norm": 0.6360165476799011,
+      "learning_rate": 0.0007868031854379977,
+      "loss": 0.89,
+      "step": 941
+    },
+    {
+      "epoch": 1.0716723549488054,
+      "grad_norm": 1.2726821899414062,
+      "learning_rate": 0.000786575654152446,
+      "loss": 2.7395,
+      "step": 942
+    },
+    {
+      "epoch": 1.0728100113765642,
+      "grad_norm": 1.024568796157837,
+      "learning_rate": 0.0007863481228668941,
+      "loss": 2.1989,
+      "step": 943
+    },
+    {
+      "epoch": 1.0739476678043232,
+      "grad_norm": 1.1119788885116577,
+      "learning_rate": 0.0007861205915813425,
+      "loss": 3.14,
+      "step": 944
+    },
+    {
+      "epoch": 1.075085324232082,
+      "grad_norm": 1.0225321054458618,
+      "learning_rate": 0.0007858930602957907,
+      "loss": 2.8281,
+      "step": 945
+    },
+    {
+      "epoch": 1.0762229806598407,
+      "grad_norm": 0.8482196927070618,
+      "learning_rate": 0.0007856655290102389,
+      "loss": 1.4616,
+      "step": 946
+    },
+    {
+      "epoch": 1.0773606370875997,
+      "grad_norm": 0.9475064277648926,
+      "learning_rate": 0.0007854379977246872,
+      "loss": 2.9296,
+      "step": 947
+    },
+    {
+      "epoch": 1.0784982935153584,
+      "grad_norm": 0.5353087186813354,
+      "learning_rate": 0.0007852104664391354,
+      "loss": 0.9475,
+      "step": 948
+    },
+    {
+      "epoch": 1.0796359499431172,
+      "grad_norm": 1.2120368480682373,
+      "learning_rate": 0.0007849829351535836,
+      "loss": 3.095,
+      "step": 949
+    },
+    {
+      "epoch": 1.080773606370876,
+      "grad_norm": 1.1500520706176758,
+      "learning_rate": 0.0007847554038680319,
+      "loss": 2.4183,
+      "step": 950
+    },
+    {
+      "epoch": 1.0819112627986347,
+      "grad_norm": 0.629219651222229,
+      "learning_rate": 0.0007845278725824802,
+      "loss": 1.1091,
+      "step": 951
+    },
+    {
+      "epoch": 1.0830489192263937,
+      "grad_norm": 0.6308894753456116,
+      "learning_rate": 0.0007843003412969284,
+      "loss": 1.0829,
+      "step": 952
+    },
+    {
+      "epoch": 1.0841865756541524,
+      "grad_norm": 1.4992165565490723,
+      "learning_rate": 0.0007840728100113766,
+      "loss": 3.7885,
+      "step": 953
+    },
+    {
+      "epoch": 1.0853242320819112,
+      "grad_norm": 1.0505317449569702,
+      "learning_rate": 0.0007838452787258248,
+      "loss": 2.2605,
+      "step": 954
+    },
+    {
+      "epoch": 1.0864618885096702,
+      "grad_norm": 0.898767352104187,
+      "learning_rate": 0.000783617747440273,
+      "loss": 3.0967,
+      "step": 955
+    },
+    {
+      "epoch": 1.087599544937429,
+      "grad_norm": 1.2503669261932373,
+      "learning_rate": 0.0007833902161547213,
+      "loss": 2.2391,
+      "step": 956
+    },
+    {
+      "epoch": 1.0887372013651877,
+      "grad_norm": 1.2384415864944458,
+      "learning_rate": 0.0007831626848691695,
+      "loss": 3.554,
+      "step": 957
+    },
+    {
+      "epoch": 1.0898748577929465,
+      "grad_norm": 1.1613341569900513,
+      "learning_rate": 0.0007829351535836177,
+      "loss": 1.7405,
+      "step": 958
+    },
+    {
+      "epoch": 1.0910125142207054,
+      "grad_norm": 0.6541109681129456,
+      "learning_rate": 0.000782707622298066,
+      "loss": 1.5068,
+      "step": 959
+    },
+    {
+      "epoch": 1.0921501706484642,
+      "grad_norm": 1.1372532844543457,
+      "learning_rate": 0.0007824800910125143,
+      "loss": 2.6415,
+      "step": 960
+    },
+    {
+      "epoch": 1.093287827076223,
+      "grad_norm": 0.944439709186554,
+      "learning_rate": 0.0007822525597269625,
+      "loss": 2.2019,
+      "step": 961
+    },
+    {
+      "epoch": 1.0944254835039817,
+      "grad_norm": 1.1232166290283203,
+      "learning_rate": 0.0007820250284414107,
+      "loss": 2.172,
+      "step": 962
+    },
+    {
+      "epoch": 1.0955631399317407,
+      "grad_norm": 1.3538726568222046,
+      "learning_rate": 0.0007817974971558589,
+      "loss": 2.0305,
+      "step": 963
+    },
+    {
+      "epoch": 1.0967007963594995,
+      "grad_norm": 0.7923217415809631,
+      "learning_rate": 0.0007815699658703072,
+      "loss": 1.2107,
+      "step": 964
+    },
+    {
+      "epoch": 1.0978384527872582,
+      "grad_norm": 0.9362553358078003,
+      "learning_rate": 0.0007813424345847554,
+      "loss": 1.9665,
+      "step": 965
+    },
+    {
+      "epoch": 1.098976109215017,
+      "grad_norm": 1.2658346891403198,
+      "learning_rate": 0.0007811149032992036,
+      "loss": 2.9011,
+      "step": 966
+    },
+    {
+      "epoch": 1.100113765642776,
+      "grad_norm": 1.526066780090332,
+      "learning_rate": 0.0007808873720136519,
+      "loss": 1.7315,
+      "step": 967
+    },
+    {
+      "epoch": 1.1012514220705347,
+      "grad_norm": 1.3018304109573364,
+      "learning_rate": 0.0007806598407281002,
+      "loss": 3.2345,
+      "step": 968
+    },
+    {
+      "epoch": 1.1023890784982935,
+      "grad_norm": 0.8176997900009155,
+      "learning_rate": 0.0007804323094425484,
+      "loss": 1.467,
+      "step": 969
+    },
+    {
+      "epoch": 1.1035267349260522,
+      "grad_norm": 0.8336053490638733,
+      "learning_rate": 0.0007802047781569967,
+      "loss": 1.4161,
+      "step": 970
+    },
+    {
+      "epoch": 1.1046643913538112,
+      "grad_norm": 0.7302894592285156,
+      "learning_rate": 0.0007799772468714449,
+      "loss": 1.3151,
+      "step": 971
+    },
+    {
+      "epoch": 1.10580204778157,
+      "grad_norm": 0.9276258945465088,
+      "learning_rate": 0.000779749715585893,
+      "loss": 2.3525,
+      "step": 972
+    },
+    {
+      "epoch": 1.1069397042093287,
+      "grad_norm": 0.8648287057876587,
+      "learning_rate": 0.0007795221843003413,
+      "loss": 1.7804,
+      "step": 973
+    },
+    {
+      "epoch": 1.1080773606370875,
+      "grad_norm": 1.4959267377853394,
+      "learning_rate": 0.0007792946530147895,
+      "loss": 2.9281,
+      "step": 974
+    },
+    {
+      "epoch": 1.1092150170648465,
+      "grad_norm": 0.7826813459396362,
+      "learning_rate": 0.0007790671217292377,
+      "loss": 0.9164,
+      "step": 975
+    },
+    {
+      "epoch": 1.1103526734926052,
+      "grad_norm": 1.157572865486145,
+      "learning_rate": 0.000778839590443686,
+      "loss": 1.9721,
+      "step": 976
+    },
+    {
+      "epoch": 1.111490329920364,
+      "grad_norm": 1.137681484222412,
+      "learning_rate": 0.0007786120591581343,
+      "loss": 1.5874,
+      "step": 977
+    },
+    {
+      "epoch": 1.1126279863481228,
+      "grad_norm": 0.9332581162452698,
+      "learning_rate": 0.0007783845278725825,
+      "loss": 2.5314,
+      "step": 978
+    },
+    {
+      "epoch": 1.1137656427758817,
+      "grad_norm": 1.1413997411727905,
+      "learning_rate": 0.0007781569965870308,
+      "loss": 3.0732,
+      "step": 979
+    },
+    {
+      "epoch": 1.1149032992036405,
+      "grad_norm": 0.7028293609619141,
+      "learning_rate": 0.000777929465301479,
+      "loss": 0.8924,
+      "step": 980
+    },
+    {
+      "epoch": 1.1160409556313993,
+      "grad_norm": 1.0388150215148926,
+      "learning_rate": 0.0007777019340159272,
+      "loss": 1.9138,
+      "step": 981
+    },
+    {
+      "epoch": 1.117178612059158,
+      "grad_norm": 0.9347336292266846,
+      "learning_rate": 0.0007774744027303754,
+      "loss": 1.5226,
+      "step": 982
+    },
+    {
+      "epoch": 1.118316268486917,
+      "grad_norm": 0.9350568056106567,
+      "learning_rate": 0.0007772468714448236,
+      "loss": 2.6978,
+      "step": 983
+    },
+    {
+      "epoch": 1.1194539249146758,
+      "grad_norm": 1.407080888748169,
+      "learning_rate": 0.0007770193401592718,
+      "loss": 1.3464,
+      "step": 984
+    },
+    {
+      "epoch": 1.1205915813424345,
+      "grad_norm": 1.077770471572876,
+      "learning_rate": 0.0007767918088737202,
+      "loss": 2.123,
+      "step": 985
+    },
+    {
+      "epoch": 1.1217292377701935,
+      "grad_norm": 1.041999101638794,
+      "learning_rate": 0.0007765642775881684,
+      "loss": 2.5526,
+      "step": 986
+    },
+    {
+      "epoch": 1.1228668941979523,
+      "grad_norm": 1.0291063785552979,
+      "learning_rate": 0.0007763367463026167,
+      "loss": 2.1948,
+      "step": 987
+    },
+    {
+      "epoch": 1.124004550625711,
+      "grad_norm": 0.9935250282287598,
+      "learning_rate": 0.0007761092150170649,
+      "loss": 3.3882,
+      "step": 988
+    },
+    {
+      "epoch": 1.1251422070534698,
+      "grad_norm": 1.035112977027893,
+      "learning_rate": 0.0007758816837315131,
+      "loss": 2.0542,
+      "step": 989
+    },
+    {
+      "epoch": 1.1262798634812285,
+      "grad_norm": 1.0226819515228271,
+      "learning_rate": 0.0007756541524459614,
+      "loss": 1.8742,
+      "step": 990
+    },
+    {
+      "epoch": 1.1274175199089875,
+      "grad_norm": 1.0159075260162354,
+      "learning_rate": 0.0007754266211604095,
+      "loss": 1.3203,
+      "step": 991
+    },
+    {
+      "epoch": 1.1285551763367463,
+      "grad_norm": 2.0587501525878906,
+      "learning_rate": 0.0007751990898748577,
+      "loss": 3.9035,
+      "step": 992
+    },
+    {
+      "epoch": 1.129692832764505,
+      "grad_norm": 0.9951366186141968,
+      "learning_rate": 0.000774971558589306,
+      "loss": 2.5265,
+      "step": 993
+    },
+    {
+      "epoch": 1.130830489192264,
+      "grad_norm": 0.7112089395523071,
+      "learning_rate": 0.0007747440273037543,
+      "loss": 1.3928,
+      "step": 994
+    },
+    {
+      "epoch": 1.1319681456200228,
+      "grad_norm": 1.0504612922668457,
+      "learning_rate": 0.0007745164960182025,
+      "loss": 2.0178,
+      "step": 995
+    },
+    {
+      "epoch": 1.1331058020477816,
+      "grad_norm": 0.7228176593780518,
+      "learning_rate": 0.0007742889647326508,
+      "loss": 1.4623,
+      "step": 996
+    },
+    {
+      "epoch": 1.1342434584755403,
+      "grad_norm": 1.1857068538665771,
+      "learning_rate": 0.000774061433447099,
+      "loss": 2.5628,
+      "step": 997
+    },
+    {
+      "epoch": 1.1353811149032993,
+      "grad_norm": 0.6408101916313171,
+      "learning_rate": 0.0007738339021615472,
+      "loss": 1.3296,
+      "step": 998
+    },
+    {
+      "epoch": 1.136518771331058,
+      "grad_norm": 0.9753408432006836,
+      "learning_rate": 0.0007736063708759955,
+      "loss": 2.0656,
+      "step": 999
+    },
+    {
+      "epoch": 1.1376564277588168,
+      "grad_norm": 1.139440894126892,
+      "learning_rate": 0.0007733788395904437,
+      "loss": 2.2056,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1387940841865756,
+      "grad_norm": 0.6815441846847534,
+      "learning_rate": 0.0007731513083048918,
+      "loss": 1.2574,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1399317406143346,
+      "grad_norm": 1.032520055770874,
+      "learning_rate": 0.0007729237770193402,
+      "loss": 2.0944,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1410693970420933,
+      "grad_norm": 0.8344050049781799,
+      "learning_rate": 0.0007726962457337884,
+      "loss": 1.5627,
+      "step": 1003
+    },
+    {
+      "epoch": 1.142207053469852,
+      "grad_norm": 1.2502180337905884,
+      "learning_rate": 0.0007724687144482366,
+      "loss": 2.0755,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1433447098976108,
+      "grad_norm": 1.008817195892334,
+      "learning_rate": 0.0007722411831626849,
+      "loss": 2.4644,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1444823663253698,
+      "grad_norm": 1.0237065553665161,
+      "learning_rate": 0.0007720136518771331,
+      "loss": 1.4251,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1456200227531286,
+      "grad_norm": 0.9443050026893616,
+      "learning_rate": 0.0007717861205915814,
+      "loss": 2.5625,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1467576791808873,
+      "grad_norm": 0.8654250502586365,
+      "learning_rate": 0.0007715585893060296,
+      "loss": 1.999,
+      "step": 1008
+    },
+    {
+      "epoch": 1.147895335608646,
+      "grad_norm": 1.082979679107666,
+      "learning_rate": 0.0007713310580204778,
+      "loss": 1.2456,
+      "step": 1009
+    },
+    {
+      "epoch": 1.149032992036405,
+      "grad_norm": 1.07938814163208,
+      "learning_rate": 0.0007711035267349262,
+      "loss": 2.0854,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1501706484641638,
+      "grad_norm": 0.6003649234771729,
+      "learning_rate": 0.0007708759954493743,
+      "loss": 1.4027,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1513083048919226,
+      "grad_norm": 1.157963514328003,
+      "learning_rate": 0.0007706484641638225,
+      "loss": 3.3311,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1524459613196814,
+      "grad_norm": 0.5869401097297668,
+      "learning_rate": 0.0007704209328782708,
+      "loss": 1.4386,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1535836177474403,
+      "grad_norm": 0.9778612852096558,
+      "learning_rate": 0.000770193401592719,
+      "loss": 2.9109,
+      "step": 1014
+    },
+    {
+      "epoch": 1.154721274175199,
+      "grad_norm": 0.9791373610496521,
+      "learning_rate": 0.0007699658703071672,
+      "loss": 1.5482,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1558589306029579,
+      "grad_norm": 1.0005147457122803,
+      "learning_rate": 0.0007697383390216155,
+      "loss": 2.9281,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1569965870307168,
+      "grad_norm": 0.6844426989555359,
+      "learning_rate": 0.0007695108077360637,
+      "loss": 0.9119,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1581342434584756,
+      "grad_norm": 0.9615586400032043,
+      "learning_rate": 0.000769283276450512,
+      "loss": 2.3095,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1592718998862344,
+      "grad_norm": 1.1434069871902466,
+      "learning_rate": 0.0007690557451649603,
+      "loss": 2.035,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1604095563139931,
+      "grad_norm": 1.1195374727249146,
+      "learning_rate": 0.0007688282138794085,
+      "loss": 2.8448,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1615472127417519,
+      "grad_norm": 0.9364283084869385,
+      "learning_rate": 0.0007686006825938566,
+      "loss": 2.0258,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1626848691695109,
+      "grad_norm": 0.37525808811187744,
+      "learning_rate": 0.0007683731513083049,
+      "loss": 0.439,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1638225255972696,
+      "grad_norm": 1.8962267637252808,
+      "learning_rate": 0.0007681456200227531,
+      "loss": 2.1773,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1649601820250284,
+      "grad_norm": 1.661497712135315,
+      "learning_rate": 0.0007679180887372013,
+      "loss": 2.577,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1660978384527874,
+      "grad_norm": 1.1258926391601562,
+      "learning_rate": 0.0007676905574516496,
+      "loss": 2.3722,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1672354948805461,
+      "grad_norm": 0.9760286808013916,
+      "learning_rate": 0.0007674630261660978,
+      "loss": 2.135,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1683731513083049,
+      "grad_norm": 0.89606112241745,
+      "learning_rate": 0.0007672354948805462,
+      "loss": 1.8733,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1695108077360636,
+      "grad_norm": 1.0964630842208862,
+      "learning_rate": 0.0007670079635949944,
+      "loss": 1.9628,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1706484641638226,
+      "grad_norm": 1.214050054550171,
+      "learning_rate": 0.0007667804323094426,
+      "loss": 2.3792,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1717861205915814,
+      "grad_norm": 1.242618203163147,
+      "learning_rate": 0.0007665529010238908,
+      "loss": 3.0571,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1729237770193401,
+      "grad_norm": 0.7527234554290771,
+      "learning_rate": 0.000766325369738339,
+      "loss": 1.7859,
+      "step": 1031
+    },
+    {
+      "epoch": 1.174061433447099,
+      "grad_norm": 0.9902629256248474,
+      "learning_rate": 0.0007660978384527872,
+      "loss": 2.1558,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1751990898748579,
+      "grad_norm": 1.109411597251892,
+      "learning_rate": 0.0007658703071672355,
+      "loss": 2.4196,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1763367463026166,
+      "grad_norm": 0.9211401343345642,
+      "learning_rate": 0.0007656427758816837,
+      "loss": 1.8503,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1774744027303754,
+      "grad_norm": 0.9370328187942505,
+      "learning_rate": 0.000765415244596132,
+      "loss": 2.7057,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1786120591581342,
+      "grad_norm": 0.8330713510513306,
+      "learning_rate": 0.0007651877133105803,
+      "loss": 1.6611,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1797497155858931,
+      "grad_norm": 0.7571448087692261,
+      "learning_rate": 0.0007649601820250285,
+      "loss": 1.2492,
+      "step": 1037
+    },
+    {
+      "epoch": 1.180887372013652,
+      "grad_norm": 1.6156978607177734,
+      "learning_rate": 0.0007647326507394767,
+      "loss": 2.814,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1820250284414107,
+      "grad_norm": 1.8085129261016846,
+      "learning_rate": 0.000764505119453925,
+      "loss": 3.3276,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1831626848691694,
+      "grad_norm": 1.0824229717254639,
+      "learning_rate": 0.0007642775881683731,
+      "loss": 2.5017,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1843003412969284,
+      "grad_norm": 0.6034184098243713,
+      "learning_rate": 0.0007640500568828213,
+      "loss": 1.304,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1854379977246872,
+      "grad_norm": 0.71455317735672,
+      "learning_rate": 0.0007638225255972696,
+      "loss": 1.8677,
+      "step": 1042
+    },
+    {
+      "epoch": 1.186575654152446,
+      "grad_norm": 0.7810432314872742,
+      "learning_rate": 0.0007635949943117178,
+      "loss": 2.6322,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1877133105802047,
+      "grad_norm": 0.7499434351921082,
+      "learning_rate": 0.0007633674630261661,
+      "loss": 1.8316,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1888509670079637,
+      "grad_norm": 0.9468162059783936,
+      "learning_rate": 0.0007631399317406144,
+      "loss": 1.5508,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1899886234357224,
+      "grad_norm": 1.3421801328659058,
+      "learning_rate": 0.0007629124004550626,
+      "loss": 2.2196,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1911262798634812,
+      "grad_norm": 0.9094765782356262,
+      "learning_rate": 0.0007626848691695109,
+      "loss": 2.1863,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1922639362912402,
+      "grad_norm": 0.902521014213562,
+      "learning_rate": 0.0007624573378839591,
+      "loss": 1.4415,
+      "step": 1048
+    },
+    {
+      "epoch": 1.193401592718999,
+      "grad_norm": 0.8511042594909668,
+      "learning_rate": 0.0007622298065984073,
+      "loss": 1.5148,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1945392491467577,
+      "grad_norm": 0.836769163608551,
+      "learning_rate": 0.0007620022753128555,
+      "loss": 1.819,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1956769055745164,
+      "grad_norm": 0.7466940879821777,
+      "learning_rate": 0.0007617747440273037,
+      "loss": 2.2535,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1968145620022752,
+      "grad_norm": 0.8963870406150818,
+      "learning_rate": 0.000761547212741752,
+      "loss": 1.8274,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1979522184300342,
+      "grad_norm": 1.2822279930114746,
+      "learning_rate": 0.0007613196814562003,
+      "loss": 2.1197,
+      "step": 1053
+    },
+    {
+      "epoch": 1.199089874857793,
+      "grad_norm": 1.0114504098892212,
+      "learning_rate": 0.0007610921501706485,
+      "loss": 3.1561,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2002275312855517,
+      "grad_norm": 0.7050504088401794,
+      "learning_rate": 0.0007608646188850967,
+      "loss": 0.8309,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2013651877133107,
+      "grad_norm": 1.2295100688934326,
+      "learning_rate": 0.000760637087599545,
+      "loss": 2.0593,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2025028441410694,
+      "grad_norm": 1.2942835092544556,
+      "learning_rate": 0.0007604095563139932,
+      "loss": 2.06,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2036405005688282,
+      "grad_norm": 0.652757465839386,
+      "learning_rate": 0.0007601820250284414,
+      "loss": 1.2173,
+      "step": 1058
+    },
+    {
+      "epoch": 1.204778156996587,
+      "grad_norm": 0.7156775593757629,
+      "learning_rate": 0.0007599544937428896,
+      "loss": 1.6357,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2059158134243457,
+      "grad_norm": 0.8230730891227722,
+      "learning_rate": 0.0007597269624573379,
+      "loss": 1.0645,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2070534698521047,
+      "grad_norm": 1.35136079788208,
+      "learning_rate": 0.0007594994311717861,
+      "loss": 3.901,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2081911262798635,
+      "grad_norm": 0.6722981333732605,
+      "learning_rate": 0.0007592718998862344,
+      "loss": 1.5358,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2093287827076222,
+      "grad_norm": 1.035335659980774,
+      "learning_rate": 0.0007590443686006826,
+      "loss": 1.7922,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2104664391353812,
+      "grad_norm": 1.9629613161087036,
+      "learning_rate": 0.0007588168373151308,
+      "loss": 3.3534,
+      "step": 1064
+    },
+    {
+      "epoch": 1.21160409556314,
+      "grad_norm": 0.9386662244796753,
+      "learning_rate": 0.0007585893060295791,
+      "loss": 1.7383,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2127417519908987,
+      "grad_norm": 1.02280592918396,
+      "learning_rate": 0.0007583617747440273,
+      "loss": 2.3867,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2138794084186575,
+      "grad_norm": 0.9958323836326599,
+      "learning_rate": 0.0007581342434584756,
+      "loss": 2.2671,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2150170648464165,
+      "grad_norm": 0.9132734537124634,
+      "learning_rate": 0.0007579067121729239,
+      "loss": 1.5166,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2161547212741752,
+      "grad_norm": 0.9900468587875366,
+      "learning_rate": 0.000757679180887372,
+      "loss": 1.9609,
+      "step": 1069
+    },
+    {
+      "epoch": 1.217292377701934,
+      "grad_norm": 0.7868292331695557,
+      "learning_rate": 0.0007574516496018203,
+      "loss": 1.2568,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2184300341296928,
+      "grad_norm": 0.7199138402938843,
+      "learning_rate": 0.0007572241183162685,
+      "loss": 1.6196,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2195676905574517,
+      "grad_norm": 1.0280647277832031,
+      "learning_rate": 0.0007569965870307167,
+      "loss": 2.3159,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2207053469852105,
+      "grad_norm": 1.435230016708374,
+      "learning_rate": 0.000756769055745165,
+      "loss": 1.9331,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2218430034129693,
+      "grad_norm": 0.982181966304779,
+      "learning_rate": 0.0007565415244596132,
+      "loss": 1.9982,
+      "step": 1074
+    },
+    {
+      "epoch": 1.222980659840728,
+      "grad_norm": 1.21809720993042,
+      "learning_rate": 0.0007563139931740614,
+      "loss": 3.0575,
+      "step": 1075
+    },
+    {
+      "epoch": 1.224118316268487,
+      "grad_norm": 1.0095391273498535,
+      "learning_rate": 0.0007560864618885098,
+      "loss": 2.1901,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2252559726962458,
+      "grad_norm": 0.8512126207351685,
+      "learning_rate": 0.000755858930602958,
+      "loss": 1.5792,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2263936291240045,
+      "grad_norm": 1.3197203874588013,
+      "learning_rate": 0.0007556313993174062,
+      "loss": 2.7239,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2275312855517633,
+      "grad_norm": 1.1738183498382568,
+      "learning_rate": 0.0007554038680318544,
+      "loss": 2.4295,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2286689419795223,
+      "grad_norm": 0.743088960647583,
+      "learning_rate": 0.0007551763367463026,
+      "loss": 1.134,
+      "step": 1080
+    },
+    {
+      "epoch": 1.229806598407281,
+      "grad_norm": 1.0175997018814087,
+      "learning_rate": 0.0007549488054607508,
+      "loss": 3.1313,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2309442548350398,
+      "grad_norm": 1.1242859363555908,
+      "learning_rate": 0.0007547212741751991,
+      "loss": 2.3403,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2320819112627985,
+      "grad_norm": 1.0247671604156494,
+      "learning_rate": 0.0007544937428896473,
+      "loss": 1.6092,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2332195676905575,
+      "grad_norm": 0.7052936553955078,
+      "learning_rate": 0.0007542662116040955,
+      "loss": 1.915,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2343572241183163,
+      "grad_norm": 1.157301425933838,
+      "learning_rate": 0.0007540386803185439,
+      "loss": 1.9053,
+      "step": 1085
+    },
+    {
+      "epoch": 1.235494880546075,
+      "grad_norm": 0.7313311100006104,
+      "learning_rate": 0.0007538111490329921,
+      "loss": 2.1036,
+      "step": 1086
+    },
+    {
+      "epoch": 1.236632536973834,
+      "grad_norm": 0.998532772064209,
+      "learning_rate": 0.0007535836177474404,
+      "loss": 1.6923,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2377701934015928,
+      "grad_norm": 0.8534971475601196,
+      "learning_rate": 0.0007533560864618886,
+      "loss": 2.4127,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2389078498293515,
+      "grad_norm": 0.6717026233673096,
+      "learning_rate": 0.0007531285551763367,
+      "loss": 1.4987,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2400455062571103,
+      "grad_norm": 0.9222111105918884,
+      "learning_rate": 0.000752901023890785,
+      "loss": 1.3934,
+      "step": 1090
+    },
+    {
+      "epoch": 1.241183162684869,
+      "grad_norm": 0.8172008395195007,
+      "learning_rate": 0.0007526734926052332,
+      "loss": 1.4544,
+      "step": 1091
+    },
+    {
+      "epoch": 1.242320819112628,
+      "grad_norm": 0.8321571946144104,
+      "learning_rate": 0.0007524459613196814,
+      "loss": 1.4483,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2434584755403868,
+      "grad_norm": 0.663096010684967,
+      "learning_rate": 0.0007522184300341298,
+      "loss": 1.1409,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2445961319681456,
+      "grad_norm": 0.871487021446228,
+      "learning_rate": 0.000751990898748578,
+      "loss": 1.6311,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2457337883959045,
+      "grad_norm": 0.6653394103050232,
+      "learning_rate": 0.0007517633674630262,
+      "loss": 0.9919,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2468714448236633,
+      "grad_norm": 1.093170404434204,
+      "learning_rate": 0.0007515358361774745,
+      "loss": 1.4075,
+      "step": 1096
+    },
+    {
+      "epoch": 1.248009101251422,
+      "grad_norm": 0.6205571293830872,
+      "learning_rate": 0.0007513083048919227,
+      "loss": 1.814,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2491467576791808,
+      "grad_norm": 1.6945722103118896,
+      "learning_rate": 0.0007510807736063708,
+      "loss": 2.4636,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2502844141069396,
+      "grad_norm": 0.8336583375930786,
+      "learning_rate": 0.0007508532423208191,
+      "loss": 1.9484,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2514220705346986,
+      "grad_norm": 0.9842813014984131,
+      "learning_rate": 0.0007506257110352673,
+      "loss": 2.2475,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2525597269624573,
+      "grad_norm": 0.6157675981521606,
+      "learning_rate": 0.0007503981797497155,
+      "loss": 1.1768,
+      "step": 1101
+    },
+    {
+      "epoch": 1.253697383390216,
+      "grad_norm": 1.6868700981140137,
+      "learning_rate": 0.0007501706484641639,
+      "loss": 1.702,
+      "step": 1102
+    },
+    {
+      "epoch": 1.254835039817975,
+      "grad_norm": 0.7272259593009949,
+      "learning_rate": 0.0007499431171786121,
+      "loss": 1.5946,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2559726962457338,
+      "grad_norm": 1.0236316919326782,
+      "learning_rate": 0.0007497155858930603,
+      "loss": 1.6014,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2571103526734926,
+      "grad_norm": 0.9424517750740051,
+      "learning_rate": 0.0007494880546075086,
+      "loss": 2.589,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2582480091012513,
+      "grad_norm": 0.8940749168395996,
+      "learning_rate": 0.0007492605233219568,
+      "loss": 2.1612,
+      "step": 1106
+    },
+    {
+      "epoch": 1.25938566552901,
+      "grad_norm": 1.2272729873657227,
+      "learning_rate": 0.000749032992036405,
+      "loss": 2.1587,
+      "step": 1107
+    },
+    {
+      "epoch": 1.260523321956769,
+      "grad_norm": 0.6935333609580994,
+      "learning_rate": 0.0007488054607508532,
+      "loss": 1.8994,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2616609783845278,
+      "grad_norm": 0.9777292013168335,
+      "learning_rate": 0.0007485779294653014,
+      "loss": 1.6954,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2627986348122868,
+      "grad_norm": 0.7491214275360107,
+      "learning_rate": 0.0007483503981797498,
+      "loss": 0.9805,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2639362912400456,
+      "grad_norm": 1.3111625909805298,
+      "learning_rate": 0.000748122866894198,
+      "loss": 3.0255,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2650739476678043,
+      "grad_norm": 0.9766296744346619,
+      "learning_rate": 0.0007478953356086462,
+      "loss": 1.5356,
+      "step": 1112
+    },
+    {
+      "epoch": 1.266211604095563,
+      "grad_norm": 0.731914758682251,
+      "learning_rate": 0.0007476678043230945,
+      "loss": 1.7081,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2673492605233219,
+      "grad_norm": 2.02380633354187,
+      "learning_rate": 0.0007474402730375427,
+      "loss": 4.2285,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2684869169510808,
+      "grad_norm": 1.3702422380447388,
+      "learning_rate": 0.0007472127417519909,
+      "loss": 1.6421,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2696245733788396,
+      "grad_norm": 1.1319351196289062,
+      "learning_rate": 0.0007469852104664392,
+      "loss": 2.3609,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2707622298065984,
+      "grad_norm": 0.7127083539962769,
+      "learning_rate": 0.0007467576791808874,
+      "loss": 1.7728,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2718998862343573,
+      "grad_norm": 1.783090353012085,
+      "learning_rate": 0.0007465301478953355,
+      "loss": 3.132,
+      "step": 1118
+    },
+    {
+      "epoch": 1.273037542662116,
+      "grad_norm": 0.9179248809814453,
+      "learning_rate": 0.0007463026166097839,
+      "loss": 2.0857,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2741751990898749,
+      "grad_norm": 0.9802193641662598,
+      "learning_rate": 0.0007460750853242321,
+      "loss": 2.608,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2753128555176336,
+      "grad_norm": 0.855622410774231,
+      "learning_rate": 0.0007458475540386803,
+      "loss": 1.9717,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2764505119453924,
+      "grad_norm": 1.1232682466506958,
+      "learning_rate": 0.0007456200227531286,
+      "loss": 2.113,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2775881683731514,
+      "grad_norm": 0.699447512626648,
+      "learning_rate": 0.0007453924914675768,
+      "loss": 1.2087,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2787258248009101,
+      "grad_norm": 1.3320637941360474,
+      "learning_rate": 0.000745164960182025,
+      "loss": 1.3834,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2798634812286689,
+      "grad_norm": 1.0277353525161743,
+      "learning_rate": 0.0007449374288964733,
+      "loss": 2.2678,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2810011376564279,
+      "grad_norm": 0.8935559391975403,
+      "learning_rate": 0.0007447098976109215,
+      "loss": 1.816,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2821387940841866,
+      "grad_norm": 1.2181432247161865,
+      "learning_rate": 0.0007444823663253698,
+      "loss": 2.2487,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2832764505119454,
+      "grad_norm": 1.007584810256958,
+      "learning_rate": 0.000744254835039818,
+      "loss": 2.1345,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2844141069397041,
+      "grad_norm": 1.0149571895599365,
+      "learning_rate": 0.0007440273037542662,
+      "loss": 2.4568,
+      "step": 1129
+    },
+    {
+      "epoch": 1.285551763367463,
+      "grad_norm": 0.6311346888542175,
+      "learning_rate": 0.0007437997724687145,
+      "loss": 1.5611,
+      "step": 1130
+    },
+    {
+      "epoch": 1.286689419795222,
+      "grad_norm": 1.3615020513534546,
+      "learning_rate": 0.0007435722411831627,
+      "loss": 2.5493,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2878270762229806,
+      "grad_norm": 0.7496081590652466,
+      "learning_rate": 0.0007433447098976109,
+      "loss": 1.5885,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2889647326507394,
+      "grad_norm": 0.9829562306404114,
+      "learning_rate": 0.0007431171786120592,
+      "loss": 1.8779,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2901023890784984,
+      "grad_norm": 1.2002413272857666,
+      "learning_rate": 0.0007428896473265074,
+      "loss": 1.9528,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2912400455062572,
+      "grad_norm": 0.7884618639945984,
+      "learning_rate": 0.0007426621160409557,
+      "loss": 0.8373,
+      "step": 1135
+    },
+    {
+      "epoch": 1.292377701934016,
+      "grad_norm": 0.734190821647644,
+      "learning_rate": 0.000742434584755404,
+      "loss": 1.6223,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2935153583617747,
+      "grad_norm": 1.1885126829147339,
+      "learning_rate": 0.0007422070534698521,
+      "loss": 2.1086,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2946530147895334,
+      "grad_norm": 1.3277819156646729,
+      "learning_rate": 0.0007419795221843003,
+      "loss": 2.6113,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2957906712172924,
+      "grad_norm": 1.1494975090026855,
+      "learning_rate": 0.0007417519908987486,
+      "loss": 2.57,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2969283276450512,
+      "grad_norm": 0.7995481491088867,
+      "learning_rate": 0.0007415244596131968,
+      "loss": 2.041,
+      "step": 1140
+    },
+    {
+      "epoch": 1.29806598407281,
+      "grad_norm": 1.1029703617095947,
+      "learning_rate": 0.000741296928327645,
+      "loss": 2.4597,
+      "step": 1141
+    },
+    {
+      "epoch": 1.299203640500569,
+      "grad_norm": 0.808023989200592,
+      "learning_rate": 0.0007410693970420933,
+      "loss": 1.195,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3003412969283277,
+      "grad_norm": 1.3540676832199097,
+      "learning_rate": 0.0007408418657565415,
+      "loss": 2.8982,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3014789533560864,
+      "grad_norm": 0.6362115740776062,
+      "learning_rate": 0.0007406143344709898,
+      "loss": 0.9607,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3026166097838452,
+      "grad_norm": 0.5855313539505005,
+      "learning_rate": 0.0007403868031854381,
+      "loss": 1.1911,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3037542662116042,
+      "grad_norm": 0.902195930480957,
+      "learning_rate": 0.0007401592718998863,
+      "loss": 1.8625,
+      "step": 1146
+    },
+    {
+      "epoch": 1.304891922639363,
+      "grad_norm": 0.9085184335708618,
+      "learning_rate": 0.0007399317406143344,
+      "loss": 2.2494,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3060295790671217,
+      "grad_norm": 0.8404464721679688,
+      "learning_rate": 0.0007397042093287827,
+      "loss": 1.7344,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3071672354948807,
+      "grad_norm": 1.5872386693954468,
+      "learning_rate": 0.0007394766780432309,
+      "loss": 2.6294,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3083048919226394,
+      "grad_norm": 0.81778484582901,
+      "learning_rate": 0.0007392491467576792,
+      "loss": 1.716,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3094425483503982,
+      "grad_norm": 0.9477559924125671,
+      "learning_rate": 0.0007390216154721274,
+      "loss": 1.7293,
+      "step": 1151
+    },
+    {
+      "epoch": 1.310580204778157,
+      "grad_norm": 1.0724713802337646,
+      "learning_rate": 0.0007387940841865757,
+      "loss": 2.1255,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3117178612059157,
+      "grad_norm": 0.7965221405029297,
+      "learning_rate": 0.000738566552901024,
+      "loss": 1.1067,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3128555176336747,
+      "grad_norm": 0.9288459420204163,
+      "learning_rate": 0.0007383390216154722,
+      "loss": 2.1202,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3139931740614335,
+      "grad_norm": 1.221725344657898,
+      "learning_rate": 0.0007381114903299204,
+      "loss": 2.4878,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3151308304891922,
+      "grad_norm": 1.0420432090759277,
+      "learning_rate": 0.0007378839590443687,
+      "loss": 1.9784,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3162684869169512,
+      "grad_norm": 1.3146973848342896,
+      "learning_rate": 0.0007376564277588168,
+      "loss": 2.385,
+      "step": 1157
+    },
+    {
+      "epoch": 1.31740614334471,
+      "grad_norm": 1.1637381315231323,
+      "learning_rate": 0.000737428896473265,
+      "loss": 1.8719,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3185437997724687,
+      "grad_norm": 1.2052867412567139,
+      "learning_rate": 0.0007372013651877133,
+      "loss": 3.4914,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3196814562002275,
+      "grad_norm": 0.9300926327705383,
+      "learning_rate": 0.0007369738339021615,
+      "loss": 2.7666,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3208191126279862,
+      "grad_norm": 1.0949327945709229,
+      "learning_rate": 0.0007367463026166098,
+      "loss": 2.3889,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3219567690557452,
+      "grad_norm": 1.1416770219802856,
+      "learning_rate": 0.0007365187713310581,
+      "loss": 2.082,
+      "step": 1162
+    },
+    {
+      "epoch": 1.323094425483504,
+      "grad_norm": 0.6161019802093506,
+      "learning_rate": 0.0007362912400455063,
+      "loss": 1.1583,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3242320819112627,
+      "grad_norm": 1.192044734954834,
+      "learning_rate": 0.0007360637087599545,
+      "loss": 3.2384,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3253697383390217,
+      "grad_norm": 0.7260434031486511,
+      "learning_rate": 0.0007358361774744028,
+      "loss": 1.4132,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3265073947667805,
+      "grad_norm": 0.8180050253868103,
+      "learning_rate": 0.0007356086461888509,
+      "loss": 1.8317,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3276450511945392,
+      "grad_norm": 0.7380133867263794,
+      "learning_rate": 0.0007353811149032991,
+      "loss": 1.4301,
+      "step": 1167
+    },
+    {
+      "epoch": 1.328782707622298,
+      "grad_norm": 0.9578267335891724,
+      "learning_rate": 0.0007351535836177474,
+      "loss": 1.9568,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3299203640500568,
+      "grad_norm": 0.9573442339897156,
+      "learning_rate": 0.0007349260523321957,
+      "loss": 2.6368,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3310580204778157,
+      "grad_norm": 0.8377273678779602,
+      "learning_rate": 0.000734698521046644,
+      "loss": 1.9496,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3321956769055745,
+      "grad_norm": 1.150334358215332,
+      "learning_rate": 0.0007344709897610922,
+      "loss": 1.8031,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 1.1262184381484985,
+      "learning_rate": 0.0007342434584755404,
+      "loss": 2.4094,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3344709897610922,
+      "grad_norm": 0.8950188159942627,
+      "learning_rate": 0.0007340159271899887,
+      "loss": 1.557,
+      "step": 1173
+    },
+    {
+      "epoch": 1.335608646188851,
+      "grad_norm": 1.0164450407028198,
+      "learning_rate": 0.0007337883959044369,
+      "loss": 2.5082,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3367463026166098,
+      "grad_norm": 0.6207723021507263,
+      "learning_rate": 0.0007335608646188851,
+      "loss": 1.6945,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3378839590443685,
+      "grad_norm": 1.3576558828353882,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 2.383,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3390216154721273,
+      "grad_norm": 1.040291428565979,
+      "learning_rate": 0.0007331058020477816,
+      "loss": 1.887,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3401592718998863,
+      "grad_norm": 1.2558809518814087,
+      "learning_rate": 0.0007328782707622298,
+      "loss": 2.6059,
+      "step": 1178
+    },
+    {
+      "epoch": 1.341296928327645,
+      "grad_norm": 1.1690802574157715,
+      "learning_rate": 0.0007326507394766781,
+      "loss": 2.0956,
+      "step": 1179
+    },
+    {
+      "epoch": 1.342434584755404,
+      "grad_norm": 0.6343255043029785,
+      "learning_rate": 0.0007324232081911263,
+      "loss": 1.1936,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3435722411831628,
+      "grad_norm": 1.210610270500183,
+      "learning_rate": 0.0007321956769055745,
+      "loss": 2.7752,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3447098976109215,
+      "grad_norm": 1.1297094821929932,
+      "learning_rate": 0.0007319681456200228,
+      "loss": 1.6202,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3458475540386803,
+      "grad_norm": 1.0680432319641113,
+      "learning_rate": 0.000731740614334471,
+      "loss": 2.5643,
+      "step": 1183
+    },
+    {
+      "epoch": 1.346985210466439,
+      "grad_norm": 1.117632508277893,
+      "learning_rate": 0.0007315130830489192,
+      "loss": 2.2324,
+      "step": 1184
+    },
+    {
+      "epoch": 1.348122866894198,
+      "grad_norm": 1.2003607749938965,
+      "learning_rate": 0.0007312855517633676,
+      "loss": 1.7135,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3492605233219568,
+      "grad_norm": 1.0429855585098267,
+      "learning_rate": 0.0007310580204778157,
+      "loss": 1.7824,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3503981797497155,
+      "grad_norm": 1.0069884061813354,
+      "learning_rate": 0.0007308304891922639,
+      "loss": 1.6408,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3515358361774745,
+      "grad_norm": 0.7664169669151306,
+      "learning_rate": 0.0007306029579067122,
+      "loss": 1.3137,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3526734926052333,
+      "grad_norm": 0.9586772322654724,
+      "learning_rate": 0.0007303754266211604,
+      "loss": 1.861,
+      "step": 1189
+    },
+    {
+      "epoch": 1.353811149032992,
+      "grad_norm": 0.6649326086044312,
+      "learning_rate": 0.0007301478953356086,
+      "loss": 1.0684,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3549488054607508,
+      "grad_norm": 1.4369438886642456,
+      "learning_rate": 0.0007299203640500569,
+      "loss": 3.2283,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3560864618885096,
+      "grad_norm": 1.1447542905807495,
+      "learning_rate": 0.0007296928327645051,
+      "loss": 2.1419,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3572241183162685,
+      "grad_norm": 0.9492262601852417,
+      "learning_rate": 0.0007294653014789535,
+      "loss": 1.9544,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3583617747440273,
+      "grad_norm": 0.8491740822792053,
+      "learning_rate": 0.0007292377701934017,
+      "loss": 1.5818,
+      "step": 1194
+    },
+    {
+      "epoch": 1.359499431171786,
+      "grad_norm": 1.0647927522659302,
+      "learning_rate": 0.0007290102389078499,
+      "loss": 2.6932,
+      "step": 1195
+    },
+    {
+      "epoch": 1.360637087599545,
+      "grad_norm": 1.270956039428711,
+      "learning_rate": 0.0007287827076222981,
+      "loss": 1.9941,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3617747440273038,
+      "grad_norm": 0.9028446674346924,
+      "learning_rate": 0.0007285551763367463,
+      "loss": 2.2445,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3629124004550626,
+      "grad_norm": 0.8218393921852112,
+      "learning_rate": 0.0007283276450511945,
+      "loss": 1.9432,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3640500568828213,
+      "grad_norm": 0.7149525284767151,
+      "learning_rate": 0.0007281001137656428,
+      "loss": 0.8669,
+      "step": 1199
+    },
+    {
+      "epoch": 1.36518771331058,
+      "grad_norm": 0.9978352189064026,
+      "learning_rate": 0.000727872582480091,
+      "loss": 1.4849,
+      "step": 1200
+    },
+    {
+      "epoch": 1.366325369738339,
+      "grad_norm": 0.9793040752410889,
+      "learning_rate": 0.0007276450511945392,
+      "loss": 1.7734,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3674630261660978,
+      "grad_norm": 0.8100583553314209,
+      "learning_rate": 0.0007274175199089876,
+      "loss": 1.7183,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3686006825938566,
+      "grad_norm": 1.051924467086792,
+      "learning_rate": 0.0007271899886234358,
+      "loss": 2.0836,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3697383390216156,
+      "grad_norm": 1.057503342628479,
+      "learning_rate": 0.000726962457337884,
+      "loss": 2.1088,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3708759954493743,
+      "grad_norm": 1.1179509162902832,
+      "learning_rate": 0.0007267349260523322,
+      "loss": 1.452,
+      "step": 1205
+    },
+    {
+      "epoch": 1.372013651877133,
+      "grad_norm": 1.2062307596206665,
+      "learning_rate": 0.0007265073947667804,
+      "loss": 2.4706,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3731513083048918,
+      "grad_norm": 1.5062495470046997,
+      "learning_rate": 0.0007262798634812286,
+      "loss": 3.7995,
+      "step": 1207
+    },
+    {
+      "epoch": 1.3742889647326506,
+      "grad_norm": 1.0101311206817627,
+      "learning_rate": 0.0007260523321956769,
+      "loss": 1.889,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3754266211604096,
+      "grad_norm": 0.5304610729217529,
+      "learning_rate": 0.0007258248009101251,
+      "loss": 0.8576,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3765642775881684,
+      "grad_norm": 0.82547926902771,
+      "learning_rate": 0.0007255972696245733,
+      "loss": 1.8134,
+      "step": 1210
+    },
+    {
+      "epoch": 1.377701934015927,
+      "grad_norm": 1.220291256904602,
+      "learning_rate": 0.0007253697383390217,
+      "loss": 2.1109,
+      "step": 1211
+    },
+    {
+      "epoch": 1.378839590443686,
+      "grad_norm": 0.7384538650512695,
+      "learning_rate": 0.0007251422070534699,
+      "loss": 1.9597,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3799772468714449,
+      "grad_norm": 1.0133798122406006,
+      "learning_rate": 0.0007249146757679182,
+      "loss": 2.7964,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3811149032992036,
+      "grad_norm": 1.00083327293396,
+      "learning_rate": 0.0007246871444823664,
+      "loss": 2.6457,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3822525597269624,
+      "grad_norm": 0.9185119271278381,
+      "learning_rate": 0.0007244596131968145,
+      "loss": 1.5573,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3833902161547214,
+      "grad_norm": 0.8959344029426575,
+      "learning_rate": 0.0007242320819112628,
+      "loss": 1.5507,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3845278725824801,
+      "grad_norm": 0.7419568300247192,
+      "learning_rate": 0.000724004550625711,
+      "loss": 1.599,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3856655290102389,
+      "grad_norm": 1.2727500200271606,
+      "learning_rate": 0.0007237770193401592,
+      "loss": 1.8304,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3868031854379979,
+      "grad_norm": 0.6277416944503784,
+      "learning_rate": 0.0007235494880546076,
+      "loss": 1.4432,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3879408418657566,
+      "grad_norm": 1.2073726654052734,
+      "learning_rate": 0.0007233219567690558,
+      "loss": 2.4847,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3890784982935154,
+      "grad_norm": 1.075276255607605,
+      "learning_rate": 0.000723094425483504,
+      "loss": 2.4881,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3902161547212741,
+      "grad_norm": 1.0907241106033325,
+      "learning_rate": 0.0007228668941979523,
+      "loss": 2.7326,
+      "step": 1222
+    },
+    {
+      "epoch": 1.391353811149033,
+      "grad_norm": 0.9791719317436218,
+      "learning_rate": 0.0007226393629124005,
+      "loss": 1.8307,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3924914675767919,
+      "grad_norm": 1.1889147758483887,
+      "learning_rate": 0.0007224118316268487,
+      "loss": 3.306,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3936291240045506,
+      "grad_norm": 0.9719458818435669,
+      "learning_rate": 0.0007221843003412969,
+      "loss": 1.7917,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3947667804323094,
+      "grad_norm": 1.136434555053711,
+      "learning_rate": 0.0007219567690557451,
+      "loss": 2.1016,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3959044368600684,
+      "grad_norm": 0.9792470335960388,
+      "learning_rate": 0.0007217292377701933,
+      "loss": 1.5415,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3970420932878271,
+      "grad_norm": 0.5439932346343994,
+      "learning_rate": 0.0007215017064846417,
+      "loss": 0.6703,
+      "step": 1228
+    },
+    {
+      "epoch": 1.398179749715586,
+      "grad_norm": 1.0243198871612549,
+      "learning_rate": 0.0007212741751990899,
+      "loss": 1.7786,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3993174061433447,
+      "grad_norm": 1.2160857915878296,
+      "learning_rate": 0.0007210466439135381,
+      "loss": 2.0586,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4004550625711034,
+      "grad_norm": 1.0249682664871216,
+      "learning_rate": 0.0007208191126279864,
+      "loss": 2.3755,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4015927189988624,
+      "grad_norm": 1.3977047204971313,
+      "learning_rate": 0.0007205915813424346,
+      "loss": 2.1492,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4027303754266212,
+      "grad_norm": 0.7847321033477783,
+      "learning_rate": 0.0007203640500568829,
+      "loss": 1.2055,
+      "step": 1233
+    },
+    {
+      "epoch": 1.40386803185438,
+      "grad_norm": 1.104148030281067,
+      "learning_rate": 0.0007201365187713311,
+      "loss": 3.1498,
+      "step": 1234
+    },
+    {
+      "epoch": 1.405005688282139,
+      "grad_norm": 0.87827068567276,
+      "learning_rate": 0.0007199089874857792,
+      "loss": 2.2748,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4061433447098977,
+      "grad_norm": 1.0710291862487793,
+      "learning_rate": 0.0007196814562002276,
+      "loss": 1.4276,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4072810011376564,
+      "grad_norm": 0.950809121131897,
+      "learning_rate": 0.0007194539249146758,
+      "loss": 1.8939,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4084186575654152,
+      "grad_norm": 1.010000228881836,
+      "learning_rate": 0.000719226393629124,
+      "loss": 1.6589,
+      "step": 1238
+    },
+    {
+      "epoch": 1.409556313993174,
+      "grad_norm": 1.1753206253051758,
+      "learning_rate": 0.0007189988623435723,
+      "loss": 2.3059,
+      "step": 1239
+    },
+    {
+      "epoch": 1.410693970420933,
+      "grad_norm": 1.1671147346496582,
+      "learning_rate": 0.0007187713310580205,
+      "loss": 2.4946,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4118316268486917,
+      "grad_norm": 0.8529374599456787,
+      "learning_rate": 0.0007185437997724687,
+      "loss": 2.0972,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4129692832764504,
+      "grad_norm": 1.1962100267410278,
+      "learning_rate": 0.000718316268486917,
+      "loss": 2.4916,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4141069397042094,
+      "grad_norm": 0.8587897419929504,
+      "learning_rate": 0.0007180887372013652,
+      "loss": 0.9147,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4152445961319682,
+      "grad_norm": 0.9558615684509277,
+      "learning_rate": 0.0007178612059158133,
+      "loss": 1.8377,
+      "step": 1244
+    },
+    {
+      "epoch": 1.416382252559727,
+      "grad_norm": 0.694108784198761,
+      "learning_rate": 0.0007176336746302617,
+      "loss": 1.4134,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4175199089874857,
+      "grad_norm": 0.855204164981842,
+      "learning_rate": 0.0007174061433447099,
+      "loss": 2.012,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4186575654152445,
+      "grad_norm": 1.4311326742172241,
+      "learning_rate": 0.0007171786120591581,
+      "loss": 3.41,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4197952218430034,
+      "grad_norm": 0.9610998630523682,
+      "learning_rate": 0.0007169510807736064,
+      "loss": 1.6547,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4209328782707622,
+      "grad_norm": 0.7780663371086121,
+      "learning_rate": 0.0007167235494880546,
+      "loss": 1.9053,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4220705346985212,
+      "grad_norm": 1.3516772985458374,
+      "learning_rate": 0.0007164960182025028,
+      "loss": 0.701,
+      "step": 1250
+    },
+    {
+      "epoch": 1.42320819112628,
+      "grad_norm": 1.596235990524292,
+      "learning_rate": 0.0007162684869169511,
+      "loss": 2.6895,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4243458475540387,
+      "grad_norm": 0.9624771475791931,
+      "learning_rate": 0.0007160409556313994,
+      "loss": 1.341,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4254835039817975,
+      "grad_norm": 0.8453911542892456,
+      "learning_rate": 0.0007158134243458477,
+      "loss": 2.0326,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4266211604095562,
+      "grad_norm": 1.0875787734985352,
+      "learning_rate": 0.0007155858930602958,
+      "loss": 1.4671,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4277588168373152,
+      "grad_norm": 1.0011261701583862,
+      "learning_rate": 0.000715358361774744,
+      "loss": 1.4844,
+      "step": 1255
+    },
+    {
+      "epoch": 1.428896473265074,
+      "grad_norm": 0.7726243734359741,
+      "learning_rate": 0.0007151308304891923,
+      "loss": 1.7404,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4300341296928327,
+      "grad_norm": 0.7656590342521667,
+      "learning_rate": 0.0007149032992036405,
+      "loss": 1.6766,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4311717861205917,
+      "grad_norm": 0.8595698475837708,
+      "learning_rate": 0.0007146757679180887,
+      "loss": 2.1565,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4323094425483505,
+      "grad_norm": 1.003932237625122,
+      "learning_rate": 0.000714448236632537,
+      "loss": 1.721,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4334470989761092,
+      "grad_norm": 0.8126673102378845,
+      "learning_rate": 0.0007142207053469852,
+      "loss": 2.0854,
+      "step": 1260
+    },
+    {
+      "epoch": 1.434584755403868,
+      "grad_norm": 0.9045354723930359,
+      "learning_rate": 0.0007139931740614335,
+      "loss": 1.5096,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4357224118316267,
+      "grad_norm": 0.923866331577301,
+      "learning_rate": 0.0007137656427758818,
+      "loss": 2.2521,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4368600682593857,
+      "grad_norm": 0.8187153339385986,
+      "learning_rate": 0.00071353811149033,
+      "loss": 2.3217,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4379977246871445,
+      "grad_norm": 1.0635052919387817,
+      "learning_rate": 0.0007133105802047781,
+      "loss": 1.8729,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4391353811149032,
+      "grad_norm": 0.8792582750320435,
+      "learning_rate": 0.0007130830489192264,
+      "loss": 1.9955,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4402730375426622,
+      "grad_norm": 1.3968185186386108,
+      "learning_rate": 0.0007128555176336746,
+      "loss": 4.4219,
+      "step": 1266
+    },
+    {
+      "epoch": 1.441410693970421,
+      "grad_norm": 0.956078827381134,
+      "learning_rate": 0.0007126279863481228,
+      "loss": 1.5945,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4425483503981797,
+      "grad_norm": 0.8696045875549316,
+      "learning_rate": 0.0007124004550625711,
+      "loss": 1.1966,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4436860068259385,
+      "grad_norm": 0.7060182690620422,
+      "learning_rate": 0.0007121729237770194,
+      "loss": 1.3995,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4448236632536973,
+      "grad_norm": 0.8254784941673279,
+      "learning_rate": 0.0007119453924914676,
+      "loss": 1.0873,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4459613196814562,
+      "grad_norm": 1.0359618663787842,
+      "learning_rate": 0.0007117178612059159,
+      "loss": 2.2489,
+      "step": 1271
+    },
+    {
+      "epoch": 1.447098976109215,
+      "grad_norm": 0.9475293159484863,
+      "learning_rate": 0.0007114903299203641,
+      "loss": 1.8545,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4482366325369738,
+      "grad_norm": 0.7588447332382202,
+      "learning_rate": 0.0007112627986348122,
+      "loss": 1.5502,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4493742889647327,
+      "grad_norm": 1.24519681930542,
+      "learning_rate": 0.0007110352673492605,
+      "loss": 2.6479,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4505119453924915,
+      "grad_norm": 0.6365463137626648,
+      "learning_rate": 0.0007108077360637087,
+      "loss": 1.404,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4516496018202503,
+      "grad_norm": 1.4220560789108276,
+      "learning_rate": 0.000710580204778157,
+      "loss": 2.4303,
+      "step": 1276
+    },
+    {
+      "epoch": 1.452787258248009,
+      "grad_norm": 0.965912938117981,
+      "learning_rate": 0.0007103526734926053,
+      "loss": 2.1538,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4539249146757678,
+      "grad_norm": 1.103785514831543,
+      "learning_rate": 0.0007101251422070535,
+      "loss": 2.8817,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4550625711035268,
+      "grad_norm": 1.0863354206085205,
+      "learning_rate": 0.0007098976109215018,
+      "loss": 3.1132,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4562002275312855,
+      "grad_norm": 0.988666296005249,
+      "learning_rate": 0.00070967007963595,
+      "loss": 1.915,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4573378839590443,
+      "grad_norm": 1.0589752197265625,
+      "learning_rate": 0.0007094425483503982,
+      "loss": 1.6478,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4584755403868033,
+      "grad_norm": 0.9732983708381653,
+      "learning_rate": 0.0007092150170648465,
+      "loss": 1.834,
+      "step": 1282
+    },
+    {
+      "epoch": 1.459613196814562,
+      "grad_norm": 0.932620644569397,
+      "learning_rate": 0.0007089874857792946,
+      "loss": 2.184,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4607508532423208,
+      "grad_norm": 1.0213489532470703,
+      "learning_rate": 0.0007087599544937428,
+      "loss": 1.9008,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4618885096700796,
+      "grad_norm": 0.8882502317428589,
+      "learning_rate": 0.0007085324232081911,
+      "loss": 1.6468,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4630261660978385,
+      "grad_norm": 0.794425904750824,
+      "learning_rate": 0.0007083048919226394,
+      "loss": 2.0429,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4641638225255973,
+      "grad_norm": 1.1039729118347168,
+      "learning_rate": 0.0007080773606370876,
+      "loss": 2.3133,
+      "step": 1287
+    },
+    {
+      "epoch": 1.465301478953356,
+      "grad_norm": 1.3886933326721191,
+      "learning_rate": 0.0007078498293515359,
+      "loss": 2.0589,
+      "step": 1288
+    },
+    {
+      "epoch": 1.466439135381115,
+      "grad_norm": 0.6157169938087463,
+      "learning_rate": 0.0007076222980659841,
+      "loss": 1.0264,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4675767918088738,
+      "grad_norm": 1.0444914102554321,
+      "learning_rate": 0.0007073947667804323,
+      "loss": 2.8021,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4687144482366326,
+      "grad_norm": 0.9021384716033936,
+      "learning_rate": 0.0007071672354948806,
+      "loss": 1.5786,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4698521046643913,
+      "grad_norm": 0.9910659790039062,
+      "learning_rate": 0.0007069397042093288,
+      "loss": 1.3046,
+      "step": 1292
+    },
+    {
+      "epoch": 1.47098976109215,
+      "grad_norm": 0.8417410254478455,
+      "learning_rate": 0.0007067121729237769,
+      "loss": 1.388,
+      "step": 1293
+    },
+    {
+      "epoch": 1.472127417519909,
+      "grad_norm": 0.8313772082328796,
+      "learning_rate": 0.0007064846416382253,
+      "loss": 1.5113,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4732650739476678,
+      "grad_norm": 0.9730493426322937,
+      "learning_rate": 0.0007062571103526735,
+      "loss": 1.3389,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4744027303754266,
+      "grad_norm": 0.9900069832801819,
+      "learning_rate": 0.0007060295790671218,
+      "loss": 2.1483,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4755403868031856,
+      "grad_norm": 0.8475415110588074,
+      "learning_rate": 0.00070580204778157,
+      "loss": 1.397,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4766780432309443,
+      "grad_norm": 1.0743128061294556,
+      "learning_rate": 0.0007055745164960182,
+      "loss": 1.8463,
+      "step": 1298
+    },
+    {
+      "epoch": 1.477815699658703,
+      "grad_norm": 1.1737279891967773,
+      "learning_rate": 0.0007053469852104665,
+      "loss": 3.4954,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4789533560864618,
+      "grad_norm": 1.0930012464523315,
+      "learning_rate": 0.0007051194539249147,
+      "loss": 2.65,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4800910125142206,
+      "grad_norm": 0.6050748229026794,
+      "learning_rate": 0.0007048919226393629,
+      "loss": 1.1043,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4812286689419796,
+      "grad_norm": 2.5956709384918213,
+      "learning_rate": 0.0007046643913538113,
+      "loss": 4.9918,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4823663253697383,
+      "grad_norm": 1.5760893821716309,
+      "learning_rate": 0.0007044368600682594,
+      "loss": 3.41,
+      "step": 1303
+    },
+    {
+      "epoch": 1.483503981797497,
+      "grad_norm": 1.2234876155853271,
+      "learning_rate": 0.0007042093287827076,
+      "loss": 2.0168,
+      "step": 1304
+    },
+    {
+      "epoch": 1.484641638225256,
+      "grad_norm": 1.5653973817825317,
+      "learning_rate": 0.0007039817974971559,
+      "loss": 3.4326,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4857792946530148,
+      "grad_norm": 0.9075149297714233,
+      "learning_rate": 0.0007037542662116041,
+      "loss": 1.5127,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4869169510807736,
+      "grad_norm": 0.8964717984199524,
+      "learning_rate": 0.0007035267349260523,
+      "loss": 1.9019,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4880546075085324,
+      "grad_norm": 0.8409413695335388,
+      "learning_rate": 0.0007032992036405006,
+      "loss": 1.4982,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4891922639362911,
+      "grad_norm": 0.7918345928192139,
+      "learning_rate": 0.0007030716723549488,
+      "loss": 1.5671,
+      "step": 1309
+    },
+    {
+      "epoch": 1.49032992036405,
+      "grad_norm": 0.9930965900421143,
+      "learning_rate": 0.000702844141069397,
+      "loss": 1.7916,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4914675767918089,
+      "grad_norm": 0.7483389377593994,
+      "learning_rate": 0.0007026166097838454,
+      "loss": 1.2254,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4926052332195676,
+      "grad_norm": 0.5501680374145508,
+      "learning_rate": 0.0007023890784982935,
+      "loss": 1.4911,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4937428896473266,
+      "grad_norm": 0.6416218280792236,
+      "learning_rate": 0.0007021615472127417,
+      "loss": 1.1327,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4948805460750854,
+      "grad_norm": 1.3638542890548706,
+      "learning_rate": 0.00070193401592719,
+      "loss": 2.1457,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4960182025028441,
+      "grad_norm": 0.625977635383606,
+      "learning_rate": 0.0007017064846416382,
+      "loss": 0.9247,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4971558589306029,
+      "grad_norm": 0.9826652407646179,
+      "learning_rate": 0.0007014789533560865,
+      "loss": 2.6433,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4982935153583616,
+      "grad_norm": 0.8866605162620544,
+      "learning_rate": 0.0007012514220705347,
+      "loss": 1.3826,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4994311717861206,
+      "grad_norm": 0.7106355428695679,
+      "learning_rate": 0.0007010238907849829,
+      "loss": 1.3616,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5005688282138794,
+      "grad_norm": 0.9405243992805481,
+      "learning_rate": 0.0007007963594994313,
+      "loss": 1.6128,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5017064846416384,
+      "grad_norm": 0.947182834148407,
+      "learning_rate": 0.0007005688282138795,
+      "loss": 3.0259,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5028441410693971,
+      "grad_norm": 0.89002925157547,
+      "learning_rate": 0.0007003412969283277,
+      "loss": 2.0293,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5039817974971559,
+      "grad_norm": 0.582564115524292,
+      "learning_rate": 0.0007001137656427759,
+      "loss": 0.5041,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5051194539249146,
+      "grad_norm": 0.961336612701416,
+      "learning_rate": 0.0006998862343572241,
+      "loss": 2.0847,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5062571103526734,
+      "grad_norm": 1.3698042631149292,
+      "learning_rate": 0.0006996587030716723,
+      "loss": 2.7407,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5073947667804322,
+      "grad_norm": 1.4105736017227173,
+      "learning_rate": 0.0006994311717861206,
+      "loss": 2.8289,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5085324232081911,
+      "grad_norm": 0.9374104142189026,
+      "learning_rate": 0.0006992036405005688,
+      "loss": 1.7586,
+      "step": 1326
+    },
+    {
+      "epoch": 1.50967007963595,
+      "grad_norm": 0.7621793746948242,
+      "learning_rate": 0.000698976109215017,
+      "loss": 1.3067,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5108077360637089,
+      "grad_norm": 1.1107509136199951,
+      "learning_rate": 0.0006987485779294654,
+      "loss": 2.591,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5119453924914676,
+      "grad_norm": 0.9593746662139893,
+      "learning_rate": 0.0006985210466439136,
+      "loss": 2.4804,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5130830489192264,
+      "grad_norm": 0.7596848011016846,
+      "learning_rate": 0.0006982935153583618,
+      "loss": 0.9857,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5142207053469852,
+      "grad_norm": 0.8829966187477112,
+      "learning_rate": 0.0006980659840728101,
+      "loss": 1.1982,
+      "step": 1331
+    },
+    {
+      "epoch": 1.515358361774744,
+      "grad_norm": 0.8279595375061035,
+      "learning_rate": 0.0006978384527872582,
+      "loss": 2.2226,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5164960182025027,
+      "grad_norm": 1.1019062995910645,
+      "learning_rate": 0.0006976109215017064,
+      "loss": 1.7367,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5176336746302617,
+      "grad_norm": 1.7979804277420044,
+      "learning_rate": 0.0006973833902161547,
+      "loss": 1.959,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5187713310580204,
+      "grad_norm": 0.8705268502235413,
+      "learning_rate": 0.0006971558589306029,
+      "loss": 2.1202,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5199089874857794,
+      "grad_norm": 1.0724592208862305,
+      "learning_rate": 0.0006969283276450513,
+      "loss": 1.7139,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5210466439135382,
+      "grad_norm": 0.9775965809822083,
+      "learning_rate": 0.0006967007963594995,
+      "loss": 2.0016,
+      "step": 1337
+    },
+    {
+      "epoch": 1.522184300341297,
+      "grad_norm": 0.9448524713516235,
+      "learning_rate": 0.0006964732650739477,
+      "loss": 1.6153,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5233219567690557,
+      "grad_norm": 1.5415128469467163,
+      "learning_rate": 0.000696245733788396,
+      "loss": 3.1016,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5244596131968144,
+      "grad_norm": 0.9991744160652161,
+      "learning_rate": 0.0006960182025028442,
+      "loss": 1.5704,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5255972696245734,
+      "grad_norm": 0.5757598280906677,
+      "learning_rate": 0.0006957906712172923,
+      "loss": 1.0319,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5267349260523322,
+      "grad_norm": 0.7392444014549255,
+      "learning_rate": 0.0006955631399317406,
+      "loss": 1.6469,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5278725824800912,
+      "grad_norm": 0.8269877433776855,
+      "learning_rate": 0.0006953356086461888,
+      "loss": 1.5095,
+      "step": 1343
+    },
+    {
+      "epoch": 1.52901023890785,
+      "grad_norm": 0.8942914605140686,
+      "learning_rate": 0.000695108077360637,
+      "loss": 1.7443,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5301478953356087,
+      "grad_norm": 0.9986234307289124,
+      "learning_rate": 0.0006948805460750854,
+      "loss": 2.1091,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5312855517633674,
+      "grad_norm": 1.5111842155456543,
+      "learning_rate": 0.0006946530147895336,
+      "loss": 2.9399,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5324232081911262,
+      "grad_norm": 0.8107113242149353,
+      "learning_rate": 0.0006944254835039818,
+      "loss": 1.5483,
+      "step": 1347
+    },
+    {
+      "epoch": 1.533560864618885,
+      "grad_norm": 0.9013845324516296,
+      "learning_rate": 0.0006941979522184301,
+      "loss": 2.0372,
+      "step": 1348
+    },
+    {
+      "epoch": 1.534698521046644,
+      "grad_norm": 0.812639057636261,
+      "learning_rate": 0.0006939704209328783,
+      "loss": 1.7029,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5358361774744027,
+      "grad_norm": 1.5611118078231812,
+      "learning_rate": 0.0006937428896473265,
+      "loss": 3.1895,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5369738339021617,
+      "grad_norm": 0.7234958410263062,
+      "learning_rate": 0.0006935153583617747,
+      "loss": 1.7994,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5381114903299204,
+      "grad_norm": 0.5042601227760315,
+      "learning_rate": 0.0006932878270762229,
+      "loss": 0.6559,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5392491467576792,
+      "grad_norm": 0.7862391471862793,
+      "learning_rate": 0.0006930602957906712,
+      "loss": 1.2773,
+      "step": 1353
+    },
+    {
+      "epoch": 1.540386803185438,
+      "grad_norm": 0.8332534432411194,
+      "learning_rate": 0.0006928327645051195,
+      "loss": 1.7685,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5415244596131967,
+      "grad_norm": 0.7258904576301575,
+      "learning_rate": 0.0006926052332195677,
+      "loss": 0.8505,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5426621160409555,
+      "grad_norm": 0.6030761003494263,
+      "learning_rate": 0.000692377701934016,
+      "loss": 1.2053,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5437997724687145,
+      "grad_norm": 1.0888185501098633,
+      "learning_rate": 0.0006921501706484642,
+      "loss": 2.0252,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5449374288964732,
+      "grad_norm": 1.2503432035446167,
+      "learning_rate": 0.0006919226393629124,
+      "loss": 3.6095,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5460750853242322,
+      "grad_norm": 0.6536726951599121,
+      "learning_rate": 0.0006916951080773607,
+      "loss": 1.171,
+      "step": 1359
+    },
+    {
+      "epoch": 1.547212741751991,
+      "grad_norm": 0.8668822646141052,
+      "learning_rate": 0.000691467576791809,
+      "loss": 2.234,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5483503981797497,
+      "grad_norm": 1.2817611694335938,
+      "learning_rate": 0.000691240045506257,
+      "loss": 1.7949,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5494880546075085,
+      "grad_norm": 0.9210306406021118,
+      "learning_rate": 0.0006910125142207054,
+      "loss": 0.6781,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5506257110352673,
+      "grad_norm": 1.0318769216537476,
+      "learning_rate": 0.0006907849829351536,
+      "loss": 2.4866,
+      "step": 1363
+    },
+    {
+      "epoch": 1.551763367463026,
+      "grad_norm": 1.0677587985992432,
+      "learning_rate": 0.0006905574516496018,
+      "loss": 2.1298,
+      "step": 1364
+    },
+    {
+      "epoch": 1.552901023890785,
+      "grad_norm": 0.6409241557121277,
+      "learning_rate": 0.0006903299203640501,
+      "loss": 1.4516,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5540386803185438,
+      "grad_norm": 1.06160569190979,
+      "learning_rate": 0.0006901023890784983,
+      "loss": 2.8236,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5551763367463027,
+      "grad_norm": 1.5889126062393188,
+      "learning_rate": 0.0006898748577929465,
+      "loss": 2.9584,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5563139931740615,
+      "grad_norm": 0.5789588093757629,
+      "learning_rate": 0.0006896473265073948,
+      "loss": 1.2789,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5574516496018203,
+      "grad_norm": 1.5852688550949097,
+      "learning_rate": 0.000689419795221843,
+      "loss": 3.2372,
+      "step": 1369
+    },
+    {
+      "epoch": 1.558589306029579,
+      "grad_norm": 0.8930657505989075,
+      "learning_rate": 0.0006891922639362913,
+      "loss": 1.2431,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5597269624573378,
+      "grad_norm": 1.1231560707092285,
+      "learning_rate": 0.0006889647326507395,
+      "loss": 2.3293,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5608646188850968,
+      "grad_norm": 1.0928031206130981,
+      "learning_rate": 0.0006887372013651877,
+      "loss": 2.5423,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5620022753128555,
+      "grad_norm": 0.9887200593948364,
+      "learning_rate": 0.0006885096700796359,
+      "loss": 2.1031,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5631399317406145,
+      "grad_norm": 1.6233134269714355,
+      "learning_rate": 0.0006882821387940842,
+      "loss": 3.366,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5642775881683733,
+      "grad_norm": 0.6267789006233215,
+      "learning_rate": 0.0006880546075085324,
+      "loss": 1.0757,
+      "step": 1375
+    },
+    {
+      "epoch": 1.565415244596132,
+      "grad_norm": 0.8272294402122498,
+      "learning_rate": 0.0006878270762229807,
+      "loss": 1.3076,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5665529010238908,
+      "grad_norm": 0.7376548051834106,
+      "learning_rate": 0.000687599544937429,
+      "loss": 1.4426,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5676905574516495,
+      "grad_norm": 1.136896014213562,
+      "learning_rate": 0.0006873720136518772,
+      "loss": 2.9297,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5688282138794083,
+      "grad_norm": 0.8306413888931274,
+      "learning_rate": 0.0006871444823663255,
+      "loss": 1.9698,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5699658703071673,
+      "grad_norm": 0.9145554900169373,
+      "learning_rate": 0.0006869169510807736,
+      "loss": 1.2156,
+      "step": 1380
+    },
+    {
+      "epoch": 1.571103526734926,
+      "grad_norm": 1.226146936416626,
+      "learning_rate": 0.0006866894197952218,
+      "loss": 2.1672,
+      "step": 1381
+    },
+    {
+      "epoch": 1.572241183162685,
+      "grad_norm": 0.6552311182022095,
+      "learning_rate": 0.0006864618885096701,
+      "loss": 0.9541,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5733788395904438,
+      "grad_norm": 0.8643866777420044,
+      "learning_rate": 0.0006862343572241183,
+      "loss": 1.3602,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5745164960182025,
+      "grad_norm": 1.0479025840759277,
+      "learning_rate": 0.0006860068259385665,
+      "loss": 2.6468,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5756541524459613,
+      "grad_norm": 1.173642635345459,
+      "learning_rate": 0.0006857792946530148,
+      "loss": 2.4588,
+      "step": 1385
+    },
+    {
+      "epoch": 1.57679180887372,
+      "grad_norm": 0.751677930355072,
+      "learning_rate": 0.0006855517633674631,
+      "loss": 0.8411,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5779294653014788,
+      "grad_norm": 1.1802860498428345,
+      "learning_rate": 0.0006853242320819113,
+      "loss": 3.0113,
+      "step": 1387
+    },
+    {
+      "epoch": 1.5790671217292378,
+      "grad_norm": 0.8296557068824768,
+      "learning_rate": 0.0006850967007963596,
+      "loss": 1.7901,
+      "step": 1388
+    },
+    {
+      "epoch": 1.5802047781569966,
+      "grad_norm": 1.1105976104736328,
+      "learning_rate": 0.0006848691695108078,
+      "loss": 1.6191,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5813424345847555,
+      "grad_norm": 1.2469074726104736,
+      "learning_rate": 0.0006846416382252559,
+      "loss": 1.4178,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5824800910125143,
+      "grad_norm": 1.089394450187683,
+      "learning_rate": 0.0006844141069397042,
+      "loss": 1.9584,
+      "step": 1391
+    },
+    {
+      "epoch": 1.583617747440273,
+      "grad_norm": 0.9064239859580994,
+      "learning_rate": 0.0006841865756541524,
+      "loss": 1.5156,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5847554038680318,
+      "grad_norm": 1.0828756093978882,
+      "learning_rate": 0.0006839590443686006,
+      "loss": 2.1486,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5858930602957906,
+      "grad_norm": 1.159847617149353,
+      "learning_rate": 0.000683731513083049,
+      "loss": 2.8738,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5870307167235493,
+      "grad_norm": 0.6444835066795349,
+      "learning_rate": 0.0006835039817974972,
+      "loss": 1.2373,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5881683731513083,
+      "grad_norm": 0.6985568404197693,
+      "learning_rate": 0.0006832764505119454,
+      "loss": 1.828,
+      "step": 1396
+    },
+    {
+      "epoch": 1.589306029579067,
+      "grad_norm": 0.9549153447151184,
+      "learning_rate": 0.0006830489192263937,
+      "loss": 1.5909,
+      "step": 1397
+    },
+    {
+      "epoch": 1.590443686006826,
+      "grad_norm": 0.720780074596405,
+      "learning_rate": 0.0006828213879408419,
+      "loss": 1.4502,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5915813424345848,
+      "grad_norm": 1.090728998184204,
+      "learning_rate": 0.0006825938566552902,
+      "loss": 2.7399,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5927189988623436,
+      "grad_norm": 1.3747398853302002,
+      "learning_rate": 0.0006823663253697383,
+      "loss": 2.4482,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5938566552901023,
+      "grad_norm": 1.0005912780761719,
+      "learning_rate": 0.0006821387940841865,
+      "loss": 1.7441,
+      "step": 1401
+    },
+    {
+      "epoch": 1.594994311717861,
+      "grad_norm": 1.298248052597046,
+      "learning_rate": 0.0006819112627986348,
+      "loss": 2.1579,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5961319681456199,
+      "grad_norm": 1.0068423748016357,
+      "learning_rate": 0.0006816837315130831,
+      "loss": 2.5003,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5972696245733788,
+      "grad_norm": 1.1366382837295532,
+      "learning_rate": 0.0006814562002275313,
+      "loss": 2.3228,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5984072810011376,
+      "grad_norm": 0.9474295973777771,
+      "learning_rate": 0.0006812286689419796,
+      "loss": 1.1316,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5995449374288966,
+      "grad_norm": 1.3815207481384277,
+      "learning_rate": 0.0006810011376564278,
+      "loss": 4.0211,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6006825938566553,
+      "grad_norm": 0.703368604183197,
+      "learning_rate": 0.000680773606370876,
+      "loss": 1.5288,
+      "step": 1407
+    },
+    {
+      "epoch": 1.601820250284414,
+      "grad_norm": 1.4243957996368408,
+      "learning_rate": 0.0006805460750853243,
+      "loss": 2.4914,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6029579067121729,
+      "grad_norm": 1.1586917638778687,
+      "learning_rate": 0.0006803185437997725,
+      "loss": 2.1698,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6040955631399316,
+      "grad_norm": 0.6247298717498779,
+      "learning_rate": 0.0006800910125142206,
+      "loss": 1.3813,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6052332195676906,
+      "grad_norm": 1.181565284729004,
+      "learning_rate": 0.000679863481228669,
+      "loss": 2.3456,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6063708759954494,
+      "grad_norm": 0.8031807541847229,
+      "learning_rate": 0.0006796359499431172,
+      "loss": 1.1918,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6075085324232083,
+      "grad_norm": 0.920711874961853,
+      "learning_rate": 0.0006794084186575654,
+      "loss": 2.328,
+      "step": 1413
+    },
+    {
+      "epoch": 1.608646188850967,
+      "grad_norm": 1.0009433031082153,
+      "learning_rate": 0.0006791808873720137,
+      "loss": 1.3054,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6097838452787259,
+      "grad_norm": 1.1435019969940186,
+      "learning_rate": 0.0006789533560864619,
+      "loss": 1.6864,
+      "step": 1415
+    },
+    {
+      "epoch": 1.6109215017064846,
+      "grad_norm": 0.9297929406166077,
+      "learning_rate": 0.0006787258248009101,
+      "loss": 2.1921,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6120591581342434,
+      "grad_norm": 0.8632849454879761,
+      "learning_rate": 0.0006784982935153584,
+      "loss": 1.8867,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6131968145620021,
+      "grad_norm": 0.7660686373710632,
+      "learning_rate": 0.0006782707622298066,
+      "loss": 1.4099,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6143344709897611,
+      "grad_norm": 1.1486921310424805,
+      "learning_rate": 0.0006780432309442548,
+      "loss": 2.032,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6154721274175199,
+      "grad_norm": 1.0095970630645752,
+      "learning_rate": 0.0006778156996587031,
+      "loss": 2.4319,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6166097838452789,
+      "grad_norm": 1.0770419836044312,
+      "learning_rate": 0.0006775881683731513,
+      "loss": 2.338,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6177474402730376,
+      "grad_norm": 0.8985020518302917,
+      "learning_rate": 0.0006773606370875996,
+      "loss": 1.3659,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6188850967007964,
+      "grad_norm": 1.1922725439071655,
+      "learning_rate": 0.0006771331058020478,
+      "loss": 0.865,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6200227531285551,
+      "grad_norm": 0.781085193157196,
+      "learning_rate": 0.000676905574516496,
+      "loss": 1.6722,
+      "step": 1424
+    },
+    {
+      "epoch": 1.621160409556314,
+      "grad_norm": 1.0233283042907715,
+      "learning_rate": 0.0006766780432309443,
+      "loss": 2.5444,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6222980659840727,
+      "grad_norm": 0.7940084934234619,
+      "learning_rate": 0.0006764505119453925,
+      "loss": 1.4521,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6234357224118316,
+      "grad_norm": 0.6902288794517517,
+      "learning_rate": 0.0006762229806598407,
+      "loss": 1.0446,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6245733788395904,
+      "grad_norm": 0.9518580436706543,
+      "learning_rate": 0.0006759954493742891,
+      "loss": 2.1233,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6257110352673494,
+      "grad_norm": 1.147662878036499,
+      "learning_rate": 0.0006757679180887372,
+      "loss": 3.0784,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6268486916951082,
+      "grad_norm": 0.7210685610771179,
+      "learning_rate": 0.0006755403868031854,
+      "loss": 1.0661,
+      "step": 1430
+    },
+    {
+      "epoch": 1.627986348122867,
+      "grad_norm": 1.2803034782409668,
+      "learning_rate": 0.0006753128555176337,
+      "loss": 1.7098,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6291240045506257,
+      "grad_norm": 1.3280972242355347,
+      "learning_rate": 0.0006750853242320819,
+      "loss": 1.756,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6302616609783844,
+      "grad_norm": 0.7079578638076782,
+      "learning_rate": 0.0006748577929465301,
+      "loss": 1.4689,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6313993174061432,
+      "grad_norm": 1.2363884449005127,
+      "learning_rate": 0.0006746302616609784,
+      "loss": 1.9533,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6325369738339022,
+      "grad_norm": 1.2230910062789917,
+      "learning_rate": 0.0006744027303754266,
+      "loss": 2.7112,
+      "step": 1435
+    },
+    {
+      "epoch": 1.633674630261661,
+      "grad_norm": 1.5893361568450928,
+      "learning_rate": 0.0006741751990898749,
+      "loss": 1.6204,
+      "step": 1436
+    },
+    {
+      "epoch": 1.63481228668942,
+      "grad_norm": 0.9474936723709106,
+      "learning_rate": 0.0006739476678043232,
+      "loss": 2.2779,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6359499431171787,
+      "grad_norm": 0.6129853129386902,
+      "learning_rate": 0.0006737201365187714,
+      "loss": 0.7177,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6370875995449374,
+      "grad_norm": 0.7504928708076477,
+      "learning_rate": 0.0006734926052332196,
+      "loss": 2.2781,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6382252559726962,
+      "grad_norm": 0.9989508986473083,
+      "learning_rate": 0.0006732650739476678,
+      "loss": 1.9803,
+      "step": 1440
+    },
+    {
+      "epoch": 1.639362912400455,
+      "grad_norm": 1.202582836151123,
+      "learning_rate": 0.000673037542662116,
+      "loss": 1.8057,
+      "step": 1441
+    },
+    {
+      "epoch": 1.640500568828214,
+      "grad_norm": 0.9157416224479675,
+      "learning_rate": 0.0006728100113765643,
+      "loss": 2.2069,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6416382252559727,
+      "grad_norm": 0.6836434602737427,
+      "learning_rate": 0.0006725824800910125,
+      "loss": 1.0477,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6427758816837317,
+      "grad_norm": 0.8547561168670654,
+      "learning_rate": 0.0006723549488054607,
+      "loss": 1.8937,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6439135381114904,
+      "grad_norm": 0.8537775874137878,
+      "learning_rate": 0.0006721274175199091,
+      "loss": 1.5731,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6450511945392492,
+      "grad_norm": 1.0358448028564453,
+      "learning_rate": 0.0006718998862343573,
+      "loss": 2.0885,
+      "step": 1446
+    },
+    {
+      "epoch": 1.646188850967008,
+      "grad_norm": 0.9270764589309692,
+      "learning_rate": 0.0006716723549488055,
+      "loss": 1.8622,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6473265073947667,
+      "grad_norm": 0.6689456701278687,
+      "learning_rate": 0.0006714448236632537,
+      "loss": 0.7905,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6484641638225255,
+      "grad_norm": 0.6682091951370239,
+      "learning_rate": 0.0006712172923777019,
+      "loss": 1.0117,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6496018202502845,
+      "grad_norm": 0.9012134671211243,
+      "learning_rate": 0.0006709897610921501,
+      "loss": 1.6618,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6507394766780432,
+      "grad_norm": 0.7726583480834961,
+      "learning_rate": 0.0006707622298065984,
+      "loss": 1.4585,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6518771331058022,
+      "grad_norm": 1.0777757167816162,
+      "learning_rate": 0.0006705346985210466,
+      "loss": 2.1442,
+      "step": 1452
+    },
+    {
+      "epoch": 1.653014789533561,
+      "grad_norm": 1.3284507989883423,
+      "learning_rate": 0.0006703071672354949,
+      "loss": 1.9266,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6541524459613197,
+      "grad_norm": 1.139455795288086,
+      "learning_rate": 0.0006700796359499432,
+      "loss": 2.3044,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6552901023890785,
+      "grad_norm": 0.736585795879364,
+      "learning_rate": 0.0006698521046643914,
+      "loss": 1.3312,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6564277588168372,
+      "grad_norm": 1.6546106338500977,
+      "learning_rate": 0.0006696245733788396,
+      "loss": 1.9867,
+      "step": 1456
+    },
+    {
+      "epoch": 1.657565415244596,
+      "grad_norm": 0.872257649898529,
+      "learning_rate": 0.0006693970420932879,
+      "loss": 2.0105,
+      "step": 1457
+    },
+    {
+      "epoch": 1.658703071672355,
+      "grad_norm": 0.9059979915618896,
+      "learning_rate": 0.000669169510807736,
+      "loss": 2.0219,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6598407281001137,
+      "grad_norm": 0.6183615326881409,
+      "learning_rate": 0.0006689419795221842,
+      "loss": 1.2641,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6609783845278727,
+      "grad_norm": 0.7358295917510986,
+      "learning_rate": 0.0006687144482366325,
+      "loss": 1.3596,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6621160409556315,
+      "grad_norm": 0.8297770023345947,
+      "learning_rate": 0.0006684869169510807,
+      "loss": 1.5614,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6632536973833902,
+      "grad_norm": 0.6983165144920349,
+      "learning_rate": 0.0006682593856655291,
+      "loss": 1.1403,
+      "step": 1462
+    },
+    {
+      "epoch": 1.664391353811149,
+      "grad_norm": 1.0305124521255493,
+      "learning_rate": 0.0006680318543799773,
+      "loss": 2.0436,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6655290102389078,
+      "grad_norm": 1.53620183467865,
+      "learning_rate": 0.0006678043230944255,
+      "loss": 2.6743,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.9701448678970337,
+      "learning_rate": 0.0006675767918088738,
+      "loss": 2.7797,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6678043230944255,
+      "grad_norm": 1.1551249027252197,
+      "learning_rate": 0.000667349260523322,
+      "loss": 1.9898,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6689419795221843,
+      "grad_norm": 1.8078079223632812,
+      "learning_rate": 0.0006671217292377702,
+      "loss": 3.4769,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6700796359499432,
+      "grad_norm": 0.9920907020568848,
+      "learning_rate": 0.0006668941979522184,
+      "loss": 1.5748,
+      "step": 1468
+    },
+    {
+      "epoch": 1.671217292377702,
+      "grad_norm": 1.0543971061706543,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 2.2308,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6723549488054608,
+      "grad_norm": 1.3774967193603516,
+      "learning_rate": 0.0006664391353811149,
+      "loss": 2.9355,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6734926052332195,
+      "grad_norm": 0.819430947303772,
+      "learning_rate": 0.0006662116040955632,
+      "loss": 1.8651,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6746302616609783,
+      "grad_norm": 0.9009912014007568,
+      "learning_rate": 0.0006659840728100114,
+      "loss": 1.771,
+      "step": 1472
+    },
+    {
+      "epoch": 1.675767918088737,
+      "grad_norm": 0.9403852224349976,
+      "learning_rate": 0.0006657565415244596,
+      "loss": 1.9841,
+      "step": 1473
+    },
+    {
+      "epoch": 1.676905574516496,
+      "grad_norm": 1.555391550064087,
+      "learning_rate": 0.0006655290102389079,
+      "loss": 2.3617,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6780432309442548,
+      "grad_norm": 1.146911382675171,
+      "learning_rate": 0.0006653014789533561,
+      "loss": 2.3203,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6791808873720138,
+      "grad_norm": 0.853667140007019,
+      "learning_rate": 0.0006650739476678043,
+      "loss": 1.2485,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6803185437997725,
+      "grad_norm": 1.278773307800293,
+      "learning_rate": 0.0006648464163822526,
+      "loss": 2.0797,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6814562002275313,
+      "grad_norm": 1.3791263103485107,
+      "learning_rate": 0.0006646188850967008,
+      "loss": 2.859,
+      "step": 1478
+    },
+    {
+      "epoch": 1.68259385665529,
+      "grad_norm": 0.9324190616607666,
+      "learning_rate": 0.000664391353811149,
+      "loss": 2.3866,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6837315130830488,
+      "grad_norm": 1.0593537092208862,
+      "learning_rate": 0.0006641638225255973,
+      "loss": 1.6384,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6848691695108078,
+      "grad_norm": 1.1699516773223877,
+      "learning_rate": 0.0006639362912400455,
+      "loss": 2.8342,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6860068259385665,
+      "grad_norm": 0.7266961336135864,
+      "learning_rate": 0.0006637087599544938,
+      "loss": 1.4487,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6871444823663255,
+      "grad_norm": 0.8684595227241516,
+      "learning_rate": 0.000663481228668942,
+      "loss": 1.6997,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6882821387940843,
+      "grad_norm": 0.958516538143158,
+      "learning_rate": 0.0006632536973833902,
+      "loss": 1.7327,
+      "step": 1484
+    },
+    {
+      "epoch": 1.689419795221843,
+      "grad_norm": 1.0229347944259644,
+      "learning_rate": 0.0006630261660978385,
+      "loss": 3.2928,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6905574516496018,
+      "grad_norm": 1.314428687095642,
+      "learning_rate": 0.0006627986348122868,
+      "loss": 1.7722,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6916951080773606,
+      "grad_norm": 0.6298933625221252,
+      "learning_rate": 0.0006625711035267349,
+      "loss": 1.2597,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6928327645051193,
+      "grad_norm": 0.8204244375228882,
+      "learning_rate": 0.0006623435722411832,
+      "loss": 1.4012,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6939704209328783,
+      "grad_norm": 0.6752439737319946,
+      "learning_rate": 0.0006621160409556314,
+      "loss": 1.3025,
+      "step": 1489
+    },
+    {
+      "epoch": 1.695108077360637,
+      "grad_norm": 0.7786659002304077,
+      "learning_rate": 0.0006618885096700796,
+      "loss": 1.601,
+      "step": 1490
+    },
+    {
+      "epoch": 1.696245733788396,
+      "grad_norm": 0.9201928377151489,
+      "learning_rate": 0.0006616609783845279,
+      "loss": 2.8528,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6973833902161548,
+      "grad_norm": 0.9727663993835449,
+      "learning_rate": 0.0006614334470989761,
+      "loss": 2.2852,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6985210466439136,
+      "grad_norm": 1.015677571296692,
+      "learning_rate": 0.0006612059158134243,
+      "loss": 2.5611,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6996587030716723,
+      "grad_norm": 0.6832883954048157,
+      "learning_rate": 0.0006609783845278727,
+      "loss": 0.8426,
+      "step": 1494
+    },
+    {
+      "epoch": 1.700796359499431,
+      "grad_norm": 1.015763759613037,
+      "learning_rate": 0.0006607508532423209,
+      "loss": 2.1979,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7019340159271898,
+      "grad_norm": 1.0312163829803467,
+      "learning_rate": 0.0006605233219567691,
+      "loss": 1.6475,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7030716723549488,
+      "grad_norm": 1.1301902532577515,
+      "learning_rate": 0.0006602957906712173,
+      "loss": 1.6602,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7042093287827076,
+      "grad_norm": 1.635214924812317,
+      "learning_rate": 0.0006600682593856655,
+      "loss": 2.0394,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7053469852104666,
+      "grad_norm": 1.2907978296279907,
+      "learning_rate": 0.0006598407281001137,
+      "loss": 2.3087,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7064846416382253,
+      "grad_norm": 0.8886736035346985,
+      "learning_rate": 0.000659613196814562,
+      "loss": 2.1899,
+      "step": 1500
+    },
+    {
+      "epoch": 1.707622298065984,
+      "grad_norm": 0.7392778992652893,
+      "learning_rate": 0.0006593856655290102,
+      "loss": 1.1168,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7087599544937428,
+      "grad_norm": 0.9279195070266724,
+      "learning_rate": 0.0006591581342434585,
+      "loss": 3.2731,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7098976109215016,
+      "grad_norm": 0.7852345705032349,
+      "learning_rate": 0.0006589306029579068,
+      "loss": 1.6468,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7110352673492604,
+      "grad_norm": 0.8137550354003906,
+      "learning_rate": 0.000658703071672355,
+      "loss": 1.7085,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7121729237770194,
+      "grad_norm": 1.1284794807434082,
+      "learning_rate": 0.0006584755403868033,
+      "loss": 1.3615,
+      "step": 1505
+    },
+    {
+      "epoch": 1.713310580204778,
+      "grad_norm": 0.9071482419967651,
+      "learning_rate": 0.0006582480091012515,
+      "loss": 1.9902,
+      "step": 1506
+    },
+    {
+      "epoch": 1.714448236632537,
+      "grad_norm": 0.6611661314964294,
+      "learning_rate": 0.0006580204778156996,
+      "loss": 1.4362,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7155858930602959,
+      "grad_norm": 0.6886146664619446,
+      "learning_rate": 0.0006577929465301479,
+      "loss": 1.5251,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7167235494880546,
+      "grad_norm": 0.859805166721344,
+      "learning_rate": 0.0006575654152445961,
+      "loss": 1.6648,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7178612059158134,
+      "grad_norm": 0.8424084782600403,
+      "learning_rate": 0.0006573378839590443,
+      "loss": 1.7239,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7189988623435721,
+      "grad_norm": 0.5435293316841125,
+      "learning_rate": 0.0006571103526734927,
+      "loss": 1.1297,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7201365187713311,
+      "grad_norm": 1.4670809507369995,
+      "learning_rate": 0.0006568828213879409,
+      "loss": 3.3636,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7212741751990899,
+      "grad_norm": 0.8184078931808472,
+      "learning_rate": 0.0006566552901023891,
+      "loss": 1.8129,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7224118316268489,
+      "grad_norm": 0.8720480799674988,
+      "learning_rate": 0.0006564277588168374,
+      "loss": 1.7082,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7235494880546076,
+      "grad_norm": 1.2240320444107056,
+      "learning_rate": 0.0006562002275312856,
+      "loss": 1.6893,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7246871444823664,
+      "grad_norm": 1.4103710651397705,
+      "learning_rate": 0.0006559726962457337,
+      "loss": 1.4781,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7258248009101251,
+      "grad_norm": 0.7082839608192444,
+      "learning_rate": 0.000655745164960182,
+      "loss": 1.6487,
+      "step": 1517
+    },
+    {
+      "epoch": 1.726962457337884,
+      "grad_norm": 1.3092166185379028,
+      "learning_rate": 0.0006555176336746302,
+      "loss": 3.0383,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7281001137656427,
+      "grad_norm": 0.6530581116676331,
+      "learning_rate": 0.0006552901023890784,
+      "loss": 1.1405,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7292377701934016,
+      "grad_norm": 0.5259117484092712,
+      "learning_rate": 0.0006550625711035268,
+      "loss": 0.9486,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7303754266211604,
+      "grad_norm": 0.8423599004745483,
+      "learning_rate": 0.000654835039817975,
+      "loss": 1.7263,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7315130830489194,
+      "grad_norm": 0.9277529716491699,
+      "learning_rate": 0.0006546075085324233,
+      "loss": 1.3625,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7326507394766781,
+      "grad_norm": 0.8496603965759277,
+      "learning_rate": 0.0006543799772468715,
+      "loss": 2.406,
+      "step": 1523
+    },
+    {
+      "epoch": 1.733788395904437,
+      "grad_norm": 1.1598669290542603,
+      "learning_rate": 0.0006541524459613197,
+      "loss": 2.4768,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7349260523321957,
+      "grad_norm": 0.9473751783370972,
+      "learning_rate": 0.000653924914675768,
+      "loss": 2.243,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7360637087599544,
+      "grad_norm": 0.8064972758293152,
+      "learning_rate": 0.0006536973833902161,
+      "loss": 1.6925,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7372013651877132,
+      "grad_norm": 1.018776297569275,
+      "learning_rate": 0.0006534698521046643,
+      "loss": 1.6876,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7383390216154722,
+      "grad_norm": 1.1636130809783936,
+      "learning_rate": 0.0006532423208191127,
+      "loss": 2.0573,
+      "step": 1528
+    },
+    {
+      "epoch": 1.739476678043231,
+      "grad_norm": 0.890720784664154,
+      "learning_rate": 0.0006530147895335609,
+      "loss": 0.8197,
+      "step": 1529
+    },
+    {
+      "epoch": 1.74061433447099,
+      "grad_norm": 0.8947886228561401,
+      "learning_rate": 0.0006527872582480091,
+      "loss": 1.5267,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7417519908987487,
+      "grad_norm": 1.0752261877059937,
+      "learning_rate": 0.0006525597269624574,
+      "loss": 2.324,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7428896473265074,
+      "grad_norm": 1.107813835144043,
+      "learning_rate": 0.0006523321956769056,
+      "loss": 2.1986,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7440273037542662,
+      "grad_norm": 0.8949710726737976,
+      "learning_rate": 0.0006521046643913538,
+      "loss": 1.3886,
+      "step": 1533
+    },
+    {
+      "epoch": 1.745164960182025,
+      "grad_norm": 1.6787731647491455,
+      "learning_rate": 0.0006518771331058021,
+      "loss": 3.7587,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7463026166097837,
+      "grad_norm": 1.2499511241912842,
+      "learning_rate": 0.0006516496018202503,
+      "loss": 2.327,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7474402730375427,
+      "grad_norm": 0.7595018148422241,
+      "learning_rate": 0.0006514220705346984,
+      "loss": 1.3373,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7485779294653014,
+      "grad_norm": 1.2214561700820923,
+      "learning_rate": 0.0006511945392491468,
+      "loss": 2.4152,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7497155858930604,
+      "grad_norm": 0.727344274520874,
+      "learning_rate": 0.000650967007963595,
+      "loss": 1.4908,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7508532423208192,
+      "grad_norm": 0.9511622786521912,
+      "learning_rate": 0.0006507394766780432,
+      "loss": 1.6895,
+      "step": 1539
+    },
+    {
+      "epoch": 1.751990898748578,
+      "grad_norm": 0.8098679184913635,
+      "learning_rate": 0.0006505119453924915,
+      "loss": 1.125,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7531285551763367,
+      "grad_norm": 0.9225071668624878,
+      "learning_rate": 0.0006502844141069397,
+      "loss": 1.8387,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7542662116040955,
+      "grad_norm": 0.999563992023468,
+      "learning_rate": 0.000650056882821388,
+      "loss": 2.7922,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7554038680318542,
+      "grad_norm": 0.722228467464447,
+      "learning_rate": 0.0006498293515358362,
+      "loss": 1.7187,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7565415244596132,
+      "grad_norm": 0.6238258481025696,
+      "learning_rate": 0.0006496018202502844,
+      "loss": 0.8305,
+      "step": 1544
+    },
+    {
+      "epoch": 1.757679180887372,
+      "grad_norm": 1.207566499710083,
+      "learning_rate": 0.0006493742889647328,
+      "loss": 2.9552,
+      "step": 1545
+    },
+    {
+      "epoch": 1.758816837315131,
+      "grad_norm": 1.0733917951583862,
+      "learning_rate": 0.0006491467576791809,
+      "loss": 1.9941,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7599544937428897,
+      "grad_norm": 0.8945872783660889,
+      "learning_rate": 0.0006489192263936291,
+      "loss": 1.6187,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7610921501706485,
+      "grad_norm": 1.0715452432632446,
+      "learning_rate": 0.0006486916951080774,
+      "loss": 3.1526,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7622298065984072,
+      "grad_norm": 0.7034804821014404,
+      "learning_rate": 0.0006484641638225256,
+      "loss": 1.6216,
+      "step": 1549
+    },
+    {
+      "epoch": 1.763367463026166,
+      "grad_norm": 1.182071328163147,
+      "learning_rate": 0.0006482366325369738,
+      "loss": 3.9126,
+      "step": 1550
+    },
+    {
+      "epoch": 1.764505119453925,
+      "grad_norm": 1.0914125442504883,
+      "learning_rate": 0.0006480091012514221,
+      "loss": 1.7731,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7656427758816837,
+      "grad_norm": 0.7337900996208191,
+      "learning_rate": 0.0006477815699658703,
+      "loss": 0.7843,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7667804323094427,
+      "grad_norm": 0.9463424682617188,
+      "learning_rate": 0.0006475540386803186,
+      "loss": 2.3052,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7679180887372015,
+      "grad_norm": 0.798876166343689,
+      "learning_rate": 0.0006473265073947669,
+      "loss": 1.7428,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7690557451649602,
+      "grad_norm": 1.0317094326019287,
+      "learning_rate": 0.000647098976109215,
+      "loss": 1.5478,
+      "step": 1555
+    },
+    {
+      "epoch": 1.770193401592719,
+      "grad_norm": 1.0436172485351562,
+      "learning_rate": 0.0006468714448236632,
+      "loss": 2.3285,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7713310580204777,
+      "grad_norm": 1.389779806137085,
+      "learning_rate": 0.0006466439135381115,
+      "loss": 3.1217,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7724687144482365,
+      "grad_norm": 0.9066593050956726,
+      "learning_rate": 0.0006464163822525597,
+      "loss": 1.5202,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7736063708759955,
+      "grad_norm": 0.7045938968658447,
+      "learning_rate": 0.0006461888509670079,
+      "loss": 2.0027,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7747440273037542,
+      "grad_norm": 1.222909688949585,
+      "learning_rate": 0.0006459613196814562,
+      "loss": 2.5047,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7758816837315132,
+      "grad_norm": 0.9253849983215332,
+      "learning_rate": 0.0006457337883959044,
+      "loss": 1.8513,
+      "step": 1561
+    },
+    {
+      "epoch": 1.777019340159272,
+      "grad_norm": 1.5743485689163208,
+      "learning_rate": 0.0006455062571103528,
+      "loss": 3.3089,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7781569965870307,
+      "grad_norm": 1.3808788061141968,
+      "learning_rate": 0.000645278725824801,
+      "loss": 2.1207,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7792946530147895,
+      "grad_norm": 0.5420240163803101,
+      "learning_rate": 0.0006450511945392492,
+      "loss": 0.9326,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7804323094425483,
+      "grad_norm": 1.006123661994934,
+      "learning_rate": 0.0006448236632536974,
+      "loss": 1.3682,
+      "step": 1565
+    },
+    {
+      "epoch": 1.781569965870307,
+      "grad_norm": 1.0188605785369873,
+      "learning_rate": 0.0006445961319681456,
+      "loss": 3.1302,
+      "step": 1566
+    },
+    {
+      "epoch": 1.782707622298066,
+      "grad_norm": 1.3520559072494507,
+      "learning_rate": 0.0006443686006825938,
+      "loss": 1.9661,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7838452787258248,
+      "grad_norm": 1.2583212852478027,
+      "learning_rate": 0.0006441410693970421,
+      "loss": 1.7267,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7849829351535837,
+      "grad_norm": 0.6001319885253906,
+      "learning_rate": 0.0006439135381114903,
+      "loss": 1.2674,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7861205915813425,
+      "grad_norm": 0.5909472703933716,
+      "learning_rate": 0.0006436860068259386,
+      "loss": 1.1398,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7872582480091013,
+      "grad_norm": 1.0126385688781738,
+      "learning_rate": 0.0006434584755403869,
+      "loss": 1.4477,
+      "step": 1571
+    },
+    {
+      "epoch": 1.78839590443686,
+      "grad_norm": 0.760152280330658,
+      "learning_rate": 0.0006432309442548351,
+      "loss": 1.0436,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7895335608646188,
+      "grad_norm": 1.2938870191574097,
+      "learning_rate": 0.0006430034129692833,
+      "loss": 2.4826,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7906712172923775,
+      "grad_norm": 0.8041395545005798,
+      "learning_rate": 0.0006427758816837316,
+      "loss": 1.7244,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7918088737201365,
+      "grad_norm": 0.9691146016120911,
+      "learning_rate": 0.0006425483503981797,
+      "loss": 1.7926,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7929465301478953,
+      "grad_norm": 1.1277107000350952,
+      "learning_rate": 0.0006423208191126279,
+      "loss": 2.1719,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7940841865756543,
+      "grad_norm": 0.8621721267700195,
+      "learning_rate": 0.0006420932878270762,
+      "loss": 2.341,
+      "step": 1577
+    },
+    {
+      "epoch": 1.795221843003413,
+      "grad_norm": 1.3702713251113892,
+      "learning_rate": 0.0006418657565415244,
+      "loss": 2.7296,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7963594994311718,
+      "grad_norm": 0.6883115768432617,
+      "learning_rate": 0.0006416382252559727,
+      "loss": 1.762,
+      "step": 1579
+    },
+    {
+      "epoch": 1.7974971558589306,
+      "grad_norm": 1.0266193151474,
+      "learning_rate": 0.000641410693970421,
+      "loss": 2.1009,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7986348122866893,
+      "grad_norm": 0.908869743347168,
+      "learning_rate": 0.0006411831626848692,
+      "loss": 1.7309,
+      "step": 1581
+    },
+    {
+      "epoch": 1.799772468714448,
+      "grad_norm": 0.8608745336532593,
+      "learning_rate": 0.0006409556313993174,
+      "loss": 1.7538,
+      "step": 1582
+    },
+    {
+      "epoch": 1.800910125142207,
+      "grad_norm": 1.5852781534194946,
+      "learning_rate": 0.0006407281001137657,
+      "loss": 2.7645,
+      "step": 1583
+    },
+    {
+      "epoch": 1.802047781569966,
+      "grad_norm": 0.8057767152786255,
+      "learning_rate": 0.0006405005688282139,
+      "loss": 1.7418,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8031854379977248,
+      "grad_norm": 1.129563570022583,
+      "learning_rate": 0.0006402730375426621,
+      "loss": 2.454,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8043230944254836,
+      "grad_norm": 0.9921088218688965,
+      "learning_rate": 0.0006400455062571103,
+      "loss": 1.8042,
+      "step": 1586
+    },
+    {
+      "epoch": 1.8054607508532423,
+      "grad_norm": 1.2487443685531616,
+      "learning_rate": 0.0006398179749715586,
+      "loss": 2.1645,
+      "step": 1587
+    },
+    {
+      "epoch": 1.806598407281001,
+      "grad_norm": 1.0499515533447266,
+      "learning_rate": 0.0006395904436860069,
+      "loss": 2.1733,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8077360637087598,
+      "grad_norm": 1.092044472694397,
+      "learning_rate": 0.0006393629124004551,
+      "loss": 2.5157,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8088737201365188,
+      "grad_norm": 0.9797663688659668,
+      "learning_rate": 0.0006391353811149033,
+      "loss": 1.3745,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8100113765642776,
+      "grad_norm": 1.0946043729782104,
+      "learning_rate": 0.0006389078498293516,
+      "loss": 2.1346,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8111490329920366,
+      "grad_norm": 1.3804898262023926,
+      "learning_rate": 0.0006386803185437998,
+      "loss": 2.088,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8122866894197953,
+      "grad_norm": 0.9063937067985535,
+      "learning_rate": 0.000638452787258248,
+      "loss": 2.0188,
+      "step": 1593
+    },
+    {
+      "epoch": 1.813424345847554,
+      "grad_norm": 0.7004676461219788,
+      "learning_rate": 0.0006382252559726962,
+      "loss": 1.5326,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8145620022753128,
+      "grad_norm": 0.7816855907440186,
+      "learning_rate": 0.0006379977246871445,
+      "loss": 1.5907,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8156996587030716,
+      "grad_norm": 1.2419103384017944,
+      "learning_rate": 0.0006377701934015927,
+      "loss": 2.683,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8168373151308304,
+      "grad_norm": 1.7206608057022095,
+      "learning_rate": 0.000637542662116041,
+      "loss": 3.0181,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8179749715585893,
+      "grad_norm": 0.7101848125457764,
+      "learning_rate": 0.0006373151308304892,
+      "loss": 2.2839,
+      "step": 1598
+    },
+    {
+      "epoch": 1.819112627986348,
+      "grad_norm": 0.9703013896942139,
+      "learning_rate": 0.0006370875995449374,
+      "loss": 1.6638,
+      "step": 1599
+    },
+    {
+      "epoch": 1.820250284414107,
+      "grad_norm": 0.7560202479362488,
+      "learning_rate": 0.0006368600682593857,
+      "loss": 1.285,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8213879408418658,
+      "grad_norm": 0.8459346294403076,
+      "learning_rate": 0.0006366325369738339,
+      "loss": 2.0187,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8225255972696246,
+      "grad_norm": 1.0976618528366089,
+      "learning_rate": 0.0006364050056882821,
+      "loss": 2.2865,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8236632536973834,
+      "grad_norm": 0.7999181747436523,
+      "learning_rate": 0.0006361774744027305,
+      "loss": 2.0923,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8248009101251421,
+      "grad_norm": 0.6172679662704468,
+      "learning_rate": 0.0006359499431171786,
+      "loss": 0.9479,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8259385665529009,
+      "grad_norm": 0.8358675241470337,
+      "learning_rate": 0.0006357224118316269,
+      "loss": 2.4008,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8270762229806599,
+      "grad_norm": 0.7997340559959412,
+      "learning_rate": 0.0006354948805460751,
+      "loss": 2.0383,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8282138794084186,
+      "grad_norm": 1.1405185461044312,
+      "learning_rate": 0.0006352673492605233,
+      "loss": 1.925,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8293515358361776,
+      "grad_norm": 0.7813712358474731,
+      "learning_rate": 0.0006350398179749716,
+      "loss": 1.7278,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8304891922639364,
+      "grad_norm": 1.3376038074493408,
+      "learning_rate": 0.0006348122866894198,
+      "loss": 2.3325,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8316268486916951,
+      "grad_norm": 1.1971509456634521,
+      "learning_rate": 0.000634584755403868,
+      "loss": 2.1967,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8327645051194539,
+      "grad_norm": 1.8342376947402954,
+      "learning_rate": 0.0006343572241183164,
+      "loss": 3.0377,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8339021615472126,
+      "grad_norm": 0.983214795589447,
+      "learning_rate": 0.0006341296928327646,
+      "loss": 1.5361,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8350398179749714,
+      "grad_norm": 0.7843421697616577,
+      "learning_rate": 0.0006339021615472128,
+      "loss": 1.4898,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8361774744027304,
+      "grad_norm": 0.8079589605331421,
+      "learning_rate": 0.000633674630261661,
+      "loss": 1.42,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8373151308304891,
+      "grad_norm": 0.9361008405685425,
+      "learning_rate": 0.0006334470989761092,
+      "loss": 2.3352,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8384527872582481,
+      "grad_norm": 0.8050186038017273,
+      "learning_rate": 0.0006332195676905574,
+      "loss": 1.2832,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8395904436860069,
+      "grad_norm": 0.6007594466209412,
+      "learning_rate": 0.0006329920364050057,
+      "loss": 0.7322,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8407281001137656,
+      "grad_norm": 1.09076988697052,
+      "learning_rate": 0.0006327645051194539,
+      "loss": 3.1086,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8418657565415244,
+      "grad_norm": 1.1623958349227905,
+      "learning_rate": 0.0006325369738339021,
+      "loss": 2.5158,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8430034129692832,
+      "grad_norm": 0.967048704624176,
+      "learning_rate": 0.0006323094425483505,
+      "loss": 1.5403,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8441410693970421,
+      "grad_norm": 0.7205291390419006,
+      "learning_rate": 0.0006320819112627987,
+      "loss": 0.9221,
+      "step": 1621
+    },
+    {
+      "epoch": 1.845278725824801,
+      "grad_norm": 0.8706921339035034,
+      "learning_rate": 0.0006318543799772469,
+      "loss": 2.1171,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8464163822525599,
+      "grad_norm": 0.8545218110084534,
+      "learning_rate": 0.0006316268486916951,
+      "loss": 1.3462,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8475540386803186,
+      "grad_norm": 1.0250085592269897,
+      "learning_rate": 0.0006313993174061433,
+      "loss": 1.8658,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8486916951080774,
+      "grad_norm": 0.8345952033996582,
+      "learning_rate": 0.0006311717861205916,
+      "loss": 2.4092,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8498293515358362,
+      "grad_norm": 0.9249033331871033,
+      "learning_rate": 0.0006309442548350398,
+      "loss": 1.7311,
+      "step": 1626
+    },
+    {
+      "epoch": 1.850967007963595,
+      "grad_norm": 1.0323681831359863,
+      "learning_rate": 0.000630716723549488,
+      "loss": 2.3136,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8521046643913537,
+      "grad_norm": 0.6797596216201782,
+      "learning_rate": 0.0006304891922639364,
+      "loss": 1.4115,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8532423208191127,
+      "grad_norm": 0.6315971612930298,
+      "learning_rate": 0.0006302616609783846,
+      "loss": 1.6303,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8543799772468714,
+      "grad_norm": 1.357260823249817,
+      "learning_rate": 0.0006300341296928328,
+      "loss": 3.4739,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8555176336746304,
+      "grad_norm": 0.8246727585792542,
+      "learning_rate": 0.0006298065984072811,
+      "loss": 1.4725,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8566552901023892,
+      "grad_norm": 1.0035685300827026,
+      "learning_rate": 0.0006295790671217293,
+      "loss": 1.9371,
+      "step": 1632
+    },
+    {
+      "epoch": 1.857792946530148,
+      "grad_norm": 0.8692108392715454,
+      "learning_rate": 0.0006293515358361774,
+      "loss": 1.5302,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8589306029579067,
+      "grad_norm": 0.7576583623886108,
+      "learning_rate": 0.0006291240045506257,
+      "loss": 1.3123,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8600682593856654,
+      "grad_norm": 1.0412949323654175,
+      "learning_rate": 0.0006288964732650739,
+      "loss": 1.8658,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8612059158134242,
+      "grad_norm": 0.6774725317955017,
+      "learning_rate": 0.0006286689419795221,
+      "loss": 1.499,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8623435722411832,
+      "grad_norm": 0.7042348980903625,
+      "learning_rate": 0.0006284414106939705,
+      "loss": 1.423,
+      "step": 1637
+    },
+    {
+      "epoch": 1.863481228668942,
+      "grad_norm": 0.8518065810203552,
+      "learning_rate": 0.0006282138794084187,
+      "loss": 1.1174,
+      "step": 1638
+    },
+    {
+      "epoch": 1.864618885096701,
+      "grad_norm": 1.1529499292373657,
+      "learning_rate": 0.0006279863481228669,
+      "loss": 2.3384,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8657565415244597,
+      "grad_norm": 0.8600414991378784,
+      "learning_rate": 0.0006277588168373152,
+      "loss": 1.7475,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8668941979522184,
+      "grad_norm": 1.1277893781661987,
+      "learning_rate": 0.0006275312855517634,
+      "loss": 3.2803,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8680318543799772,
+      "grad_norm": 1.1876293420791626,
+      "learning_rate": 0.0006273037542662116,
+      "loss": 2.062,
+      "step": 1642
+    },
+    {
+      "epoch": 1.869169510807736,
+      "grad_norm": 0.8083938360214233,
+      "learning_rate": 0.0006270762229806598,
+      "loss": 1.4945,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8703071672354947,
+      "grad_norm": 1.136566400527954,
+      "learning_rate": 0.000626848691695108,
+      "loss": 2.2583,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8714448236632537,
+      "grad_norm": 0.8216238617897034,
+      "learning_rate": 0.0006266211604095564,
+      "loss": 1.1567,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8725824800910125,
+      "grad_norm": 0.8726761341094971,
+      "learning_rate": 0.0006263936291240046,
+      "loss": 1.4604,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8737201365187715,
+      "grad_norm": 0.7124577164649963,
+      "learning_rate": 0.0006261660978384528,
+      "loss": 1.8203,
+      "step": 1647
+    },
+    {
+      "epoch": 1.8748577929465302,
+      "grad_norm": 0.7909445762634277,
+      "learning_rate": 0.0006259385665529011,
+      "loss": 1.5815,
+      "step": 1648
+    },
+    {
+      "epoch": 1.875995449374289,
+      "grad_norm": 1.0055829286575317,
+      "learning_rate": 0.0006257110352673493,
+      "loss": 1.5233,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8771331058020477,
+      "grad_norm": 0.6920607089996338,
+      "learning_rate": 0.0006254835039817975,
+      "loss": 0.8769,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8782707622298065,
+      "grad_norm": 0.8368102312088013,
+      "learning_rate": 0.0006252559726962458,
+      "loss": 1.9653,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8794084186575652,
+      "grad_norm": 1.1881768703460693,
+      "learning_rate": 0.000625028441410694,
+      "loss": 2.2672,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8805460750853242,
+      "grad_norm": 0.7409395575523376,
+      "learning_rate": 0.0006248009101251421,
+      "loss": 1.6343,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8816837315130832,
+      "grad_norm": 1.2841352224349976,
+      "learning_rate": 0.0006245733788395905,
+      "loss": 1.9493,
+      "step": 1654
+    },
+    {
+      "epoch": 1.882821387940842,
+      "grad_norm": 0.8569640517234802,
+      "learning_rate": 0.0006243458475540387,
+      "loss": 1.9457,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8839590443686007,
+      "grad_norm": 1.32857346534729,
+      "learning_rate": 0.0006241183162684869,
+      "loss": 1.7589,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8850967007963595,
+      "grad_norm": 0.8553183674812317,
+      "learning_rate": 0.0006238907849829352,
+      "loss": 1.2652,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8862343572241183,
+      "grad_norm": 0.9133054614067078,
+      "learning_rate": 0.0006236632536973834,
+      "loss": 1.5618,
+      "step": 1658
+    },
+    {
+      "epoch": 1.887372013651877,
+      "grad_norm": 1.1053630113601685,
+      "learning_rate": 0.0006234357224118316,
+      "loss": 1.4316,
+      "step": 1659
+    },
+    {
+      "epoch": 1.888509670079636,
+      "grad_norm": 1.3917018175125122,
+      "learning_rate": 0.0006232081911262799,
+      "loss": 3.7291,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8896473265073948,
+      "grad_norm": 0.6863465905189514,
+      "learning_rate": 0.0006229806598407281,
+      "loss": 1.2208,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8907849829351537,
+      "grad_norm": 1.6601629257202148,
+      "learning_rate": 0.0006227531285551762,
+      "loss": 3.3405,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8919226393629125,
+      "grad_norm": 0.6873295903205872,
+      "learning_rate": 0.0006225255972696246,
+      "loss": 1.66,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8930602957906713,
+      "grad_norm": 1.2501400709152222,
+      "learning_rate": 0.0006222980659840728,
+      "loss": 2.2344,
+      "step": 1664
+    },
+    {
+      "epoch": 1.89419795221843,
+      "grad_norm": 1.2724485397338867,
+      "learning_rate": 0.000622070534698521,
+      "loss": 1.8736,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8953356086461888,
+      "grad_norm": 1.0667139291763306,
+      "learning_rate": 0.0006218430034129693,
+      "loss": 3.2257,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8964732650739475,
+      "grad_norm": 0.7864385843276978,
+      "learning_rate": 0.0006216154721274175,
+      "loss": 1.8004,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8976109215017065,
+      "grad_norm": 0.8352164030075073,
+      "learning_rate": 0.0006213879408418658,
+      "loss": 1.4674,
+      "step": 1668
+    },
+    {
+      "epoch": 1.8987485779294653,
+      "grad_norm": 0.8850026726722717,
+      "learning_rate": 0.000621160409556314,
+      "loss": 1.7806,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8998862343572243,
+      "grad_norm": 0.7337632775306702,
+      "learning_rate": 0.0006209328782707623,
+      "loss": 1.2096,
+      "step": 1670
+    },
+    {
+      "epoch": 1.901023890784983,
+      "grad_norm": 0.9589748978614807,
+      "learning_rate": 0.0006207053469852106,
+      "loss": 2.1104,
+      "step": 1671
+    },
+    {
+      "epoch": 1.9021615472127418,
+      "grad_norm": 1.047741174697876,
+      "learning_rate": 0.0006204778156996587,
+      "loss": 2.5989,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9032992036405005,
+      "grad_norm": 1.0804181098937988,
+      "learning_rate": 0.0006202502844141069,
+      "loss": 1.5729,
+      "step": 1673
+    },
+    {
+      "epoch": 1.9044368600682593,
+      "grad_norm": 0.99363774061203,
+      "learning_rate": 0.0006200227531285552,
+      "loss": 2.5205,
+      "step": 1674
+    },
+    {
+      "epoch": 1.905574516496018,
+      "grad_norm": 1.1013110876083374,
+      "learning_rate": 0.0006197952218430034,
+      "loss": 1.4929,
+      "step": 1675
+    },
+    {
+      "epoch": 1.906712172923777,
+      "grad_norm": 1.1436536312103271,
+      "learning_rate": 0.0006195676905574516,
+      "loss": 1.503,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9078498293515358,
+      "grad_norm": 0.9888268113136292,
+      "learning_rate": 0.0006193401592718999,
+      "loss": 1.8105,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9089874857792948,
+      "grad_norm": 0.5987884998321533,
+      "learning_rate": 0.0006191126279863481,
+      "loss": 0.6153,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9101251422070535,
+      "grad_norm": 1.2569278478622437,
+      "learning_rate": 0.0006188850967007964,
+      "loss": 2.58,
+      "step": 1679
+    },
+    {
+      "epoch": 1.9112627986348123,
+      "grad_norm": 1.7216945886611938,
+      "learning_rate": 0.0006186575654152447,
+      "loss": 3.3363,
+      "step": 1680
+    },
+    {
+      "epoch": 1.912400455062571,
+      "grad_norm": 0.8490168452262878,
+      "learning_rate": 0.0006184300341296929,
+      "loss": 1.6139,
+      "step": 1681
+    },
+    {
+      "epoch": 1.9135381114903298,
+      "grad_norm": 0.8663358688354492,
+      "learning_rate": 0.000618202502844141,
+      "loss": 1.0861,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9146757679180886,
+      "grad_norm": 1.2262353897094727,
+      "learning_rate": 0.0006179749715585893,
+      "loss": 2.4656,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9158134243458476,
+      "grad_norm": 0.8106787800788879,
+      "learning_rate": 0.0006177474402730375,
+      "loss": 0.7479,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9169510807736063,
+      "grad_norm": 0.8843865394592285,
+      "learning_rate": 0.0006175199089874857,
+      "loss": 1.8057,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9180887372013653,
+      "grad_norm": 1.0910826921463013,
+      "learning_rate": 0.000617292377701934,
+      "loss": 2.1269,
+      "step": 1686
+    },
+    {
+      "epoch": 1.919226393629124,
+      "grad_norm": 0.8909353613853455,
+      "learning_rate": 0.0006170648464163823,
+      "loss": 1.4374,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9203640500568828,
+      "grad_norm": 0.838852047920227,
+      "learning_rate": 0.0006168373151308306,
+      "loss": 1.455,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9215017064846416,
+      "grad_norm": 1.186137080192566,
+      "learning_rate": 0.0006166097838452788,
+      "loss": 2.9913,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9226393629124003,
+      "grad_norm": 0.947559654712677,
+      "learning_rate": 0.000616382252559727,
+      "loss": 2.5864,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9237770193401593,
+      "grad_norm": 1.1154701709747314,
+      "learning_rate": 0.0006161547212741753,
+      "loss": 1.7531,
+      "step": 1691
+    },
+    {
+      "epoch": 1.924914675767918,
+      "grad_norm": 0.9917047023773193,
+      "learning_rate": 0.0006159271899886234,
+      "loss": 1.847,
+      "step": 1692
+    },
+    {
+      "epoch": 1.926052332195677,
+      "grad_norm": 0.8470612168312073,
+      "learning_rate": 0.0006156996587030716,
+      "loss": 1.9226,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9271899886234358,
+      "grad_norm": 1.0250667333602905,
+      "learning_rate": 0.0006154721274175199,
+      "loss": 1.3656,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9283276450511946,
+      "grad_norm": 1.122154951095581,
+      "learning_rate": 0.0006152445961319682,
+      "loss": 1.7061,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9294653014789533,
+      "grad_norm": 0.7498093843460083,
+      "learning_rate": 0.0006150170648464164,
+      "loss": 1.86,
+      "step": 1696
+    },
+    {
+      "epoch": 1.930602957906712,
+      "grad_norm": 1.526132583618164,
+      "learning_rate": 0.0006147895335608647,
+      "loss": 1.9259,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9317406143344709,
+      "grad_norm": 0.7942246794700623,
+      "learning_rate": 0.0006145620022753129,
+      "loss": 2.1063,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9328782707622298,
+      "grad_norm": 0.7092151045799255,
+      "learning_rate": 0.0006143344709897611,
+      "loss": 1.3752,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9340159271899886,
+      "grad_norm": 1.0814862251281738,
+      "learning_rate": 0.0006141069397042094,
+      "loss": 1.3951,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9351535836177476,
+      "grad_norm": 0.8918007016181946,
+      "learning_rate": 0.0006138794084186575,
+      "loss": 1.4203,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9362912400455063,
+      "grad_norm": 1.2497609853744507,
+      "learning_rate": 0.0006136518771331057,
+      "loss": 2.2451,
+      "step": 1702
+    },
+    {
+      "epoch": 1.937428896473265,
+      "grad_norm": 1.0966827869415283,
+      "learning_rate": 0.000613424345847554,
+      "loss": 2.617,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9385665529010239,
+      "grad_norm": 0.6311344504356384,
+      "learning_rate": 0.0006131968145620023,
+      "loss": 0.9732,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9397042093287826,
+      "grad_norm": 0.7735137939453125,
+      "learning_rate": 0.0006129692832764505,
+      "loss": 1.6126,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9408418657565414,
+      "grad_norm": 0.8205373287200928,
+      "learning_rate": 0.0006127417519908988,
+      "loss": 0.8851,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9419795221843004,
+      "grad_norm": 1.8471014499664307,
+      "learning_rate": 0.000612514220705347,
+      "loss": 3.0294,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9431171786120591,
+      "grad_norm": 0.8841797113418579,
+      "learning_rate": 0.0006122866894197953,
+      "loss": 2.4339,
+      "step": 1708
+    },
+    {
+      "epoch": 1.944254835039818,
+      "grad_norm": 1.112204670906067,
+      "learning_rate": 0.0006120591581342435,
+      "loss": 3.4836,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9453924914675769,
+      "grad_norm": 0.7821568250656128,
+      "learning_rate": 0.0006118316268486917,
+      "loss": 1.1741,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9465301478953356,
+      "grad_norm": 0.9497551918029785,
+      "learning_rate": 0.0006116040955631399,
+      "loss": 1.5836,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9476678043230944,
+      "grad_norm": 0.9662843942642212,
+      "learning_rate": 0.0006113765642775882,
+      "loss": 2.4538,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9488054607508531,
+      "grad_norm": 0.8117369413375854,
+      "learning_rate": 0.0006111490329920364,
+      "loss": 2.783,
+      "step": 1713
+    },
+    {
+      "epoch": 1.949943117178612,
+      "grad_norm": 0.7812523245811462,
+      "learning_rate": 0.0006109215017064847,
+      "loss": 1.5216,
+      "step": 1714
+    },
+    {
+      "epoch": 1.9510807736063709,
+      "grad_norm": 0.7647354006767273,
+      "learning_rate": 0.0006106939704209329,
+      "loss": 1.8356,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9522184300341296,
+      "grad_norm": 1.348557949066162,
+      "learning_rate": 0.0006104664391353811,
+      "loss": 2.3718,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9533560864618886,
+      "grad_norm": 0.9883414506912231,
+      "learning_rate": 0.0006102389078498294,
+      "loss": 1.9512,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9544937428896474,
+      "grad_norm": 1.112888216972351,
+      "learning_rate": 0.0006100113765642776,
+      "loss": 1.9945,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9556313993174061,
+      "grad_norm": 1.1321780681610107,
+      "learning_rate": 0.0006097838452787258,
+      "loss": 1.7481,
+      "step": 1719
+    },
+    {
+      "epoch": 1.956769055745165,
+      "grad_norm": 1.408277153968811,
+      "learning_rate": 0.0006095563139931742,
+      "loss": 2.5749,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9579067121729237,
+      "grad_norm": 0.9839853644371033,
+      "learning_rate": 0.0006093287827076223,
+      "loss": 1.2224,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9590443686006824,
+      "grad_norm": 0.9178591966629028,
+      "learning_rate": 0.0006091012514220705,
+      "loss": 1.5087,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9601820250284414,
+      "grad_norm": 1.2959963083267212,
+      "learning_rate": 0.0006088737201365188,
+      "loss": 3.2079,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9613196814562004,
+      "grad_norm": 0.9269609451293945,
+      "learning_rate": 0.000608646188850967,
+      "loss": 1.5689,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9624573378839592,
+      "grad_norm": 0.8281601071357727,
+      "learning_rate": 0.0006084186575654152,
+      "loss": 1.2933,
+      "step": 1725
+    },
+    {
+      "epoch": 1.963594994311718,
+      "grad_norm": 1.0037777423858643,
+      "learning_rate": 0.0006081911262798635,
+      "loss": 1.7339,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9647326507394767,
+      "grad_norm": 1.0925790071487427,
+      "learning_rate": 0.0006079635949943117,
+      "loss": 2.8919,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9658703071672354,
+      "grad_norm": 0.9971732497215271,
+      "learning_rate": 0.00060773606370876,
+      "loss": 2.1831,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9670079635949942,
+      "grad_norm": 1.161362648010254,
+      "learning_rate": 0.0006075085324232083,
+      "loss": 3.2354,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9681456200227532,
+      "grad_norm": 1.0122507810592651,
+      "learning_rate": 0.0006072810011376564,
+      "loss": 2.0288,
+      "step": 1730
+    },
+    {
+      "epoch": 1.969283276450512,
+      "grad_norm": 0.47560903429985046,
+      "learning_rate": 0.0006070534698521047,
+      "loss": 0.5176,
+      "step": 1731
+    },
+    {
+      "epoch": 1.970420932878271,
+      "grad_norm": 0.6199741363525391,
+      "learning_rate": 0.0006068259385665529,
+      "loss": 1.2013,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9715585893060297,
+      "grad_norm": 1.1044777631759644,
+      "learning_rate": 0.0006065984072810011,
+      "loss": 1.7702,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9726962457337884,
+      "grad_norm": 0.9448471069335938,
+      "learning_rate": 0.0006063708759954494,
+      "loss": 1.9484,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9738339021615472,
+      "grad_norm": 0.9100543856620789,
+      "learning_rate": 0.0006061433447098976,
+      "loss": 2.0936,
+      "step": 1735
+    },
+    {
+      "epoch": 1.974971558589306,
+      "grad_norm": 0.6803526282310486,
+      "learning_rate": 0.0006059158134243458,
+      "loss": 1.1829,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9761092150170647,
+      "grad_norm": 1.1807910203933716,
+      "learning_rate": 0.0006056882821387942,
+      "loss": 2.4637,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9772468714448237,
+      "grad_norm": 0.9801465272903442,
+      "learning_rate": 0.0006054607508532424,
+      "loss": 1.9998,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9783845278725825,
+      "grad_norm": 0.987866222858429,
+      "learning_rate": 0.0006052332195676906,
+      "loss": 1.2759,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9795221843003414,
+      "grad_norm": 1.0122973918914795,
+      "learning_rate": 0.0006050056882821388,
+      "loss": 0.9901,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9806598407281002,
+      "grad_norm": 0.9722836017608643,
+      "learning_rate": 0.000604778156996587,
+      "loss": 2.178,
+      "step": 1741
+    },
+    {
+      "epoch": 1.981797497155859,
+      "grad_norm": 0.9257890582084656,
+      "learning_rate": 0.0006045506257110352,
+      "loss": 2.1119,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9829351535836177,
+      "grad_norm": 0.6407757997512817,
+      "learning_rate": 0.0006043230944254835,
+      "loss": 1.2141,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9840728100113765,
+      "grad_norm": 0.7565338015556335,
+      "learning_rate": 0.0006040955631399317,
+      "loss": 1.7172,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9852104664391352,
+      "grad_norm": 0.7070271372795105,
+      "learning_rate": 0.0006038680318543799,
+      "loss": 0.856,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9863481228668942,
+      "grad_norm": 1.3683280944824219,
+      "learning_rate": 0.0006036405005688283,
+      "loss": 3.7194,
+      "step": 1746
+    },
+    {
+      "epoch": 1.987485779294653,
+      "grad_norm": 0.8713628649711609,
+      "learning_rate": 0.0006034129692832765,
+      "loss": 2.4061,
+      "step": 1747
+    },
+    {
+      "epoch": 1.988623435722412,
+      "grad_norm": 1.6695564985275269,
+      "learning_rate": 0.0006031854379977248,
+      "loss": 3.2265,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9897610921501707,
+      "grad_norm": 1.007358431816101,
+      "learning_rate": 0.000602957906712173,
+      "loss": 2.1454,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9908987485779295,
+      "grad_norm": 1.256252408027649,
+      "learning_rate": 0.0006027303754266211,
+      "loss": 2.2461,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9920364050056882,
+      "grad_norm": 1.4839099645614624,
+      "learning_rate": 0.0006025028441410694,
+      "loss": 3.5124,
+      "step": 1751
+    },
+    {
+      "epoch": 1.993174061433447,
+      "grad_norm": 0.8481181263923645,
+      "learning_rate": 0.0006022753128555176,
+      "loss": 1.5822,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9943117178612058,
+      "grad_norm": 1.6936454772949219,
+      "learning_rate": 0.0006020477815699658,
+      "loss": 2.7639,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9954493742889647,
+      "grad_norm": 1.1564828157424927,
+      "learning_rate": 0.0006018202502844142,
+      "loss": 2.4716,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9965870307167235,
+      "grad_norm": 1.2504764795303345,
+      "learning_rate": 0.0006015927189988624,
+      "loss": 1.9542,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9977246871444825,
+      "grad_norm": 0.9221636652946472,
+      "learning_rate": 0.0006013651877133106,
+      "loss": 2.1488,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9988623435722412,
+      "grad_norm": 1.1251505613327026,
+      "learning_rate": 0.0006011376564277589,
+      "loss": 1.9403,
+      "step": 1757
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7302814722061157,
+      "learning_rate": 0.0006009101251422071,
+      "loss": 1.4557,
+      "step": 1758
+    },
+    {
+      "epoch": 2.0,
+      "eval_f1": 0.8901,
+      "eval_gen_len": 49.6091,
+      "eval_loss": 1.861061930656433,
+      "eval_precision": 0.8889,
+      "eval_recall": 0.8915,
+      "eval_rouge1": 0.4491,
+      "eval_rouge2": 0.2031,
+      "eval_rougeL": 0.3721,
+      "eval_rougeLsum": 0.4148,
+      "eval_runtime": 28.0457,
+      "eval_samples_per_second": 3.922,
+      "eval_steps_per_second": 0.499,
+      "step": 1758
+    },
+    {
+      "epoch": 2.0011376564277588,
+      "grad_norm": 0.7474573850631714,
+      "learning_rate": 0.0006006825938566553,
+      "loss": 0.9735,
+      "step": 1759
+    },
+    {
+      "epoch": 2.0022753128555175,
+      "grad_norm": 1.003105640411377,
+      "learning_rate": 0.0006004550625711035,
+      "loss": 1.6829,
+      "step": 1760
+    },
+    {
+      "epoch": 2.0034129692832763,
+      "grad_norm": 0.7681793570518494,
+      "learning_rate": 0.0006002275312855517,
+      "loss": 1.6531,
+      "step": 1761
+    },
+    {
+      "epoch": 2.0045506257110355,
+      "grad_norm": 1.0925558805465698,
+      "learning_rate": 0.0006,
+      "loss": 2.754,
+      "step": 1762
+    },
+    {
+      "epoch": 2.0056882821387942,
+      "grad_norm": 0.8437767028808594,
+      "learning_rate": 0.0005997724687144483,
+      "loss": 1.3487,
+      "step": 1763
+    },
+    {
+      "epoch": 2.006825938566553,
+      "grad_norm": 0.6217951774597168,
+      "learning_rate": 0.0005995449374288965,
+      "loss": 1.3183,
+      "step": 1764
+    },
+    {
+      "epoch": 2.0079635949943118,
+      "grad_norm": 0.9381899237632751,
+      "learning_rate": 0.0005993174061433447,
+      "loss": 1.5027,
+      "step": 1765
+    },
+    {
+      "epoch": 2.0091012514220705,
+      "grad_norm": 0.801108181476593,
+      "learning_rate": 0.000599089874857793,
+      "loss": 1.207,
+      "step": 1766
+    },
+    {
+      "epoch": 2.0102389078498293,
+      "grad_norm": 0.9824976325035095,
+      "learning_rate": 0.0005988623435722412,
+      "loss": 2.4301,
+      "step": 1767
+    },
+    {
+      "epoch": 2.011376564277588,
+      "grad_norm": 0.7496545910835266,
+      "learning_rate": 0.0005986348122866895,
+      "loss": 1.25,
+      "step": 1768
+    },
+    {
+      "epoch": 2.012514220705347,
+      "grad_norm": 0.8144867420196533,
+      "learning_rate": 0.0005984072810011376,
+      "loss": 1.9912,
+      "step": 1769
+    },
+    {
+      "epoch": 2.013651877133106,
+      "grad_norm": 1.2619274854660034,
+      "learning_rate": 0.0005981797497155858,
+      "loss": 2.0556,
+      "step": 1770
+    },
+    {
+      "epoch": 2.0147895335608648,
+      "grad_norm": 0.8984837532043457,
+      "learning_rate": 0.0005979522184300342,
+      "loss": 1.3548,
+      "step": 1771
+    },
+    {
+      "epoch": 2.0159271899886235,
+      "grad_norm": 0.9129440784454346,
+      "learning_rate": 0.0005977246871444824,
+      "loss": 2.3292,
+      "step": 1772
+    },
+    {
+      "epoch": 2.0170648464163823,
+      "grad_norm": 0.7606117129325867,
+      "learning_rate": 0.0005974971558589306,
+      "loss": 1.3862,
+      "step": 1773
+    },
+    {
+      "epoch": 2.018202502844141,
+      "grad_norm": 1.030834674835205,
+      "learning_rate": 0.0005972696245733789,
+      "loss": 1.4272,
+      "step": 1774
+    },
+    {
+      "epoch": 2.0193401592719,
+      "grad_norm": 0.6545241475105286,
+      "learning_rate": 0.0005970420932878271,
+      "loss": 1.1461,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0204778156996586,
+      "grad_norm": 0.8098762631416321,
+      "learning_rate": 0.0005968145620022753,
+      "loss": 1.6309,
+      "step": 1776
+    },
+    {
+      "epoch": 2.0216154721274173,
+      "grad_norm": 1.2966359853744507,
+      "learning_rate": 0.0005965870307167236,
+      "loss": 2.2563,
+      "step": 1777
+    },
+    {
+      "epoch": 2.0227531285551765,
+      "grad_norm": 1.0538915395736694,
+      "learning_rate": 0.0005963594994311718,
+      "loss": 2.8999,
+      "step": 1778
+    },
+    {
+      "epoch": 2.0238907849829353,
+      "grad_norm": 0.788987934589386,
+      "learning_rate": 0.00059613196814562,
+      "loss": 1.5423,
+      "step": 1779
+    },
+    {
+      "epoch": 2.025028441410694,
+      "grad_norm": 0.9566778540611267,
+      "learning_rate": 0.0005959044368600683,
+      "loss": 2.4709,
+      "step": 1780
+    },
+    {
+      "epoch": 2.026166097838453,
+      "grad_norm": 1.1112850904464722,
+      "learning_rate": 0.0005956769055745165,
+      "loss": 2.3085,
+      "step": 1781
+    },
+    {
+      "epoch": 2.0273037542662116,
+      "grad_norm": 1.074602484703064,
+      "learning_rate": 0.0005954493742889647,
+      "loss": 1.9328,
+      "step": 1782
+    },
+    {
+      "epoch": 2.0284414106939703,
+      "grad_norm": 1.017351508140564,
+      "learning_rate": 0.000595221843003413,
+      "loss": 2.1152,
+      "step": 1783
+    },
+    {
+      "epoch": 2.029579067121729,
+      "grad_norm": 1.069661021232605,
+      "learning_rate": 0.0005949943117178612,
+      "loss": 1.8133,
+      "step": 1784
+    },
+    {
+      "epoch": 2.030716723549488,
+      "grad_norm": 1.099387764930725,
+      "learning_rate": 0.0005947667804323094,
+      "loss": 2.207,
+      "step": 1785
+    },
+    {
+      "epoch": 2.031854379977247,
+      "grad_norm": 1.02996027469635,
+      "learning_rate": 0.0005945392491467577,
+      "loss": 1.9207,
+      "step": 1786
+    },
+    {
+      "epoch": 2.032992036405006,
+      "grad_norm": 0.7691861391067505,
+      "learning_rate": 0.000594311717861206,
+      "loss": 1.2884,
+      "step": 1787
+    },
+    {
+      "epoch": 2.0341296928327646,
+      "grad_norm": 0.9716812968254089,
+      "learning_rate": 0.0005940841865756542,
+      "loss": 1.4235,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0352673492605233,
+      "grad_norm": 0.9133804440498352,
+      "learning_rate": 0.0005938566552901024,
+      "loss": 1.654,
+      "step": 1789
+    },
+    {
+      "epoch": 2.036405005688282,
+      "grad_norm": 1.1487979888916016,
+      "learning_rate": 0.0005936291240045506,
+      "loss": 2.4215,
+      "step": 1790
+    },
+    {
+      "epoch": 2.037542662116041,
+      "grad_norm": 0.8201186060905457,
+      "learning_rate": 0.0005934015927189989,
+      "loss": 1.775,
+      "step": 1791
+    },
+    {
+      "epoch": 2.0386803185437996,
+      "grad_norm": 1.0376436710357666,
+      "learning_rate": 0.0005931740614334471,
+      "loss": 2.3957,
+      "step": 1792
+    },
+    {
+      "epoch": 2.039817974971559,
+      "grad_norm": 0.9257051944732666,
+      "learning_rate": 0.0005929465301478953,
+      "loss": 2.02,
+      "step": 1793
+    },
+    {
+      "epoch": 2.0409556313993176,
+      "grad_norm": 1.158453106880188,
+      "learning_rate": 0.0005927189988623436,
+      "loss": 2.3353,
+      "step": 1794
+    },
+    {
+      "epoch": 2.0420932878270763,
+      "grad_norm": 0.9140053987503052,
+      "learning_rate": 0.0005924914675767918,
+      "loss": 1.6694,
+      "step": 1795
+    },
+    {
+      "epoch": 2.043230944254835,
+      "grad_norm": 0.8473706245422363,
+      "learning_rate": 0.0005922639362912401,
+      "loss": 1.4332,
+      "step": 1796
+    },
+    {
+      "epoch": 2.044368600682594,
+      "grad_norm": 1.5226987600326538,
+      "learning_rate": 0.0005920364050056884,
+      "loss": 2.6597,
+      "step": 1797
+    },
+    {
+      "epoch": 2.0455062571103526,
+      "grad_norm": 1.0220167636871338,
+      "learning_rate": 0.0005918088737201365,
+      "loss": 1.9809,
+      "step": 1798
+    },
+    {
+      "epoch": 2.0466439135381114,
+      "grad_norm": 0.8180732727050781,
+      "learning_rate": 0.0005915813424345847,
+      "loss": 1.5031,
+      "step": 1799
+    },
+    {
+      "epoch": 2.04778156996587,
+      "grad_norm": 1.1611244678497314,
+      "learning_rate": 0.000591353811149033,
+      "loss": 1.2286,
+      "step": 1800
+    },
+    {
+      "epoch": 2.0489192263936293,
+      "grad_norm": 1.1958653926849365,
+      "learning_rate": 0.0005911262798634812,
+      "loss": 2.6489,
+      "step": 1801
+    },
+    {
+      "epoch": 2.050056882821388,
+      "grad_norm": 1.252683401107788,
+      "learning_rate": 0.0005908987485779294,
+      "loss": 2.6139,
+      "step": 1802
+    },
+    {
+      "epoch": 2.051194539249147,
+      "grad_norm": 1.3542306423187256,
+      "learning_rate": 0.0005906712172923777,
+      "loss": 2.6136,
+      "step": 1803
+    },
+    {
+      "epoch": 2.0523321956769056,
+      "grad_norm": 1.3260592222213745,
+      "learning_rate": 0.000590443686006826,
+      "loss": 3.5016,
+      "step": 1804
+    },
+    {
+      "epoch": 2.0534698521046644,
+      "grad_norm": 1.2958813905715942,
+      "learning_rate": 0.0005902161547212742,
+      "loss": 3.2267,
+      "step": 1805
+    },
+    {
+      "epoch": 2.054607508532423,
+      "grad_norm": 0.963817834854126,
+      "learning_rate": 0.0005899886234357225,
+      "loss": 2.2298,
+      "step": 1806
+    },
+    {
+      "epoch": 2.055745164960182,
+      "grad_norm": 0.8188225626945496,
+      "learning_rate": 0.0005897610921501707,
+      "loss": 1.3832,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0568828213879407,
+      "grad_norm": 1.3529433012008667,
+      "learning_rate": 0.0005895335608646188,
+      "loss": 1.5602,
+      "step": 1808
+    },
+    {
+      "epoch": 2.0580204778157,
+      "grad_norm": 1.1197400093078613,
+      "learning_rate": 0.0005893060295790671,
+      "loss": 2.5839,
+      "step": 1809
+    },
+    {
+      "epoch": 2.0591581342434586,
+      "grad_norm": 0.9456532001495361,
+      "learning_rate": 0.0005890784982935153,
+      "loss": 2.6459,
+      "step": 1810
+    },
+    {
+      "epoch": 2.0602957906712174,
+      "grad_norm": 0.923581063747406,
+      "learning_rate": 0.0005888509670079636,
+      "loss": 1.9121,
+      "step": 1811
+    },
+    {
+      "epoch": 2.061433447098976,
+      "grad_norm": 1.202931523323059,
+      "learning_rate": 0.0005886234357224119,
+      "loss": 1.8609,
+      "step": 1812
+    },
+    {
+      "epoch": 2.062571103526735,
+      "grad_norm": 1.0954371690750122,
+      "learning_rate": 0.0005883959044368601,
+      "loss": 1.6039,
+      "step": 1813
+    },
+    {
+      "epoch": 2.0637087599544937,
+      "grad_norm": 0.8433867692947388,
+      "learning_rate": 0.0005881683731513084,
+      "loss": 1.2853,
+      "step": 1814
+    },
+    {
+      "epoch": 2.0648464163822524,
+      "grad_norm": 0.7211450934410095,
+      "learning_rate": 0.0005879408418657566,
+      "loss": 0.9241,
+      "step": 1815
+    },
+    {
+      "epoch": 2.065984072810011,
+      "grad_norm": 1.0169728994369507,
+      "learning_rate": 0.0005877133105802048,
+      "loss": 3.0501,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0671217292377704,
+      "grad_norm": 1.1246399879455566,
+      "learning_rate": 0.0005874857792946531,
+      "loss": 2.3138,
+      "step": 1817
+    },
+    {
+      "epoch": 2.068259385665529,
+      "grad_norm": 0.945504367351532,
+      "learning_rate": 0.0005872582480091012,
+      "loss": 1.3534,
+      "step": 1818
+    },
+    {
+      "epoch": 2.069397042093288,
+      "grad_norm": 0.8048107624053955,
+      "learning_rate": 0.0005870307167235494,
+      "loss": 2.2219,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0705346985210467,
+      "grad_norm": 0.7122802734375,
+      "learning_rate": 0.0005868031854379977,
+      "loss": 1.239,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0716723549488054,
+      "grad_norm": 0.5502296686172485,
+      "learning_rate": 0.000586575654152446,
+      "loss": 1.1367,
+      "step": 1821
+    },
+    {
+      "epoch": 2.072810011376564,
+      "grad_norm": 0.9720338582992554,
+      "learning_rate": 0.0005863481228668942,
+      "loss": 2.4818,
+      "step": 1822
+    },
+    {
+      "epoch": 2.073947667804323,
+      "grad_norm": 1.1907950639724731,
+      "learning_rate": 0.0005861205915813425,
+      "loss": 3.3499,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0750853242320817,
+      "grad_norm": 0.7875828146934509,
+      "learning_rate": 0.0005858930602957907,
+      "loss": 1.2593,
+      "step": 1824
+    },
+    {
+      "epoch": 2.076222980659841,
+      "grad_norm": 1.0819391012191772,
+      "learning_rate": 0.0005856655290102389,
+      "loss": 2.0526,
+      "step": 1825
+    },
+    {
+      "epoch": 2.0773606370875997,
+      "grad_norm": 0.7829857468605042,
+      "learning_rate": 0.0005854379977246872,
+      "loss": 1.7224,
+      "step": 1826
+    },
+    {
+      "epoch": 2.0784982935153584,
+      "grad_norm": 0.6612402200698853,
+      "learning_rate": 0.0005852104664391354,
+      "loss": 1.3579,
+      "step": 1827
+    },
+    {
+      "epoch": 2.079635949943117,
+      "grad_norm": 0.6340951323509216,
+      "learning_rate": 0.0005849829351535835,
+      "loss": 1.2102,
+      "step": 1828
+    },
+    {
+      "epoch": 2.080773606370876,
+      "grad_norm": 0.6889867186546326,
+      "learning_rate": 0.0005847554038680319,
+      "loss": 1.0524,
+      "step": 1829
+    },
+    {
+      "epoch": 2.0819112627986347,
+      "grad_norm": 1.2241733074188232,
+      "learning_rate": 0.0005845278725824801,
+      "loss": 1.7329,
+      "step": 1830
+    },
+    {
+      "epoch": 2.0830489192263935,
+      "grad_norm": 0.9656223654747009,
+      "learning_rate": 0.0005843003412969284,
+      "loss": 1.5006,
+      "step": 1831
+    },
+    {
+      "epoch": 2.0841865756541527,
+      "grad_norm": 0.9914301037788391,
+      "learning_rate": 0.0005840728100113766,
+      "loss": 1.72,
+      "step": 1832
+    },
+    {
+      "epoch": 2.0853242320819114,
+      "grad_norm": 0.6369336843490601,
+      "learning_rate": 0.0005838452787258248,
+      "loss": 1.4671,
+      "step": 1833
+    },
+    {
+      "epoch": 2.08646188850967,
+      "grad_norm": 1.0386950969696045,
+      "learning_rate": 0.0005836177474402731,
+      "loss": 1.8412,
+      "step": 1834
+    },
+    {
+      "epoch": 2.087599544937429,
+      "grad_norm": 0.9129205346107483,
+      "learning_rate": 0.0005833902161547213,
+      "loss": 1.8206,
+      "step": 1835
+    },
+    {
+      "epoch": 2.0887372013651877,
+      "grad_norm": 0.8238604664802551,
+      "learning_rate": 0.0005831626848691695,
+      "loss": 1.8884,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0898748577929465,
+      "grad_norm": 1.2888492345809937,
+      "learning_rate": 0.0005829351535836177,
+      "loss": 2.121,
+      "step": 1837
+    },
+    {
+      "epoch": 2.091012514220705,
+      "grad_norm": 1.0287847518920898,
+      "learning_rate": 0.000582707622298066,
+      "loss": 1.878,
+      "step": 1838
+    },
+    {
+      "epoch": 2.092150170648464,
+      "grad_norm": 1.340661883354187,
+      "learning_rate": 0.0005824800910125142,
+      "loss": 2.5264,
+      "step": 1839
+    },
+    {
+      "epoch": 2.093287827076223,
+      "grad_norm": 0.8161119818687439,
+      "learning_rate": 0.0005822525597269625,
+      "loss": 1.8105,
+      "step": 1840
+    },
+    {
+      "epoch": 2.094425483503982,
+      "grad_norm": 1.0750125646591187,
+      "learning_rate": 0.0005820250284414107,
+      "loss": 2.3553,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0955631399317407,
+      "grad_norm": 1.0218100547790527,
+      "learning_rate": 0.0005817974971558589,
+      "loss": 1.8317,
+      "step": 1842
+    },
+    {
+      "epoch": 2.0967007963594995,
+      "grad_norm": 0.5259703397750854,
+      "learning_rate": 0.0005815699658703072,
+      "loss": 0.7726,
+      "step": 1843
+    },
+    {
+      "epoch": 2.0978384527872582,
+      "grad_norm": 0.9108942151069641,
+      "learning_rate": 0.0005813424345847554,
+      "loss": 1.7108,
+      "step": 1844
+    },
+    {
+      "epoch": 2.098976109215017,
+      "grad_norm": 1.2494343519210815,
+      "learning_rate": 0.0005811149032992036,
+      "loss": 2.0056,
+      "step": 1845
+    },
+    {
+      "epoch": 2.1001137656427757,
+      "grad_norm": 0.6544501185417175,
+      "learning_rate": 0.000580887372013652,
+      "loss": 1.1255,
+      "step": 1846
+    },
+    {
+      "epoch": 2.1012514220705345,
+      "grad_norm": 1.4156732559204102,
+      "learning_rate": 0.0005806598407281001,
+      "loss": 2.7314,
+      "step": 1847
+    },
+    {
+      "epoch": 2.1023890784982937,
+      "grad_norm": 0.6912670731544495,
+      "learning_rate": 0.0005804323094425483,
+      "loss": 1.3614,
+      "step": 1848
+    },
+    {
+      "epoch": 2.1035267349260525,
+      "grad_norm": 0.7118052840232849,
+      "learning_rate": 0.0005802047781569966,
+      "loss": 1.9744,
+      "step": 1849
+    },
+    {
+      "epoch": 2.1046643913538112,
+      "grad_norm": 0.9591397047042847,
+      "learning_rate": 0.0005799772468714448,
+      "loss": 0.8785,
+      "step": 1850
+    },
+    {
+      "epoch": 2.10580204778157,
+      "grad_norm": 0.7862758040428162,
+      "learning_rate": 0.0005797497155858931,
+      "loss": 1.5104,
+      "step": 1851
+    },
+    {
+      "epoch": 2.1069397042093287,
+      "grad_norm": 0.8719160556793213,
+      "learning_rate": 0.0005795221843003413,
+      "loss": 1.3436,
+      "step": 1852
+    },
+    {
+      "epoch": 2.1080773606370875,
+      "grad_norm": 0.7223889231681824,
+      "learning_rate": 0.0005792946530147895,
+      "loss": 1.116,
+      "step": 1853
+    },
+    {
+      "epoch": 2.1092150170648463,
+      "grad_norm": 1.0977421998977661,
+      "learning_rate": 0.0005790671217292379,
+      "loss": 1.5358,
+      "step": 1854
+    },
+    {
+      "epoch": 2.110352673492605,
+      "grad_norm": 1.2012536525726318,
+      "learning_rate": 0.0005788395904436861,
+      "loss": 1.9212,
+      "step": 1855
+    },
+    {
+      "epoch": 2.1114903299203642,
+      "grad_norm": 1.2642079591751099,
+      "learning_rate": 0.0005786120591581343,
+      "loss": 2.3365,
+      "step": 1856
+    },
+    {
+      "epoch": 2.112627986348123,
+      "grad_norm": 0.7523776888847351,
+      "learning_rate": 0.0005783845278725825,
+      "loss": 1.2154,
+      "step": 1857
+    },
+    {
+      "epoch": 2.1137656427758817,
+      "grad_norm": 0.4518072307109833,
+      "learning_rate": 0.0005781569965870307,
+      "loss": 0.5069,
+      "step": 1858
+    },
+    {
+      "epoch": 2.1149032992036405,
+      "grad_norm": 0.6998793482780457,
+      "learning_rate": 0.0005779294653014789,
+      "loss": 1.7925,
+      "step": 1859
+    },
+    {
+      "epoch": 2.1160409556313993,
+      "grad_norm": 1.0769786834716797,
+      "learning_rate": 0.0005777019340159272,
+      "loss": 2.0462,
+      "step": 1860
+    },
+    {
+      "epoch": 2.117178612059158,
+      "grad_norm": 1.269291877746582,
+      "learning_rate": 0.0005774744027303754,
+      "loss": 1.5268,
+      "step": 1861
+    },
+    {
+      "epoch": 2.118316268486917,
+      "grad_norm": 1.7893264293670654,
+      "learning_rate": 0.0005772468714448236,
+      "loss": 3.6841,
+      "step": 1862
+    },
+    {
+      "epoch": 2.1194539249146755,
+      "grad_norm": 1.0378164052963257,
+      "learning_rate": 0.000577019340159272,
+      "loss": 1.8547,
+      "step": 1863
+    },
+    {
+      "epoch": 2.1205915813424348,
+      "grad_norm": 0.983869731426239,
+      "learning_rate": 0.0005767918088737202,
+      "loss": 2.1582,
+      "step": 1864
+    },
+    {
+      "epoch": 2.1217292377701935,
+      "grad_norm": 0.6995578408241272,
+      "learning_rate": 0.0005765642775881684,
+      "loss": 1.7941,
+      "step": 1865
+    },
+    {
+      "epoch": 2.1228668941979523,
+      "grad_norm": 0.8206987380981445,
+      "learning_rate": 0.0005763367463026167,
+      "loss": 1.4095,
+      "step": 1866
+    },
+    {
+      "epoch": 2.124004550625711,
+      "grad_norm": 0.9261349439620972,
+      "learning_rate": 0.0005761092150170648,
+      "loss": 1.2816,
+      "step": 1867
+    },
+    {
+      "epoch": 2.12514220705347,
+      "grad_norm": 0.971121609210968,
+      "learning_rate": 0.000575881683731513,
+      "loss": 1.9692,
+      "step": 1868
+    },
+    {
+      "epoch": 2.1262798634812285,
+      "grad_norm": 0.8004586100578308,
+      "learning_rate": 0.0005756541524459613,
+      "loss": 1.4504,
+      "step": 1869
+    },
+    {
+      "epoch": 2.1274175199089873,
+      "grad_norm": 1.1579177379608154,
+      "learning_rate": 0.0005754266211604095,
+      "loss": 1.8514,
+      "step": 1870
+    },
+    {
+      "epoch": 2.1285551763367465,
+      "grad_norm": 0.6660380959510803,
+      "learning_rate": 0.0005751990898748578,
+      "loss": 1.2882,
+      "step": 1871
+    },
+    {
+      "epoch": 2.1296928327645053,
+      "grad_norm": 1.0082716941833496,
+      "learning_rate": 0.0005749715585893061,
+      "loss": 1.9409,
+      "step": 1872
+    },
+    {
+      "epoch": 2.130830489192264,
+      "grad_norm": 0.6915638446807861,
+      "learning_rate": 0.0005747440273037543,
+      "loss": 1.6384,
+      "step": 1873
+    },
+    {
+      "epoch": 2.131968145620023,
+      "grad_norm": 0.91794753074646,
+      "learning_rate": 0.0005745164960182026,
+      "loss": 2.066,
+      "step": 1874
+    },
+    {
+      "epoch": 2.1331058020477816,
+      "grad_norm": 0.7902770638465881,
+      "learning_rate": 0.0005742889647326508,
+      "loss": 1.3539,
+      "step": 1875
+    },
+    {
+      "epoch": 2.1342434584755403,
+      "grad_norm": 0.8815358281135559,
+      "learning_rate": 0.0005740614334470989,
+      "loss": 1.9951,
+      "step": 1876
+    },
+    {
+      "epoch": 2.135381114903299,
+      "grad_norm": 2.684619665145874,
+      "learning_rate": 0.0005738339021615472,
+      "loss": 2.3747,
+      "step": 1877
+    },
+    {
+      "epoch": 2.136518771331058,
+      "grad_norm": 0.864995539188385,
+      "learning_rate": 0.0005736063708759954,
+      "loss": 1.2905,
+      "step": 1878
+    },
+    {
+      "epoch": 2.137656427758817,
+      "grad_norm": 1.144092082977295,
+      "learning_rate": 0.0005733788395904436,
+      "loss": 3.1334,
+      "step": 1879
+    },
+    {
+      "epoch": 2.138794084186576,
+      "grad_norm": 0.8507954478263855,
+      "learning_rate": 0.000573151308304892,
+      "loss": 2.1402,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1399317406143346,
+      "grad_norm": 0.9215604662895203,
+      "learning_rate": 0.0005729237770193402,
+      "loss": 1.339,
+      "step": 1881
+    },
+    {
+      "epoch": 2.1410693970420933,
+      "grad_norm": 1.1672780513763428,
+      "learning_rate": 0.0005726962457337884,
+      "loss": 1.6453,
+      "step": 1882
+    },
+    {
+      "epoch": 2.142207053469852,
+      "grad_norm": 1.1848803758621216,
+      "learning_rate": 0.0005724687144482367,
+      "loss": 2.445,
+      "step": 1883
+    },
+    {
+      "epoch": 2.143344709897611,
+      "grad_norm": 1.0215684175491333,
+      "learning_rate": 0.0005722411831626849,
+      "loss": 1.3356,
+      "step": 1884
+    },
+    {
+      "epoch": 2.1444823663253696,
+      "grad_norm": 0.9936044216156006,
+      "learning_rate": 0.0005720136518771331,
+      "loss": 1.9894,
+      "step": 1885
+    },
+    {
+      "epoch": 2.1456200227531284,
+      "grad_norm": 1.2094396352767944,
+      "learning_rate": 0.0005717861205915813,
+      "loss": 1.5998,
+      "step": 1886
+    },
+    {
+      "epoch": 2.1467576791808876,
+      "grad_norm": 1.1404001712799072,
+      "learning_rate": 0.0005715585893060295,
+      "loss": 2.1319,
+      "step": 1887
+    },
+    {
+      "epoch": 2.1478953356086463,
+      "grad_norm": 0.656340479850769,
+      "learning_rate": 0.0005713310580204778,
+      "loss": 1.1715,
+      "step": 1888
+    },
+    {
+      "epoch": 2.149032992036405,
+      "grad_norm": 1.03688645362854,
+      "learning_rate": 0.0005711035267349261,
+      "loss": 2.4935,
+      "step": 1889
+    },
+    {
+      "epoch": 2.150170648464164,
+      "grad_norm": 0.8806471228599548,
+      "learning_rate": 0.0005708759954493743,
+      "loss": 1.9498,
+      "step": 1890
+    },
+    {
+      "epoch": 2.1513083048919226,
+      "grad_norm": 0.9819498062133789,
+      "learning_rate": 0.0005706484641638225,
+      "loss": 1.7589,
+      "step": 1891
+    },
+    {
+      "epoch": 2.1524459613196814,
+      "grad_norm": 0.9967643618583679,
+      "learning_rate": 0.0005704209328782708,
+      "loss": 1.479,
+      "step": 1892
+    },
+    {
+      "epoch": 2.15358361774744,
+      "grad_norm": 0.8531262278556824,
+      "learning_rate": 0.000570193401592719,
+      "loss": 2.3837,
+      "step": 1893
+    },
+    {
+      "epoch": 2.1547212741751993,
+      "grad_norm": 0.7077489495277405,
+      "learning_rate": 0.0005699658703071673,
+      "loss": 0.9478,
+      "step": 1894
+    },
+    {
+      "epoch": 2.155858930602958,
+      "grad_norm": 0.4791664183139801,
+      "learning_rate": 0.0005697383390216155,
+      "loss": 0.6543,
+      "step": 1895
+    },
+    {
+      "epoch": 2.156996587030717,
+      "grad_norm": 1.3306792974472046,
+      "learning_rate": 0.0005695108077360637,
+      "loss": 2.6389,
+      "step": 1896
+    },
+    {
+      "epoch": 2.1581342434584756,
+      "grad_norm": 0.8933519721031189,
+      "learning_rate": 0.000569283276450512,
+      "loss": 1.1383,
+      "step": 1897
+    },
+    {
+      "epoch": 2.1592718998862344,
+      "grad_norm": 1.3570899963378906,
+      "learning_rate": 0.0005690557451649602,
+      "loss": 1.7026,
+      "step": 1898
+    },
+    {
+      "epoch": 2.160409556313993,
+      "grad_norm": 0.8122515678405762,
+      "learning_rate": 0.0005688282138794084,
+      "loss": 1.8717,
+      "step": 1899
+    },
+    {
+      "epoch": 2.161547212741752,
+      "grad_norm": 1.0075249671936035,
+      "learning_rate": 0.0005686006825938567,
+      "loss": 1.9019,
+      "step": 1900
+    },
+    {
+      "epoch": 2.1626848691695106,
+      "grad_norm": 1.1624641418457031,
+      "learning_rate": 0.0005683731513083049,
+      "loss": 1.4538,
+      "step": 1901
+    },
+    {
+      "epoch": 2.1638225255972694,
+      "grad_norm": 1.6518383026123047,
+      "learning_rate": 0.0005681456200227531,
+      "loss": 3.6701,
+      "step": 1902
+    },
+    {
+      "epoch": 2.1649601820250286,
+      "grad_norm": 0.6541008353233337,
+      "learning_rate": 0.0005679180887372014,
+      "loss": 1.3688,
+      "step": 1903
+    },
+    {
+      "epoch": 2.1660978384527874,
+      "grad_norm": 1.05472993850708,
+      "learning_rate": 0.0005676905574516497,
+      "loss": 2.1368,
+      "step": 1904
+    },
+    {
+      "epoch": 2.167235494880546,
+      "grad_norm": 1.1374725103378296,
+      "learning_rate": 0.0005674630261660978,
+      "loss": 3.3545,
+      "step": 1905
+    },
+    {
+      "epoch": 2.168373151308305,
+      "grad_norm": 1.0366812944412231,
+      "learning_rate": 0.0005672354948805461,
+      "loss": 2.4839,
+      "step": 1906
+    },
+    {
+      "epoch": 2.1695108077360636,
+      "grad_norm": 0.8595036268234253,
+      "learning_rate": 0.0005670079635949943,
+      "loss": 1.784,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1706484641638224,
+      "grad_norm": 1.023034930229187,
+      "learning_rate": 0.0005667804323094425,
+      "loss": 1.3478,
+      "step": 1908
+    },
+    {
+      "epoch": 2.171786120591581,
+      "grad_norm": 0.741868257522583,
+      "learning_rate": 0.0005665529010238908,
+      "loss": 1.0789,
+      "step": 1909
+    },
+    {
+      "epoch": 2.1729237770193404,
+      "grad_norm": 1.1272695064544678,
+      "learning_rate": 0.000566325369738339,
+      "loss": 2.037,
+      "step": 1910
+    },
+    {
+      "epoch": 2.174061433447099,
+      "grad_norm": 1.0772629976272583,
+      "learning_rate": 0.0005660978384527872,
+      "loss": 2.3381,
+      "step": 1911
+    },
+    {
+      "epoch": 2.175199089874858,
+      "grad_norm": 1.0794868469238281,
+      "learning_rate": 0.0005658703071672356,
+      "loss": 2.9036,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1763367463026166,
+      "grad_norm": 0.9306978583335876,
+      "learning_rate": 0.0005656427758816838,
+      "loss": 2.0821,
+      "step": 1913
+    },
+    {
+      "epoch": 2.1774744027303754,
+      "grad_norm": 0.88816237449646,
+      "learning_rate": 0.0005654152445961321,
+      "loss": 1.7599,
+      "step": 1914
+    },
+    {
+      "epoch": 2.178612059158134,
+      "grad_norm": 1.169001817703247,
+      "learning_rate": 0.0005651877133105802,
+      "loss": 2.1915,
+      "step": 1915
+    },
+    {
+      "epoch": 2.179749715585893,
+      "grad_norm": 1.0751813650131226,
+      "learning_rate": 0.0005649601820250284,
+      "loss": 1.6451,
+      "step": 1916
+    },
+    {
+      "epoch": 2.1808873720136517,
+      "grad_norm": 1.062515377998352,
+      "learning_rate": 0.0005647326507394767,
+      "loss": 2.4742,
+      "step": 1917
+    },
+    {
+      "epoch": 2.182025028441411,
+      "grad_norm": 0.9888359308242798,
+      "learning_rate": 0.0005645051194539249,
+      "loss": 3.1331,
+      "step": 1918
+    },
+    {
+      "epoch": 2.1831626848691696,
+      "grad_norm": 0.8473877310752869,
+      "learning_rate": 0.0005642775881683731,
+      "loss": 1.4678,
+      "step": 1919
+    },
+    {
+      "epoch": 2.1843003412969284,
+      "grad_norm": 0.85085129737854,
+      "learning_rate": 0.0005640500568828214,
+      "loss": 2.127,
+      "step": 1920
+    },
+    {
+      "epoch": 2.185437997724687,
+      "grad_norm": 1.2935888767242432,
+      "learning_rate": 0.0005638225255972697,
+      "loss": 3.5023,
+      "step": 1921
+    },
+    {
+      "epoch": 2.186575654152446,
+      "grad_norm": 0.8884962201118469,
+      "learning_rate": 0.0005635949943117179,
+      "loss": 0.8726,
+      "step": 1922
+    },
+    {
+      "epoch": 2.1877133105802047,
+      "grad_norm": 0.9087686538696289,
+      "learning_rate": 0.0005633674630261662,
+      "loss": 1.3675,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1888509670079634,
+      "grad_norm": 0.6281200647354126,
+      "learning_rate": 0.0005631399317406144,
+      "loss": 0.7575,
+      "step": 1924
+    },
+    {
+      "epoch": 2.189988623435722,
+      "grad_norm": 1.0639797449111938,
+      "learning_rate": 0.0005629124004550625,
+      "loss": 2.3185,
+      "step": 1925
+    },
+    {
+      "epoch": 2.1911262798634814,
+      "grad_norm": 0.596128523349762,
+      "learning_rate": 0.0005626848691695108,
+      "loss": 0.8454,
+      "step": 1926
+    },
+    {
+      "epoch": 2.19226393629124,
+      "grad_norm": 0.8989168405532837,
+      "learning_rate": 0.000562457337883959,
+      "loss": 0.7302,
+      "step": 1927
+    },
+    {
+      "epoch": 2.193401592718999,
+      "grad_norm": 0.7336655855178833,
+      "learning_rate": 0.0005622298065984072,
+      "loss": 1.2145,
+      "step": 1928
+    },
+    {
+      "epoch": 2.1945392491467577,
+      "grad_norm": 1.077013611793518,
+      "learning_rate": 0.0005620022753128556,
+      "loss": 2.4185,
+      "step": 1929
+    },
+    {
+      "epoch": 2.1956769055745164,
+      "grad_norm": 0.8258651494979858,
+      "learning_rate": 0.0005617747440273038,
+      "loss": 1.4536,
+      "step": 1930
+    },
+    {
+      "epoch": 2.196814562002275,
+      "grad_norm": 0.569733202457428,
+      "learning_rate": 0.000561547212741752,
+      "loss": 0.7994,
+      "step": 1931
+    },
+    {
+      "epoch": 2.197952218430034,
+      "grad_norm": 0.843547523021698,
+      "learning_rate": 0.0005613196814562003,
+      "loss": 1.3272,
+      "step": 1932
+    },
+    {
+      "epoch": 2.199089874857793,
+      "grad_norm": 0.8367027044296265,
+      "learning_rate": 0.0005610921501706485,
+      "loss": 1.625,
+      "step": 1933
+    },
+    {
+      "epoch": 2.200227531285552,
+      "grad_norm": 1.0637445449829102,
+      "learning_rate": 0.0005608646188850968,
+      "loss": 1.3352,
+      "step": 1934
+    },
+    {
+      "epoch": 2.2013651877133107,
+      "grad_norm": 1.4886406660079956,
+      "learning_rate": 0.0005606370875995449,
+      "loss": 2.3982,
+      "step": 1935
+    },
+    {
+      "epoch": 2.2025028441410694,
+      "grad_norm": 1.015264630317688,
+      "learning_rate": 0.0005604095563139931,
+      "loss": 2.7429,
+      "step": 1936
+    },
+    {
+      "epoch": 2.203640500568828,
+      "grad_norm": 1.2239853143692017,
+      "learning_rate": 0.0005601820250284414,
+      "loss": 1.8034,
+      "step": 1937
+    },
+    {
+      "epoch": 2.204778156996587,
+      "grad_norm": 1.1284427642822266,
+      "learning_rate": 0.0005599544937428897,
+      "loss": 2.232,
+      "step": 1938
+    },
+    {
+      "epoch": 2.2059158134243457,
+      "grad_norm": 1.118551254272461,
+      "learning_rate": 0.0005597269624573379,
+      "loss": 1.9726,
+      "step": 1939
+    },
+    {
+      "epoch": 2.2070534698521045,
+      "grad_norm": 1.115544080734253,
+      "learning_rate": 0.0005594994311717862,
+      "loss": 2.3797,
+      "step": 1940
+    },
+    {
+      "epoch": 2.2081911262798632,
+      "grad_norm": 0.8640685677528381,
+      "learning_rate": 0.0005592718998862344,
+      "loss": 2.5765,
+      "step": 1941
+    },
+    {
+      "epoch": 2.2093287827076225,
+      "grad_norm": 1.0405633449554443,
+      "learning_rate": 0.0005590443686006826,
+      "loss": 1.6865,
+      "step": 1942
+    },
+    {
+      "epoch": 2.210466439135381,
+      "grad_norm": 1.6807098388671875,
+      "learning_rate": 0.0005588168373151309,
+      "loss": 2.5395,
+      "step": 1943
+    },
+    {
+      "epoch": 2.21160409556314,
+      "grad_norm": 0.9843060374259949,
+      "learning_rate": 0.000558589306029579,
+      "loss": 2.0961,
+      "step": 1944
+    },
+    {
+      "epoch": 2.2127417519908987,
+      "grad_norm": 0.995924174785614,
+      "learning_rate": 0.0005583617747440272,
+      "loss": 2.0971,
+      "step": 1945
+    },
+    {
+      "epoch": 2.2138794084186575,
+      "grad_norm": 0.8533014059066772,
+      "learning_rate": 0.0005581342434584756,
+      "loss": 1.5151,
+      "step": 1946
+    },
+    {
+      "epoch": 2.2150170648464163,
+      "grad_norm": 0.9610128402709961,
+      "learning_rate": 0.0005579067121729238,
+      "loss": 2.239,
+      "step": 1947
+    },
+    {
+      "epoch": 2.216154721274175,
+      "grad_norm": 0.799690306186676,
+      "learning_rate": 0.000557679180887372,
+      "loss": 1.5567,
+      "step": 1948
+    },
+    {
+      "epoch": 2.217292377701934,
+      "grad_norm": 0.8226889371871948,
+      "learning_rate": 0.0005574516496018203,
+      "loss": 2.6057,
+      "step": 1949
+    },
+    {
+      "epoch": 2.218430034129693,
+      "grad_norm": 1.314974308013916,
+      "learning_rate": 0.0005572241183162685,
+      "loss": 2.2834,
+      "step": 1950
+    },
+    {
+      "epoch": 2.2195676905574517,
+      "grad_norm": 0.9259127974510193,
+      "learning_rate": 0.0005569965870307167,
+      "loss": 1.9402,
+      "step": 1951
+    },
+    {
+      "epoch": 2.2207053469852105,
+      "grad_norm": 1.1179683208465576,
+      "learning_rate": 0.000556769055745165,
+      "loss": 2.6231,
+      "step": 1952
+    },
+    {
+      "epoch": 2.2218430034129693,
+      "grad_norm": 1.0641530752182007,
+      "learning_rate": 0.0005565415244596132,
+      "loss": 2.055,
+      "step": 1953
+    },
+    {
+      "epoch": 2.222980659840728,
+      "grad_norm": 0.7080785036087036,
+      "learning_rate": 0.0005563139931740613,
+      "loss": 0.7283,
+      "step": 1954
+    },
+    {
+      "epoch": 2.2241183162684868,
+      "grad_norm": 1.3224704265594482,
+      "learning_rate": 0.0005560864618885097,
+      "loss": 2.189,
+      "step": 1955
+    },
+    {
+      "epoch": 2.2252559726962455,
+      "grad_norm": 0.9138402938842773,
+      "learning_rate": 0.0005558589306029579,
+      "loss": 1.9759,
+      "step": 1956
+    },
+    {
+      "epoch": 2.2263936291240047,
+      "grad_norm": 1.1304935216903687,
+      "learning_rate": 0.0005556313993174062,
+      "loss": 2.2992,
+      "step": 1957
+    },
+    {
+      "epoch": 2.2275312855517635,
+      "grad_norm": 0.7692627906799316,
+      "learning_rate": 0.0005554038680318544,
+      "loss": 1.195,
+      "step": 1958
+    },
+    {
+      "epoch": 2.2286689419795223,
+      "grad_norm": 0.9814881086349487,
+      "learning_rate": 0.0005551763367463026,
+      "loss": 1.7418,
+      "step": 1959
+    },
+    {
+      "epoch": 2.229806598407281,
+      "grad_norm": 1.449527621269226,
+      "learning_rate": 0.0005549488054607509,
+      "loss": 2.8254,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2309442548350398,
+      "grad_norm": 1.177649974822998,
+      "learning_rate": 0.0005547212741751991,
+      "loss": 2.1793,
+      "step": 1961
+    },
+    {
+      "epoch": 2.2320819112627985,
+      "grad_norm": 1.1293201446533203,
+      "learning_rate": 0.0005544937428896473,
+      "loss": 3.0604,
+      "step": 1962
+    },
+    {
+      "epoch": 2.2332195676905573,
+      "grad_norm": 0.9934133887290955,
+      "learning_rate": 0.0005542662116040957,
+      "loss": 1.8564,
+      "step": 1963
+    },
+    {
+      "epoch": 2.234357224118316,
+      "grad_norm": 0.8571220636367798,
+      "learning_rate": 0.0005540386803185438,
+      "loss": 1.4268,
+      "step": 1964
+    },
+    {
+      "epoch": 2.2354948805460753,
+      "grad_norm": 0.84453284740448,
+      "learning_rate": 0.000553811149032992,
+      "loss": 1.4253,
+      "step": 1965
+    },
+    {
+      "epoch": 2.236632536973834,
+      "grad_norm": 0.842588484287262,
+      "learning_rate": 0.0005535836177474403,
+      "loss": 2.4052,
+      "step": 1966
+    },
+    {
+      "epoch": 2.2377701934015928,
+      "grad_norm": 1.4743871688842773,
+      "learning_rate": 0.0005533560864618885,
+      "loss": 2.9152,
+      "step": 1967
+    },
+    {
+      "epoch": 2.2389078498293515,
+      "grad_norm": 0.6935710310935974,
+      "learning_rate": 0.0005531285551763367,
+      "loss": 1.2375,
+      "step": 1968
+    },
+    {
+      "epoch": 2.2400455062571103,
+      "grad_norm": 1.2612910270690918,
+      "learning_rate": 0.000552901023890785,
+      "loss": 1.9998,
+      "step": 1969
+    },
+    {
+      "epoch": 2.241183162684869,
+      "grad_norm": 0.8181163668632507,
+      "learning_rate": 0.0005526734926052332,
+      "loss": 1.3744,
+      "step": 1970
+    },
+    {
+      "epoch": 2.242320819112628,
+      "grad_norm": 1.1267778873443604,
+      "learning_rate": 0.0005524459613196815,
+      "loss": 2.3809,
+      "step": 1971
+    },
+    {
+      "epoch": 2.243458475540387,
+      "grad_norm": 1.200745701789856,
+      "learning_rate": 0.0005522184300341298,
+      "loss": 2.4499,
+      "step": 1972
+    },
+    {
+      "epoch": 2.244596131968146,
+      "grad_norm": 1.04634690284729,
+      "learning_rate": 0.0005519908987485779,
+      "loss": 1.1126,
+      "step": 1973
+    },
+    {
+      "epoch": 2.2457337883959045,
+      "grad_norm": 0.9528573751449585,
+      "learning_rate": 0.0005517633674630261,
+      "loss": 2.4237,
+      "step": 1974
+    },
+    {
+      "epoch": 2.2468714448236633,
+      "grad_norm": 1.3052282333374023,
+      "learning_rate": 0.0005515358361774744,
+      "loss": 2.5489,
+      "step": 1975
+    },
+    {
+      "epoch": 2.248009101251422,
+      "grad_norm": 1.095739722251892,
+      "learning_rate": 0.0005513083048919226,
+      "loss": 1.1754,
+      "step": 1976
+    },
+    {
+      "epoch": 2.249146757679181,
+      "grad_norm": 1.1007702350616455,
+      "learning_rate": 0.0005510807736063709,
+      "loss": 1.7626,
+      "step": 1977
+    },
+    {
+      "epoch": 2.2502844141069396,
+      "grad_norm": 0.6495097279548645,
+      "learning_rate": 0.0005508532423208191,
+      "loss": 1.7669,
+      "step": 1978
+    },
+    {
+      "epoch": 2.2514220705346983,
+      "grad_norm": 0.9295183420181274,
+      "learning_rate": 0.0005506257110352673,
+      "loss": 1.9161,
+      "step": 1979
+    },
+    {
+      "epoch": 2.252559726962457,
+      "grad_norm": 1.0629956722259521,
+      "learning_rate": 0.0005503981797497157,
+      "loss": 2.2528,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2536973833902163,
+      "grad_norm": 0.6436116695404053,
+      "learning_rate": 0.0005501706484641639,
+      "loss": 1.2159,
+      "step": 1981
+    },
+    {
+      "epoch": 2.254835039817975,
+      "grad_norm": 1.0467777252197266,
+      "learning_rate": 0.0005499431171786121,
+      "loss": 2.0886,
+      "step": 1982
+    },
+    {
+      "epoch": 2.255972696245734,
+      "grad_norm": 0.7404427528381348,
+      "learning_rate": 0.0005497155858930603,
+      "loss": 1.4913,
+      "step": 1983
+    },
+    {
+      "epoch": 2.2571103526734926,
+      "grad_norm": 1.0175822973251343,
+      "learning_rate": 0.0005494880546075085,
+      "loss": 1.9654,
+      "step": 1984
+    },
+    {
+      "epoch": 2.2582480091012513,
+      "grad_norm": 0.930240273475647,
+      "learning_rate": 0.0005492605233219567,
+      "loss": 1.7043,
+      "step": 1985
+    },
+    {
+      "epoch": 2.25938566552901,
+      "grad_norm": 2.17866587638855,
+      "learning_rate": 0.000549032992036405,
+      "loss": 3.0881,
+      "step": 1986
+    },
+    {
+      "epoch": 2.260523321956769,
+      "grad_norm": 1.5989547967910767,
+      "learning_rate": 0.0005488054607508532,
+      "loss": 3.0951,
+      "step": 1987
+    },
+    {
+      "epoch": 2.261660978384528,
+      "grad_norm": 1.6310112476348877,
+      "learning_rate": 0.0005485779294653015,
+      "loss": 3.687,
+      "step": 1988
+    },
+    {
+      "epoch": 2.262798634812287,
+      "grad_norm": 0.682499885559082,
+      "learning_rate": 0.0005483503981797498,
+      "loss": 1.5835,
+      "step": 1989
+    },
+    {
+      "epoch": 2.2639362912400456,
+      "grad_norm": 1.0656623840332031,
+      "learning_rate": 0.000548122866894198,
+      "loss": 2.0308,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2650739476678043,
+      "grad_norm": 0.7635136842727661,
+      "learning_rate": 0.0005478953356086462,
+      "loss": 2.4743,
+      "step": 1991
+    },
+    {
+      "epoch": 2.266211604095563,
+      "grad_norm": 0.7663008570671082,
+      "learning_rate": 0.0005476678043230945,
+      "loss": 2.0004,
+      "step": 1992
+    },
+    {
+      "epoch": 2.267349260523322,
+      "grad_norm": 0.5305643677711487,
+      "learning_rate": 0.0005474402730375426,
+      "loss": 0.7431,
+      "step": 1993
+    },
+    {
+      "epoch": 2.2684869169510806,
+      "grad_norm": 1.2221653461456299,
+      "learning_rate": 0.0005472127417519908,
+      "loss": 2.3999,
+      "step": 1994
+    },
+    {
+      "epoch": 2.26962457337884,
+      "grad_norm": 1.0990016460418701,
+      "learning_rate": 0.0005469852104664391,
+      "loss": 3.3276,
+      "step": 1995
+    },
+    {
+      "epoch": 2.2707622298065986,
+      "grad_norm": 1.074623703956604,
+      "learning_rate": 0.0005467576791808873,
+      "loss": 2.2739,
+      "step": 1996
+    },
+    {
+      "epoch": 2.2718998862343573,
+      "grad_norm": 0.7581028342247009,
+      "learning_rate": 0.0005465301478953357,
+      "loss": 1.9877,
+      "step": 1997
+    },
+    {
+      "epoch": 2.273037542662116,
+      "grad_norm": 1.2376879453659058,
+      "learning_rate": 0.0005463026166097839,
+      "loss": 2.5045,
+      "step": 1998
+    },
+    {
+      "epoch": 2.274175199089875,
+      "grad_norm": 0.8391635417938232,
+      "learning_rate": 0.0005460750853242321,
+      "loss": 1.414,
+      "step": 1999
+    },
+    {
+      "epoch": 2.2753128555176336,
+      "grad_norm": 0.917779803276062,
+      "learning_rate": 0.0005458475540386804,
+      "loss": 1.5652,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2764505119453924,
+      "grad_norm": 0.9148494005203247,
+      "learning_rate": 0.0005456200227531286,
+      "loss": 0.9553,
+      "step": 2001
+    },
+    {
+      "epoch": 2.277588168373151,
+      "grad_norm": 1.6205681562423706,
+      "learning_rate": 0.0005453924914675768,
+      "loss": 3.2662,
+      "step": 2002
+    },
+    {
+      "epoch": 2.27872582480091,
+      "grad_norm": 0.5104256272315979,
+      "learning_rate": 0.000545164960182025,
+      "loss": 1.1526,
+      "step": 2003
+    },
+    {
+      "epoch": 2.279863481228669,
+      "grad_norm": 0.9608463644981384,
+      "learning_rate": 0.0005449374288964732,
+      "loss": 1.3495,
+      "step": 2004
+    },
+    {
+      "epoch": 2.281001137656428,
+      "grad_norm": 0.8729520440101624,
+      "learning_rate": 0.0005447098976109215,
+      "loss": 1.6019,
+      "step": 2005
+    },
+    {
+      "epoch": 2.2821387940841866,
+      "grad_norm": 0.7411619424819946,
+      "learning_rate": 0.0005444823663253698,
+      "loss": 1.5559,
+      "step": 2006
+    },
+    {
+      "epoch": 2.2832764505119454,
+      "grad_norm": 0.7767372727394104,
+      "learning_rate": 0.000544254835039818,
+      "loss": 2.0135,
+      "step": 2007
+    },
+    {
+      "epoch": 2.284414106939704,
+      "grad_norm": 0.5355758666992188,
+      "learning_rate": 0.0005440273037542662,
+      "loss": 1.1862,
+      "step": 2008
+    },
+    {
+      "epoch": 2.285551763367463,
+      "grad_norm": 0.719421923160553,
+      "learning_rate": 0.0005437997724687145,
+      "loss": 1.1892,
+      "step": 2009
+    },
+    {
+      "epoch": 2.2866894197952217,
+      "grad_norm": 0.7264088988304138,
+      "learning_rate": 0.0005435722411831627,
+      "loss": 1.5477,
+      "step": 2010
+    },
+    {
+      "epoch": 2.287827076222981,
+      "grad_norm": 0.8898872137069702,
+      "learning_rate": 0.0005433447098976109,
+      "loss": 1.5906,
+      "step": 2011
+    },
+    {
+      "epoch": 2.2889647326507396,
+      "grad_norm": 0.6888718605041504,
+      "learning_rate": 0.0005431171786120591,
+      "loss": 1.1981,
+      "step": 2012
+    },
+    {
+      "epoch": 2.2901023890784984,
+      "grad_norm": 0.6405702829360962,
+      "learning_rate": 0.0005428896473265074,
+      "loss": 0.9176,
+      "step": 2013
+    },
+    {
+      "epoch": 2.291240045506257,
+      "grad_norm": 1.125874638557434,
+      "learning_rate": 0.0005426621160409556,
+      "loss": 1.9476,
+      "step": 2014
+    },
+    {
+      "epoch": 2.292377701934016,
+      "grad_norm": 0.9443691372871399,
+      "learning_rate": 0.0005424345847554039,
+      "loss": 1.1386,
+      "step": 2015
+    },
+    {
+      "epoch": 2.2935153583617747,
+      "grad_norm": 1.1550638675689697,
+      "learning_rate": 0.0005422070534698521,
+      "loss": 2.695,
+      "step": 2016
+    },
+    {
+      "epoch": 2.2946530147895334,
+      "grad_norm": 0.9223167896270752,
+      "learning_rate": 0.0005419795221843004,
+      "loss": 1.962,
+      "step": 2017
+    },
+    {
+      "epoch": 2.295790671217292,
+      "grad_norm": 0.9546693563461304,
+      "learning_rate": 0.0005417519908987486,
+      "loss": 1.9184,
+      "step": 2018
+    },
+    {
+      "epoch": 2.296928327645051,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0005415244596131968,
+      "loss": 2.3646,
+      "step": 2019
+    },
+    {
+      "epoch": 2.29806598407281,
+      "grad_norm": 0.7954057455062866,
+      "learning_rate": 0.0005412969283276451,
+      "loss": 1.0092,
+      "step": 2020
+    },
+    {
+      "epoch": 2.299203640500569,
+      "grad_norm": 0.9007298946380615,
+      "learning_rate": 0.0005410693970420934,
+      "loss": 1.6872,
+      "step": 2021
+    },
+    {
+      "epoch": 2.3003412969283277,
+      "grad_norm": 0.9581443071365356,
+      "learning_rate": 0.0005408418657565415,
+      "loss": 2.0143,
+      "step": 2022
+    },
+    {
+      "epoch": 2.3014789533560864,
+      "grad_norm": 0.9659209251403809,
+      "learning_rate": 0.0005406143344709898,
+      "loss": 1.2962,
+      "step": 2023
+    },
+    {
+      "epoch": 2.302616609783845,
+      "grad_norm": 0.974143385887146,
+      "learning_rate": 0.000540386803185438,
+      "loss": 2.7885,
+      "step": 2024
+    },
+    {
+      "epoch": 2.303754266211604,
+      "grad_norm": 1.0462844371795654,
+      "learning_rate": 0.0005401592718998862,
+      "loss": 2.1409,
+      "step": 2025
+    },
+    {
+      "epoch": 2.3048919226393627,
+      "grad_norm": 0.8965214490890503,
+      "learning_rate": 0.0005399317406143345,
+      "loss": 2.2281,
+      "step": 2026
+    },
+    {
+      "epoch": 2.306029579067122,
+      "grad_norm": 1.1970421075820923,
+      "learning_rate": 0.0005397042093287827,
+      "loss": 2.378,
+      "step": 2027
+    },
+    {
+      "epoch": 2.3071672354948807,
+      "grad_norm": 1.012830138206482,
+      "learning_rate": 0.0005394766780432309,
+      "loss": 1.4127,
+      "step": 2028
+    },
+    {
+      "epoch": 2.3083048919226394,
+      "grad_norm": 0.8547312021255493,
+      "learning_rate": 0.0005392491467576793,
+      "loss": 2.5707,
+      "step": 2029
+    },
+    {
+      "epoch": 2.309442548350398,
+      "grad_norm": 1.0248013734817505,
+      "learning_rate": 0.0005390216154721275,
+      "loss": 1.1376,
+      "step": 2030
+    },
+    {
+      "epoch": 2.310580204778157,
+      "grad_norm": 1.2407397031784058,
+      "learning_rate": 0.0005387940841865757,
+      "loss": 2.5036,
+      "step": 2031
+    },
+    {
+      "epoch": 2.3117178612059157,
+      "grad_norm": 1.169358730316162,
+      "learning_rate": 0.0005385665529010239,
+      "loss": 2.4767,
+      "step": 2032
+    },
+    {
+      "epoch": 2.3128555176336745,
+      "grad_norm": 0.552113950252533,
+      "learning_rate": 0.0005383390216154721,
+      "loss": 0.8825,
+      "step": 2033
+    },
+    {
+      "epoch": 2.3139931740614337,
+      "grad_norm": 1.2916892766952515,
+      "learning_rate": 0.0005381114903299203,
+      "loss": 2.3583,
+      "step": 2034
+    },
+    {
+      "epoch": 2.3151308304891924,
+      "grad_norm": 0.9264313578605652,
+      "learning_rate": 0.0005378839590443686,
+      "loss": 1.945,
+      "step": 2035
+    },
+    {
+      "epoch": 2.316268486916951,
+      "grad_norm": 0.6379786729812622,
+      "learning_rate": 0.0005376564277588168,
+      "loss": 0.7919,
+      "step": 2036
+    },
+    {
+      "epoch": 2.31740614334471,
+      "grad_norm": 1.199393630027771,
+      "learning_rate": 0.0005374288964732651,
+      "loss": 2.3252,
+      "step": 2037
+    },
+    {
+      "epoch": 2.3185437997724687,
+      "grad_norm": 1.0265052318572998,
+      "learning_rate": 0.0005372013651877134,
+      "loss": 1.6382,
+      "step": 2038
+    },
+    {
+      "epoch": 2.3196814562002275,
+      "grad_norm": 1.808610439300537,
+      "learning_rate": 0.0005369738339021616,
+      "loss": 3.1174,
+      "step": 2039
+    },
+    {
+      "epoch": 2.3208191126279862,
+      "grad_norm": 0.7785828113555908,
+      "learning_rate": 0.0005367463026166099,
+      "loss": 1.7553,
+      "step": 2040
+    },
+    {
+      "epoch": 2.321956769055745,
+      "grad_norm": 1.1902951002120972,
+      "learning_rate": 0.0005365187713310581,
+      "loss": 1.9243,
+      "step": 2041
+    },
+    {
+      "epoch": 2.3230944254835038,
+      "grad_norm": 0.8798537254333496,
+      "learning_rate": 0.0005362912400455062,
+      "loss": 1.8127,
+      "step": 2042
+    },
+    {
+      "epoch": 2.324232081911263,
+      "grad_norm": 1.0299949645996094,
+      "learning_rate": 0.0005360637087599545,
+      "loss": 2.0079,
+      "step": 2043
+    },
+    {
+      "epoch": 2.3253697383390217,
+      "grad_norm": 0.8465110063552856,
+      "learning_rate": 0.0005358361774744027,
+      "loss": 1.6008,
+      "step": 2044
+    },
+    {
+      "epoch": 2.3265073947667805,
+      "grad_norm": 1.031044363975525,
+      "learning_rate": 0.0005356086461888509,
+      "loss": 1.6429,
+      "step": 2045
+    },
+    {
+      "epoch": 2.3276450511945392,
+      "grad_norm": 0.9194596409797668,
+      "learning_rate": 0.0005353811149032993,
+      "loss": 1.5493,
+      "step": 2046
+    },
+    {
+      "epoch": 2.328782707622298,
+      "grad_norm": 0.8609010577201843,
+      "learning_rate": 0.0005351535836177475,
+      "loss": 1.4473,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3299203640500568,
+      "grad_norm": 1.3532382249832153,
+      "learning_rate": 0.0005349260523321957,
+      "loss": 4.2047,
+      "step": 2048
+    },
+    {
+      "epoch": 2.3310580204778155,
+      "grad_norm": 0.9629133939743042,
+      "learning_rate": 0.000534698521046644,
+      "loss": 1.1077,
+      "step": 2049
+    },
+    {
+      "epoch": 2.3321956769055747,
+      "grad_norm": 0.8413902521133423,
+      "learning_rate": 0.0005344709897610922,
+      "loss": 1.5364,
+      "step": 2050
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 1.0402729511260986,
+      "learning_rate": 0.0005342434584755403,
+      "loss": 2.7655,
+      "step": 2051
+    },
+    {
+      "epoch": 2.3344709897610922,
+      "grad_norm": 0.9589785933494568,
+      "learning_rate": 0.0005340159271899886,
+      "loss": 1.5259,
+      "step": 2052
+    },
+    {
+      "epoch": 2.335608646188851,
+      "grad_norm": 0.6738478541374207,
+      "learning_rate": 0.0005337883959044368,
+      "loss": 1.5741,
+      "step": 2053
+    },
+    {
+      "epoch": 2.3367463026166098,
+      "grad_norm": 1.0174530744552612,
+      "learning_rate": 0.000533560864618885,
+      "loss": 3.3126,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3378839590443685,
+      "grad_norm": 0.9035469889640808,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 1.6059,
+      "step": 2055
+    },
+    {
+      "epoch": 2.3390216154721273,
+      "grad_norm": 1.1631959676742554,
+      "learning_rate": 0.0005331058020477816,
+      "loss": 1.7973,
+      "step": 2056
+    },
+    {
+      "epoch": 2.3401592718998865,
+      "grad_norm": 0.8147276043891907,
+      "learning_rate": 0.0005328782707622299,
+      "loss": 1.3416,
+      "step": 2057
+    },
+    {
+      "epoch": 2.3412969283276452,
+      "grad_norm": 1.1491146087646484,
+      "learning_rate": 0.0005326507394766781,
+      "loss": 3.0711,
+      "step": 2058
+    },
+    {
+      "epoch": 2.342434584755404,
+      "grad_norm": 0.7472861409187317,
+      "learning_rate": 0.0005324232081911263,
+      "loss": 1.4432,
+      "step": 2059
+    },
+    {
+      "epoch": 2.3435722411831628,
+      "grad_norm": 1.2172683477401733,
+      "learning_rate": 0.0005321956769055746,
+      "loss": 1.5058,
+      "step": 2060
+    },
+    {
+      "epoch": 2.3447098976109215,
+      "grad_norm": 1.0745410919189453,
+      "learning_rate": 0.0005319681456200227,
+      "loss": 1.6888,
+      "step": 2061
+    },
+    {
+      "epoch": 2.3458475540386803,
+      "grad_norm": 0.8925970792770386,
+      "learning_rate": 0.0005317406143344709,
+      "loss": 1.4089,
+      "step": 2062
+    },
+    {
+      "epoch": 2.346985210466439,
+      "grad_norm": 0.8472179770469666,
+      "learning_rate": 0.0005315130830489193,
+      "loss": 1.4942,
+      "step": 2063
+    },
+    {
+      "epoch": 2.348122866894198,
+      "grad_norm": 0.4184499979019165,
+      "learning_rate": 0.0005312855517633675,
+      "loss": 0.6314,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3492605233219566,
+      "grad_norm": 1.262465238571167,
+      "learning_rate": 0.0005310580204778157,
+      "loss": 2.0469,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3503981797497158,
+      "grad_norm": 1.097849726676941,
+      "learning_rate": 0.000530830489192264,
+      "loss": 2.8422,
+      "step": 2066
+    },
+    {
+      "epoch": 2.3515358361774745,
+      "grad_norm": 1.6128443479537964,
+      "learning_rate": 0.0005306029579067122,
+      "loss": 1.9792,
+      "step": 2067
+    },
+    {
+      "epoch": 2.3526734926052333,
+      "grad_norm": 0.8861138820648193,
+      "learning_rate": 0.0005303754266211604,
+      "loss": 1.7751,
+      "step": 2068
+    },
+    {
+      "epoch": 2.353811149032992,
+      "grad_norm": 0.8302369713783264,
+      "learning_rate": 0.0005301478953356087,
+      "loss": 1.8749,
+      "step": 2069
+    },
+    {
+      "epoch": 2.354948805460751,
+      "grad_norm": 0.9765578508377075,
+      "learning_rate": 0.0005299203640500569,
+      "loss": 2.7383,
+      "step": 2070
+    },
+    {
+      "epoch": 2.3560864618885096,
+      "grad_norm": 1.0690255165100098,
+      "learning_rate": 0.000529692832764505,
+      "loss": 1.6116,
+      "step": 2071
+    },
+    {
+      "epoch": 2.3572241183162683,
+      "grad_norm": 0.96379154920578,
+      "learning_rate": 0.0005294653014789534,
+      "loss": 1.7744,
+      "step": 2072
+    },
+    {
+      "epoch": 2.3583617747440275,
+      "grad_norm": 0.5584749579429626,
+      "learning_rate": 0.0005292377701934016,
+      "loss": 1.0452,
+      "step": 2073
+    },
+    {
+      "epoch": 2.3594994311717863,
+      "grad_norm": 1.5071784257888794,
+      "learning_rate": 0.0005290102389078498,
+      "loss": 3.7007,
+      "step": 2074
+    },
+    {
+      "epoch": 2.360637087599545,
+      "grad_norm": 0.7578591108322144,
+      "learning_rate": 0.0005287827076222981,
+      "loss": 0.9569,
+      "step": 2075
+    },
+    {
+      "epoch": 2.361774744027304,
+      "grad_norm": 0.8119309544563293,
+      "learning_rate": 0.0005285551763367463,
+      "loss": 1.6168,
+      "step": 2076
+    },
+    {
+      "epoch": 2.3629124004550626,
+      "grad_norm": 0.9481661319732666,
+      "learning_rate": 0.0005283276450511945,
+      "loss": 2.9747,
+      "step": 2077
+    },
+    {
+      "epoch": 2.3640500568828213,
+      "grad_norm": 0.9913211464881897,
+      "learning_rate": 0.0005281001137656428,
+      "loss": 2.35,
+      "step": 2078
+    },
+    {
+      "epoch": 2.36518771331058,
+      "grad_norm": 1.1122124195098877,
+      "learning_rate": 0.000527872582480091,
+      "loss": 2.0505,
+      "step": 2079
+    },
+    {
+      "epoch": 2.366325369738339,
+      "grad_norm": 0.8721759915351868,
+      "learning_rate": 0.0005276450511945393,
+      "loss": 1.6231,
+      "step": 2080
+    },
+    {
+      "epoch": 2.3674630261660976,
+      "grad_norm": 0.8457727432250977,
+      "learning_rate": 0.0005274175199089875,
+      "loss": 1.6846,
+      "step": 2081
+    },
+    {
+      "epoch": 2.368600682593857,
+      "grad_norm": 1.0846689939498901,
+      "learning_rate": 0.0005271899886234357,
+      "loss": 1.7606,
+      "step": 2082
+    },
+    {
+      "epoch": 2.3697383390216156,
+      "grad_norm": 0.9165617823600769,
+      "learning_rate": 0.000526962457337884,
+      "loss": 1.3047,
+      "step": 2083
+    },
+    {
+      "epoch": 2.3708759954493743,
+      "grad_norm": 1.1322617530822754,
+      "learning_rate": 0.0005267349260523322,
+      "loss": 1.1403,
+      "step": 2084
+    },
+    {
+      "epoch": 2.372013651877133,
+      "grad_norm": 1.6939765214920044,
+      "learning_rate": 0.0005265073947667804,
+      "loss": 3.1601,
+      "step": 2085
+    },
+    {
+      "epoch": 2.373151308304892,
+      "grad_norm": 1.0793713331222534,
+      "learning_rate": 0.0005262798634812287,
+      "loss": 2.0837,
+      "step": 2086
+    },
+    {
+      "epoch": 2.3742889647326506,
+      "grad_norm": 0.8238543272018433,
+      "learning_rate": 0.0005260523321956769,
+      "loss": 1.438,
+      "step": 2087
+    },
+    {
+      "epoch": 2.3754266211604094,
+      "grad_norm": 0.932790994644165,
+      "learning_rate": 0.0005258248009101252,
+      "loss": 1.9309,
+      "step": 2088
+    },
+    {
+      "epoch": 2.3765642775881686,
+      "grad_norm": 0.6618528366088867,
+      "learning_rate": 0.0005255972696245735,
+      "loss": 1.4155,
+      "step": 2089
+    },
+    {
+      "epoch": 2.3777019340159273,
+      "grad_norm": 0.7635231018066406,
+      "learning_rate": 0.0005253697383390216,
+      "loss": 1.3532,
+      "step": 2090
+    },
+    {
+      "epoch": 2.378839590443686,
+      "grad_norm": 0.9490572810173035,
+      "learning_rate": 0.0005251422070534698,
+      "loss": 2.3154,
+      "step": 2091
+    },
+    {
+      "epoch": 2.379977246871445,
+      "grad_norm": 1.2259362936019897,
+      "learning_rate": 0.0005249146757679181,
+      "loss": 2.4494,
+      "step": 2092
+    },
+    {
+      "epoch": 2.3811149032992036,
+      "grad_norm": 0.7659657001495361,
+      "learning_rate": 0.0005246871444823663,
+      "loss": 1.1013,
+      "step": 2093
+    },
+    {
+      "epoch": 2.3822525597269624,
+      "grad_norm": 1.1086747646331787,
+      "learning_rate": 0.0005244596131968145,
+      "loss": 2.263,
+      "step": 2094
+    },
+    {
+      "epoch": 2.383390216154721,
+      "grad_norm": 0.9705541729927063,
+      "learning_rate": 0.0005242320819112628,
+      "loss": 1.4336,
+      "step": 2095
+    },
+    {
+      "epoch": 2.3845278725824803,
+      "grad_norm": 0.5771173238754272,
+      "learning_rate": 0.000524004550625711,
+      "loss": 0.8406,
+      "step": 2096
+    },
+    {
+      "epoch": 2.385665529010239,
+      "grad_norm": 0.424938440322876,
+      "learning_rate": 0.0005237770193401593,
+      "loss": 0.5156,
+      "step": 2097
+    },
+    {
+      "epoch": 2.386803185437998,
+      "grad_norm": 0.7281388640403748,
+      "learning_rate": 0.0005235494880546076,
+      "loss": 0.8201,
+      "step": 2098
+    },
+    {
+      "epoch": 2.3879408418657566,
+      "grad_norm": 1.5582374334335327,
+      "learning_rate": 0.0005233219567690558,
+      "loss": 3.5289,
+      "step": 2099
+    },
+    {
+      "epoch": 2.3890784982935154,
+      "grad_norm": 1.2791839838027954,
+      "learning_rate": 0.000523094425483504,
+      "loss": 1.6449,
+      "step": 2100
+    },
+    {
+      "epoch": 2.390216154721274,
+      "grad_norm": 1.1435104608535767,
+      "learning_rate": 0.0005228668941979522,
+      "loss": 2.1264,
+      "step": 2101
+    },
+    {
+      "epoch": 2.391353811149033,
+      "grad_norm": 1.234621286392212,
+      "learning_rate": 0.0005226393629124004,
+      "loss": 1.0853,
+      "step": 2102
+    },
+    {
+      "epoch": 2.3924914675767917,
+      "grad_norm": 0.9681219458580017,
+      "learning_rate": 0.0005224118316268487,
+      "loss": 2.4709,
+      "step": 2103
+    },
+    {
+      "epoch": 2.3936291240045504,
+      "grad_norm": 1.064215064048767,
+      "learning_rate": 0.0005221843003412969,
+      "loss": 2.7476,
+      "step": 2104
+    },
+    {
+      "epoch": 2.3947667804323096,
+      "grad_norm": 1.3225258588790894,
+      "learning_rate": 0.0005219567690557452,
+      "loss": 2.7113,
+      "step": 2105
+    },
+    {
+      "epoch": 2.3959044368600684,
+      "grad_norm": 1.2326604127883911,
+      "learning_rate": 0.0005217292377701935,
+      "loss": 1.7622,
+      "step": 2106
+    },
+    {
+      "epoch": 2.397042093287827,
+      "grad_norm": 0.7115911841392517,
+      "learning_rate": 0.0005215017064846417,
+      "loss": 1.8257,
+      "step": 2107
+    },
+    {
+      "epoch": 2.398179749715586,
+      "grad_norm": 0.762967050075531,
+      "learning_rate": 0.0005212741751990899,
+      "loss": 1.5422,
+      "step": 2108
+    },
+    {
+      "epoch": 2.3993174061433447,
+      "grad_norm": 1.2912230491638184,
+      "learning_rate": 0.0005210466439135382,
+      "loss": 2.3056,
+      "step": 2109
+    },
+    {
+      "epoch": 2.4004550625711034,
+      "grad_norm": 0.8842760920524597,
+      "learning_rate": 0.0005208191126279863,
+      "loss": 1.9213,
+      "step": 2110
+    },
+    {
+      "epoch": 2.401592718998862,
+      "grad_norm": 1.1768587827682495,
+      "learning_rate": 0.0005205915813424345,
+      "loss": 2.635,
+      "step": 2111
+    },
+    {
+      "epoch": 2.4027303754266214,
+      "grad_norm": 0.7537290453910828,
+      "learning_rate": 0.0005203640500568828,
+      "loss": 1.4866,
+      "step": 2112
+    },
+    {
+      "epoch": 2.40386803185438,
+      "grad_norm": 0.5235282182693481,
+      "learning_rate": 0.000520136518771331,
+      "loss": 1.3188,
+      "step": 2113
+    },
+    {
+      "epoch": 2.405005688282139,
+      "grad_norm": 1.8466582298278809,
+      "learning_rate": 0.0005199089874857793,
+      "loss": 3.5333,
+      "step": 2114
+    },
+    {
+      "epoch": 2.4061433447098977,
+      "grad_norm": 2.3424501419067383,
+      "learning_rate": 0.0005196814562002276,
+      "loss": 1.6052,
+      "step": 2115
+    },
+    {
+      "epoch": 2.4072810011376564,
+      "grad_norm": 1.0736968517303467,
+      "learning_rate": 0.0005194539249146758,
+      "loss": 1.6249,
+      "step": 2116
+    },
+    {
+      "epoch": 2.408418657565415,
+      "grad_norm": 1.0152912139892578,
+      "learning_rate": 0.000519226393629124,
+      "loss": 2.1684,
+      "step": 2117
+    },
+    {
+      "epoch": 2.409556313993174,
+      "grad_norm": 0.896634578704834,
+      "learning_rate": 0.0005189988623435723,
+      "loss": 1.4528,
+      "step": 2118
+    },
+    {
+      "epoch": 2.4106939704209327,
+      "grad_norm": 0.5919451713562012,
+      "learning_rate": 0.0005187713310580204,
+      "loss": 1.2616,
+      "step": 2119
+    },
+    {
+      "epoch": 2.4118316268486915,
+      "grad_norm": 0.9945755004882812,
+      "learning_rate": 0.0005185437997724687,
+      "loss": 1.9749,
+      "step": 2120
+    },
+    {
+      "epoch": 2.4129692832764507,
+      "grad_norm": 0.6944431066513062,
+      "learning_rate": 0.000518316268486917,
+      "loss": 1.538,
+      "step": 2121
+    },
+    {
+      "epoch": 2.4141069397042094,
+      "grad_norm": 1.0082086324691772,
+      "learning_rate": 0.0005180887372013652,
+      "loss": 1.2937,
+      "step": 2122
+    },
+    {
+      "epoch": 2.415244596131968,
+      "grad_norm": 0.8779463768005371,
+      "learning_rate": 0.0005178612059158135,
+      "loss": 2.8401,
+      "step": 2123
+    },
+    {
+      "epoch": 2.416382252559727,
+      "grad_norm": 0.729810357093811,
+      "learning_rate": 0.0005176336746302617,
+      "loss": 2.1615,
+      "step": 2124
+    },
+    {
+      "epoch": 2.4175199089874857,
+      "grad_norm": 0.5567348003387451,
+      "learning_rate": 0.0005174061433447099,
+      "loss": 1.0664,
+      "step": 2125
+    },
+    {
+      "epoch": 2.4186575654152445,
+      "grad_norm": 0.7598175406455994,
+      "learning_rate": 0.0005171786120591582,
+      "loss": 1.6606,
+      "step": 2126
+    },
+    {
+      "epoch": 2.419795221843003,
+      "grad_norm": 0.9565114378929138,
+      "learning_rate": 0.0005169510807736064,
+      "loss": 1.97,
+      "step": 2127
+    },
+    {
+      "epoch": 2.4209328782707624,
+      "grad_norm": 1.6986292600631714,
+      "learning_rate": 0.0005167235494880546,
+      "loss": 3.6617,
+      "step": 2128
+    },
+    {
+      "epoch": 2.422070534698521,
+      "grad_norm": 0.8568170666694641,
+      "learning_rate": 0.0005164960182025028,
+      "loss": 2.0927,
+      "step": 2129
+    },
+    {
+      "epoch": 2.42320819112628,
+      "grad_norm": 1.218289852142334,
+      "learning_rate": 0.000516268486916951,
+      "loss": 2.0588,
+      "step": 2130
+    },
+    {
+      "epoch": 2.4243458475540387,
+      "grad_norm": 1.114538550376892,
+      "learning_rate": 0.0005160409556313993,
+      "loss": 2.6901,
+      "step": 2131
+    },
+    {
+      "epoch": 2.4254835039817975,
+      "grad_norm": 0.9539201855659485,
+      "learning_rate": 0.0005158134243458476,
+      "loss": 1.3669,
+      "step": 2132
+    },
+    {
+      "epoch": 2.426621160409556,
+      "grad_norm": 1.0585428476333618,
+      "learning_rate": 0.0005155858930602958,
+      "loss": 2.5264,
+      "step": 2133
+    },
+    {
+      "epoch": 2.427758816837315,
+      "grad_norm": 0.5750988125801086,
+      "learning_rate": 0.000515358361774744,
+      "loss": 1.1382,
+      "step": 2134
+    },
+    {
+      "epoch": 2.428896473265074,
+      "grad_norm": 1.5301157236099243,
+      "learning_rate": 0.0005151308304891923,
+      "loss": 2.8573,
+      "step": 2135
+    },
+    {
+      "epoch": 2.430034129692833,
+      "grad_norm": 0.6881375312805176,
+      "learning_rate": 0.0005149032992036405,
+      "loss": 1.4282,
+      "step": 2136
+    },
+    {
+      "epoch": 2.4311717861205917,
+      "grad_norm": 0.642068088054657,
+      "learning_rate": 0.0005146757679180887,
+      "loss": 0.8719,
+      "step": 2137
+    },
+    {
+      "epoch": 2.4323094425483505,
+      "grad_norm": 0.9515458941459656,
+      "learning_rate": 0.0005144482366325371,
+      "loss": 1.7866,
+      "step": 2138
+    },
+    {
+      "epoch": 2.4334470989761092,
+      "grad_norm": 1.4139736890792847,
+      "learning_rate": 0.0005142207053469852,
+      "loss": 2.3178,
+      "step": 2139
+    },
+    {
+      "epoch": 2.434584755403868,
+      "grad_norm": 0.7378571629524231,
+      "learning_rate": 0.0005139931740614334,
+      "loss": 1.5993,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4357224118316267,
+      "grad_norm": 1.243595004081726,
+      "learning_rate": 0.0005137656427758817,
+      "loss": 2.0773,
+      "step": 2141
+    },
+    {
+      "epoch": 2.4368600682593855,
+      "grad_norm": 0.7528473734855652,
+      "learning_rate": 0.0005135381114903299,
+      "loss": 1.8747,
+      "step": 2142
+    },
+    {
+      "epoch": 2.4379977246871443,
+      "grad_norm": 0.9805837273597717,
+      "learning_rate": 0.0005133105802047782,
+      "loss": 2.3093,
+      "step": 2143
+    },
+    {
+      "epoch": 2.4391353811149035,
+      "grad_norm": 1.0196789503097534,
+      "learning_rate": 0.0005130830489192264,
+      "loss": 1.6322,
+      "step": 2144
+    },
+    {
+      "epoch": 2.4402730375426622,
+      "grad_norm": 1.2920022010803223,
+      "learning_rate": 0.0005128555176336746,
+      "loss": 1.0212,
+      "step": 2145
+    },
+    {
+      "epoch": 2.441410693970421,
+      "grad_norm": 1.036919116973877,
+      "learning_rate": 0.000512627986348123,
+      "loss": 2.4244,
+      "step": 2146
+    },
+    {
+      "epoch": 2.4425483503981797,
+      "grad_norm": 0.7487674951553345,
+      "learning_rate": 0.0005124004550625712,
+      "loss": 1.2761,
+      "step": 2147
+    },
+    {
+      "epoch": 2.4436860068259385,
+      "grad_norm": 0.9197911024093628,
+      "learning_rate": 0.0005121729237770194,
+      "loss": 2.1063,
+      "step": 2148
+    },
+    {
+      "epoch": 2.4448236632536973,
+      "grad_norm": 0.9425987005233765,
+      "learning_rate": 0.0005119453924914676,
+      "loss": 1.6781,
+      "step": 2149
+    },
+    {
+      "epoch": 2.445961319681456,
+      "grad_norm": 1.1719930171966553,
+      "learning_rate": 0.0005117178612059158,
+      "loss": 2.0511,
+      "step": 2150
+    },
+    {
+      "epoch": 2.4470989761092152,
+      "grad_norm": 0.8875211477279663,
+      "learning_rate": 0.000511490329920364,
+      "loss": 1.6565,
+      "step": 2151
+    },
+    {
+      "epoch": 2.448236632536974,
+      "grad_norm": 0.61222243309021,
+      "learning_rate": 0.0005112627986348123,
+      "loss": 1.1201,
+      "step": 2152
+    },
+    {
+      "epoch": 2.4493742889647327,
+      "grad_norm": 1.5771725177764893,
+      "learning_rate": 0.0005110352673492605,
+      "loss": 3.2926,
+      "step": 2153
+    },
+    {
+      "epoch": 2.4505119453924915,
+      "grad_norm": 0.8051725029945374,
+      "learning_rate": 0.0005108077360637087,
+      "loss": 1.0753,
+      "step": 2154
+    },
+    {
+      "epoch": 2.4516496018202503,
+      "grad_norm": 0.7018311619758606,
+      "learning_rate": 0.0005105802047781571,
+      "loss": 0.9978,
+      "step": 2155
+    },
+    {
+      "epoch": 2.452787258248009,
+      "grad_norm": 1.0658284425735474,
+      "learning_rate": 0.0005103526734926053,
+      "loss": 3.1038,
+      "step": 2156
+    },
+    {
+      "epoch": 2.453924914675768,
+      "grad_norm": 0.890017032623291,
+      "learning_rate": 0.0005101251422070535,
+      "loss": 1.3806,
+      "step": 2157
+    },
+    {
+      "epoch": 2.4550625711035265,
+      "grad_norm": 1.4623593091964722,
+      "learning_rate": 0.0005098976109215017,
+      "loss": 3.3608,
+      "step": 2158
+    },
+    {
+      "epoch": 2.4562002275312853,
+      "grad_norm": 1.3333830833435059,
+      "learning_rate": 0.0005096700796359499,
+      "loss": 1.9364,
+      "step": 2159
+    },
+    {
+      "epoch": 2.4573378839590445,
+      "grad_norm": 1.0337042808532715,
+      "learning_rate": 0.0005094425483503981,
+      "loss": 1.6893,
+      "step": 2160
+    },
+    {
+      "epoch": 2.4584755403868033,
+      "grad_norm": 1.2401093244552612,
+      "learning_rate": 0.0005092150170648464,
+      "loss": 1.8306,
+      "step": 2161
+    },
+    {
+      "epoch": 2.459613196814562,
+      "grad_norm": 0.6069979071617126,
+      "learning_rate": 0.0005089874857792946,
+      "loss": 1.5413,
+      "step": 2162
+    },
+    {
+      "epoch": 2.460750853242321,
+      "grad_norm": 1.1139168739318848,
+      "learning_rate": 0.000508759954493743,
+      "loss": 2.6418,
+      "step": 2163
+    },
+    {
+      "epoch": 2.4618885096700796,
+      "grad_norm": 0.7371734976768494,
+      "learning_rate": 0.0005085324232081912,
+      "loss": 1.3743,
+      "step": 2164
+    },
+    {
+      "epoch": 2.4630261660978383,
+      "grad_norm": 1.3759161233901978,
+      "learning_rate": 0.0005083048919226394,
+      "loss": 2.3257,
+      "step": 2165
+    },
+    {
+      "epoch": 2.464163822525597,
+      "grad_norm": 1.0357613563537598,
+      "learning_rate": 0.0005080773606370877,
+      "loss": 1.6746,
+      "step": 2166
+    },
+    {
+      "epoch": 2.4653014789533563,
+      "grad_norm": 0.5313310623168945,
+      "learning_rate": 0.0005078498293515359,
+      "loss": 0.8946,
+      "step": 2167
+    },
+    {
+      "epoch": 2.466439135381115,
+      "grad_norm": 0.9679710268974304,
+      "learning_rate": 0.000507622298065984,
+      "loss": 1.906,
+      "step": 2168
+    },
+    {
+      "epoch": 2.467576791808874,
+      "grad_norm": 0.7682181596755981,
+      "learning_rate": 0.0005073947667804323,
+      "loss": 1.9641,
+      "step": 2169
+    },
+    {
+      "epoch": 2.4687144482366326,
+      "grad_norm": 0.6322567462921143,
+      "learning_rate": 0.0005071672354948805,
+      "loss": 0.5872,
+      "step": 2170
+    },
+    {
+      "epoch": 2.4698521046643913,
+      "grad_norm": 0.6293168663978577,
+      "learning_rate": 0.0005069397042093287,
+      "loss": 1.1534,
+      "step": 2171
+    },
+    {
+      "epoch": 2.47098976109215,
+      "grad_norm": 0.9249051213264465,
+      "learning_rate": 0.0005067121729237771,
+      "loss": 2.3496,
+      "step": 2172
+    },
+    {
+      "epoch": 2.472127417519909,
+      "grad_norm": 0.8873506784439087,
+      "learning_rate": 0.0005064846416382253,
+      "loss": 1.5146,
+      "step": 2173
+    },
+    {
+      "epoch": 2.473265073947668,
+      "grad_norm": 0.8709667325019836,
+      "learning_rate": 0.0005062571103526735,
+      "loss": 1.8368,
+      "step": 2174
+    },
+    {
+      "epoch": 2.474402730375427,
+      "grad_norm": 0.8933395147323608,
+      "learning_rate": 0.0005060295790671218,
+      "loss": 1.5841,
+      "step": 2175
+    },
+    {
+      "epoch": 2.4755403868031856,
+      "grad_norm": 0.8149117231369019,
+      "learning_rate": 0.00050580204778157,
+      "loss": 1.6464,
+      "step": 2176
+    },
+    {
+      "epoch": 2.4766780432309443,
+      "grad_norm": 0.8341697454452515,
+      "learning_rate": 0.0005055745164960182,
+      "loss": 1.1955,
+      "step": 2177
+    },
+    {
+      "epoch": 2.477815699658703,
+      "grad_norm": 1.2008378505706787,
+      "learning_rate": 0.0005053469852104664,
+      "loss": 2.0201,
+      "step": 2178
+    },
+    {
+      "epoch": 2.478953356086462,
+      "grad_norm": 1.0396666526794434,
+      "learning_rate": 0.0005051194539249146,
+      "loss": 1.9521,
+      "step": 2179
+    },
+    {
+      "epoch": 2.4800910125142206,
+      "grad_norm": 0.7898449897766113,
+      "learning_rate": 0.0005048919226393628,
+      "loss": 1.1885,
+      "step": 2180
+    },
+    {
+      "epoch": 2.4812286689419794,
+      "grad_norm": 0.9955923557281494,
+      "learning_rate": 0.0005046643913538112,
+      "loss": 1.8145,
+      "step": 2181
+    },
+    {
+      "epoch": 2.482366325369738,
+      "grad_norm": 1.1414570808410645,
+      "learning_rate": 0.0005044368600682594,
+      "loss": 2.2897,
+      "step": 2182
+    },
+    {
+      "epoch": 2.4835039817974973,
+      "grad_norm": 1.0973477363586426,
+      "learning_rate": 0.0005042093287827077,
+      "loss": 2.2097,
+      "step": 2183
+    },
+    {
+      "epoch": 2.484641638225256,
+      "grad_norm": 0.955952525138855,
+      "learning_rate": 0.0005039817974971559,
+      "loss": 1.6976,
+      "step": 2184
+    },
+    {
+      "epoch": 2.485779294653015,
+      "grad_norm": 0.9474056959152222,
+      "learning_rate": 0.0005037542662116041,
+      "loss": 1.5724,
+      "step": 2185
+    },
+    {
+      "epoch": 2.4869169510807736,
+      "grad_norm": 0.7748664021492004,
+      "learning_rate": 0.0005035267349260524,
+      "loss": 2.1302,
+      "step": 2186
+    },
+    {
+      "epoch": 2.4880546075085324,
+      "grad_norm": 1.1004105806350708,
+      "learning_rate": 0.0005032992036405005,
+      "loss": 1.9072,
+      "step": 2187
+    },
+    {
+      "epoch": 2.489192263936291,
+      "grad_norm": 0.8701795935630798,
+      "learning_rate": 0.0005030716723549487,
+      "loss": 1.2504,
+      "step": 2188
+    },
+    {
+      "epoch": 2.49032992036405,
+      "grad_norm": 1.0033451318740845,
+      "learning_rate": 0.0005028441410693971,
+      "loss": 2.6241,
+      "step": 2189
+    },
+    {
+      "epoch": 2.491467576791809,
+      "grad_norm": 0.9840807318687439,
+      "learning_rate": 0.0005026166097838453,
+      "loss": 1.5594,
+      "step": 2190
+    },
+    {
+      "epoch": 2.492605233219568,
+      "grad_norm": 1.0026658773422241,
+      "learning_rate": 0.0005023890784982935,
+      "loss": 2.1102,
+      "step": 2191
+    },
+    {
+      "epoch": 2.4937428896473266,
+      "grad_norm": 0.8774160146713257,
+      "learning_rate": 0.0005021615472127418,
+      "loss": 2.5087,
+      "step": 2192
+    },
+    {
+      "epoch": 2.4948805460750854,
+      "grad_norm": 0.7475762963294983,
+      "learning_rate": 0.00050193401592719,
+      "loss": 0.8467,
+      "step": 2193
+    },
+    {
+      "epoch": 2.496018202502844,
+      "grad_norm": 1.0985372066497803,
+      "learning_rate": 0.0005017064846416382,
+      "loss": 2.0641,
+      "step": 2194
+    },
+    {
+      "epoch": 2.497155858930603,
+      "grad_norm": 0.8731891512870789,
+      "learning_rate": 0.0005014789533560865,
+      "loss": 2.1272,
+      "step": 2195
+    },
+    {
+      "epoch": 2.4982935153583616,
+      "grad_norm": 0.9103354215621948,
+      "learning_rate": 0.0005012514220705347,
+      "loss": 2.0572,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4994311717861204,
+      "grad_norm": 1.041029691696167,
+      "learning_rate": 0.0005010238907849828,
+      "loss": 1.9749,
+      "step": 2197
+    },
+    {
+      "epoch": 2.500568828213879,
+      "grad_norm": 1.2097538709640503,
+      "learning_rate": 0.0005007963594994312,
+      "loss": 1.671,
+      "step": 2198
+    },
+    {
+      "epoch": 2.5017064846416384,
+      "grad_norm": 0.9088473916053772,
+      "learning_rate": 0.0005005688282138794,
+      "loss": 2.0683,
+      "step": 2199
+    },
+    {
+      "epoch": 2.502844141069397,
+      "grad_norm": 0.728599488735199,
+      "learning_rate": 0.0005003412969283276,
+      "loss": 1.4432,
+      "step": 2200
+    },
+    {
+      "epoch": 2.503981797497156,
+      "grad_norm": 0.7432947754859924,
+      "learning_rate": 0.0005001137656427759,
+      "loss": 1.3989,
+      "step": 2201
+    },
+    {
+      "epoch": 2.5051194539249146,
+      "grad_norm": 0.9259990453720093,
+      "learning_rate": 0.0004998862343572241,
+      "loss": 2.1398,
+      "step": 2202
+    },
+    {
+      "epoch": 2.5062571103526734,
+      "grad_norm": 1.316038966178894,
+      "learning_rate": 0.0004996587030716724,
+      "loss": 2.2743,
+      "step": 2203
+    },
+    {
+      "epoch": 2.507394766780432,
+      "grad_norm": 0.8661054372787476,
+      "learning_rate": 0.0004994311717861205,
+      "loss": 1.706,
+      "step": 2204
+    },
+    {
+      "epoch": 2.508532423208191,
+      "grad_norm": 0.8994172215461731,
+      "learning_rate": 0.0004992036405005689,
+      "loss": 1.6767,
+      "step": 2205
+    },
+    {
+      "epoch": 2.50967007963595,
+      "grad_norm": 0.683313250541687,
+      "learning_rate": 0.0004989761092150171,
+      "loss": 0.8057,
+      "step": 2206
+    },
+    {
+      "epoch": 2.510807736063709,
+      "grad_norm": 1.4154108762741089,
+      "learning_rate": 0.0004987485779294653,
+      "loss": 1.5263,
+      "step": 2207
+    },
+    {
+      "epoch": 2.5119453924914676,
+      "grad_norm": 1.0941131114959717,
+      "learning_rate": 0.0004985210466439136,
+      "loss": 1.7871,
+      "step": 2208
+    },
+    {
+      "epoch": 2.5130830489192264,
+      "grad_norm": 0.8073767423629761,
+      "learning_rate": 0.0004982935153583618,
+      "loss": 1.7233,
+      "step": 2209
+    },
+    {
+      "epoch": 2.514220705346985,
+      "grad_norm": 1.2724852561950684,
+      "learning_rate": 0.00049806598407281,
+      "loss": 2.4039,
+      "step": 2210
+    },
+    {
+      "epoch": 2.515358361774744,
+      "grad_norm": 0.6580117344856262,
+      "learning_rate": 0.0004978384527872582,
+      "loss": 1.5502,
+      "step": 2211
+    },
+    {
+      "epoch": 2.5164960182025027,
+      "grad_norm": 0.9353324174880981,
+      "learning_rate": 0.0004976109215017065,
+      "loss": 1.7699,
+      "step": 2212
+    },
+    {
+      "epoch": 2.517633674630262,
+      "grad_norm": 0.7029836177825928,
+      "learning_rate": 0.0004973833902161547,
+      "loss": 1.4815,
+      "step": 2213
+    },
+    {
+      "epoch": 2.51877133105802,
+      "grad_norm": 0.8808972239494324,
+      "learning_rate": 0.000497155858930603,
+      "loss": 1.9639,
+      "step": 2214
+    },
+    {
+      "epoch": 2.5199089874857794,
+      "grad_norm": 0.7552292943000793,
+      "learning_rate": 0.0004969283276450512,
+      "loss": 0.9553,
+      "step": 2215
+    },
+    {
+      "epoch": 2.521046643913538,
+      "grad_norm": 0.8695642948150635,
+      "learning_rate": 0.0004967007963594995,
+      "loss": 1.8042,
+      "step": 2216
+    },
+    {
+      "epoch": 2.522184300341297,
+      "grad_norm": 1.5711946487426758,
+      "learning_rate": 0.0004964732650739477,
+      "loss": 2.2919,
+      "step": 2217
+    },
+    {
+      "epoch": 2.5233219567690557,
+      "grad_norm": 1.2966930866241455,
+      "learning_rate": 0.0004962457337883959,
+      "loss": 2.9262,
+      "step": 2218
+    },
+    {
+      "epoch": 2.5244596131968144,
+      "grad_norm": 0.7041467428207397,
+      "learning_rate": 0.0004960182025028441,
+      "loss": 1.2383,
+      "step": 2219
+    },
+    {
+      "epoch": 2.5255972696245736,
+      "grad_norm": 0.8301080465316772,
+      "learning_rate": 0.0004957906712172923,
+      "loss": 1.9915,
+      "step": 2220
+    },
+    {
+      "epoch": 2.526734926052332,
+      "grad_norm": 0.8538893461227417,
+      "learning_rate": 0.0004955631399317406,
+      "loss": 2.2926,
+      "step": 2221
+    },
+    {
+      "epoch": 2.527872582480091,
+      "grad_norm": 1.1304670572280884,
+      "learning_rate": 0.0004953356086461889,
+      "loss": 2.2345,
+      "step": 2222
+    },
+    {
+      "epoch": 2.52901023890785,
+      "grad_norm": 1.1222370862960815,
+      "learning_rate": 0.0004951080773606372,
+      "loss": 1.9092,
+      "step": 2223
+    },
+    {
+      "epoch": 2.5301478953356087,
+      "grad_norm": 1.1593031883239746,
+      "learning_rate": 0.0004948805460750853,
+      "loss": 2.5001,
+      "step": 2224
+    },
+    {
+      "epoch": 2.5312855517633674,
+      "grad_norm": 0.9046427011489868,
+      "learning_rate": 0.0004946530147895336,
+      "loss": 1.8159,
+      "step": 2225
+    },
+    {
+      "epoch": 2.532423208191126,
+      "grad_norm": 0.9191851615905762,
+      "learning_rate": 0.0004944254835039818,
+      "loss": 2.3244,
+      "step": 2226
+    },
+    {
+      "epoch": 2.533560864618885,
+      "grad_norm": 0.8399704098701477,
+      "learning_rate": 0.00049419795221843,
+      "loss": 2.1157,
+      "step": 2227
+    },
+    {
+      "epoch": 2.5346985210466437,
+      "grad_norm": 1.3385767936706543,
+      "learning_rate": 0.0004939704209328783,
+      "loss": 2.495,
+      "step": 2228
+    },
+    {
+      "epoch": 2.535836177474403,
+      "grad_norm": 0.9084081053733826,
+      "learning_rate": 0.0004937428896473265,
+      "loss": 1.222,
+      "step": 2229
+    },
+    {
+      "epoch": 2.5369738339021617,
+      "grad_norm": 1.0885717868804932,
+      "learning_rate": 0.0004935153583617748,
+      "loss": 2.0583,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5381114903299204,
+      "grad_norm": 0.6400699019432068,
+      "learning_rate": 0.000493287827076223,
+      "loss": 0.7095,
+      "step": 2231
+    },
+    {
+      "epoch": 2.539249146757679,
+      "grad_norm": 0.8817832469940186,
+      "learning_rate": 0.0004930602957906713,
+      "loss": 2.0175,
+      "step": 2232
+    },
+    {
+      "epoch": 2.540386803185438,
+      "grad_norm": 0.5931545495986938,
+      "learning_rate": 0.0004928327645051195,
+      "loss": 1.6168,
+      "step": 2233
+    },
+    {
+      "epoch": 2.5415244596131967,
+      "grad_norm": 1.0301717519760132,
+      "learning_rate": 0.0004926052332195677,
+      "loss": 1.5534,
+      "step": 2234
+    },
+    {
+      "epoch": 2.5426621160409555,
+      "grad_norm": 0.6240630149841309,
+      "learning_rate": 0.0004923777019340159,
+      "loss": 0.7599,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5437997724687147,
+      "grad_norm": 0.7773973345756531,
+      "learning_rate": 0.0004921501706484642,
+      "loss": 2.37,
+      "step": 2236
+    },
+    {
+      "epoch": 2.544937428896473,
+      "grad_norm": 0.7759304046630859,
+      "learning_rate": 0.0004919226393629124,
+      "loss": 1.0116,
+      "step": 2237
+    },
+    {
+      "epoch": 2.546075085324232,
+      "grad_norm": 1.2322027683258057,
+      "learning_rate": 0.0004916951080773606,
+      "loss": 3.2444,
+      "step": 2238
+    },
+    {
+      "epoch": 2.547212741751991,
+      "grad_norm": 0.8914145231246948,
+      "learning_rate": 0.0004914675767918089,
+      "loss": 1.5857,
+      "step": 2239
+    },
+    {
+      "epoch": 2.5483503981797497,
+      "grad_norm": 0.9285094738006592,
+      "learning_rate": 0.0004912400455062571,
+      "loss": 1.8327,
+      "step": 2240
+    },
+    {
+      "epoch": 2.5494880546075085,
+      "grad_norm": 1.1583962440490723,
+      "learning_rate": 0.0004910125142207054,
+      "loss": 2.0139,
+      "step": 2241
+    },
+    {
+      "epoch": 2.5506257110352673,
+      "grad_norm": 1.8207440376281738,
+      "learning_rate": 0.0004907849829351536,
+      "loss": 1.8943,
+      "step": 2242
+    },
+    {
+      "epoch": 2.551763367463026,
+      "grad_norm": 1.2174469232559204,
+      "learning_rate": 0.0004905574516496018,
+      "loss": 2.1719,
+      "step": 2243
+    },
+    {
+      "epoch": 2.5529010238907848,
+      "grad_norm": 1.4703840017318726,
+      "learning_rate": 0.00049032992036405,
+      "loss": 3.2219,
+      "step": 2244
+    },
+    {
+      "epoch": 2.554038680318544,
+      "grad_norm": 0.9278658628463745,
+      "learning_rate": 0.0004901023890784983,
+      "loss": 1.4814,
+      "step": 2245
+    },
+    {
+      "epoch": 2.5551763367463027,
+      "grad_norm": 0.5682125687599182,
+      "learning_rate": 0.0004898748577929465,
+      "loss": 0.9685,
+      "step": 2246
+    },
+    {
+      "epoch": 2.5563139931740615,
+      "grad_norm": 0.5616304278373718,
+      "learning_rate": 0.0004896473265073948,
+      "loss": 1.0696,
+      "step": 2247
+    },
+    {
+      "epoch": 2.5574516496018203,
+      "grad_norm": 0.9043198227882385,
+      "learning_rate": 0.000489419795221843,
+      "loss": 1.2706,
+      "step": 2248
+    },
+    {
+      "epoch": 2.558589306029579,
+      "grad_norm": 1.1037359237670898,
+      "learning_rate": 0.0004891922639362913,
+      "loss": 1.8091,
+      "step": 2249
+    },
+    {
+      "epoch": 2.5597269624573378,
+      "grad_norm": 1.0690373182296753,
+      "learning_rate": 0.0004889647326507395,
+      "loss": 1.6337,
+      "step": 2250
+    },
+    {
+      "epoch": 2.5608646188850965,
+      "grad_norm": 0.9125576019287109,
+      "learning_rate": 0.0004887372013651877,
+      "loss": 2.0998,
+      "step": 2251
+    },
+    {
+      "epoch": 2.5620022753128557,
+      "grad_norm": 1.3516008853912354,
+      "learning_rate": 0.000488509670079636,
+      "loss": 2.317,
+      "step": 2252
+    },
+    {
+      "epoch": 2.5631399317406145,
+      "grad_norm": 1.6707299947738647,
+      "learning_rate": 0.0004882821387940841,
+      "loss": 2.8554,
+      "step": 2253
+    },
+    {
+      "epoch": 2.5642775881683733,
+      "grad_norm": 0.7115156054496765,
+      "learning_rate": 0.00048805460750853244,
+      "loss": 0.7969,
+      "step": 2254
+    },
+    {
+      "epoch": 2.565415244596132,
+      "grad_norm": 0.7895395755767822,
+      "learning_rate": 0.00048782707622298065,
+      "loss": 1.3132,
+      "step": 2255
+    },
+    {
+      "epoch": 2.5665529010238908,
+      "grad_norm": 1.3991349935531616,
+      "learning_rate": 0.0004875995449374289,
+      "loss": 2.6118,
+      "step": 2256
+    },
+    {
+      "epoch": 2.5676905574516495,
+      "grad_norm": 1.0328587293624878,
+      "learning_rate": 0.0004873720136518772,
+      "loss": 1.8474,
+      "step": 2257
+    },
+    {
+      "epoch": 2.5688282138794083,
+      "grad_norm": 0.8418103456497192,
+      "learning_rate": 0.00048714448236632533,
+      "loss": 1.9257,
+      "step": 2258
+    },
+    {
+      "epoch": 2.5699658703071675,
+      "grad_norm": 1.1641095876693726,
+      "learning_rate": 0.0004869169510807736,
+      "loss": 1.6591,
+      "step": 2259
+    },
+    {
+      "epoch": 2.571103526734926,
+      "grad_norm": 1.1286218166351318,
+      "learning_rate": 0.00048668941979522186,
+      "loss": 1.7781,
+      "step": 2260
+    },
+    {
+      "epoch": 2.572241183162685,
+      "grad_norm": 0.9085434675216675,
+      "learning_rate": 0.0004864618885096701,
+      "loss": 1.7569,
+      "step": 2261
+    },
+    {
+      "epoch": 2.573378839590444,
+      "grad_norm": 0.6849757432937622,
+      "learning_rate": 0.00048623435722411833,
+      "loss": 1.2772,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5745164960182025,
+      "grad_norm": 0.6630122661590576,
+      "learning_rate": 0.00048600682593856654,
+      "loss": 1.5381,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5756541524459613,
+      "grad_norm": 0.7607198357582092,
+      "learning_rate": 0.0004857792946530148,
+      "loss": 1.3695,
+      "step": 2264
+    },
+    {
+      "epoch": 2.57679180887372,
+      "grad_norm": 1.6203835010528564,
+      "learning_rate": 0.000485551763367463,
+      "loss": 2.5078,
+      "step": 2265
+    },
+    {
+      "epoch": 2.577929465301479,
+      "grad_norm": 1.9112815856933594,
+      "learning_rate": 0.0004853242320819113,
+      "loss": 1.0513,
+      "step": 2266
+    },
+    {
+      "epoch": 2.5790671217292376,
+      "grad_norm": 1.3533493280410767,
+      "learning_rate": 0.00048509670079635955,
+      "loss": 3.0494,
+      "step": 2267
+    },
+    {
+      "epoch": 2.580204778156997,
+      "grad_norm": 0.7832985520362854,
+      "learning_rate": 0.0004848691695108077,
+      "loss": 1.0865,
+      "step": 2268
+    },
+    {
+      "epoch": 2.5813424345847555,
+      "grad_norm": 0.8799665570259094,
+      "learning_rate": 0.00048464163822525597,
+      "loss": 2.473,
+      "step": 2269
+    },
+    {
+      "epoch": 2.5824800910125143,
+      "grad_norm": 1.135883092880249,
+      "learning_rate": 0.00048441410693970423,
+      "loss": 2.009,
+      "step": 2270
+    },
+    {
+      "epoch": 2.583617747440273,
+      "grad_norm": 0.740875780582428,
+      "learning_rate": 0.0004841865756541525,
+      "loss": 1.5996,
+      "step": 2271
+    },
+    {
+      "epoch": 2.584755403868032,
+      "grad_norm": 1.1426676511764526,
+      "learning_rate": 0.00048395904436860065,
+      "loss": 1.7545,
+      "step": 2272
+    },
+    {
+      "epoch": 2.5858930602957906,
+      "grad_norm": 1.1677348613739014,
+      "learning_rate": 0.0004837315130830489,
+      "loss": 3.6223,
+      "step": 2273
+    },
+    {
+      "epoch": 2.5870307167235493,
+      "grad_norm": 1.1222909688949585,
+      "learning_rate": 0.0004835039817974972,
+      "loss": 2.0058,
+      "step": 2274
+    },
+    {
+      "epoch": 2.5881683731513085,
+      "grad_norm": 0.7786708474159241,
+      "learning_rate": 0.0004832764505119454,
+      "loss": 1.3458,
+      "step": 2275
+    },
+    {
+      "epoch": 2.589306029579067,
+      "grad_norm": 0.6702552437782288,
+      "learning_rate": 0.00048304891922639365,
+      "loss": 1.3789,
+      "step": 2276
+    },
+    {
+      "epoch": 2.590443686006826,
+      "grad_norm": 1.0909732580184937,
+      "learning_rate": 0.00048282138794084186,
+      "loss": 2.0698,
+      "step": 2277
+    },
+    {
+      "epoch": 2.591581342434585,
+      "grad_norm": 0.9485477209091187,
+      "learning_rate": 0.00048259385665529007,
+      "loss": 1.8122,
+      "step": 2278
+    },
+    {
+      "epoch": 2.5927189988623436,
+      "grad_norm": 1.065976619720459,
+      "learning_rate": 0.00048236632536973834,
+      "loss": 2.2232,
+      "step": 2279
+    },
+    {
+      "epoch": 2.5938566552901023,
+      "grad_norm": 0.547852635383606,
+      "learning_rate": 0.0004821387940841866,
+      "loss": 1.5199,
+      "step": 2280
+    },
+    {
+      "epoch": 2.594994311717861,
+      "grad_norm": 0.7793285250663757,
+      "learning_rate": 0.00048191126279863486,
+      "loss": 1.6383,
+      "step": 2281
+    },
+    {
+      "epoch": 2.59613196814562,
+      "grad_norm": 1.2793906927108765,
+      "learning_rate": 0.000481683731513083,
+      "loss": 1.7647,
+      "step": 2282
+    },
+    {
+      "epoch": 2.5972696245733786,
+      "grad_norm": 0.7028161883354187,
+      "learning_rate": 0.0004814562002275313,
+      "loss": 0.9584,
+      "step": 2283
+    },
+    {
+      "epoch": 2.598407281001138,
+      "grad_norm": 0.725344717502594,
+      "learning_rate": 0.00048122866894197955,
+      "loss": 1.2296,
+      "step": 2284
+    },
+    {
+      "epoch": 2.5995449374288966,
+      "grad_norm": 0.9674602746963501,
+      "learning_rate": 0.00048100113765642776,
+      "loss": 2.275,
+      "step": 2285
+    },
+    {
+      "epoch": 2.6006825938566553,
+      "grad_norm": 0.7748560905456543,
+      "learning_rate": 0.000480773606370876,
+      "loss": 1.7323,
+      "step": 2286
+    },
+    {
+      "epoch": 2.601820250284414,
+      "grad_norm": 1.1546324491500854,
+      "learning_rate": 0.00048054607508532423,
+      "loss": 2.054,
+      "step": 2287
+    },
+    {
+      "epoch": 2.602957906712173,
+      "grad_norm": 0.7889047861099243,
+      "learning_rate": 0.00048031854379977244,
+      "loss": 1.787,
+      "step": 2288
+    },
+    {
+      "epoch": 2.6040955631399316,
+      "grad_norm": 1.0750888586044312,
+      "learning_rate": 0.0004800910125142207,
+      "loss": 2.1864,
+      "step": 2289
+    },
+    {
+      "epoch": 2.6052332195676904,
+      "grad_norm": 0.9969229698181152,
+      "learning_rate": 0.00047986348122866897,
+      "loss": 1.7515,
+      "step": 2290
+    },
+    {
+      "epoch": 2.6063708759954496,
+      "grad_norm": 0.8853392004966736,
+      "learning_rate": 0.00047963594994311723,
+      "loss": 1.1181,
+      "step": 2291
+    },
+    {
+      "epoch": 2.6075085324232083,
+      "grad_norm": 0.7410064935684204,
+      "learning_rate": 0.0004794084186575654,
+      "loss": 1.8175,
+      "step": 2292
+    },
+    {
+      "epoch": 2.608646188850967,
+      "grad_norm": 1.1709070205688477,
+      "learning_rate": 0.00047918088737201365,
+      "loss": 1.8607,
+      "step": 2293
+    },
+    {
+      "epoch": 2.609783845278726,
+      "grad_norm": 1.0198490619659424,
+      "learning_rate": 0.0004789533560864619,
+      "loss": 1.6995,
+      "step": 2294
+    },
+    {
+      "epoch": 2.6109215017064846,
+      "grad_norm": 1.765076994895935,
+      "learning_rate": 0.00047872582480091013,
+      "loss": 3.519,
+      "step": 2295
+    },
+    {
+      "epoch": 2.6120591581342434,
+      "grad_norm": 1.3096511363983154,
+      "learning_rate": 0.0004784982935153584,
+      "loss": 2.7398,
+      "step": 2296
+    },
+    {
+      "epoch": 2.613196814562002,
+      "grad_norm": 2.6314849853515625,
+      "learning_rate": 0.0004782707622298066,
+      "loss": 5.0637,
+      "step": 2297
+    },
+    {
+      "epoch": 2.6143344709897613,
+      "grad_norm": 0.6920953392982483,
+      "learning_rate": 0.0004780432309442548,
+      "loss": 1.1274,
+      "step": 2298
+    },
+    {
+      "epoch": 2.6154721274175197,
+      "grad_norm": 1.1631900072097778,
+      "learning_rate": 0.0004778156996587031,
+      "loss": 1.7851,
+      "step": 2299
+    },
+    {
+      "epoch": 2.616609783845279,
+      "grad_norm": 0.7960054874420166,
+      "learning_rate": 0.00047758816837315134,
+      "loss": 1.2316,
+      "step": 2300
+    },
+    {
+      "epoch": 2.6177474402730376,
+      "grad_norm": 1.066870927810669,
+      "learning_rate": 0.0004773606370875996,
+      "loss": 2.3944,
+      "step": 2301
+    },
+    {
+      "epoch": 2.6188850967007964,
+      "grad_norm": 0.7467948198318481,
+      "learning_rate": 0.00047713310580204776,
+      "loss": 1.1561,
+      "step": 2302
+    },
+    {
+      "epoch": 2.620022753128555,
+      "grad_norm": 0.7254666686058044,
+      "learning_rate": 0.000476905574516496,
+      "loss": 1.0797,
+      "step": 2303
+    },
+    {
+      "epoch": 2.621160409556314,
+      "grad_norm": 0.6046991944313049,
+      "learning_rate": 0.0004766780432309443,
+      "loss": 0.9486,
+      "step": 2304
+    },
+    {
+      "epoch": 2.6222980659840727,
+      "grad_norm": 1.0409255027770996,
+      "learning_rate": 0.0004764505119453925,
+      "loss": 1.9647,
+      "step": 2305
+    },
+    {
+      "epoch": 2.6234357224118314,
+      "grad_norm": 1.2402981519699097,
+      "learning_rate": 0.00047622298065984076,
+      "loss": 2.1702,
+      "step": 2306
+    },
+    {
+      "epoch": 2.6245733788395906,
+      "grad_norm": 1.274269700050354,
+      "learning_rate": 0.00047599544937428897,
+      "loss": 2.9181,
+      "step": 2307
+    },
+    {
+      "epoch": 2.6257110352673494,
+      "grad_norm": 0.724886953830719,
+      "learning_rate": 0.0004757679180887372,
+      "loss": 1.357,
+      "step": 2308
+    },
+    {
+      "epoch": 2.626848691695108,
+      "grad_norm": 0.9972879886627197,
+      "learning_rate": 0.00047554038680318545,
+      "loss": 2.0593,
+      "step": 2309
+    },
+    {
+      "epoch": 2.627986348122867,
+      "grad_norm": 0.8259227275848389,
+      "learning_rate": 0.0004753128555176337,
+      "loss": 1.7083,
+      "step": 2310
+    },
+    {
+      "epoch": 2.6291240045506257,
+      "grad_norm": 1.0254158973693848,
+      "learning_rate": 0.0004750853242320819,
+      "loss": 2.4944,
+      "step": 2311
+    },
+    {
+      "epoch": 2.6302616609783844,
+      "grad_norm": 0.8054444193840027,
+      "learning_rate": 0.00047485779294653013,
+      "loss": 1.1276,
+      "step": 2312
+    },
+    {
+      "epoch": 2.631399317406143,
+      "grad_norm": 0.7392967939376831,
+      "learning_rate": 0.0004746302616609784,
+      "loss": 1.3377,
+      "step": 2313
+    },
+    {
+      "epoch": 2.6325369738339024,
+      "grad_norm": 0.9780520796775818,
+      "learning_rate": 0.00047440273037542666,
+      "loss": 2.8929,
+      "step": 2314
+    },
+    {
+      "epoch": 2.6336746302616607,
+      "grad_norm": 1.0950572490692139,
+      "learning_rate": 0.00047417519908987487,
+      "loss": 2.1541,
+      "step": 2315
+    },
+    {
+      "epoch": 2.63481228668942,
+      "grad_norm": 0.9937504529953003,
+      "learning_rate": 0.0004739476678043231,
+      "loss": 1.8749,
+      "step": 2316
+    },
+    {
+      "epoch": 2.6359499431171787,
+      "grad_norm": 0.6954947710037231,
+      "learning_rate": 0.00047372013651877134,
+      "loss": 1.8055,
+      "step": 2317
+    },
+    {
+      "epoch": 2.6370875995449374,
+      "grad_norm": 0.8226844668388367,
+      "learning_rate": 0.00047349260523321955,
+      "loss": 1.5377,
+      "step": 2318
+    },
+    {
+      "epoch": 2.638225255972696,
+      "grad_norm": 1.9866377115249634,
+      "learning_rate": 0.0004732650739476678,
+      "loss": 4.2124,
+      "step": 2319
+    },
+    {
+      "epoch": 2.639362912400455,
+      "grad_norm": 1.1584763526916504,
+      "learning_rate": 0.0004730375426621161,
+      "loss": 1.5185,
+      "step": 2320
+    },
+    {
+      "epoch": 2.640500568828214,
+      "grad_norm": 0.9889481067657471,
+      "learning_rate": 0.0004728100113765643,
+      "loss": 2.2838,
+      "step": 2321
+    },
+    {
+      "epoch": 2.6416382252559725,
+      "grad_norm": 0.9309085011482239,
+      "learning_rate": 0.0004725824800910125,
+      "loss": 1.152,
+      "step": 2322
+    },
+    {
+      "epoch": 2.6427758816837317,
+      "grad_norm": 1.0950933694839478,
+      "learning_rate": 0.00047235494880546076,
+      "loss": 2.3803,
+      "step": 2323
+    },
+    {
+      "epoch": 2.6439135381114904,
+      "grad_norm": 0.8817667365074158,
+      "learning_rate": 0.00047212741751990903,
+      "loss": 1.4906,
+      "step": 2324
+    },
+    {
+      "epoch": 2.645051194539249,
+      "grad_norm": 0.8397241830825806,
+      "learning_rate": 0.00047189988623435724,
+      "loss": 1.8036,
+      "step": 2325
+    },
+    {
+      "epoch": 2.646188850967008,
+      "grad_norm": 1.4048362970352173,
+      "learning_rate": 0.00047167235494880545,
+      "loss": 2.0251,
+      "step": 2326
+    },
+    {
+      "epoch": 2.6473265073947667,
+      "grad_norm": 1.4513121843338013,
+      "learning_rate": 0.0004714448236632537,
+      "loss": 1.8078,
+      "step": 2327
+    },
+    {
+      "epoch": 2.6484641638225255,
+      "grad_norm": 0.9570140838623047,
+      "learning_rate": 0.0004712172923777019,
+      "loss": 1.8617,
+      "step": 2328
+    },
+    {
+      "epoch": 2.6496018202502842,
+      "grad_norm": 0.8819069862365723,
+      "learning_rate": 0.0004709897610921502,
+      "loss": 1.8745,
+      "step": 2329
+    },
+    {
+      "epoch": 2.6507394766780434,
+      "grad_norm": 1.094341516494751,
+      "learning_rate": 0.00047076222980659845,
+      "loss": 1.788,
+      "step": 2330
+    },
+    {
+      "epoch": 2.651877133105802,
+      "grad_norm": 0.832409143447876,
+      "learning_rate": 0.00047053469852104666,
+      "loss": 2.2568,
+      "step": 2331
+    },
+    {
+      "epoch": 2.653014789533561,
+      "grad_norm": 0.9308044910430908,
+      "learning_rate": 0.00047030716723549487,
+      "loss": 2.3719,
+      "step": 2332
+    },
+    {
+      "epoch": 2.6541524459613197,
+      "grad_norm": 1.2896952629089355,
+      "learning_rate": 0.00047007963594994313,
+      "loss": 3.7022,
+      "step": 2333
+    },
+    {
+      "epoch": 2.6552901023890785,
+      "grad_norm": 1.0395236015319824,
+      "learning_rate": 0.0004698521046643914,
+      "loss": 2.8605,
+      "step": 2334
+    },
+    {
+      "epoch": 2.6564277588168372,
+      "grad_norm": 0.8033049702644348,
+      "learning_rate": 0.0004696245733788396,
+      "loss": 1.5591,
+      "step": 2335
+    },
+    {
+      "epoch": 2.657565415244596,
+      "grad_norm": 0.798548698425293,
+      "learning_rate": 0.0004693970420932878,
+      "loss": 1.158,
+      "step": 2336
+    },
+    {
+      "epoch": 2.658703071672355,
+      "grad_norm": 0.7860295176506042,
+      "learning_rate": 0.0004691695108077361,
+      "loss": 2.112,
+      "step": 2337
+    },
+    {
+      "epoch": 2.6598407281001135,
+      "grad_norm": 0.8449381589889526,
+      "learning_rate": 0.0004689419795221843,
+      "loss": 1.5287,
+      "step": 2338
+    },
+    {
+      "epoch": 2.6609783845278727,
+      "grad_norm": 1.0243216753005981,
+      "learning_rate": 0.00046871444823663256,
+      "loss": 1.6321,
+      "step": 2339
+    },
+    {
+      "epoch": 2.6621160409556315,
+      "grad_norm": 0.9676476716995239,
+      "learning_rate": 0.0004684869169510808,
+      "loss": 2.4557,
+      "step": 2340
+    },
+    {
+      "epoch": 2.6632536973833902,
+      "grad_norm": 0.9129576683044434,
+      "learning_rate": 0.000468259385665529,
+      "loss": 1.8287,
+      "step": 2341
+    },
+    {
+      "epoch": 2.664391353811149,
+      "grad_norm": 1.1413075923919678,
+      "learning_rate": 0.00046803185437997724,
+      "loss": 2.0987,
+      "step": 2342
+    },
+    {
+      "epoch": 2.6655290102389078,
+      "grad_norm": 0.7088634967803955,
+      "learning_rate": 0.0004678043230944255,
+      "loss": 1.6769,
+      "step": 2343
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.8204455971717834,
+      "learning_rate": 0.00046757679180887377,
+      "loss": 2.0439,
+      "step": 2344
+    },
+    {
+      "epoch": 2.6678043230944253,
+      "grad_norm": 0.7846645712852478,
+      "learning_rate": 0.0004673492605233219,
+      "loss": 1.4144,
+      "step": 2345
+    },
+    {
+      "epoch": 2.6689419795221845,
+      "grad_norm": 0.8308724761009216,
+      "learning_rate": 0.0004671217292377702,
+      "loss": 1.2093,
+      "step": 2346
+    },
+    {
+      "epoch": 2.6700796359499432,
+      "grad_norm": 1.0567412376403809,
+      "learning_rate": 0.00046689419795221845,
+      "loss": 1.8646,
+      "step": 2347
+    },
+    {
+      "epoch": 2.671217292377702,
+      "grad_norm": 0.986167311668396,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 2.4904,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6723549488054608,
+      "grad_norm": 1.8298397064208984,
+      "learning_rate": 0.0004664391353811149,
+      "loss": 2.9794,
+      "step": 2349
+    },
+    {
+      "epoch": 2.6734926052332195,
+      "grad_norm": 0.6814413070678711,
+      "learning_rate": 0.00046621160409556314,
+      "loss": 1.7104,
+      "step": 2350
+    },
+    {
+      "epoch": 2.6746302616609783,
+      "grad_norm": 1.0616414546966553,
+      "learning_rate": 0.00046598407281001135,
+      "loss": 2.269,
+      "step": 2351
+    },
+    {
+      "epoch": 2.675767918088737,
+      "grad_norm": 1.0702776908874512,
+      "learning_rate": 0.0004657565415244596,
+      "loss": 1.9667,
+      "step": 2352
+    },
+    {
+      "epoch": 2.6769055745164962,
+      "grad_norm": 1.4715807437896729,
+      "learning_rate": 0.0004655290102389079,
+      "loss": 3.1964,
+      "step": 2353
+    },
+    {
+      "epoch": 2.6780432309442546,
+      "grad_norm": 0.9601611495018005,
+      "learning_rate": 0.00046530147895335614,
+      "loss": 2.5208,
+      "step": 2354
+    },
+    {
+      "epoch": 2.6791808873720138,
+      "grad_norm": 1.083862543106079,
+      "learning_rate": 0.0004650739476678043,
+      "loss": 1.8693,
+      "step": 2355
+    },
+    {
+      "epoch": 2.6803185437997725,
+      "grad_norm": 1.272933006286621,
+      "learning_rate": 0.00046484641638225256,
+      "loss": 3.148,
+      "step": 2356
+    },
+    {
+      "epoch": 2.6814562002275313,
+      "grad_norm": 1.0518128871917725,
+      "learning_rate": 0.0004646188850967008,
+      "loss": 1.2337,
+      "step": 2357
+    },
+    {
+      "epoch": 2.68259385665529,
+      "grad_norm": 0.9924670457839966,
+      "learning_rate": 0.00046439135381114903,
+      "loss": 1.4861,
+      "step": 2358
+    },
+    {
+      "epoch": 2.683731513083049,
+      "grad_norm": 1.0956393480300903,
+      "learning_rate": 0.0004641638225255973,
+      "loss": 1.7977,
+      "step": 2359
+    },
+    {
+      "epoch": 2.684869169510808,
+      "grad_norm": 1.3490567207336426,
+      "learning_rate": 0.0004639362912400455,
+      "loss": 2.5556,
+      "step": 2360
+    },
+    {
+      "epoch": 2.6860068259385663,
+      "grad_norm": 0.8191376328468323,
+      "learning_rate": 0.0004637087599544937,
+      "loss": 1.7643,
+      "step": 2361
+    },
+    {
+      "epoch": 2.6871444823663255,
+      "grad_norm": 0.9714041948318481,
+      "learning_rate": 0.000463481228668942,
+      "loss": 1.9565,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6882821387940843,
+      "grad_norm": 1.045387625694275,
+      "learning_rate": 0.00046325369738339024,
+      "loss": 2.746,
+      "step": 2363
+    },
+    {
+      "epoch": 2.689419795221843,
+      "grad_norm": 0.934880793094635,
+      "learning_rate": 0.0004630261660978385,
+      "loss": 2.6093,
+      "step": 2364
+    },
+    {
+      "epoch": 2.690557451649602,
+      "grad_norm": 0.7764321565628052,
+      "learning_rate": 0.00046279863481228666,
+      "loss": 1.0081,
+      "step": 2365
+    },
+    {
+      "epoch": 2.6916951080773606,
+      "grad_norm": 0.7609637379646301,
+      "learning_rate": 0.00046257110352673493,
+      "loss": 1.7661,
+      "step": 2366
+    },
+    {
+      "epoch": 2.6928327645051193,
+      "grad_norm": 0.8447152376174927,
+      "learning_rate": 0.0004623435722411832,
+      "loss": 1.5554,
+      "step": 2367
+    },
+    {
+      "epoch": 2.693970420932878,
+      "grad_norm": 0.76449054479599,
+      "learning_rate": 0.0004621160409556314,
+      "loss": 1.6179,
+      "step": 2368
+    },
+    {
+      "epoch": 2.6951080773606373,
+      "grad_norm": 0.7846193313598633,
+      "learning_rate": 0.00046188850967007967,
+      "loss": 1.0255,
+      "step": 2369
+    },
+    {
+      "epoch": 2.696245733788396,
+      "grad_norm": 0.8696131706237793,
+      "learning_rate": 0.0004616609783845279,
+      "loss": 2.0406,
+      "step": 2370
+    },
+    {
+      "epoch": 2.697383390216155,
+      "grad_norm": 0.9892042279243469,
+      "learning_rate": 0.0004614334470989761,
+      "loss": 2.7178,
+      "step": 2371
+    },
+    {
+      "epoch": 2.6985210466439136,
+      "grad_norm": 1.1036131381988525,
+      "learning_rate": 0.00046120591581342435,
+      "loss": 1.5833,
+      "step": 2372
+    },
+    {
+      "epoch": 2.6996587030716723,
+      "grad_norm": 0.9095990061759949,
+      "learning_rate": 0.0004609783845278726,
+      "loss": 2.3233,
+      "step": 2373
+    },
+    {
+      "epoch": 2.700796359499431,
+      "grad_norm": 1.0550446510314941,
+      "learning_rate": 0.0004607508532423209,
+      "loss": 1.5132,
+      "step": 2374
+    },
+    {
+      "epoch": 2.70193401592719,
+      "grad_norm": 0.984180748462677,
+      "learning_rate": 0.00046052332195676903,
+      "loss": 2.2976,
+      "step": 2375
+    },
+    {
+      "epoch": 2.703071672354949,
+      "grad_norm": 0.9732262492179871,
+      "learning_rate": 0.0004602957906712173,
+      "loss": 1.767,
+      "step": 2376
+    },
+    {
+      "epoch": 2.7042093287827074,
+      "grad_norm": 0.858201265335083,
+      "learning_rate": 0.00046006825938566556,
+      "loss": 1.3972,
+      "step": 2377
+    },
+    {
+      "epoch": 2.7053469852104666,
+      "grad_norm": 0.9151699542999268,
+      "learning_rate": 0.00045984072810011377,
+      "loss": 1.821,
+      "step": 2378
+    },
+    {
+      "epoch": 2.7064846416382253,
+      "grad_norm": 1.1489654779434204,
+      "learning_rate": 0.000459613196814562,
+      "loss": 1.3073,
+      "step": 2379
+    },
+    {
+      "epoch": 2.707622298065984,
+      "grad_norm": 1.2800509929656982,
+      "learning_rate": 0.00045938566552901025,
+      "loss": 1.9937,
+      "step": 2380
+    },
+    {
+      "epoch": 2.708759954493743,
+      "grad_norm": 0.6855766177177429,
+      "learning_rate": 0.00045915813424345846,
+      "loss": 0.7626,
+      "step": 2381
+    },
+    {
+      "epoch": 2.7098976109215016,
+      "grad_norm": 0.8592435717582703,
+      "learning_rate": 0.0004589306029579067,
+      "loss": 1.4657,
+      "step": 2382
+    },
+    {
+      "epoch": 2.7110352673492604,
+      "grad_norm": 0.6695002913475037,
+      "learning_rate": 0.000458703071672355,
+      "loss": 0.964,
+      "step": 2383
+    },
+    {
+      "epoch": 2.712172923777019,
+      "grad_norm": 0.6979312300682068,
+      "learning_rate": 0.0004584755403868032,
+      "loss": 1.4597,
+      "step": 2384
+    },
+    {
+      "epoch": 2.7133105802047783,
+      "grad_norm": 1.0145376920700073,
+      "learning_rate": 0.0004582480091012514,
+      "loss": 2.5069,
+      "step": 2385
+    },
+    {
+      "epoch": 2.714448236632537,
+      "grad_norm": 0.6209183931350708,
+      "learning_rate": 0.00045802047781569967,
+      "loss": 1.0564,
+      "step": 2386
+    },
+    {
+      "epoch": 2.715585893060296,
+      "grad_norm": 0.9510672688484192,
+      "learning_rate": 0.00045779294653014793,
+      "loss": 1.5424,
+      "step": 2387
+    },
+    {
+      "epoch": 2.7167235494880546,
+      "grad_norm": 0.911859393119812,
+      "learning_rate": 0.00045756541524459614,
+      "loss": 1.6315,
+      "step": 2388
+    },
+    {
+      "epoch": 2.7178612059158134,
+      "grad_norm": 0.7897265553474426,
+      "learning_rate": 0.00045733788395904435,
+      "loss": 1.3671,
+      "step": 2389
+    },
+    {
+      "epoch": 2.718998862343572,
+      "grad_norm": 0.8800843358039856,
+      "learning_rate": 0.0004571103526734926,
+      "loss": 2.1943,
+      "step": 2390
+    },
+    {
+      "epoch": 2.720136518771331,
+      "grad_norm": 1.3472900390625,
+      "learning_rate": 0.0004568828213879408,
+      "loss": 2.8666,
+      "step": 2391
+    },
+    {
+      "epoch": 2.72127417519909,
+      "grad_norm": 1.0880565643310547,
+      "learning_rate": 0.0004566552901023891,
+      "loss": 1.939,
+      "step": 2392
+    },
+    {
+      "epoch": 2.722411831626849,
+      "grad_norm": 0.8288098573684692,
+      "learning_rate": 0.00045642775881683735,
+      "loss": 2.334,
+      "step": 2393
+    },
+    {
+      "epoch": 2.7235494880546076,
+      "grad_norm": 1.0421987771987915,
+      "learning_rate": 0.00045620022753128556,
+      "loss": 2.0377,
+      "step": 2394
+    },
+    {
+      "epoch": 2.7246871444823664,
+      "grad_norm": 1.3530818223953247,
+      "learning_rate": 0.0004559726962457338,
+      "loss": 3.6316,
+      "step": 2395
+    },
+    {
+      "epoch": 2.725824800910125,
+      "grad_norm": 0.9477719068527222,
+      "learning_rate": 0.00045574516496018204,
+      "loss": 2.0223,
+      "step": 2396
+    },
+    {
+      "epoch": 2.726962457337884,
+      "grad_norm": 1.1411749124526978,
+      "learning_rate": 0.0004555176336746303,
+      "loss": 2.1665,
+      "step": 2397
+    },
+    {
+      "epoch": 2.7281001137656427,
+      "grad_norm": 1.3905211687088013,
+      "learning_rate": 0.0004552901023890785,
+      "loss": 1.8378,
+      "step": 2398
+    },
+    {
+      "epoch": 2.729237770193402,
+      "grad_norm": 0.7103641629219055,
+      "learning_rate": 0.0004550625711035267,
+      "loss": 1.6213,
+      "step": 2399
+    },
+    {
+      "epoch": 2.73037542662116,
+      "grad_norm": 0.7716681361198425,
+      "learning_rate": 0.000454835039817975,
+      "loss": 1.0399,
+      "step": 2400
+    },
+    {
+      "epoch": 2.7315130830489194,
+      "grad_norm": 0.6330393552780151,
+      "learning_rate": 0.0004546075085324232,
+      "loss": 1.2756,
+      "step": 2401
+    },
+    {
+      "epoch": 2.732650739476678,
+      "grad_norm": 1.137176513671875,
+      "learning_rate": 0.00045437997724687146,
+      "loss": 3.3,
+      "step": 2402
+    },
+    {
+      "epoch": 2.733788395904437,
+      "grad_norm": 0.9734787940979004,
+      "learning_rate": 0.0004541524459613197,
+      "loss": 1.4292,
+      "step": 2403
+    },
+    {
+      "epoch": 2.7349260523321957,
+      "grad_norm": 0.896617591381073,
+      "learning_rate": 0.00045392491467576793,
+      "loss": 0.7397,
+      "step": 2404
+    },
+    {
+      "epoch": 2.7360637087599544,
+      "grad_norm": 1.0114847421646118,
+      "learning_rate": 0.00045369738339021614,
+      "loss": 1.7089,
+      "step": 2405
+    },
+    {
+      "epoch": 2.737201365187713,
+      "grad_norm": 0.9761648774147034,
+      "learning_rate": 0.0004534698521046644,
+      "loss": 1.9788,
+      "step": 2406
+    },
+    {
+      "epoch": 2.738339021615472,
+      "grad_norm": 1.1160939931869507,
+      "learning_rate": 0.00045324232081911267,
+      "loss": 2.4602,
+      "step": 2407
+    },
+    {
+      "epoch": 2.739476678043231,
+      "grad_norm": 0.9416034817695618,
+      "learning_rate": 0.0004530147895335609,
+      "loss": 1.6489,
+      "step": 2408
+    },
+    {
+      "epoch": 2.74061433447099,
+      "grad_norm": 0.9150708913803101,
+      "learning_rate": 0.0004527872582480091,
+      "loss": 1.4331,
+      "step": 2409
+    },
+    {
+      "epoch": 2.7417519908987487,
+      "grad_norm": 1.14388108253479,
+      "learning_rate": 0.00045255972696245736,
+      "loss": 1.9662,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7428896473265074,
+      "grad_norm": 0.8412722945213318,
+      "learning_rate": 0.00045233219567690557,
+      "loss": 1.2538,
+      "step": 2411
+    },
+    {
+      "epoch": 2.744027303754266,
+      "grad_norm": 1.3321762084960938,
+      "learning_rate": 0.00045210466439135383,
+      "loss": 1.693,
+      "step": 2412
+    },
+    {
+      "epoch": 2.745164960182025,
+      "grad_norm": 0.9153639674186707,
+      "learning_rate": 0.00045187713310580204,
+      "loss": 1.625,
+      "step": 2413
+    },
+    {
+      "epoch": 2.7463026166097837,
+      "grad_norm": 0.8170965313911438,
+      "learning_rate": 0.0004516496018202503,
+      "loss": 1.6253,
+      "step": 2414
+    },
+    {
+      "epoch": 2.747440273037543,
+      "grad_norm": 0.7869139909744263,
+      "learning_rate": 0.0004514220705346985,
+      "loss": 1.4511,
+      "step": 2415
+    },
+    {
+      "epoch": 2.748577929465301,
+      "grad_norm": 1.0545377731323242,
+      "learning_rate": 0.0004511945392491468,
+      "loss": 1.6514,
+      "step": 2416
+    },
+    {
+      "epoch": 2.7497155858930604,
+      "grad_norm": 1.045305848121643,
+      "learning_rate": 0.00045096700796359504,
+      "loss": 1.2496,
+      "step": 2417
+    },
+    {
+      "epoch": 2.750853242320819,
+      "grad_norm": 1.186232566833496,
+      "learning_rate": 0.0004507394766780432,
+      "loss": 2.2593,
+      "step": 2418
+    },
+    {
+      "epoch": 2.751990898748578,
+      "grad_norm": 0.8955073356628418,
+      "learning_rate": 0.00045051194539249146,
+      "loss": 1.1633,
+      "step": 2419
+    },
+    {
+      "epoch": 2.7531285551763367,
+      "grad_norm": 1.0408909320831299,
+      "learning_rate": 0.0004502844141069397,
+      "loss": 1.6361,
+      "step": 2420
+    },
+    {
+      "epoch": 2.7542662116040955,
+      "grad_norm": 1.312453031539917,
+      "learning_rate": 0.00045005688282138794,
+      "loss": 1.7388,
+      "step": 2421
+    },
+    {
+      "epoch": 2.755403868031854,
+      "grad_norm": 1.0657882690429688,
+      "learning_rate": 0.0004498293515358362,
+      "loss": 2.0169,
+      "step": 2422
+    },
+    {
+      "epoch": 2.756541524459613,
+      "grad_norm": 1.0726020336151123,
+      "learning_rate": 0.0004496018202502844,
+      "loss": 1.5911,
+      "step": 2423
+    },
+    {
+      "epoch": 2.757679180887372,
+      "grad_norm": 0.9591920375823975,
+      "learning_rate": 0.0004493742889647327,
+      "loss": 1.6334,
+      "step": 2424
+    },
+    {
+      "epoch": 2.758816837315131,
+      "grad_norm": 1.259514570236206,
+      "learning_rate": 0.0004491467576791809,
+      "loss": 2.5346,
+      "step": 2425
+    },
+    {
+      "epoch": 2.7599544937428897,
+      "grad_norm": 0.8274584412574768,
+      "learning_rate": 0.00044891922639362915,
+      "loss": 1.5429,
+      "step": 2426
+    },
+    {
+      "epoch": 2.7610921501706485,
+      "grad_norm": 1.121607780456543,
+      "learning_rate": 0.0004486916951080774,
+      "loss": 1.9823,
+      "step": 2427
+    },
+    {
+      "epoch": 2.7622298065984072,
+      "grad_norm": 1.2291669845581055,
+      "learning_rate": 0.00044846416382252557,
+      "loss": 1.9587,
+      "step": 2428
+    },
+    {
+      "epoch": 2.763367463026166,
+      "grad_norm": 0.8511943817138672,
+      "learning_rate": 0.00044823663253697383,
+      "loss": 1.1893,
+      "step": 2429
+    },
+    {
+      "epoch": 2.7645051194539247,
+      "grad_norm": 0.9760825037956238,
+      "learning_rate": 0.0004480091012514221,
+      "loss": 1.1563,
+      "step": 2430
+    },
+    {
+      "epoch": 2.765642775881684,
+      "grad_norm": 0.7539849877357483,
+      "learning_rate": 0.0004477815699658703,
+      "loss": 1.6624,
+      "step": 2431
+    },
+    {
+      "epoch": 2.7667804323094427,
+      "grad_norm": 1.0250943899154663,
+      "learning_rate": 0.00044755403868031857,
+      "loss": 1.4936,
+      "step": 2432
+    },
+    {
+      "epoch": 2.7679180887372015,
+      "grad_norm": 0.6538174152374268,
+      "learning_rate": 0.0004473265073947668,
+      "loss": 1.3009,
+      "step": 2433
+    },
+    {
+      "epoch": 2.7690557451649602,
+      "grad_norm": 0.6914688944816589,
+      "learning_rate": 0.00044709897610921504,
+      "loss": 1.0585,
+      "step": 2434
+    },
+    {
+      "epoch": 2.770193401592719,
+      "grad_norm": 0.9097804427146912,
+      "learning_rate": 0.00044687144482366325,
+      "loss": 1.5146,
+      "step": 2435
+    },
+    {
+      "epoch": 2.7713310580204777,
+      "grad_norm": 1.8249222040176392,
+      "learning_rate": 0.0004466439135381115,
+      "loss": 4.1139,
+      "step": 2436
+    },
+    {
+      "epoch": 2.7724687144482365,
+      "grad_norm": 0.8782042860984802,
+      "learning_rate": 0.0004464163822525598,
+      "loss": 1.4278,
+      "step": 2437
+    },
+    {
+      "epoch": 2.7736063708759957,
+      "grad_norm": 0.8963881134986877,
+      "learning_rate": 0.00044618885096700794,
+      "loss": 0.8318,
+      "step": 2438
+    },
+    {
+      "epoch": 2.774744027303754,
+      "grad_norm": 0.768720805644989,
+      "learning_rate": 0.0004459613196814562,
+      "loss": 1.2816,
+      "step": 2439
+    },
+    {
+      "epoch": 2.7758816837315132,
+      "grad_norm": 1.2364832162857056,
+      "learning_rate": 0.00044573378839590447,
+      "loss": 2.7189,
+      "step": 2440
+    },
+    {
+      "epoch": 2.777019340159272,
+      "grad_norm": 0.8980220556259155,
+      "learning_rate": 0.0004455062571103527,
+      "loss": 2.1147,
+      "step": 2441
+    },
+    {
+      "epoch": 2.7781569965870307,
+      "grad_norm": 1.1556415557861328,
+      "learning_rate": 0.00044527872582480094,
+      "loss": 2.5734,
+      "step": 2442
+    },
+    {
+      "epoch": 2.7792946530147895,
+      "grad_norm": 1.0572619438171387,
+      "learning_rate": 0.00044505119453924915,
+      "loss": 1.7883,
+      "step": 2443
+    },
+    {
+      "epoch": 2.7804323094425483,
+      "grad_norm": 1.2387428283691406,
+      "learning_rate": 0.00044482366325369736,
+      "loss": 1.9794,
+      "step": 2444
+    },
+    {
+      "epoch": 2.781569965870307,
+      "grad_norm": 0.6518829464912415,
+      "learning_rate": 0.0004445961319681456,
+      "loss": 1.567,
+      "step": 2445
+    },
+    {
+      "epoch": 2.782707622298066,
+      "grad_norm": 0.9689300060272217,
+      "learning_rate": 0.0004443686006825939,
+      "loss": 1.8925,
+      "step": 2446
+    },
+    {
+      "epoch": 2.783845278725825,
+      "grad_norm": 1.2153396606445312,
+      "learning_rate": 0.00044414106939704215,
+      "loss": 2.3077,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7849829351535837,
+      "grad_norm": 0.9673851728439331,
+      "learning_rate": 0.0004439135381114903,
+      "loss": 1.3106,
+      "step": 2448
+    },
+    {
+      "epoch": 2.7861205915813425,
+      "grad_norm": 1.2174962759017944,
+      "learning_rate": 0.00044368600682593857,
+      "loss": 2.8822,
+      "step": 2449
+    },
+    {
+      "epoch": 2.7872582480091013,
+      "grad_norm": 1.0069944858551025,
+      "learning_rate": 0.00044345847554038684,
+      "loss": 1.4753,
+      "step": 2450
+    },
+    {
+      "epoch": 2.78839590443686,
+      "grad_norm": 1.794924020767212,
+      "learning_rate": 0.00044323094425483504,
+      "loss": 3.3568,
+      "step": 2451
+    },
+    {
+      "epoch": 2.789533560864619,
+      "grad_norm": 1.3522320985794067,
+      "learning_rate": 0.00044300341296928325,
+      "loss": 2.2246,
+      "step": 2452
+    },
+    {
+      "epoch": 2.7906712172923775,
+      "grad_norm": 1.126507043838501,
+      "learning_rate": 0.0004427758816837315,
+      "loss": 2.1909,
+      "step": 2453
+    },
+    {
+      "epoch": 2.7918088737201368,
+      "grad_norm": 0.840438961982727,
+      "learning_rate": 0.00044254835039817973,
+      "loss": 1.315,
+      "step": 2454
+    },
+    {
+      "epoch": 2.792946530147895,
+      "grad_norm": 7.848939418792725,
+      "learning_rate": 0.000442320819112628,
+      "loss": 2.2406,
+      "step": 2455
+    },
+    {
+      "epoch": 2.7940841865756543,
+      "grad_norm": 1.4858578443527222,
+      "learning_rate": 0.00044209328782707626,
+      "loss": 3.2399,
+      "step": 2456
+    },
+    {
+      "epoch": 2.795221843003413,
+      "grad_norm": 0.7680813670158386,
+      "learning_rate": 0.00044186575654152447,
+      "loss": 1.4417,
+      "step": 2457
+    },
+    {
+      "epoch": 2.796359499431172,
+      "grad_norm": 1.2522087097167969,
+      "learning_rate": 0.0004416382252559727,
+      "loss": 2.4295,
+      "step": 2458
+    },
+    {
+      "epoch": 2.7974971558589306,
+      "grad_norm": 1.1698752641677856,
+      "learning_rate": 0.00044141069397042094,
+      "loss": 1.9101,
+      "step": 2459
+    },
+    {
+      "epoch": 2.7986348122866893,
+      "grad_norm": 0.5076554417610168,
+      "learning_rate": 0.0004411831626848692,
+      "loss": 0.9025,
+      "step": 2460
+    },
+    {
+      "epoch": 2.799772468714448,
+      "grad_norm": 0.7442627549171448,
+      "learning_rate": 0.0004409556313993174,
+      "loss": 1.3259,
+      "step": 2461
+    },
+    {
+      "epoch": 2.800910125142207,
+      "grad_norm": 1.1652323007583618,
+      "learning_rate": 0.0004407281001137656,
+      "loss": 1.3523,
+      "step": 2462
+    },
+    {
+      "epoch": 2.802047781569966,
+      "grad_norm": 1.1334046125411987,
+      "learning_rate": 0.0004405005688282139,
+      "loss": 2.4571,
+      "step": 2463
+    },
+    {
+      "epoch": 2.803185437997725,
+      "grad_norm": 1.446699857711792,
+      "learning_rate": 0.0004402730375426621,
+      "loss": 3.8358,
+      "step": 2464
+    },
+    {
+      "epoch": 2.8043230944254836,
+      "grad_norm": 0.8421689867973328,
+      "learning_rate": 0.00044004550625711036,
+      "loss": 2.1174,
+      "step": 2465
+    },
+    {
+      "epoch": 2.8054607508532423,
+      "grad_norm": 1.1578481197357178,
+      "learning_rate": 0.0004398179749715586,
+      "loss": 2.3561,
+      "step": 2466
+    },
+    {
+      "epoch": 2.806598407281001,
+      "grad_norm": 1.4153037071228027,
+      "learning_rate": 0.00043959044368600684,
+      "loss": 3.3699,
+      "step": 2467
+    },
+    {
+      "epoch": 2.80773606370876,
+      "grad_norm": 0.9994482398033142,
+      "learning_rate": 0.00043936291240045505,
+      "loss": 2.3389,
+      "step": 2468
+    },
+    {
+      "epoch": 2.8088737201365186,
+      "grad_norm": 1.0240802764892578,
+      "learning_rate": 0.0004391353811149033,
+      "loss": 1.4459,
+      "step": 2469
+    },
+    {
+      "epoch": 2.810011376564278,
+      "grad_norm": 0.505115807056427,
+      "learning_rate": 0.0004389078498293516,
+      "loss": 0.5616,
+      "step": 2470
+    },
+    {
+      "epoch": 2.8111490329920366,
+      "grad_norm": 1.041799545288086,
+      "learning_rate": 0.0004386803185437998,
+      "loss": 1.6856,
+      "step": 2471
+    },
+    {
+      "epoch": 2.8122866894197953,
+      "grad_norm": 0.7291197776794434,
+      "learning_rate": 0.000438452787258248,
+      "loss": 1.8971,
+      "step": 2472
+    },
+    {
+      "epoch": 2.813424345847554,
+      "grad_norm": 0.7052469253540039,
+      "learning_rate": 0.00043822525597269626,
+      "loss": 1.6314,
+      "step": 2473
+    },
+    {
+      "epoch": 2.814562002275313,
+      "grad_norm": 1.3250932693481445,
+      "learning_rate": 0.00043799772468714447,
+      "loss": 1.6973,
+      "step": 2474
+    },
+    {
+      "epoch": 2.8156996587030716,
+      "grad_norm": 0.8435229659080505,
+      "learning_rate": 0.00043777019340159273,
+      "loss": 1.387,
+      "step": 2475
+    },
+    {
+      "epoch": 2.8168373151308304,
+      "grad_norm": 0.8370683193206787,
+      "learning_rate": 0.000437542662116041,
+      "loss": 1.1863,
+      "step": 2476
+    },
+    {
+      "epoch": 2.8179749715585896,
+      "grad_norm": 0.940768301486969,
+      "learning_rate": 0.0004373151308304892,
+      "loss": 1.4706,
+      "step": 2477
+    },
+    {
+      "epoch": 2.819112627986348,
+      "grad_norm": 0.8866887092590332,
+      "learning_rate": 0.0004370875995449374,
+      "loss": 2.093,
+      "step": 2478
+    },
+    {
+      "epoch": 2.820250284414107,
+      "grad_norm": 0.6502683162689209,
+      "learning_rate": 0.0004368600682593857,
+      "loss": 1.17,
+      "step": 2479
+    },
+    {
+      "epoch": 2.821387940841866,
+      "grad_norm": 1.3090890645980835,
+      "learning_rate": 0.00043663253697383394,
+      "loss": 2.7351,
+      "step": 2480
+    },
+    {
+      "epoch": 2.8225255972696246,
+      "grad_norm": 0.6929295063018799,
+      "learning_rate": 0.00043640500568828215,
+      "loss": 1.0882,
+      "step": 2481
+    },
+    {
+      "epoch": 2.8236632536973834,
+      "grad_norm": 1.296984314918518,
+      "learning_rate": 0.00043617747440273036,
+      "loss": 2.5549,
+      "step": 2482
+    },
+    {
+      "epoch": 2.824800910125142,
+      "grad_norm": 1.1640862226486206,
+      "learning_rate": 0.00043594994311717863,
+      "loss": 2.2164,
+      "step": 2483
+    },
+    {
+      "epoch": 2.825938566552901,
+      "grad_norm": 0.8431540727615356,
+      "learning_rate": 0.00043572241183162684,
+      "loss": 1.8785,
+      "step": 2484
+    },
+    {
+      "epoch": 2.8270762229806596,
+      "grad_norm": 1.0226455926895142,
+      "learning_rate": 0.0004354948805460751,
+      "loss": 2.7847,
+      "step": 2485
+    },
+    {
+      "epoch": 2.828213879408419,
+      "grad_norm": 0.9944515824317932,
+      "learning_rate": 0.0004352673492605233,
+      "loss": 3.0653,
+      "step": 2486
+    },
+    {
+      "epoch": 2.8293515358361776,
+      "grad_norm": 0.989396333694458,
+      "learning_rate": 0.0004350398179749716,
+      "loss": 1.5644,
+      "step": 2487
+    },
+    {
+      "epoch": 2.8304891922639364,
+      "grad_norm": 1.1219333410263062,
+      "learning_rate": 0.0004348122866894198,
+      "loss": 3.0597,
+      "step": 2488
+    },
+    {
+      "epoch": 2.831626848691695,
+      "grad_norm": 0.9987985491752625,
+      "learning_rate": 0.00043458475540386805,
+      "loss": 1.918,
+      "step": 2489
+    },
+    {
+      "epoch": 2.832764505119454,
+      "grad_norm": 1.3687235116958618,
+      "learning_rate": 0.0004343572241183163,
+      "loss": 2.0344,
+      "step": 2490
+    },
+    {
+      "epoch": 2.8339021615472126,
+      "grad_norm": 0.5967879891395569,
+      "learning_rate": 0.00043412969283276447,
+      "loss": 1.1413,
+      "step": 2491
+    },
+    {
+      "epoch": 2.8350398179749714,
+      "grad_norm": 0.7929881811141968,
+      "learning_rate": 0.00043390216154721273,
+      "loss": 1.3015,
+      "step": 2492
+    },
+    {
+      "epoch": 2.8361774744027306,
+      "grad_norm": 1.2926918268203735,
+      "learning_rate": 0.000433674630261661,
+      "loss": 2.332,
+      "step": 2493
+    },
+    {
+      "epoch": 2.837315130830489,
+      "grad_norm": 0.7639207243919373,
+      "learning_rate": 0.0004334470989761092,
+      "loss": 1.7233,
+      "step": 2494
+    },
+    {
+      "epoch": 2.838452787258248,
+      "grad_norm": 0.7680888772010803,
+      "learning_rate": 0.00043321956769055747,
+      "loss": 0.8933,
+      "step": 2495
+    },
+    {
+      "epoch": 2.839590443686007,
+      "grad_norm": 0.9917737245559692,
+      "learning_rate": 0.0004329920364050057,
+      "loss": 1.8064,
+      "step": 2496
+    },
+    {
+      "epoch": 2.8407281001137656,
+      "grad_norm": 0.8714781403541565,
+      "learning_rate": 0.00043276450511945395,
+      "loss": 1.688,
+      "step": 2497
+    },
+    {
+      "epoch": 2.8418657565415244,
+      "grad_norm": 1.283627986907959,
+      "learning_rate": 0.00043253697383390216,
+      "loss": 1.9438,
+      "step": 2498
+    },
+    {
+      "epoch": 2.843003412969283,
+      "grad_norm": 0.9072063565254211,
+      "learning_rate": 0.0004323094425483504,
+      "loss": 1.5175,
+      "step": 2499
+    },
+    {
+      "epoch": 2.8441410693970424,
+      "grad_norm": 0.9820964336395264,
+      "learning_rate": 0.0004320819112627987,
+      "loss": 3.0533,
+      "step": 2500
+    },
+    {
+      "epoch": 2.8452787258248007,
+      "grad_norm": 1.6198686361312866,
+      "learning_rate": 0.00043185437997724684,
+      "loss": 2.8954,
+      "step": 2501
+    },
+    {
+      "epoch": 2.84641638225256,
+      "grad_norm": 1.0898877382278442,
+      "learning_rate": 0.0004316268486916951,
+      "loss": 1.7376,
+      "step": 2502
+    },
+    {
+      "epoch": 2.8475540386803186,
+      "grad_norm": 0.8086861968040466,
+      "learning_rate": 0.00043139931740614337,
+      "loss": 1.8419,
+      "step": 2503
+    },
+    {
+      "epoch": 2.8486916951080774,
+      "grad_norm": 0.9705652594566345,
+      "learning_rate": 0.0004311717861205916,
+      "loss": 1.2042,
+      "step": 2504
+    },
+    {
+      "epoch": 2.849829351535836,
+      "grad_norm": 0.7838239669799805,
+      "learning_rate": 0.00043094425483503984,
+      "loss": 0.9004,
+      "step": 2505
+    },
+    {
+      "epoch": 2.850967007963595,
+      "grad_norm": 3.2648749351501465,
+      "learning_rate": 0.00043071672354948805,
+      "loss": 1.1682,
+      "step": 2506
+    },
+    {
+      "epoch": 2.8521046643913537,
+      "grad_norm": 1.287819743156433,
+      "learning_rate": 0.0004304891922639363,
+      "loss": 2.8992,
+      "step": 2507
+    },
+    {
+      "epoch": 2.8532423208191124,
+      "grad_norm": 1.7003014087677002,
+      "learning_rate": 0.0004302616609783845,
+      "loss": 1.2319,
+      "step": 2508
+    },
+    {
+      "epoch": 2.8543799772468716,
+      "grad_norm": 0.8558884859085083,
+      "learning_rate": 0.0004300341296928328,
+      "loss": 2.2637,
+      "step": 2509
+    },
+    {
+      "epoch": 2.8555176336746304,
+      "grad_norm": 0.9409870505332947,
+      "learning_rate": 0.00042980659840728105,
+      "loss": 2.3024,
+      "step": 2510
+    },
+    {
+      "epoch": 2.856655290102389,
+      "grad_norm": 0.627686619758606,
+      "learning_rate": 0.0004295790671217292,
+      "loss": 1.1552,
+      "step": 2511
+    },
+    {
+      "epoch": 2.857792946530148,
+      "grad_norm": 1.0397640466690063,
+      "learning_rate": 0.0004293515358361775,
+      "loss": 1.6213,
+      "step": 2512
+    },
+    {
+      "epoch": 2.8589306029579067,
+      "grad_norm": 1.128299593925476,
+      "learning_rate": 0.00042912400455062574,
+      "loss": 2.0099,
+      "step": 2513
+    },
+    {
+      "epoch": 2.8600682593856654,
+      "grad_norm": 1.486132025718689,
+      "learning_rate": 0.00042889647326507395,
+      "loss": 2.9052,
+      "step": 2514
+    },
+    {
+      "epoch": 2.861205915813424,
+      "grad_norm": 1.4445245265960693,
+      "learning_rate": 0.0004286689419795222,
+      "loss": 2.075,
+      "step": 2515
+    },
+    {
+      "epoch": 2.8623435722411834,
+      "grad_norm": 1.2654812335968018,
+      "learning_rate": 0.0004284414106939704,
+      "loss": 0.5957,
+      "step": 2516
+    },
+    {
+      "epoch": 2.8634812286689417,
+      "grad_norm": 0.7981293201446533,
+      "learning_rate": 0.0004282138794084187,
+      "loss": 1.9943,
+      "step": 2517
+    },
+    {
+      "epoch": 2.864618885096701,
+      "grad_norm": 0.5515188574790955,
+      "learning_rate": 0.0004279863481228669,
+      "loss": 0.8496,
+      "step": 2518
+    },
+    {
+      "epoch": 2.8657565415244597,
+      "grad_norm": 1.4269306659698486,
+      "learning_rate": 0.00042775881683731516,
+      "loss": 4.1156,
+      "step": 2519
+    },
+    {
+      "epoch": 2.8668941979522184,
+      "grad_norm": 0.7809433341026306,
+      "learning_rate": 0.00042753128555176337,
+      "loss": 1.3538,
+      "step": 2520
+    },
+    {
+      "epoch": 2.868031854379977,
+      "grad_norm": 0.7941915988922119,
+      "learning_rate": 0.0004273037542662116,
+      "loss": 1.7591,
+      "step": 2521
+    },
+    {
+      "epoch": 2.869169510807736,
+      "grad_norm": 1.2019193172454834,
+      "learning_rate": 0.00042707622298065984,
+      "loss": 2.0824,
+      "step": 2522
+    },
+    {
+      "epoch": 2.8703071672354947,
+      "grad_norm": 0.727832555770874,
+      "learning_rate": 0.0004268486916951081,
+      "loss": 1.8108,
+      "step": 2523
+    },
+    {
+      "epoch": 2.8714448236632535,
+      "grad_norm": 0.7144151926040649,
+      "learning_rate": 0.0004266211604095563,
+      "loss": 1.6019,
+      "step": 2524
+    },
+    {
+      "epoch": 2.8725824800910127,
+      "grad_norm": 1.0367114543914795,
+      "learning_rate": 0.00042639362912400453,
+      "loss": 1.4144,
+      "step": 2525
+    },
+    {
+      "epoch": 2.8737201365187715,
+      "grad_norm": 0.9273185729980469,
+      "learning_rate": 0.0004261660978384528,
+      "loss": 1.5532,
+      "step": 2526
+    },
+    {
+      "epoch": 2.87485779294653,
+      "grad_norm": 0.8667803406715393,
+      "learning_rate": 0.00042593856655290106,
+      "loss": 2.2255,
+      "step": 2527
+    },
+    {
+      "epoch": 2.875995449374289,
+      "grad_norm": 1.1647021770477295,
+      "learning_rate": 0.00042571103526734927,
+      "loss": 1.6445,
+      "step": 2528
+    },
+    {
+      "epoch": 2.8771331058020477,
+      "grad_norm": 0.8458011150360107,
+      "learning_rate": 0.00042548350398179753,
+      "loss": 1.6222,
+      "step": 2529
+    },
+    {
+      "epoch": 2.8782707622298065,
+      "grad_norm": 0.6948020458221436,
+      "learning_rate": 0.00042525597269624574,
+      "loss": 0.8429,
+      "step": 2530
+    },
+    {
+      "epoch": 2.8794084186575652,
+      "grad_norm": 1.5468206405639648,
+      "learning_rate": 0.00042502844141069395,
+      "loss": 2.3449,
+      "step": 2531
+    },
+    {
+      "epoch": 2.8805460750853245,
+      "grad_norm": 1.0919640064239502,
+      "learning_rate": 0.0004248009101251422,
+      "loss": 2.7566,
+      "step": 2532
+    },
+    {
+      "epoch": 2.881683731513083,
+      "grad_norm": 0.9144538044929504,
+      "learning_rate": 0.0004245733788395905,
+      "loss": 1.4095,
+      "step": 2533
+    },
+    {
+      "epoch": 2.882821387940842,
+      "grad_norm": 0.6683197617530823,
+      "learning_rate": 0.0004243458475540387,
+      "loss": 1.396,
+      "step": 2534
+    },
+    {
+      "epoch": 2.8839590443686007,
+      "grad_norm": 1.1698358058929443,
+      "learning_rate": 0.0004241183162684869,
+      "loss": 2.6583,
+      "step": 2535
+    },
+    {
+      "epoch": 2.8850967007963595,
+      "grad_norm": 0.9804813265800476,
+      "learning_rate": 0.00042389078498293516,
+      "loss": 1.8973,
+      "step": 2536
+    },
+    {
+      "epoch": 2.8862343572241183,
+      "grad_norm": 0.8138146996498108,
+      "learning_rate": 0.0004236632536973834,
+      "loss": 1.2295,
+      "step": 2537
+    },
+    {
+      "epoch": 2.887372013651877,
+      "grad_norm": 0.9173576831817627,
+      "learning_rate": 0.00042343572241183164,
+      "loss": 1.904,
+      "step": 2538
+    },
+    {
+      "epoch": 2.888509670079636,
+      "grad_norm": 0.9176083207130432,
+      "learning_rate": 0.0004232081911262799,
+      "loss": 3.0666,
+      "step": 2539
+    },
+    {
+      "epoch": 2.8896473265073945,
+      "grad_norm": 0.901013195514679,
+      "learning_rate": 0.0004229806598407281,
+      "loss": 2.0642,
+      "step": 2540
+    },
+    {
+      "epoch": 2.8907849829351537,
+      "grad_norm": 0.8971734642982483,
+      "learning_rate": 0.0004227531285551763,
+      "loss": 1.586,
+      "step": 2541
+    },
+    {
+      "epoch": 2.8919226393629125,
+      "grad_norm": 0.9048395752906799,
+      "learning_rate": 0.0004225255972696246,
+      "loss": 2.0343,
+      "step": 2542
+    },
+    {
+      "epoch": 2.8930602957906713,
+      "grad_norm": 1.1656171083450317,
+      "learning_rate": 0.00042229806598407285,
+      "loss": 2.5376,
+      "step": 2543
+    },
+    {
+      "epoch": 2.89419795221843,
+      "grad_norm": 0.655851423740387,
+      "learning_rate": 0.00042207053469852106,
+      "loss": 1.4965,
+      "step": 2544
+    },
+    {
+      "epoch": 2.8953356086461888,
+      "grad_norm": 0.8759918212890625,
+      "learning_rate": 0.00042184300341296927,
+      "loss": 2.1066,
+      "step": 2545
+    },
+    {
+      "epoch": 2.8964732650739475,
+      "grad_norm": 0.8949549794197083,
+      "learning_rate": 0.00042161547212741753,
+      "loss": 1.6719,
+      "step": 2546
+    },
+    {
+      "epoch": 2.8976109215017063,
+      "grad_norm": 1.0389626026153564,
+      "learning_rate": 0.00042138794084186574,
+      "loss": 2.4444,
+      "step": 2547
+    },
+    {
+      "epoch": 2.8987485779294655,
+      "grad_norm": 1.5342673063278198,
+      "learning_rate": 0.000421160409556314,
+      "loss": 2.5499,
+      "step": 2548
+    },
+    {
+      "epoch": 2.8998862343572243,
+      "grad_norm": 1.0685155391693115,
+      "learning_rate": 0.00042093287827076227,
+      "loss": 2.8287,
+      "step": 2549
+    },
+    {
+      "epoch": 2.901023890784983,
+      "grad_norm": 1.428017020225525,
+      "learning_rate": 0.0004207053469852105,
+      "loss": 3.5992,
+      "step": 2550
+    },
+    {
+      "epoch": 2.9021615472127418,
+      "grad_norm": 0.7409979701042175,
+      "learning_rate": 0.0004204778156996587,
+      "loss": 1.7332,
+      "step": 2551
+    },
+    {
+      "epoch": 2.9032992036405005,
+      "grad_norm": 1.2421164512634277,
+      "learning_rate": 0.00042025028441410695,
+      "loss": 1.5578,
+      "step": 2552
+    },
+    {
+      "epoch": 2.9044368600682593,
+      "grad_norm": 0.9279178380966187,
+      "learning_rate": 0.0004200227531285552,
+      "loss": 1.1623,
+      "step": 2553
+    },
+    {
+      "epoch": 2.905574516496018,
+      "grad_norm": 0.8222157955169678,
+      "learning_rate": 0.0004197952218430034,
+      "loss": 1.5473,
+      "step": 2554
+    },
+    {
+      "epoch": 2.9067121729237773,
+      "grad_norm": 0.7585321664810181,
+      "learning_rate": 0.00041956769055745164,
+      "loss": 1.3996,
+      "step": 2555
+    },
+    {
+      "epoch": 2.9078498293515356,
+      "grad_norm": 0.9835622310638428,
+      "learning_rate": 0.0004193401592718999,
+      "loss": 1.7122,
+      "step": 2556
+    },
+    {
+      "epoch": 2.908987485779295,
+      "grad_norm": 0.6884040236473083,
+      "learning_rate": 0.0004191126279863481,
+      "loss": 1.533,
+      "step": 2557
+    },
+    {
+      "epoch": 2.9101251422070535,
+      "grad_norm": 1.0069630146026611,
+      "learning_rate": 0.0004188850967007964,
+      "loss": 1.7265,
+      "step": 2558
+    },
+    {
+      "epoch": 2.9112627986348123,
+      "grad_norm": 0.984389066696167,
+      "learning_rate": 0.0004186575654152446,
+      "loss": 1.6568,
+      "step": 2559
+    },
+    {
+      "epoch": 2.912400455062571,
+      "grad_norm": 0.7369997501373291,
+      "learning_rate": 0.00041843003412969285,
+      "loss": 1.8428,
+      "step": 2560
+    },
+    {
+      "epoch": 2.91353811149033,
+      "grad_norm": 1.4680728912353516,
+      "learning_rate": 0.00041820250284414106,
+      "loss": 3.2181,
+      "step": 2561
+    },
+    {
+      "epoch": 2.9146757679180886,
+      "grad_norm": 0.7019348740577698,
+      "learning_rate": 0.0004179749715585893,
+      "loss": 1.529,
+      "step": 2562
+    },
+    {
+      "epoch": 2.9158134243458473,
+      "grad_norm": 0.6902445554733276,
+      "learning_rate": 0.0004177474402730376,
+      "loss": 0.6651,
+      "step": 2563
+    },
+    {
+      "epoch": 2.9169510807736065,
+      "grad_norm": 1.133726716041565,
+      "learning_rate": 0.00041751990898748574,
+      "loss": 3.7021,
+      "step": 2564
+    },
+    {
+      "epoch": 2.9180887372013653,
+      "grad_norm": 0.9320973753929138,
+      "learning_rate": 0.000417292377701934,
+      "loss": 2.1056,
+      "step": 2565
+    },
+    {
+      "epoch": 2.919226393629124,
+      "grad_norm": 0.9788163900375366,
+      "learning_rate": 0.00041706484641638227,
+      "loss": 1.8839,
+      "step": 2566
+    },
+    {
+      "epoch": 2.920364050056883,
+      "grad_norm": 1.0894557237625122,
+      "learning_rate": 0.0004168373151308305,
+      "loss": 2.4257,
+      "step": 2567
+    },
+    {
+      "epoch": 2.9215017064846416,
+      "grad_norm": 0.7440232038497925,
+      "learning_rate": 0.00041660978384527875,
+      "loss": 1.6213,
+      "step": 2568
+    },
+    {
+      "epoch": 2.9226393629124003,
+      "grad_norm": 1.0889848470687866,
+      "learning_rate": 0.00041638225255972696,
+      "loss": 1.632,
+      "step": 2569
+    },
+    {
+      "epoch": 2.923777019340159,
+      "grad_norm": 0.9316622018814087,
+      "learning_rate": 0.0004161547212741752,
+      "loss": 1.5941,
+      "step": 2570
+    },
+    {
+      "epoch": 2.9249146757679183,
+      "grad_norm": 1.1574180126190186,
+      "learning_rate": 0.00041592718998862343,
+      "loss": 2.5757,
+      "step": 2571
+    },
+    {
+      "epoch": 2.926052332195677,
+      "grad_norm": 0.629701554775238,
+      "learning_rate": 0.0004156996587030717,
+      "loss": 1.1369,
+      "step": 2572
+    },
+    {
+      "epoch": 2.927189988623436,
+      "grad_norm": 0.4679393172264099,
+      "learning_rate": 0.00041547212741751996,
+      "loss": 0.8413,
+      "step": 2573
+    },
+    {
+      "epoch": 2.9283276450511946,
+      "grad_norm": 0.8145558834075928,
+      "learning_rate": 0.0004152445961319681,
+      "loss": 1.3481,
+      "step": 2574
+    },
+    {
+      "epoch": 2.9294653014789533,
+      "grad_norm": 0.5542713403701782,
+      "learning_rate": 0.0004150170648464164,
+      "loss": 1.2693,
+      "step": 2575
+    },
+    {
+      "epoch": 2.930602957906712,
+      "grad_norm": 1.2317445278167725,
+      "learning_rate": 0.00041478953356086464,
+      "loss": 2.8908,
+      "step": 2576
+    },
+    {
+      "epoch": 2.931740614334471,
+      "grad_norm": 0.9596227407455444,
+      "learning_rate": 0.00041456200227531285,
+      "loss": 1.3738,
+      "step": 2577
+    },
+    {
+      "epoch": 2.93287827076223,
+      "grad_norm": 0.988106369972229,
+      "learning_rate": 0.0004143344709897611,
+      "loss": 2.4601,
+      "step": 2578
+    },
+    {
+      "epoch": 2.9340159271899884,
+      "grad_norm": 0.9745345115661621,
+      "learning_rate": 0.0004141069397042093,
+      "loss": 1.9739,
+      "step": 2579
+    },
+    {
+      "epoch": 2.9351535836177476,
+      "grad_norm": 0.8387298583984375,
+      "learning_rate": 0.0004138794084186576,
+      "loss": 1.6605,
+      "step": 2580
+    },
+    {
+      "epoch": 2.9362912400455063,
+      "grad_norm": 0.811417818069458,
+      "learning_rate": 0.0004136518771331058,
+      "loss": 0.7618,
+      "step": 2581
+    },
+    {
+      "epoch": 2.937428896473265,
+      "grad_norm": 0.8037904500961304,
+      "learning_rate": 0.00041342434584755406,
+      "loss": 1.1208,
+      "step": 2582
+    },
+    {
+      "epoch": 2.938566552901024,
+      "grad_norm": 1.0455831289291382,
+      "learning_rate": 0.00041319681456200233,
+      "loss": 2.6639,
+      "step": 2583
+    },
+    {
+      "epoch": 2.9397042093287826,
+      "grad_norm": 0.9639081358909607,
+      "learning_rate": 0.0004129692832764505,
+      "loss": 1.7122,
+      "step": 2584
+    },
+    {
+      "epoch": 2.9408418657565414,
+      "grad_norm": 0.9406579732894897,
+      "learning_rate": 0.00041274175199089875,
+      "loss": 1.7419,
+      "step": 2585
+    },
+    {
+      "epoch": 2.9419795221843,
+      "grad_norm": 1.0303080081939697,
+      "learning_rate": 0.000412514220705347,
+      "loss": 1.1925,
+      "step": 2586
+    },
+    {
+      "epoch": 2.9431171786120593,
+      "grad_norm": 1.262969970703125,
+      "learning_rate": 0.0004122866894197952,
+      "loss": 3.077,
+      "step": 2587
+    },
+    {
+      "epoch": 2.944254835039818,
+      "grad_norm": 1.7101789712905884,
+      "learning_rate": 0.0004120591581342435,
+      "loss": 3.1926,
+      "step": 2588
+    },
+    {
+      "epoch": 2.945392491467577,
+      "grad_norm": 0.583219587802887,
+      "learning_rate": 0.0004118316268486917,
+      "loss": 1.4564,
+      "step": 2589
+    },
+    {
+      "epoch": 2.9465301478953356,
+      "grad_norm": 0.916775107383728,
+      "learning_rate": 0.00041160409556313996,
+      "loss": 1.9524,
+      "step": 2590
+    },
+    {
+      "epoch": 2.9476678043230944,
+      "grad_norm": 0.6460449695587158,
+      "learning_rate": 0.00041137656427758817,
+      "loss": 0.8583,
+      "step": 2591
+    },
+    {
+      "epoch": 2.948805460750853,
+      "grad_norm": 0.9714045524597168,
+      "learning_rate": 0.00041114903299203643,
+      "loss": 1.8316,
+      "step": 2592
+    },
+    {
+      "epoch": 2.949943117178612,
+      "grad_norm": 0.7467246651649475,
+      "learning_rate": 0.00041092150170648464,
+      "loss": 1.1244,
+      "step": 2593
+    },
+    {
+      "epoch": 2.951080773606371,
+      "grad_norm": 1.166478157043457,
+      "learning_rate": 0.00041069397042093285,
+      "loss": 2.589,
+      "step": 2594
+    },
+    {
+      "epoch": 2.9522184300341294,
+      "grad_norm": 1.1131619215011597,
+      "learning_rate": 0.0004104664391353811,
+      "loss": 2.0701,
+      "step": 2595
+    },
+    {
+      "epoch": 2.9533560864618886,
+      "grad_norm": 0.7802785038948059,
+      "learning_rate": 0.0004102389078498294,
+      "loss": 1.5037,
+      "step": 2596
+    },
+    {
+      "epoch": 2.9544937428896474,
+      "grad_norm": 1.3113210201263428,
+      "learning_rate": 0.0004100113765642776,
+      "loss": 3.5825,
+      "step": 2597
+    },
+    {
+      "epoch": 2.955631399317406,
+      "grad_norm": 0.9257882833480835,
+      "learning_rate": 0.0004097838452787258,
+      "loss": 2.0172,
+      "step": 2598
+    },
+    {
+      "epoch": 2.956769055745165,
+      "grad_norm": 0.6490365862846375,
+      "learning_rate": 0.00040955631399317407,
+      "loss": 1.0444,
+      "step": 2599
+    },
+    {
+      "epoch": 2.9579067121729237,
+      "grad_norm": 0.7082270979881287,
+      "learning_rate": 0.00040932878270762233,
+      "loss": 1.6777,
+      "step": 2600
+    },
+    {
+      "epoch": 2.9590443686006824,
+      "grad_norm": 1.1460232734680176,
+      "learning_rate": 0.00040910125142207054,
+      "loss": 2.843,
+      "step": 2601
+    },
+    {
+      "epoch": 2.960182025028441,
+      "grad_norm": 1.1202770471572876,
+      "learning_rate": 0.0004088737201365188,
+      "loss": 1.8192,
+      "step": 2602
+    },
+    {
+      "epoch": 2.9613196814562004,
+      "grad_norm": 1.5068840980529785,
+      "learning_rate": 0.000408646188850967,
+      "loss": 3.6918,
+      "step": 2603
+    },
+    {
+      "epoch": 2.962457337883959,
+      "grad_norm": 0.9092023968696594,
+      "learning_rate": 0.0004084186575654152,
+      "loss": 1.8649,
+      "step": 2604
+    },
+    {
+      "epoch": 2.963594994311718,
+      "grad_norm": 1.0940500497817993,
+      "learning_rate": 0.0004081911262798635,
+      "loss": 2.9138,
+      "step": 2605
+    },
+    {
+      "epoch": 2.9647326507394767,
+      "grad_norm": 1.2195347547531128,
+      "learning_rate": 0.00040796359499431175,
+      "loss": 2.0279,
+      "step": 2606
+    },
+    {
+      "epoch": 2.9658703071672354,
+      "grad_norm": 0.7400806546211243,
+      "learning_rate": 0.00040773606370875996,
+      "loss": 1.3149,
+      "step": 2607
+    },
+    {
+      "epoch": 2.967007963594994,
+      "grad_norm": 1.0603524446487427,
+      "learning_rate": 0.00040750853242320817,
+      "loss": 2.151,
+      "step": 2608
+    },
+    {
+      "epoch": 2.968145620022753,
+      "grad_norm": 0.9565656185150146,
+      "learning_rate": 0.00040728100113765644,
+      "loss": 1.4545,
+      "step": 2609
+    },
+    {
+      "epoch": 2.969283276450512,
+      "grad_norm": 0.9198881387710571,
+      "learning_rate": 0.0004070534698521047,
+      "loss": 1.7001,
+      "step": 2610
+    },
+    {
+      "epoch": 2.970420932878271,
+      "grad_norm": 0.9140946865081787,
+      "learning_rate": 0.0004068259385665529,
+      "loss": 1.5544,
+      "step": 2611
+    },
+    {
+      "epoch": 2.9715585893060297,
+      "grad_norm": 0.6245988607406616,
+      "learning_rate": 0.0004065984072810012,
+      "loss": 1.3737,
+      "step": 2612
+    },
+    {
+      "epoch": 2.9726962457337884,
+      "grad_norm": 0.9493893980979919,
+      "learning_rate": 0.0004063708759954494,
+      "loss": 1.7512,
+      "step": 2613
+    },
+    {
+      "epoch": 2.973833902161547,
+      "grad_norm": 1.0149247646331787,
+      "learning_rate": 0.0004061433447098976,
+      "loss": 1.3255,
+      "step": 2614
+    },
+    {
+      "epoch": 2.974971558589306,
+      "grad_norm": 1.1951128244400024,
+      "learning_rate": 0.00040591581342434586,
+      "loss": 2.2548,
+      "step": 2615
+    },
+    {
+      "epoch": 2.9761092150170647,
+      "grad_norm": 1.160248875617981,
+      "learning_rate": 0.0004056882821387941,
+      "loss": 1.9538,
+      "step": 2616
+    },
+    {
+      "epoch": 2.977246871444824,
+      "grad_norm": 1.113005518913269,
+      "learning_rate": 0.00040546075085324233,
+      "loss": 1.813,
+      "step": 2617
+    },
+    {
+      "epoch": 2.9783845278725822,
+      "grad_norm": 0.6714508533477783,
+      "learning_rate": 0.00040523321956769054,
+      "loss": 1.2078,
+      "step": 2618
+    },
+    {
+      "epoch": 2.9795221843003414,
+      "grad_norm": 0.9738563299179077,
+      "learning_rate": 0.0004050056882821388,
+      "loss": 1.5833,
+      "step": 2619
+    },
+    {
+      "epoch": 2.9806598407281,
+      "grad_norm": 0.7636871933937073,
+      "learning_rate": 0.00040477815699658707,
+      "loss": 1.7431,
+      "step": 2620
+    },
+    {
+      "epoch": 2.981797497155859,
+      "grad_norm": 0.6886274814605713,
+      "learning_rate": 0.0004045506257110353,
+      "loss": 1.346,
+      "step": 2621
+    },
+    {
+      "epoch": 2.9829351535836177,
+      "grad_norm": 0.7128561735153198,
+      "learning_rate": 0.00040432309442548354,
+      "loss": 1.2031,
+      "step": 2622
+    },
+    {
+      "epoch": 2.9840728100113765,
+      "grad_norm": 1.0730477571487427,
+      "learning_rate": 0.00040409556313993175,
+      "loss": 1.9424,
+      "step": 2623
+    },
+    {
+      "epoch": 2.9852104664391352,
+      "grad_norm": 0.5753147006034851,
+      "learning_rate": 0.00040386803185437996,
+      "loss": 0.5761,
+      "step": 2624
+    },
+    {
+      "epoch": 2.986348122866894,
+      "grad_norm": 1.2759993076324463,
+      "learning_rate": 0.00040364050056882823,
+      "loss": 2.9886,
+      "step": 2625
+    },
+    {
+      "epoch": 2.987485779294653,
+      "grad_norm": 0.916204571723938,
+      "learning_rate": 0.0004034129692832765,
+      "loss": 1.5574,
+      "step": 2626
+    },
+    {
+      "epoch": 2.988623435722412,
+      "grad_norm": 0.940258264541626,
+      "learning_rate": 0.00040318543799772465,
+      "loss": 1.8689,
+      "step": 2627
+    },
+    {
+      "epoch": 2.9897610921501707,
+      "grad_norm": 0.5174016952514648,
+      "learning_rate": 0.0004029579067121729,
+      "loss": 0.8162,
+      "step": 2628
+    },
+    {
+      "epoch": 2.9908987485779295,
+      "grad_norm": 0.779630720615387,
+      "learning_rate": 0.0004027303754266212,
+      "loss": 1.2215,
+      "step": 2629
+    },
+    {
+      "epoch": 2.9920364050056882,
+      "grad_norm": 1.0173307657241821,
+      "learning_rate": 0.00040250284414106944,
+      "loss": 1.6225,
+      "step": 2630
+    },
+    {
+      "epoch": 2.993174061433447,
+      "grad_norm": 0.9133301377296448,
+      "learning_rate": 0.00040227531285551765,
+      "loss": 2.143,
+      "step": 2631
+    },
+    {
+      "epoch": 2.9943117178612058,
+      "grad_norm": 0.818585991859436,
+      "learning_rate": 0.00040204778156996586,
+      "loss": 2.071,
+      "step": 2632
+    },
+    {
+      "epoch": 2.995449374288965,
+      "grad_norm": 1.6560077667236328,
+      "learning_rate": 0.0004018202502844141,
+      "loss": 2.403,
+      "step": 2633
+    },
+    {
+      "epoch": 2.9965870307167233,
+      "grad_norm": 0.7060096859931946,
+      "learning_rate": 0.00040159271899886233,
+      "loss": 0.9656,
+      "step": 2634
+    },
+    {
+      "epoch": 2.9977246871444825,
+      "grad_norm": 0.7223090529441833,
+      "learning_rate": 0.0004013651877133106,
+      "loss": 1.7089,
+      "step": 2635
+    },
+    {
+      "epoch": 2.9988623435722412,
+      "grad_norm": 0.6294040679931641,
+      "learning_rate": 0.00040113765642775886,
+      "loss": 1.2626,
+      "step": 2636
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.931922972202301,
+      "learning_rate": 0.000400910125142207,
+      "loss": 1.8149,
+      "step": 2637
+    },
+    {
+      "epoch": 3.0,
+      "eval_f1": 0.8905,
+      "eval_gen_len": 49.5636,
+      "eval_loss": 1.8385756015777588,
+      "eval_precision": 0.889,
+      "eval_recall": 0.8923,
+      "eval_rouge1": 0.4436,
+      "eval_rouge2": 0.2001,
+      "eval_rougeL": 0.3707,
+      "eval_rougeLsum": 0.4092,
+      "eval_runtime": 28.0556,
+      "eval_samples_per_second": 3.921,
+      "eval_steps_per_second": 0.499,
+      "step": 2637
+    },
+    {
+      "epoch": 3.0011376564277588,
+      "grad_norm": 0.7378818988800049,
+      "learning_rate": 0.0004006825938566553,
+      "loss": 1.5501,
+      "step": 2638
+    },
+    {
+      "epoch": 3.0022753128555175,
+      "grad_norm": 0.9810494780540466,
+      "learning_rate": 0.00040045506257110355,
+      "loss": 2.0296,
+      "step": 2639
+    },
+    {
+      "epoch": 3.0034129692832763,
+      "grad_norm": 1.0457974672317505,
+      "learning_rate": 0.0004002275312855518,
+      "loss": 1.9824,
+      "step": 2640
+    },
+    {
+      "epoch": 3.0045506257110355,
+      "grad_norm": 1.2151912450790405,
+      "learning_rate": 0.0004,
+      "loss": 2.0357,
+      "step": 2641
+    },
+    {
+      "epoch": 3.0056882821387942,
+      "grad_norm": 0.9331043362617493,
+      "learning_rate": 0.00039977246871444823,
+      "loss": 1.1641,
+      "step": 2642
+    },
+    {
+      "epoch": 3.006825938566553,
+      "grad_norm": 0.8802271485328674,
+      "learning_rate": 0.0003995449374288965,
+      "loss": 1.4499,
+      "step": 2643
+    },
+    {
+      "epoch": 3.0079635949943118,
+      "grad_norm": 0.8039838075637817,
+      "learning_rate": 0.0003993174061433447,
+      "loss": 1.3322,
+      "step": 2644
+    },
+    {
+      "epoch": 3.0091012514220705,
+      "grad_norm": 1.716158390045166,
+      "learning_rate": 0.00039908987485779297,
+      "loss": 1.9503,
+      "step": 2645
+    },
+    {
+      "epoch": 3.0102389078498293,
+      "grad_norm": 0.589878261089325,
+      "learning_rate": 0.00039886234357224123,
+      "loss": 1.2121,
+      "step": 2646
+    },
+    {
+      "epoch": 3.011376564277588,
+      "grad_norm": 0.6693535447120667,
+      "learning_rate": 0.0003986348122866894,
+      "loss": 1.5157,
+      "step": 2647
+    },
+    {
+      "epoch": 3.012514220705347,
+      "grad_norm": 1.0818902254104614,
+      "learning_rate": 0.00039840728100113765,
+      "loss": 2.445,
+      "step": 2648
+    },
+    {
+      "epoch": 3.013651877133106,
+      "grad_norm": 1.072434663772583,
+      "learning_rate": 0.0003981797497155859,
+      "loss": 3.0144,
+      "step": 2649
+    },
+    {
+      "epoch": 3.0147895335608648,
+      "grad_norm": 1.0067790746688843,
+      "learning_rate": 0.0003979522184300341,
+      "loss": 1.1798,
+      "step": 2650
+    },
+    {
+      "epoch": 3.0159271899886235,
+      "grad_norm": 0.8552153706550598,
+      "learning_rate": 0.0003977246871444824,
+      "loss": 1.7862,
+      "step": 2651
+    },
+    {
+      "epoch": 3.0170648464163823,
+      "grad_norm": 0.9946853518486023,
+      "learning_rate": 0.0003974971558589306,
+      "loss": 2.5064,
+      "step": 2652
+    },
+    {
+      "epoch": 3.018202502844141,
+      "grad_norm": 0.8925816416740417,
+      "learning_rate": 0.00039726962457337886,
+      "loss": 1.3693,
+      "step": 2653
+    },
+    {
+      "epoch": 3.0193401592719,
+      "grad_norm": 1.815544605255127,
+      "learning_rate": 0.0003970420932878271,
+      "loss": 3.0686,
+      "step": 2654
+    },
+    {
+      "epoch": 3.0204778156996586,
+      "grad_norm": 1.122719407081604,
+      "learning_rate": 0.00039681456200227534,
+      "loss": 2.3256,
+      "step": 2655
+    },
+    {
+      "epoch": 3.0216154721274173,
+      "grad_norm": 1.784497618675232,
+      "learning_rate": 0.0003965870307167236,
+      "loss": 3.1539,
+      "step": 2656
+    },
+    {
+      "epoch": 3.0227531285551765,
+      "grad_norm": 1.0257576704025269,
+      "learning_rate": 0.00039635949943117176,
+      "loss": 1.8041,
+      "step": 2657
+    },
+    {
+      "epoch": 3.0238907849829353,
+      "grad_norm": 0.9659282565116882,
+      "learning_rate": 0.00039613196814562,
+      "loss": 1.4202,
+      "step": 2658
+    },
+    {
+      "epoch": 3.025028441410694,
+      "grad_norm": 1.3800609111785889,
+      "learning_rate": 0.0003959044368600683,
+      "loss": 1.9181,
+      "step": 2659
+    },
+    {
+      "epoch": 3.026166097838453,
+      "grad_norm": 1.056583285331726,
+      "learning_rate": 0.0003956769055745165,
+      "loss": 1.6186,
+      "step": 2660
+    },
+    {
+      "epoch": 3.0273037542662116,
+      "grad_norm": 0.7123137712478638,
+      "learning_rate": 0.0003954493742889647,
+      "loss": 1.2721,
+      "step": 2661
+    },
+    {
+      "epoch": 3.0284414106939703,
+      "grad_norm": 0.8692189455032349,
+      "learning_rate": 0.00039522184300341297,
+      "loss": 1.8621,
+      "step": 2662
+    },
+    {
+      "epoch": 3.029579067121729,
+      "grad_norm": 0.8391653895378113,
+      "learning_rate": 0.00039499431171786123,
+      "loss": 1.0275,
+      "step": 2663
+    },
+    {
+      "epoch": 3.030716723549488,
+      "grad_norm": 1.1637401580810547,
+      "learning_rate": 0.00039476678043230944,
+      "loss": 2.4589,
+      "step": 2664
+    },
+    {
+      "epoch": 3.031854379977247,
+      "grad_norm": 0.9150156378746033,
+      "learning_rate": 0.0003945392491467577,
+      "loss": 1.5759,
+      "step": 2665
+    },
+    {
+      "epoch": 3.032992036405006,
+      "grad_norm": 1.1568349599838257,
+      "learning_rate": 0.0003943117178612059,
+      "loss": 2.138,
+      "step": 2666
+    },
+    {
+      "epoch": 3.0341296928327646,
+      "grad_norm": 1.1453518867492676,
+      "learning_rate": 0.0003940841865756541,
+      "loss": 2.3239,
+      "step": 2667
+    },
+    {
+      "epoch": 3.0352673492605233,
+      "grad_norm": 0.8770859837532043,
+      "learning_rate": 0.0003938566552901024,
+      "loss": 1.523,
+      "step": 2668
+    },
+    {
+      "epoch": 3.036405005688282,
+      "grad_norm": 1.0244057178497314,
+      "learning_rate": 0.00039362912400455065,
+      "loss": 1.7684,
+      "step": 2669
+    },
+    {
+      "epoch": 3.037542662116041,
+      "grad_norm": 1.0480972528457642,
+      "learning_rate": 0.00039340159271899886,
+      "loss": 1.5302,
+      "step": 2670
+    },
+    {
+      "epoch": 3.0386803185437996,
+      "grad_norm": 1.1314648389816284,
+      "learning_rate": 0.0003931740614334471,
+      "loss": 1.6721,
+      "step": 2671
+    },
+    {
+      "epoch": 3.039817974971559,
+      "grad_norm": 0.7867646813392639,
+      "learning_rate": 0.00039294653014789534,
+      "loss": 1.3925,
+      "step": 2672
+    },
+    {
+      "epoch": 3.0409556313993176,
+      "grad_norm": 1.450878381729126,
+      "learning_rate": 0.0003927189988623436,
+      "loss": 2.2478,
+      "step": 2673
+    },
+    {
+      "epoch": 3.0420932878270763,
+      "grad_norm": 0.7771217226982117,
+      "learning_rate": 0.0003924914675767918,
+      "loss": 1.3815,
+      "step": 2674
+    },
+    {
+      "epoch": 3.043230944254835,
+      "grad_norm": 1.2466049194335938,
+      "learning_rate": 0.0003922639362912401,
+      "loss": 3.2891,
+      "step": 2675
+    },
+    {
+      "epoch": 3.044368600682594,
+      "grad_norm": 1.7478383779525757,
+      "learning_rate": 0.0003920364050056883,
+      "loss": 2.8603,
+      "step": 2676
+    },
+    {
+      "epoch": 3.0455062571103526,
+      "grad_norm": 0.8281832933425903,
+      "learning_rate": 0.0003918088737201365,
+      "loss": 1.5477,
+      "step": 2677
+    },
+    {
+      "epoch": 3.0466439135381114,
+      "grad_norm": 0.695612370967865,
+      "learning_rate": 0.00039158134243458476,
+      "loss": 1.4521,
+      "step": 2678
+    },
+    {
+      "epoch": 3.04778156996587,
+      "grad_norm": 0.9778900742530823,
+      "learning_rate": 0.000391353811149033,
+      "loss": 1.341,
+      "step": 2679
+    },
+    {
+      "epoch": 3.0489192263936293,
+      "grad_norm": 1.0386836528778076,
+      "learning_rate": 0.00039112627986348123,
+      "loss": 1.7095,
+      "step": 2680
+    },
+    {
+      "epoch": 3.050056882821388,
+      "grad_norm": 0.9944807291030884,
+      "learning_rate": 0.00039089874857792944,
+      "loss": 1.9036,
+      "step": 2681
+    },
+    {
+      "epoch": 3.051194539249147,
+      "grad_norm": 0.8040305972099304,
+      "learning_rate": 0.0003906712172923777,
+      "loss": 1.1083,
+      "step": 2682
+    },
+    {
+      "epoch": 3.0523321956769056,
+      "grad_norm": 1.0549193620681763,
+      "learning_rate": 0.00039044368600682597,
+      "loss": 2.558,
+      "step": 2683
+    },
+    {
+      "epoch": 3.0534698521046644,
+      "grad_norm": 0.7308775782585144,
+      "learning_rate": 0.0003902161547212742,
+      "loss": 1.073,
+      "step": 2684
+    },
+    {
+      "epoch": 3.054607508532423,
+      "grad_norm": 0.532346785068512,
+      "learning_rate": 0.00038998862343572245,
+      "loss": 0.8228,
+      "step": 2685
+    },
+    {
+      "epoch": 3.055745164960182,
+      "grad_norm": 0.9749463200569153,
+      "learning_rate": 0.00038976109215017066,
+      "loss": 1.6924,
+      "step": 2686
+    },
+    {
+      "epoch": 3.0568828213879407,
+      "grad_norm": 0.9695119857788086,
+      "learning_rate": 0.00038953356086461887,
+      "loss": 1.8108,
+      "step": 2687
+    },
+    {
+      "epoch": 3.0580204778157,
+      "grad_norm": 0.9351180195808411,
+      "learning_rate": 0.00038930602957906713,
+      "loss": 2.2291,
+      "step": 2688
+    },
+    {
+      "epoch": 3.0591581342434586,
+      "grad_norm": 0.8033020496368408,
+      "learning_rate": 0.0003890784982935154,
+      "loss": 1.8916,
+      "step": 2689
+    },
+    {
+      "epoch": 3.0602957906712174,
+      "grad_norm": 1.3790737390518188,
+      "learning_rate": 0.0003888509670079636,
+      "loss": 1.9922,
+      "step": 2690
+    },
+    {
+      "epoch": 3.061433447098976,
+      "grad_norm": 1.2332922220230103,
+      "learning_rate": 0.0003886234357224118,
+      "loss": 2.345,
+      "step": 2691
+    },
+    {
+      "epoch": 3.062571103526735,
+      "grad_norm": 1.38813316822052,
+      "learning_rate": 0.0003883959044368601,
+      "loss": 2.8993,
+      "step": 2692
+    },
+    {
+      "epoch": 3.0637087599544937,
+      "grad_norm": 0.952480137348175,
+      "learning_rate": 0.00038816837315130834,
+      "loss": 1.8394,
+      "step": 2693
+    },
+    {
+      "epoch": 3.0648464163822524,
+      "grad_norm": 0.8586364984512329,
+      "learning_rate": 0.00038794084186575655,
+      "loss": 1.1771,
+      "step": 2694
+    },
+    {
+      "epoch": 3.065984072810011,
+      "grad_norm": 0.9498509764671326,
+      "learning_rate": 0.00038771331058020476,
+      "loss": 1.4771,
+      "step": 2695
+    },
+    {
+      "epoch": 3.0671217292377704,
+      "grad_norm": 0.9688740968704224,
+      "learning_rate": 0.000387485779294653,
+      "loss": 2.1815,
+      "step": 2696
+    },
+    {
+      "epoch": 3.068259385665529,
+      "grad_norm": 2.0778191089630127,
+      "learning_rate": 0.00038725824800910124,
+      "loss": 2.46,
+      "step": 2697
+    },
+    {
+      "epoch": 3.069397042093288,
+      "grad_norm": 1.1978789567947388,
+      "learning_rate": 0.0003870307167235495,
+      "loss": 2.347,
+      "step": 2698
+    },
+    {
+      "epoch": 3.0705346985210467,
+      "grad_norm": 1.4718050956726074,
+      "learning_rate": 0.00038680318543799776,
+      "loss": 2.3554,
+      "step": 2699
+    },
+    {
+      "epoch": 3.0716723549488054,
+      "grad_norm": 0.9731530547142029,
+      "learning_rate": 0.0003865756541524459,
+      "loss": 1.4802,
+      "step": 2700
+    },
+    {
+      "epoch": 3.072810011376564,
+      "grad_norm": 1.0412006378173828,
+      "learning_rate": 0.0003863481228668942,
+      "loss": 1.8537,
+      "step": 2701
+    },
+    {
+      "epoch": 3.073947667804323,
+      "grad_norm": 0.7110769748687744,
+      "learning_rate": 0.00038612059158134245,
+      "loss": 1.6169,
+      "step": 2702
+    },
+    {
+      "epoch": 3.0750853242320817,
+      "grad_norm": 1.7731373310089111,
+      "learning_rate": 0.0003858930602957907,
+      "loss": 3.4582,
+      "step": 2703
+    },
+    {
+      "epoch": 3.076222980659841,
+      "grad_norm": 1.136021614074707,
+      "learning_rate": 0.0003856655290102389,
+      "loss": 2.161,
+      "step": 2704
+    },
+    {
+      "epoch": 3.0773606370875997,
+      "grad_norm": 1.2377517223358154,
+      "learning_rate": 0.00038543799772468713,
+      "loss": 1.9528,
+      "step": 2705
+    },
+    {
+      "epoch": 3.0784982935153584,
+      "grad_norm": 0.7984117865562439,
+      "learning_rate": 0.0003852104664391354,
+      "loss": 1.0635,
+      "step": 2706
+    },
+    {
+      "epoch": 3.079635949943117,
+      "grad_norm": 1.1283124685287476,
+      "learning_rate": 0.0003849829351535836,
+      "loss": 2.1958,
+      "step": 2707
+    },
+    {
+      "epoch": 3.080773606370876,
+      "grad_norm": 1.1983591318130493,
+      "learning_rate": 0.00038475540386803187,
+      "loss": 2.5032,
+      "step": 2708
+    },
+    {
+      "epoch": 3.0819112627986347,
+      "grad_norm": 0.8128067255020142,
+      "learning_rate": 0.00038452787258248013,
+      "loss": 1.566,
+      "step": 2709
+    },
+    {
+      "epoch": 3.0830489192263935,
+      "grad_norm": 1.4617652893066406,
+      "learning_rate": 0.0003843003412969283,
+      "loss": 2.3352,
+      "step": 2710
+    },
+    {
+      "epoch": 3.0841865756541527,
+      "grad_norm": 0.9528759717941284,
+      "learning_rate": 0.00038407281001137655,
+      "loss": 1.7667,
+      "step": 2711
+    },
+    {
+      "epoch": 3.0853242320819114,
+      "grad_norm": 1.0481518507003784,
+      "learning_rate": 0.0003838452787258248,
+      "loss": 1.9046,
+      "step": 2712
+    },
+    {
+      "epoch": 3.08646188850967,
+      "grad_norm": 0.9186555743217468,
+      "learning_rate": 0.0003836177474402731,
+      "loss": 1.8621,
+      "step": 2713
+    },
+    {
+      "epoch": 3.087599544937429,
+      "grad_norm": 0.6513910293579102,
+      "learning_rate": 0.0003833902161547213,
+      "loss": 1.2062,
+      "step": 2714
+    },
+    {
+      "epoch": 3.0887372013651877,
+      "grad_norm": 1.1539534330368042,
+      "learning_rate": 0.0003831626848691695,
+      "loss": 2.0368,
+      "step": 2715
+    },
+    {
+      "epoch": 3.0898748577929465,
+      "grad_norm": 1.2739286422729492,
+      "learning_rate": 0.00038293515358361777,
+      "loss": 2.7124,
+      "step": 2716
+    },
+    {
+      "epoch": 3.091012514220705,
+      "grad_norm": 0.796596348285675,
+      "learning_rate": 0.000382707622298066,
+      "loss": 1.5446,
+      "step": 2717
+    },
+    {
+      "epoch": 3.092150170648464,
+      "grad_norm": 1.144640564918518,
+      "learning_rate": 0.00038248009101251424,
+      "loss": 1.3855,
+      "step": 2718
+    },
+    {
+      "epoch": 3.093287827076223,
+      "grad_norm": 0.6495816111564636,
+      "learning_rate": 0.0003822525597269625,
+      "loss": 1.5521,
+      "step": 2719
+    },
+    {
+      "epoch": 3.094425483503982,
+      "grad_norm": 1.0980159044265747,
+      "learning_rate": 0.00038202502844141066,
+      "loss": 1.5025,
+      "step": 2720
+    },
+    {
+      "epoch": 3.0955631399317407,
+      "grad_norm": 0.8496591448783875,
+      "learning_rate": 0.0003817974971558589,
+      "loss": 2.1075,
+      "step": 2721
+    },
+    {
+      "epoch": 3.0967007963594995,
+      "grad_norm": 0.9507794380187988,
+      "learning_rate": 0.0003815699658703072,
+      "loss": 1.8178,
+      "step": 2722
+    },
+    {
+      "epoch": 3.0978384527872582,
+      "grad_norm": 1.0450421571731567,
+      "learning_rate": 0.00038134243458475545,
+      "loss": 2.2859,
+      "step": 2723
+    },
+    {
+      "epoch": 3.098976109215017,
+      "grad_norm": 0.8103886246681213,
+      "learning_rate": 0.00038111490329920366,
+      "loss": 1.7053,
+      "step": 2724
+    },
+    {
+      "epoch": 3.1001137656427757,
+      "grad_norm": 1.1811316013336182,
+      "learning_rate": 0.00038088737201365187,
+      "loss": 1.5576,
+      "step": 2725
+    },
+    {
+      "epoch": 3.1012514220705345,
+      "grad_norm": 1.117814064025879,
+      "learning_rate": 0.00038065984072810014,
+      "loss": 2.484,
+      "step": 2726
+    },
+    {
+      "epoch": 3.1023890784982937,
+      "grad_norm": 1.1703466176986694,
+      "learning_rate": 0.00038043230944254835,
+      "loss": 2.0287,
+      "step": 2727
+    },
+    {
+      "epoch": 3.1035267349260525,
+      "grad_norm": 1.159627914428711,
+      "learning_rate": 0.0003802047781569966,
+      "loss": 2.4316,
+      "step": 2728
+    },
+    {
+      "epoch": 3.1046643913538112,
+      "grad_norm": 0.6235855221748352,
+      "learning_rate": 0.0003799772468714448,
+      "loss": 1.3779,
+      "step": 2729
+    },
+    {
+      "epoch": 3.10580204778157,
+      "grad_norm": 0.8967525959014893,
+      "learning_rate": 0.00037974971558589303,
+      "loss": 1.8938,
+      "step": 2730
+    },
+    {
+      "epoch": 3.1069397042093287,
+      "grad_norm": 0.7541645169258118,
+      "learning_rate": 0.0003795221843003413,
+      "loss": 1.6706,
+      "step": 2731
+    },
+    {
+      "epoch": 3.1080773606370875,
+      "grad_norm": 0.7087587714195251,
+      "learning_rate": 0.00037929465301478956,
+      "loss": 1.185,
+      "step": 2732
+    },
+    {
+      "epoch": 3.1092150170648463,
+      "grad_norm": 0.4740487337112427,
+      "learning_rate": 0.0003790671217292378,
+      "loss": 0.5638,
+      "step": 2733
+    },
+    {
+      "epoch": 3.110352673492605,
+      "grad_norm": 0.6750523447990417,
+      "learning_rate": 0.000378839590443686,
+      "loss": 1.064,
+      "step": 2734
+    },
+    {
+      "epoch": 3.1114903299203642,
+      "grad_norm": 0.6977342367172241,
+      "learning_rate": 0.00037861205915813424,
+      "loss": 1.721,
+      "step": 2735
+    },
+    {
+      "epoch": 3.112627986348123,
+      "grad_norm": 0.5952358841896057,
+      "learning_rate": 0.0003783845278725825,
+      "loss": 0.7465,
+      "step": 2736
+    },
+    {
+      "epoch": 3.1137656427758817,
+      "grad_norm": 1.1384859085083008,
+      "learning_rate": 0.0003781569965870307,
+      "loss": 2.1374,
+      "step": 2737
+    },
+    {
+      "epoch": 3.1149032992036405,
+      "grad_norm": 1.135901927947998,
+      "learning_rate": 0.000377929465301479,
+      "loss": 3.0381,
+      "step": 2738
+    },
+    {
+      "epoch": 3.1160409556313993,
+      "grad_norm": 1.3090107440948486,
+      "learning_rate": 0.0003777019340159272,
+      "loss": 3.5599,
+      "step": 2739
+    },
+    {
+      "epoch": 3.117178612059158,
+      "grad_norm": 0.7160171866416931,
+      "learning_rate": 0.0003774744027303754,
+      "loss": 0.7931,
+      "step": 2740
+    },
+    {
+      "epoch": 3.118316268486917,
+      "grad_norm": 0.9017439484596252,
+      "learning_rate": 0.00037724687144482366,
+      "loss": 1.3222,
+      "step": 2741
+    },
+    {
+      "epoch": 3.1194539249146755,
+      "grad_norm": 0.7213937640190125,
+      "learning_rate": 0.00037701934015927193,
+      "loss": 1.8957,
+      "step": 2742
+    },
+    {
+      "epoch": 3.1205915813424348,
+      "grad_norm": 1.6694979667663574,
+      "learning_rate": 0.0003767918088737202,
+      "loss": 1.9735,
+      "step": 2743
+    },
+    {
+      "epoch": 3.1217292377701935,
+      "grad_norm": 1.073201298713684,
+      "learning_rate": 0.00037656427758816835,
+      "loss": 1.2988,
+      "step": 2744
+    },
+    {
+      "epoch": 3.1228668941979523,
+      "grad_norm": 0.7072137594223022,
+      "learning_rate": 0.0003763367463026166,
+      "loss": 1.4237,
+      "step": 2745
+    },
+    {
+      "epoch": 3.124004550625711,
+      "grad_norm": 1.140438437461853,
+      "learning_rate": 0.0003761092150170649,
+      "loss": 2.3338,
+      "step": 2746
+    },
+    {
+      "epoch": 3.12514220705347,
+      "grad_norm": 0.9373031258583069,
+      "learning_rate": 0.0003758816837315131,
+      "loss": 1.7434,
+      "step": 2747
+    },
+    {
+      "epoch": 3.1262798634812285,
+      "grad_norm": 1.1408966779708862,
+      "learning_rate": 0.00037565415244596135,
+      "loss": 2.5215,
+      "step": 2748
+    },
+    {
+      "epoch": 3.1274175199089873,
+      "grad_norm": 0.9613400101661682,
+      "learning_rate": 0.00037542662116040956,
+      "loss": 1.7157,
+      "step": 2749
+    },
+    {
+      "epoch": 3.1285551763367465,
+      "grad_norm": 1.0001013278961182,
+      "learning_rate": 0.00037519908987485777,
+      "loss": 1.3756,
+      "step": 2750
+    },
+    {
+      "epoch": 3.1296928327645053,
+      "grad_norm": 1.740805983543396,
+      "learning_rate": 0.00037497155858930603,
+      "loss": 2.2071,
+      "step": 2751
+    },
+    {
+      "epoch": 3.130830489192264,
+      "grad_norm": 0.9777475595474243,
+      "learning_rate": 0.0003747440273037543,
+      "loss": 1.8485,
+      "step": 2752
+    },
+    {
+      "epoch": 3.131968145620023,
+      "grad_norm": 1.1642980575561523,
+      "learning_rate": 0.0003745164960182025,
+      "loss": 1.2435,
+      "step": 2753
+    },
+    {
+      "epoch": 3.1331058020477816,
+      "grad_norm": 1.1870895624160767,
+      "learning_rate": 0.0003742889647326507,
+      "loss": 1.3834,
+      "step": 2754
+    },
+    {
+      "epoch": 3.1342434584755403,
+      "grad_norm": 0.9124128818511963,
+      "learning_rate": 0.000374061433447099,
+      "loss": 1.534,
+      "step": 2755
+    },
+    {
+      "epoch": 3.135381114903299,
+      "grad_norm": 0.6941938400268555,
+      "learning_rate": 0.00037383390216154725,
+      "loss": 1.5597,
+      "step": 2756
+    },
+    {
+      "epoch": 3.136518771331058,
+      "grad_norm": 0.9864102602005005,
+      "learning_rate": 0.00037360637087599546,
+      "loss": 1.5537,
+      "step": 2757
+    },
+    {
+      "epoch": 3.137656427758817,
+      "grad_norm": 0.6157810688018799,
+      "learning_rate": 0.0003733788395904437,
+      "loss": 1.1024,
+      "step": 2758
+    },
+    {
+      "epoch": 3.138794084186576,
+      "grad_norm": 0.9289652109146118,
+      "learning_rate": 0.00037315130830489193,
+      "loss": 1.2505,
+      "step": 2759
+    },
+    {
+      "epoch": 3.1399317406143346,
+      "grad_norm": 0.6562715768814087,
+      "learning_rate": 0.00037292377701934014,
+      "loss": 1.1522,
+      "step": 2760
+    },
+    {
+      "epoch": 3.1410693970420933,
+      "grad_norm": 0.7173290252685547,
+      "learning_rate": 0.0003726962457337884,
+      "loss": 1.3983,
+      "step": 2761
+    },
+    {
+      "epoch": 3.142207053469852,
+      "grad_norm": 0.8549534678459167,
+      "learning_rate": 0.00037246871444823667,
+      "loss": 1.4353,
+      "step": 2762
+    },
+    {
+      "epoch": 3.143344709897611,
+      "grad_norm": 1.8511520624160767,
+      "learning_rate": 0.0003722411831626849,
+      "loss": 2.9588,
+      "step": 2763
+    },
+    {
+      "epoch": 3.1444823663253696,
+      "grad_norm": 1.1187710762023926,
+      "learning_rate": 0.0003720136518771331,
+      "loss": 1.8577,
+      "step": 2764
+    },
+    {
+      "epoch": 3.1456200227531284,
+      "grad_norm": 0.9971904158592224,
+      "learning_rate": 0.00037178612059158135,
+      "loss": 2.0944,
+      "step": 2765
+    },
+    {
+      "epoch": 3.1467576791808876,
+      "grad_norm": 1.0542211532592773,
+      "learning_rate": 0.0003715585893060296,
+      "loss": 1.5842,
+      "step": 2766
+    },
+    {
+      "epoch": 3.1478953356086463,
+      "grad_norm": 0.8797990083694458,
+      "learning_rate": 0.0003713310580204778,
+      "loss": 2.6677,
+      "step": 2767
+    },
+    {
+      "epoch": 3.149032992036405,
+      "grad_norm": 1.0683292150497437,
+      "learning_rate": 0.00037110352673492604,
+      "loss": 1.4515,
+      "step": 2768
+    },
+    {
+      "epoch": 3.150170648464164,
+      "grad_norm": 0.9214059114456177,
+      "learning_rate": 0.0003708759954493743,
+      "loss": 2.1592,
+      "step": 2769
+    },
+    {
+      "epoch": 3.1513083048919226,
+      "grad_norm": 1.1661478281021118,
+      "learning_rate": 0.0003706484641638225,
+      "loss": 2.3451,
+      "step": 2770
+    },
+    {
+      "epoch": 3.1524459613196814,
+      "grad_norm": 0.8435385823249817,
+      "learning_rate": 0.0003704209328782708,
+      "loss": 1.3745,
+      "step": 2771
+    },
+    {
+      "epoch": 3.15358361774744,
+      "grad_norm": 0.646821916103363,
+      "learning_rate": 0.00037019340159271904,
+      "loss": 1.2736,
+      "step": 2772
+    },
+    {
+      "epoch": 3.1547212741751993,
+      "grad_norm": 1.5238386392593384,
+      "learning_rate": 0.0003699658703071672,
+      "loss": 2.0629,
+      "step": 2773
+    },
+    {
+      "epoch": 3.155858930602958,
+      "grad_norm": 0.9329938292503357,
+      "learning_rate": 0.00036973833902161546,
+      "loss": 2.6737,
+      "step": 2774
+    },
+    {
+      "epoch": 3.156996587030717,
+      "grad_norm": 0.9424443244934082,
+      "learning_rate": 0.0003695108077360637,
+      "loss": 1.814,
+      "step": 2775
+    },
+    {
+      "epoch": 3.1581342434584756,
+      "grad_norm": 1.060135006904602,
+      "learning_rate": 0.000369283276450512,
+      "loss": 2.2409,
+      "step": 2776
+    },
+    {
+      "epoch": 3.1592718998862344,
+      "grad_norm": 1.1120651960372925,
+      "learning_rate": 0.0003690557451649602,
+      "loss": 2.7252,
+      "step": 2777
+    },
+    {
+      "epoch": 3.160409556313993,
+      "grad_norm": 1.0458694696426392,
+      "learning_rate": 0.0003688282138794084,
+      "loss": 2.2002,
+      "step": 2778
+    },
+    {
+      "epoch": 3.161547212741752,
+      "grad_norm": 0.7672396898269653,
+      "learning_rate": 0.00036860068259385667,
+      "loss": 1.5965,
+      "step": 2779
+    },
+    {
+      "epoch": 3.1626848691695106,
+      "grad_norm": 0.9745705127716064,
+      "learning_rate": 0.0003683731513083049,
+      "loss": 2.8014,
+      "step": 2780
+    },
+    {
+      "epoch": 3.1638225255972694,
+      "grad_norm": 1.4451996088027954,
+      "learning_rate": 0.00036814562002275314,
+      "loss": 3.0917,
+      "step": 2781
+    },
+    {
+      "epoch": 3.1649601820250286,
+      "grad_norm": 0.7620997428894043,
+      "learning_rate": 0.0003679180887372014,
+      "loss": 1.4155,
+      "step": 2782
+    },
+    {
+      "epoch": 3.1660978384527874,
+      "grad_norm": 0.8162498474121094,
+      "learning_rate": 0.00036769055745164956,
+      "loss": 2.5782,
+      "step": 2783
+    },
+    {
+      "epoch": 3.167235494880546,
+      "grad_norm": 0.8843737244606018,
+      "learning_rate": 0.00036746302616609783,
+      "loss": 1.6138,
+      "step": 2784
+    },
+    {
+      "epoch": 3.168373151308305,
+      "grad_norm": 0.6327600479125977,
+      "learning_rate": 0.0003672354948805461,
+      "loss": 1.0368,
+      "step": 2785
+    },
+    {
+      "epoch": 3.1695108077360636,
+      "grad_norm": 1.3573901653289795,
+      "learning_rate": 0.00036700796359499436,
+      "loss": 3.1572,
+      "step": 2786
+    },
+    {
+      "epoch": 3.1706484641638224,
+      "grad_norm": 0.9209976196289062,
+      "learning_rate": 0.00036678043230944257,
+      "loss": 1.9979,
+      "step": 2787
+    },
+    {
+      "epoch": 3.171786120591581,
+      "grad_norm": 1.0654752254486084,
+      "learning_rate": 0.0003665529010238908,
+      "loss": 1.1564,
+      "step": 2788
+    },
+    {
+      "epoch": 3.1729237770193404,
+      "grad_norm": 0.8304941654205322,
+      "learning_rate": 0.00036632536973833904,
+      "loss": 0.804,
+      "step": 2789
+    },
+    {
+      "epoch": 3.174061433447099,
+      "grad_norm": 0.7033611536026001,
+      "learning_rate": 0.00036609783845278725,
+      "loss": 1.2548,
+      "step": 2790
+    },
+    {
+      "epoch": 3.175199089874858,
+      "grad_norm": 0.7177807688713074,
+      "learning_rate": 0.0003658703071672355,
+      "loss": 0.925,
+      "step": 2791
+    },
+    {
+      "epoch": 3.1763367463026166,
+      "grad_norm": 0.7976208925247192,
+      "learning_rate": 0.0003656427758816838,
+      "loss": 1.8367,
+      "step": 2792
+    },
+    {
+      "epoch": 3.1774744027303754,
+      "grad_norm": 1.0865761041641235,
+      "learning_rate": 0.00036541524459613193,
+      "loss": 2.0069,
+      "step": 2793
+    },
+    {
+      "epoch": 3.178612059158134,
+      "grad_norm": 1.0845462083816528,
+      "learning_rate": 0.0003651877133105802,
+      "loss": 2.4428,
+      "step": 2794
+    },
+    {
+      "epoch": 3.179749715585893,
+      "grad_norm": 1.635365605354309,
+      "learning_rate": 0.00036496018202502846,
+      "loss": 2.376,
+      "step": 2795
+    },
+    {
+      "epoch": 3.1808873720136517,
+      "grad_norm": 0.9851645231246948,
+      "learning_rate": 0.0003647326507394767,
+      "loss": 1.7787,
+      "step": 2796
+    },
+    {
+      "epoch": 3.182025028441411,
+      "grad_norm": 0.8619524836540222,
+      "learning_rate": 0.00036450511945392494,
+      "loss": 0.8745,
+      "step": 2797
+    },
+    {
+      "epoch": 3.1831626848691696,
+      "grad_norm": 0.8011617064476013,
+      "learning_rate": 0.00036427758816837315,
+      "loss": 1.2395,
+      "step": 2798
+    },
+    {
+      "epoch": 3.1843003412969284,
+      "grad_norm": 1.1883291006088257,
+      "learning_rate": 0.0003640500568828214,
+      "loss": 2.4134,
+      "step": 2799
+    },
+    {
+      "epoch": 3.185437997724687,
+      "grad_norm": 0.9594371914863586,
+      "learning_rate": 0.0003638225255972696,
+      "loss": 1.8051,
+      "step": 2800
+    },
+    {
+      "epoch": 3.186575654152446,
+      "grad_norm": 0.9812491536140442,
+      "learning_rate": 0.0003635949943117179,
+      "loss": 1.9984,
+      "step": 2801
+    },
+    {
+      "epoch": 3.1877133105802047,
+      "grad_norm": 1.5427114963531494,
+      "learning_rate": 0.0003633674630261661,
+      "loss": 2.8443,
+      "step": 2802
+    },
+    {
+      "epoch": 3.1888509670079634,
+      "grad_norm": 1.0131993293762207,
+      "learning_rate": 0.0003631399317406143,
+      "loss": 1.4077,
+      "step": 2803
+    },
+    {
+      "epoch": 3.189988623435722,
+      "grad_norm": 1.21848464012146,
+      "learning_rate": 0.00036291240045506257,
+      "loss": 2.6364,
+      "step": 2804
+    },
+    {
+      "epoch": 3.1911262798634814,
+      "grad_norm": 0.8747435212135315,
+      "learning_rate": 0.00036268486916951083,
+      "loss": 1.4387,
+      "step": 2805
+    },
+    {
+      "epoch": 3.19226393629124,
+      "grad_norm": 1.2929614782333374,
+      "learning_rate": 0.0003624573378839591,
+      "loss": 2.2714,
+      "step": 2806
+    },
+    {
+      "epoch": 3.193401592718999,
+      "grad_norm": 0.6196789145469666,
+      "learning_rate": 0.00036222980659840725,
+      "loss": 1.4344,
+      "step": 2807
+    },
+    {
+      "epoch": 3.1945392491467577,
+      "grad_norm": 1.1467669010162354,
+      "learning_rate": 0.0003620022753128555,
+      "loss": 2.1119,
+      "step": 2808
+    },
+    {
+      "epoch": 3.1956769055745164,
+      "grad_norm": 0.781956136226654,
+      "learning_rate": 0.0003617747440273038,
+      "loss": 1.4887,
+      "step": 2809
+    },
+    {
+      "epoch": 3.196814562002275,
+      "grad_norm": 1.177976369857788,
+      "learning_rate": 0.000361547212741752,
+      "loss": 3.2043,
+      "step": 2810
+    },
+    {
+      "epoch": 3.197952218430034,
+      "grad_norm": 0.8673222064971924,
+      "learning_rate": 0.00036131968145620025,
+      "loss": 1.0446,
+      "step": 2811
+    },
+    {
+      "epoch": 3.199089874857793,
+      "grad_norm": 0.8764348030090332,
+      "learning_rate": 0.00036109215017064846,
+      "loss": 2.2418,
+      "step": 2812
+    },
+    {
+      "epoch": 3.200227531285552,
+      "grad_norm": 1.8115644454956055,
+      "learning_rate": 0.0003608646188850967,
+      "loss": 1.0493,
+      "step": 2813
+    },
+    {
+      "epoch": 3.2013651877133107,
+      "grad_norm": 1.0345537662506104,
+      "learning_rate": 0.00036063708759954494,
+      "loss": 2.0083,
+      "step": 2814
+    },
+    {
+      "epoch": 3.2025028441410694,
+      "grad_norm": 0.7956832647323608,
+      "learning_rate": 0.0003604095563139932,
+      "loss": 1.377,
+      "step": 2815
+    },
+    {
+      "epoch": 3.203640500568828,
+      "grad_norm": 0.7763450145721436,
+      "learning_rate": 0.00036018202502844147,
+      "loss": 1.4905,
+      "step": 2816
+    },
+    {
+      "epoch": 3.204778156996587,
+      "grad_norm": 1.0623788833618164,
+      "learning_rate": 0.0003599544937428896,
+      "loss": 1.4396,
+      "step": 2817
+    },
+    {
+      "epoch": 3.2059158134243457,
+      "grad_norm": 0.8401424288749695,
+      "learning_rate": 0.0003597269624573379,
+      "loss": 1.7506,
+      "step": 2818
+    },
+    {
+      "epoch": 3.2070534698521045,
+      "grad_norm": 1.2517905235290527,
+      "learning_rate": 0.00035949943117178615,
+      "loss": 1.8421,
+      "step": 2819
+    },
+    {
+      "epoch": 3.2081911262798632,
+      "grad_norm": 0.9095095992088318,
+      "learning_rate": 0.00035927189988623436,
+      "loss": 1.069,
+      "step": 2820
+    },
+    {
+      "epoch": 3.2093287827076225,
+      "grad_norm": 1.1191394329071045,
+      "learning_rate": 0.0003590443686006826,
+      "loss": 2.1787,
+      "step": 2821
+    },
+    {
+      "epoch": 3.210466439135381,
+      "grad_norm": 0.9273049235343933,
+      "learning_rate": 0.00035881683731513083,
+      "loss": 2.2313,
+      "step": 2822
+    },
+    {
+      "epoch": 3.21160409556314,
+      "grad_norm": 1.0636650323867798,
+      "learning_rate": 0.00035858930602957904,
+      "loss": 1.5293,
+      "step": 2823
+    },
+    {
+      "epoch": 3.2127417519908987,
+      "grad_norm": 1.2871456146240234,
+      "learning_rate": 0.0003583617747440273,
+      "loss": 3.0545,
+      "step": 2824
+    },
+    {
+      "epoch": 3.2138794084186575,
+      "grad_norm": 1.2373828887939453,
+      "learning_rate": 0.00035813424345847557,
+      "loss": 2.0834,
+      "step": 2825
+    },
+    {
+      "epoch": 3.2150170648464163,
+      "grad_norm": 0.7762901186943054,
+      "learning_rate": 0.00035790671217292384,
+      "loss": 1.4895,
+      "step": 2826
+    },
+    {
+      "epoch": 3.216154721274175,
+      "grad_norm": 1.2859349250793457,
+      "learning_rate": 0.000357679180887372,
+      "loss": 1.9091,
+      "step": 2827
+    },
+    {
+      "epoch": 3.217292377701934,
+      "grad_norm": 1.2489643096923828,
+      "learning_rate": 0.00035745164960182026,
+      "loss": 1.831,
+      "step": 2828
+    },
+    {
+      "epoch": 3.218430034129693,
+      "grad_norm": 0.9940024018287659,
+      "learning_rate": 0.0003572241183162685,
+      "loss": 2.157,
+      "step": 2829
+    },
+    {
+      "epoch": 3.2195676905574517,
+      "grad_norm": 1.1232733726501465,
+      "learning_rate": 0.00035699658703071673,
+      "loss": 2.9384,
+      "step": 2830
+    },
+    {
+      "epoch": 3.2207053469852105,
+      "grad_norm": 0.7718859910964966,
+      "learning_rate": 0.000356769055745165,
+      "loss": 1.8196,
+      "step": 2831
+    },
+    {
+      "epoch": 3.2218430034129693,
+      "grad_norm": 1.089641809463501,
+      "learning_rate": 0.0003565415244596132,
+      "loss": 1.8792,
+      "step": 2832
+    },
+    {
+      "epoch": 3.222980659840728,
+      "grad_norm": 1.2740263938903809,
+      "learning_rate": 0.0003563139931740614,
+      "loss": 2.5414,
+      "step": 2833
+    },
+    {
+      "epoch": 3.2241183162684868,
+      "grad_norm": 0.9137701392173767,
+      "learning_rate": 0.0003560864618885097,
+      "loss": 1.3666,
+      "step": 2834
+    },
+    {
+      "epoch": 3.2252559726962455,
+      "grad_norm": 0.6575736403465271,
+      "learning_rate": 0.00035585893060295794,
+      "loss": 1.2568,
+      "step": 2835
+    },
+    {
+      "epoch": 3.2263936291240047,
+      "grad_norm": 0.9433609843254089,
+      "learning_rate": 0.0003556313993174061,
+      "loss": 1.4612,
+      "step": 2836
+    },
+    {
+      "epoch": 3.2275312855517635,
+      "grad_norm": 0.8380080461502075,
+      "learning_rate": 0.00035540386803185436,
+      "loss": 1.1819,
+      "step": 2837
+    },
+    {
+      "epoch": 3.2286689419795223,
+      "grad_norm": 0.7855201363563538,
+      "learning_rate": 0.0003551763367463026,
+      "loss": 1.8921,
+      "step": 2838
+    },
+    {
+      "epoch": 3.229806598407281,
+      "grad_norm": 1.5746235847473145,
+      "learning_rate": 0.0003549488054607509,
+      "loss": 3.0391,
+      "step": 2839
+    },
+    {
+      "epoch": 3.2309442548350398,
+      "grad_norm": 1.1520813703536987,
+      "learning_rate": 0.0003547212741751991,
+      "loss": 2.1142,
+      "step": 2840
+    },
+    {
+      "epoch": 3.2320819112627985,
+      "grad_norm": 0.8569504618644714,
+      "learning_rate": 0.0003544937428896473,
+      "loss": 2.0138,
+      "step": 2841
+    },
+    {
+      "epoch": 3.2332195676905573,
+      "grad_norm": 1.0433871746063232,
+      "learning_rate": 0.0003542662116040956,
+      "loss": 1.673,
+      "step": 2842
+    },
+    {
+      "epoch": 3.234357224118316,
+      "grad_norm": 1.01398766040802,
+      "learning_rate": 0.0003540386803185438,
+      "loss": 1.9582,
+      "step": 2843
+    },
+    {
+      "epoch": 3.2354948805460753,
+      "grad_norm": 1.0465679168701172,
+      "learning_rate": 0.00035381114903299205,
+      "loss": 1.9819,
+      "step": 2844
+    },
+    {
+      "epoch": 3.236632536973834,
+      "grad_norm": 0.787959635257721,
+      "learning_rate": 0.0003535836177474403,
+      "loss": 1.639,
+      "step": 2845
+    },
+    {
+      "epoch": 3.2377701934015928,
+      "grad_norm": 1.155023455619812,
+      "learning_rate": 0.00035335608646188847,
+      "loss": 1.8541,
+      "step": 2846
+    },
+    {
+      "epoch": 3.2389078498293515,
+      "grad_norm": 1.1908761262893677,
+      "learning_rate": 0.00035312855517633673,
+      "loss": 1.9333,
+      "step": 2847
+    },
+    {
+      "epoch": 3.2400455062571103,
+      "grad_norm": 0.9262527227401733,
+      "learning_rate": 0.000352901023890785,
+      "loss": 1.3209,
+      "step": 2848
+    },
+    {
+      "epoch": 3.241183162684869,
+      "grad_norm": 0.8731938004493713,
+      "learning_rate": 0.00035267349260523326,
+      "loss": 2.2036,
+      "step": 2849
+    },
+    {
+      "epoch": 3.242320819112628,
+      "grad_norm": 0.6273211240768433,
+      "learning_rate": 0.00035244596131968147,
+      "loss": 0.9153,
+      "step": 2850
+    },
+    {
+      "epoch": 3.243458475540387,
+      "grad_norm": 1.0732612609863281,
+      "learning_rate": 0.0003522184300341297,
+      "loss": 2.286,
+      "step": 2851
+    },
+    {
+      "epoch": 3.244596131968146,
+      "grad_norm": 1.1554096937179565,
+      "learning_rate": 0.00035199089874857794,
+      "loss": 2.0289,
+      "step": 2852
+    },
+    {
+      "epoch": 3.2457337883959045,
+      "grad_norm": 0.8074318170547485,
+      "learning_rate": 0.00035176336746302615,
+      "loss": 1.8181,
+      "step": 2853
+    },
+    {
+      "epoch": 3.2468714448236633,
+      "grad_norm": 1.392298936843872,
+      "learning_rate": 0.0003515358361774744,
+      "loss": 3.4711,
+      "step": 2854
+    },
+    {
+      "epoch": 3.248009101251422,
+      "grad_norm": 0.9060192108154297,
+      "learning_rate": 0.0003513083048919227,
+      "loss": 2.0728,
+      "step": 2855
+    },
+    {
+      "epoch": 3.249146757679181,
+      "grad_norm": 1.2438005208969116,
+      "learning_rate": 0.00035108077360637084,
+      "loss": 1.4773,
+      "step": 2856
+    },
+    {
+      "epoch": 3.2502844141069396,
+      "grad_norm": 1.0764005184173584,
+      "learning_rate": 0.0003508532423208191,
+      "loss": 1.5244,
+      "step": 2857
+    },
+    {
+      "epoch": 3.2514220705346983,
+      "grad_norm": 0.8813822269439697,
+      "learning_rate": 0.00035062571103526737,
+      "loss": 1.2567,
+      "step": 2858
+    },
+    {
+      "epoch": 3.252559726962457,
+      "grad_norm": 0.6739155054092407,
+      "learning_rate": 0.00035039817974971563,
+      "loss": 0.682,
+      "step": 2859
+    },
+    {
+      "epoch": 3.2536973833902163,
+      "grad_norm": 1.2803453207015991,
+      "learning_rate": 0.00035017064846416384,
+      "loss": 1.5126,
+      "step": 2860
+    },
+    {
+      "epoch": 3.254835039817975,
+      "grad_norm": 1.1614502668380737,
+      "learning_rate": 0.00034994311717861205,
+      "loss": 3.0873,
+      "step": 2861
+    },
+    {
+      "epoch": 3.255972696245734,
+      "grad_norm": 0.8287785053253174,
+      "learning_rate": 0.0003497155858930603,
+      "loss": 1.3949,
+      "step": 2862
+    },
+    {
+      "epoch": 3.2571103526734926,
+      "grad_norm": 0.8668496608734131,
+      "learning_rate": 0.0003494880546075085,
+      "loss": 1.5943,
+      "step": 2863
+    },
+    {
+      "epoch": 3.2582480091012513,
+      "grad_norm": 0.8574039936065674,
+      "learning_rate": 0.0003492605233219568,
+      "loss": 1.8113,
+      "step": 2864
+    },
+    {
+      "epoch": 3.25938566552901,
+      "grad_norm": 0.8691564202308655,
+      "learning_rate": 0.00034903299203640505,
+      "loss": 1.5579,
+      "step": 2865
+    },
+    {
+      "epoch": 3.260523321956769,
+      "grad_norm": 0.948115885257721,
+      "learning_rate": 0.0003488054607508532,
+      "loss": 1.4459,
+      "step": 2866
+    },
+    {
+      "epoch": 3.261660978384528,
+      "grad_norm": 1.410564661026001,
+      "learning_rate": 0.00034857792946530147,
+      "loss": 3.3305,
+      "step": 2867
+    },
+    {
+      "epoch": 3.262798634812287,
+      "grad_norm": 0.904012143611908,
+      "learning_rate": 0.00034835039817974973,
+      "loss": 1.7655,
+      "step": 2868
+    },
+    {
+      "epoch": 3.2639362912400456,
+      "grad_norm": 0.9044657945632935,
+      "learning_rate": 0.000348122866894198,
+      "loss": 1.2666,
+      "step": 2869
+    },
+    {
+      "epoch": 3.2650739476678043,
+      "grad_norm": 0.8898612856864929,
+      "learning_rate": 0.00034789533560864615,
+      "loss": 2.0107,
+      "step": 2870
+    },
+    {
+      "epoch": 3.266211604095563,
+      "grad_norm": 0.8412598967552185,
+      "learning_rate": 0.0003476678043230944,
+      "loss": 1.4021,
+      "step": 2871
+    },
+    {
+      "epoch": 3.267349260523322,
+      "grad_norm": 1.2648974657058716,
+      "learning_rate": 0.0003474402730375427,
+      "loss": 3.4206,
+      "step": 2872
+    },
+    {
+      "epoch": 3.2684869169510806,
+      "grad_norm": 0.9399024248123169,
+      "learning_rate": 0.0003472127417519909,
+      "loss": 2.8845,
+      "step": 2873
+    },
+    {
+      "epoch": 3.26962457337884,
+      "grad_norm": 0.9316064119338989,
+      "learning_rate": 0.00034698521046643916,
+      "loss": 2.0801,
+      "step": 2874
+    },
+    {
+      "epoch": 3.2707622298065986,
+      "grad_norm": 0.9166231155395508,
+      "learning_rate": 0.00034675767918088737,
+      "loss": 1.661,
+      "step": 2875
+    },
+    {
+      "epoch": 3.2718998862343573,
+      "grad_norm": 1.0802944898605347,
+      "learning_rate": 0.0003465301478953356,
+      "loss": 3.3163,
+      "step": 2876
+    },
+    {
+      "epoch": 3.273037542662116,
+      "grad_norm": 1.2445770502090454,
+      "learning_rate": 0.00034630261660978384,
+      "loss": 2.9616,
+      "step": 2877
+    },
+    {
+      "epoch": 3.274175199089875,
+      "grad_norm": 0.8100591897964478,
+      "learning_rate": 0.0003460750853242321,
+      "loss": 1.1971,
+      "step": 2878
+    },
+    {
+      "epoch": 3.2753128555176336,
+      "grad_norm": 0.9066864252090454,
+      "learning_rate": 0.00034584755403868037,
+      "loss": 1.7813,
+      "step": 2879
+    },
+    {
+      "epoch": 3.2764505119453924,
+      "grad_norm": 1.0590927600860596,
+      "learning_rate": 0.0003456200227531285,
+      "loss": 2.0441,
+      "step": 2880
+    },
+    {
+      "epoch": 3.277588168373151,
+      "grad_norm": 0.6991727352142334,
+      "learning_rate": 0.0003453924914675768,
+      "loss": 1.2761,
+      "step": 2881
+    },
+    {
+      "epoch": 3.27872582480091,
+      "grad_norm": 0.6339994668960571,
+      "learning_rate": 0.00034516496018202505,
+      "loss": 1.0168,
+      "step": 2882
+    },
+    {
+      "epoch": 3.279863481228669,
+      "grad_norm": 0.630440890789032,
+      "learning_rate": 0.00034493742889647326,
+      "loss": 1.3744,
+      "step": 2883
+    },
+    {
+      "epoch": 3.281001137656428,
+      "grad_norm": 0.8100972771644592,
+      "learning_rate": 0.0003447098976109215,
+      "loss": 1.9117,
+      "step": 2884
+    },
+    {
+      "epoch": 3.2821387940841866,
+      "grad_norm": 0.9163875579833984,
+      "learning_rate": 0.00034448236632536974,
+      "loss": 1.1392,
+      "step": 2885
+    },
+    {
+      "epoch": 3.2832764505119454,
+      "grad_norm": 0.9157561659812927,
+      "learning_rate": 0.00034425483503981795,
+      "loss": 1.9776,
+      "step": 2886
+    },
+    {
+      "epoch": 3.284414106939704,
+      "grad_norm": 0.7969180345535278,
+      "learning_rate": 0.0003440273037542662,
+      "loss": 1.4908,
+      "step": 2887
+    },
+    {
+      "epoch": 3.285551763367463,
+      "grad_norm": 0.7929701209068298,
+      "learning_rate": 0.0003437997724687145,
+      "loss": 1.6476,
+      "step": 2888
+    },
+    {
+      "epoch": 3.2866894197952217,
+      "grad_norm": 1.0109394788742065,
+      "learning_rate": 0.00034357224118316274,
+      "loss": 2.3243,
+      "step": 2889
+    },
+    {
+      "epoch": 3.287827076222981,
+      "grad_norm": 1.095586895942688,
+      "learning_rate": 0.0003433447098976109,
+      "loss": 2.0614,
+      "step": 2890
+    },
+    {
+      "epoch": 3.2889647326507396,
+      "grad_norm": 0.8181962966918945,
+      "learning_rate": 0.00034311717861205916,
+      "loss": 1.0926,
+      "step": 2891
+    },
+    {
+      "epoch": 3.2901023890784984,
+      "grad_norm": 0.5682795643806458,
+      "learning_rate": 0.0003428896473265074,
+      "loss": 0.5625,
+      "step": 2892
+    },
+    {
+      "epoch": 3.291240045506257,
+      "grad_norm": 0.8557625412940979,
+      "learning_rate": 0.00034266211604095563,
+      "loss": 1.8372,
+      "step": 2893
+    },
+    {
+      "epoch": 3.292377701934016,
+      "grad_norm": 0.6423472166061401,
+      "learning_rate": 0.0003424345847554039,
+      "loss": 1.3594,
+      "step": 2894
+    },
+    {
+      "epoch": 3.2935153583617747,
+      "grad_norm": 1.044715404510498,
+      "learning_rate": 0.0003422070534698521,
+      "loss": 2.7601,
+      "step": 2895
+    },
+    {
+      "epoch": 3.2946530147895334,
+      "grad_norm": 0.7515947222709656,
+      "learning_rate": 0.0003419795221843003,
+      "loss": 1.1865,
+      "step": 2896
+    },
+    {
+      "epoch": 3.295790671217292,
+      "grad_norm": 0.9097882509231567,
+      "learning_rate": 0.0003417519908987486,
+      "loss": 1.4461,
+      "step": 2897
+    },
+    {
+      "epoch": 3.296928327645051,
+      "grad_norm": 0.7172821760177612,
+      "learning_rate": 0.00034152445961319684,
+      "loss": 2.2885,
+      "step": 2898
+    },
+    {
+      "epoch": 3.29806598407281,
+      "grad_norm": 1.1029298305511475,
+      "learning_rate": 0.0003412969283276451,
+      "loss": 1.5583,
+      "step": 2899
+    },
+    {
+      "epoch": 3.299203640500569,
+      "grad_norm": 0.853276252746582,
+      "learning_rate": 0.00034106939704209326,
+      "loss": 1.4709,
+      "step": 2900
+    },
+    {
+      "epoch": 3.3003412969283277,
+      "grad_norm": 0.8551457524299622,
+      "learning_rate": 0.00034084186575654153,
+      "loss": 2.4956,
+      "step": 2901
+    },
+    {
+      "epoch": 3.3014789533560864,
+      "grad_norm": 0.909735381603241,
+      "learning_rate": 0.0003406143344709898,
+      "loss": 2.0525,
+      "step": 2902
+    },
+    {
+      "epoch": 3.302616609783845,
+      "grad_norm": 1.0006370544433594,
+      "learning_rate": 0.000340386803185438,
+      "loss": 1.6306,
+      "step": 2903
+    },
+    {
+      "epoch": 3.303754266211604,
+      "grad_norm": 0.5882608294487,
+      "learning_rate": 0.00034015927189988627,
+      "loss": 1.3989,
+      "step": 2904
+    },
+    {
+      "epoch": 3.3048919226393627,
+      "grad_norm": 1.2735271453857422,
+      "learning_rate": 0.0003399317406143345,
+      "loss": 2.394,
+      "step": 2905
+    },
+    {
+      "epoch": 3.306029579067122,
+      "grad_norm": 0.9738203287124634,
+      "learning_rate": 0.0003397042093287827,
+      "loss": 1.951,
+      "step": 2906
+    },
+    {
+      "epoch": 3.3071672354948807,
+      "grad_norm": 0.5416101217269897,
+      "learning_rate": 0.00033947667804323095,
+      "loss": 1.0747,
+      "step": 2907
+    },
+    {
+      "epoch": 3.3083048919226394,
+      "grad_norm": 0.9188492298126221,
+      "learning_rate": 0.0003392491467576792,
+      "loss": 1.4549,
+      "step": 2908
+    },
+    {
+      "epoch": 3.309442548350398,
+      "grad_norm": 1.4744441509246826,
+      "learning_rate": 0.0003390216154721274,
+      "loss": 2.4909,
+      "step": 2909
+    },
+    {
+      "epoch": 3.310580204778157,
+      "grad_norm": 0.6633999347686768,
+      "learning_rate": 0.00033879408418657563,
+      "loss": 1.4089,
+      "step": 2910
+    },
+    {
+      "epoch": 3.3117178612059157,
+      "grad_norm": 0.6373212933540344,
+      "learning_rate": 0.0003385665529010239,
+      "loss": 1.4588,
+      "step": 2911
+    },
+    {
+      "epoch": 3.3128555176336745,
+      "grad_norm": 0.5806794166564941,
+      "learning_rate": 0.00033833902161547216,
+      "loss": 1.3158,
+      "step": 2912
+    },
+    {
+      "epoch": 3.3139931740614337,
+      "grad_norm": 1.0303038358688354,
+      "learning_rate": 0.00033811149032992037,
+      "loss": 1.6245,
+      "step": 2913
+    },
+    {
+      "epoch": 3.3151308304891924,
+      "grad_norm": 1.0633960962295532,
+      "learning_rate": 0.0003378839590443686,
+      "loss": 1.9088,
+      "step": 2914
+    },
+    {
+      "epoch": 3.316268486916951,
+      "grad_norm": 0.9028981328010559,
+      "learning_rate": 0.00033765642775881685,
+      "loss": 1.64,
+      "step": 2915
+    },
+    {
+      "epoch": 3.31740614334471,
+      "grad_norm": 0.886370062828064,
+      "learning_rate": 0.00033742889647326506,
+      "loss": 1.292,
+      "step": 2916
+    },
+    {
+      "epoch": 3.3185437997724687,
+      "grad_norm": 1.6007177829742432,
+      "learning_rate": 0.0003372013651877133,
+      "loss": 3.1382,
+      "step": 2917
+    },
+    {
+      "epoch": 3.3196814562002275,
+      "grad_norm": 1.1357500553131104,
+      "learning_rate": 0.0003369738339021616,
+      "loss": 2.1361,
+      "step": 2918
+    },
+    {
+      "epoch": 3.3208191126279862,
+      "grad_norm": 0.9156873822212219,
+      "learning_rate": 0.0003367463026166098,
+      "loss": 1.3616,
+      "step": 2919
+    },
+    {
+      "epoch": 3.321956769055745,
+      "grad_norm": 1.0556998252868652,
+      "learning_rate": 0.000336518771331058,
+      "loss": 2.6077,
+      "step": 2920
+    },
+    {
+      "epoch": 3.3230944254835038,
+      "grad_norm": 1.1073150634765625,
+      "learning_rate": 0.00033629124004550627,
+      "loss": 1.6265,
+      "step": 2921
+    },
+    {
+      "epoch": 3.324232081911263,
+      "grad_norm": 0.6052758097648621,
+      "learning_rate": 0.00033606370875995453,
+      "loss": 1.0629,
+      "step": 2922
+    },
+    {
+      "epoch": 3.3253697383390217,
+      "grad_norm": 0.9209448099136353,
+      "learning_rate": 0.00033583617747440274,
+      "loss": 1.3336,
+      "step": 2923
+    },
+    {
+      "epoch": 3.3265073947667805,
+      "grad_norm": 0.5461276769638062,
+      "learning_rate": 0.00033560864618885095,
+      "loss": 1.1149,
+      "step": 2924
+    },
+    {
+      "epoch": 3.3276450511945392,
+      "grad_norm": 1.0375051498413086,
+      "learning_rate": 0.0003353811149032992,
+      "loss": 1.4975,
+      "step": 2925
+    },
+    {
+      "epoch": 3.328782707622298,
+      "grad_norm": 0.9948595762252808,
+      "learning_rate": 0.0003351535836177474,
+      "loss": 1.7277,
+      "step": 2926
+    },
+    {
+      "epoch": 3.3299203640500568,
+      "grad_norm": 1.711958408355713,
+      "learning_rate": 0.0003349260523321957,
+      "loss": 2.7451,
+      "step": 2927
+    },
+    {
+      "epoch": 3.3310580204778155,
+      "grad_norm": 0.9110103845596313,
+      "learning_rate": 0.00033469852104664395,
+      "loss": 1.8131,
+      "step": 2928
+    },
+    {
+      "epoch": 3.3321956769055747,
+      "grad_norm": 0.9176352620124817,
+      "learning_rate": 0.0003344709897610921,
+      "loss": 1.4978,
+      "step": 2929
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.8127102255821228,
+      "learning_rate": 0.0003342434584755404,
+      "loss": 1.8899,
+      "step": 2930
+    },
+    {
+      "epoch": 3.3344709897610922,
+      "grad_norm": 1.1340645551681519,
+      "learning_rate": 0.00033401592718998864,
+      "loss": 1.3567,
+      "step": 2931
+    },
+    {
+      "epoch": 3.335608646188851,
+      "grad_norm": 1.366903305053711,
+      "learning_rate": 0.0003337883959044369,
+      "loss": 3.0076,
+      "step": 2932
+    },
+    {
+      "epoch": 3.3367463026166098,
+      "grad_norm": 0.8700219392776489,
+      "learning_rate": 0.0003335608646188851,
+      "loss": 1.2597,
+      "step": 2933
+    },
+    {
+      "epoch": 3.3378839590443685,
+      "grad_norm": 0.9277002215385437,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 2.267,
+      "step": 2934
+    },
+    {
+      "epoch": 3.3390216154721273,
+      "grad_norm": 0.5930815935134888,
+      "learning_rate": 0.0003331058020477816,
+      "loss": 1.3837,
+      "step": 2935
+    },
+    {
+      "epoch": 3.3401592718998865,
+      "grad_norm": 1.0349963903427124,
+      "learning_rate": 0.0003328782707622298,
+      "loss": 2.4195,
+      "step": 2936
+    },
+    {
+      "epoch": 3.3412969283276452,
+      "grad_norm": 1.2994656562805176,
+      "learning_rate": 0.00033265073947667806,
+      "loss": 1.4095,
+      "step": 2937
+    },
+    {
+      "epoch": 3.342434584755404,
+      "grad_norm": 0.7547191381454468,
+      "learning_rate": 0.0003324232081911263,
+      "loss": 1.6748,
+      "step": 2938
+    },
+    {
+      "epoch": 3.3435722411831628,
+      "grad_norm": 0.9032368659973145,
+      "learning_rate": 0.0003321956769055745,
+      "loss": 2.3666,
+      "step": 2939
+    },
+    {
+      "epoch": 3.3447098976109215,
+      "grad_norm": 0.8295652866363525,
+      "learning_rate": 0.00033196814562002274,
+      "loss": 1.4128,
+      "step": 2940
+    },
+    {
+      "epoch": 3.3458475540386803,
+      "grad_norm": 0.6656214594841003,
+      "learning_rate": 0.000331740614334471,
+      "loss": 0.8634,
+      "step": 2941
+    },
+    {
+      "epoch": 3.346985210466439,
+      "grad_norm": 0.9207725524902344,
+      "learning_rate": 0.00033151308304891927,
+      "loss": 2.3504,
+      "step": 2942
+    },
+    {
+      "epoch": 3.348122866894198,
+      "grad_norm": 0.7400831580162048,
+      "learning_rate": 0.00033128555176336743,
+      "loss": 0.9548,
+      "step": 2943
+    },
+    {
+      "epoch": 3.3492605233219566,
+      "grad_norm": 0.9420154690742493,
+      "learning_rate": 0.0003310580204778157,
+      "loss": 2.347,
+      "step": 2944
+    },
+    {
+      "epoch": 3.3503981797497158,
+      "grad_norm": 0.9879665970802307,
+      "learning_rate": 0.00033083048919226396,
+      "loss": 1.6713,
+      "step": 2945
+    },
+    {
+      "epoch": 3.3515358361774745,
+      "grad_norm": 0.8822195529937744,
+      "learning_rate": 0.00033060295790671217,
+      "loss": 0.9356,
+      "step": 2946
+    },
+    {
+      "epoch": 3.3526734926052333,
+      "grad_norm": 0.959632933139801,
+      "learning_rate": 0.00033037542662116043,
+      "loss": 2.099,
+      "step": 2947
+    },
+    {
+      "epoch": 3.353811149032992,
+      "grad_norm": 0.8753674626350403,
+      "learning_rate": 0.00033014789533560864,
+      "loss": 1.0478,
+      "step": 2948
+    },
+    {
+      "epoch": 3.354948805460751,
+      "grad_norm": 0.9367252588272095,
+      "learning_rate": 0.00032992036405005685,
+      "loss": 2.0897,
+      "step": 2949
+    },
+    {
+      "epoch": 3.3560864618885096,
+      "grad_norm": 1.059227705001831,
+      "learning_rate": 0.0003296928327645051,
+      "loss": 2.1907,
+      "step": 2950
+    },
+    {
+      "epoch": 3.3572241183162683,
+      "grad_norm": 1.2451213598251343,
+      "learning_rate": 0.0003294653014789534,
+      "loss": 2.9248,
+      "step": 2951
+    },
+    {
+      "epoch": 3.3583617747440275,
+      "grad_norm": 1.4228962659835815,
+      "learning_rate": 0.00032923777019340164,
+      "loss": 4.301,
+      "step": 2952
+    },
+    {
+      "epoch": 3.3594994311717863,
+      "grad_norm": 1.0747414827346802,
+      "learning_rate": 0.0003290102389078498,
+      "loss": 1.7888,
+      "step": 2953
+    },
+    {
+      "epoch": 3.360637087599545,
+      "grad_norm": 1.38980233669281,
+      "learning_rate": 0.00032878270762229806,
+      "loss": 0.5853,
+      "step": 2954
+    },
+    {
+      "epoch": 3.361774744027304,
+      "grad_norm": 0.793297290802002,
+      "learning_rate": 0.0003285551763367463,
+      "loss": 1.1242,
+      "step": 2955
+    },
+    {
+      "epoch": 3.3629124004550626,
+      "grad_norm": 2.7201013565063477,
+      "learning_rate": 0.00032832764505119454,
+      "loss": 4.6065,
+      "step": 2956
+    },
+    {
+      "epoch": 3.3640500568828213,
+      "grad_norm": 1.3290942907333374,
+      "learning_rate": 0.0003281001137656428,
+      "loss": 2.6127,
+      "step": 2957
+    },
+    {
+      "epoch": 3.36518771331058,
+      "grad_norm": 1.104690670967102,
+      "learning_rate": 0.000327872582480091,
+      "loss": 1.6713,
+      "step": 2958
+    },
+    {
+      "epoch": 3.366325369738339,
+      "grad_norm": 0.6908164620399475,
+      "learning_rate": 0.0003276450511945392,
+      "loss": 1.0,
+      "step": 2959
+    },
+    {
+      "epoch": 3.3674630261660976,
+      "grad_norm": 0.8555911779403687,
+      "learning_rate": 0.0003274175199089875,
+      "loss": 1.6721,
+      "step": 2960
+    },
+    {
+      "epoch": 3.368600682593857,
+      "grad_norm": 0.8147174119949341,
+      "learning_rate": 0.00032718998862343575,
+      "loss": 2.1911,
+      "step": 2961
+    },
+    {
+      "epoch": 3.3697383390216156,
+      "grad_norm": 1.1406668424606323,
+      "learning_rate": 0.000326962457337884,
+      "loss": 2.4286,
+      "step": 2962
+    },
+    {
+      "epoch": 3.3708759954493743,
+      "grad_norm": 0.8895596861839294,
+      "learning_rate": 0.00032673492605233217,
+      "loss": 1.8801,
+      "step": 2963
+    },
+    {
+      "epoch": 3.372013651877133,
+      "grad_norm": 0.6777999401092529,
+      "learning_rate": 0.00032650739476678043,
+      "loss": 0.8933,
+      "step": 2964
+    },
+    {
+      "epoch": 3.373151308304892,
+      "grad_norm": 0.8110511898994446,
+      "learning_rate": 0.0003262798634812287,
+      "loss": 1.6234,
+      "step": 2965
+    },
+    {
+      "epoch": 3.3742889647326506,
+      "grad_norm": 1.3193098306655884,
+      "learning_rate": 0.0003260523321956769,
+      "loss": 2.0755,
+      "step": 2966
+    },
+    {
+      "epoch": 3.3754266211604094,
+      "grad_norm": 1.0191433429718018,
+      "learning_rate": 0.00032582480091012517,
+      "loss": 2.6676,
+      "step": 2967
+    },
+    {
+      "epoch": 3.3765642775881686,
+      "grad_norm": 0.5699923038482666,
+      "learning_rate": 0.0003255972696245734,
+      "loss": 1.0331,
+      "step": 2968
+    },
+    {
+      "epoch": 3.3777019340159273,
+      "grad_norm": 1.2073493003845215,
+      "learning_rate": 0.0003253697383390216,
+      "loss": 3.2836,
+      "step": 2969
+    },
+    {
+      "epoch": 3.378839590443686,
+      "grad_norm": 0.5660185217857361,
+      "learning_rate": 0.00032514220705346985,
+      "loss": 0.8925,
+      "step": 2970
+    },
+    {
+      "epoch": 3.379977246871445,
+      "grad_norm": 1.1539686918258667,
+      "learning_rate": 0.0003249146757679181,
+      "loss": 1.6701,
+      "step": 2971
+    },
+    {
+      "epoch": 3.3811149032992036,
+      "grad_norm": 1.223471999168396,
+      "learning_rate": 0.0003246871444823664,
+      "loss": 2.0681,
+      "step": 2972
+    },
+    {
+      "epoch": 3.3822525597269624,
+      "grad_norm": 0.7714402675628662,
+      "learning_rate": 0.00032445961319681454,
+      "loss": 1.9618,
+      "step": 2973
+    },
+    {
+      "epoch": 3.383390216154721,
+      "grad_norm": 1.1004797220230103,
+      "learning_rate": 0.0003242320819112628,
+      "loss": 1.4624,
+      "step": 2974
+    },
+    {
+      "epoch": 3.3845278725824803,
+      "grad_norm": 0.7591771483421326,
+      "learning_rate": 0.00032400455062571107,
+      "loss": 1.1837,
+      "step": 2975
+    },
+    {
+      "epoch": 3.385665529010239,
+      "grad_norm": 0.9769864082336426,
+      "learning_rate": 0.0003237770193401593,
+      "loss": 1.6725,
+      "step": 2976
+    },
+    {
+      "epoch": 3.386803185437998,
+      "grad_norm": 1.0868338346481323,
+      "learning_rate": 0.0003235494880546075,
+      "loss": 2.3017,
+      "step": 2977
+    },
+    {
+      "epoch": 3.3879408418657566,
+      "grad_norm": 0.943532407283783,
+      "learning_rate": 0.00032332195676905575,
+      "loss": 2.1185,
+      "step": 2978
+    },
+    {
+      "epoch": 3.3890784982935154,
+      "grad_norm": 1.0216169357299805,
+      "learning_rate": 0.00032309442548350396,
+      "loss": 1.9361,
+      "step": 2979
+    },
+    {
+      "epoch": 3.390216154721274,
+      "grad_norm": 0.9502822756767273,
+      "learning_rate": 0.0003228668941979522,
+      "loss": 0.9628,
+      "step": 2980
+    },
+    {
+      "epoch": 3.391353811149033,
+      "grad_norm": 0.6996874213218689,
+      "learning_rate": 0.0003226393629124005,
+      "loss": 1.223,
+      "step": 2981
+    },
+    {
+      "epoch": 3.3924914675767917,
+      "grad_norm": 0.9376258254051208,
+      "learning_rate": 0.0003224118316268487,
+      "loss": 1.4156,
+      "step": 2982
+    },
+    {
+      "epoch": 3.3936291240045504,
+      "grad_norm": 1.000174880027771,
+      "learning_rate": 0.0003221843003412969,
+      "loss": 2.2908,
+      "step": 2983
+    },
+    {
+      "epoch": 3.3947667804323096,
+      "grad_norm": 1.1017203330993652,
+      "learning_rate": 0.00032195676905574517,
+      "loss": 2.837,
+      "step": 2984
+    },
+    {
+      "epoch": 3.3959044368600684,
+      "grad_norm": 0.7534080147743225,
+      "learning_rate": 0.00032172923777019344,
+      "loss": 1.6282,
+      "step": 2985
+    },
+    {
+      "epoch": 3.397042093287827,
+      "grad_norm": 0.8684095144271851,
+      "learning_rate": 0.00032150170648464165,
+      "loss": 2.2301,
+      "step": 2986
+    },
+    {
+      "epoch": 3.398179749715586,
+      "grad_norm": 1.0283501148223877,
+      "learning_rate": 0.00032127417519908986,
+      "loss": 2.0298,
+      "step": 2987
+    },
+    {
+      "epoch": 3.3993174061433447,
+      "grad_norm": 0.8715803623199463,
+      "learning_rate": 0.0003210466439135381,
+      "loss": 1.5656,
+      "step": 2988
+    },
+    {
+      "epoch": 3.4004550625711034,
+      "grad_norm": 1.0342568159103394,
+      "learning_rate": 0.00032081911262798633,
+      "loss": 1.8496,
+      "step": 2989
+    },
+    {
+      "epoch": 3.401592718998862,
+      "grad_norm": 1.0820322036743164,
+      "learning_rate": 0.0003205915813424346,
+      "loss": 1.4579,
+      "step": 2990
+    },
+    {
+      "epoch": 3.4027303754266214,
+      "grad_norm": 0.9718009233474731,
+      "learning_rate": 0.00032036405005688286,
+      "loss": 1.8414,
+      "step": 2991
+    },
+    {
+      "epoch": 3.40386803185438,
+      "grad_norm": 0.6825365424156189,
+      "learning_rate": 0.00032013651877133107,
+      "loss": 1.4311,
+      "step": 2992
+    },
+    {
+      "epoch": 3.405005688282139,
+      "grad_norm": 0.7811647057533264,
+      "learning_rate": 0.0003199089874857793,
+      "loss": 0.7493,
+      "step": 2993
+    },
+    {
+      "epoch": 3.4061433447098977,
+      "grad_norm": 0.8294082283973694,
+      "learning_rate": 0.00031968145620022754,
+      "loss": 1.5319,
+      "step": 2994
+    },
+    {
+      "epoch": 3.4072810011376564,
+      "grad_norm": 1.365054965019226,
+      "learning_rate": 0.0003194539249146758,
+      "loss": 3.0346,
+      "step": 2995
+    },
+    {
+      "epoch": 3.408418657565415,
+      "grad_norm": 1.0830689668655396,
+      "learning_rate": 0.000319226393629124,
+      "loss": 2.3795,
+      "step": 2996
+    },
+    {
+      "epoch": 3.409556313993174,
+      "grad_norm": 1.158379077911377,
+      "learning_rate": 0.0003189988623435722,
+      "loss": 1.6544,
+      "step": 2997
+    },
+    {
+      "epoch": 3.4106939704209327,
+      "grad_norm": 0.8690700531005859,
+      "learning_rate": 0.0003187713310580205,
+      "loss": 1.473,
+      "step": 2998
+    },
+    {
+      "epoch": 3.4118316268486915,
+      "grad_norm": 0.7095939517021179,
+      "learning_rate": 0.0003185437997724687,
+      "loss": 1.4625,
+      "step": 2999
+    },
+    {
+      "epoch": 3.4129692832764507,
+      "grad_norm": 1.2548210620880127,
+      "learning_rate": 0.00031831626848691696,
+      "loss": 1.8051,
+      "step": 3000
+    },
+    {
+      "epoch": 3.4141069397042094,
+      "grad_norm": 1.4674617052078247,
+      "learning_rate": 0.00031808873720136523,
+      "loss": 1.9178,
+      "step": 3001
+    },
+    {
+      "epoch": 3.415244596131968,
+      "grad_norm": 0.9363518953323364,
+      "learning_rate": 0.00031786120591581344,
+      "loss": 2.3612,
+      "step": 3002
+    },
+    {
+      "epoch": 3.416382252559727,
+      "grad_norm": 1.0488542318344116,
+      "learning_rate": 0.00031763367463026165,
+      "loss": 2.0528,
+      "step": 3003
+    },
+    {
+      "epoch": 3.4175199089874857,
+      "grad_norm": 1.0611392259597778,
+      "learning_rate": 0.0003174061433447099,
+      "loss": 1.2209,
+      "step": 3004
+    },
+    {
+      "epoch": 3.4186575654152445,
+      "grad_norm": 0.7108457088470459,
+      "learning_rate": 0.0003171786120591582,
+      "loss": 2.2781,
+      "step": 3005
+    },
+    {
+      "epoch": 3.419795221843003,
+      "grad_norm": 0.7717037200927734,
+      "learning_rate": 0.0003169510807736064,
+      "loss": 1.8069,
+      "step": 3006
+    },
+    {
+      "epoch": 3.4209328782707624,
+      "grad_norm": 1.1115652322769165,
+      "learning_rate": 0.0003167235494880546,
+      "loss": 2.1257,
+      "step": 3007
+    },
+    {
+      "epoch": 3.422070534698521,
+      "grad_norm": 1.1573059558868408,
+      "learning_rate": 0.00031649601820250286,
+      "loss": 1.5691,
+      "step": 3008
+    },
+    {
+      "epoch": 3.42320819112628,
+      "grad_norm": 0.7534975409507751,
+      "learning_rate": 0.00031626848691695107,
+      "loss": 1.6702,
+      "step": 3009
+    },
+    {
+      "epoch": 3.4243458475540387,
+      "grad_norm": 0.8538755178451538,
+      "learning_rate": 0.00031604095563139933,
+      "loss": 1.8829,
+      "step": 3010
+    },
+    {
+      "epoch": 3.4254835039817975,
+      "grad_norm": 1.041550636291504,
+      "learning_rate": 0.00031581342434584754,
+      "loss": 1.8357,
+      "step": 3011
+    },
+    {
+      "epoch": 3.426621160409556,
+      "grad_norm": 0.9445088505744934,
+      "learning_rate": 0.0003155858930602958,
+      "loss": 2.0357,
+      "step": 3012
+    },
+    {
+      "epoch": 3.427758816837315,
+      "grad_norm": 0.8078857064247131,
+      "learning_rate": 0.000315358361774744,
+      "loss": 1.6423,
+      "step": 3013
+    },
+    {
+      "epoch": 3.428896473265074,
+      "grad_norm": 1.6331647634506226,
+      "learning_rate": 0.0003151308304891923,
+      "loss": 2.6264,
+      "step": 3014
+    },
+    {
+      "epoch": 3.430034129692833,
+      "grad_norm": 1.0118826627731323,
+      "learning_rate": 0.00031490329920364055,
+      "loss": 2.5419,
+      "step": 3015
+    },
+    {
+      "epoch": 3.4311717861205917,
+      "grad_norm": 1.1630852222442627,
+      "learning_rate": 0.0003146757679180887,
+      "loss": 2.345,
+      "step": 3016
+    },
+    {
+      "epoch": 3.4323094425483505,
+      "grad_norm": 1.082171082496643,
+      "learning_rate": 0.00031444823663253697,
+      "loss": 1.6889,
+      "step": 3017
+    },
+    {
+      "epoch": 3.4334470989761092,
+      "grad_norm": 0.9755992889404297,
+      "learning_rate": 0.00031422070534698523,
+      "loss": 1.599,
+      "step": 3018
+    },
+    {
+      "epoch": 3.434584755403868,
+      "grad_norm": 1.2539499998092651,
+      "learning_rate": 0.00031399317406143344,
+      "loss": 3.2968,
+      "step": 3019
+    },
+    {
+      "epoch": 3.4357224118316267,
+      "grad_norm": 0.8818562030792236,
+      "learning_rate": 0.0003137656427758817,
+      "loss": 2.6973,
+      "step": 3020
+    },
+    {
+      "epoch": 3.4368600682593855,
+      "grad_norm": 0.900875449180603,
+      "learning_rate": 0.0003135381114903299,
+      "loss": 1.6828,
+      "step": 3021
+    },
+    {
+      "epoch": 3.4379977246871443,
+      "grad_norm": 0.7886406779289246,
+      "learning_rate": 0.0003133105802047782,
+      "loss": 2.1497,
+      "step": 3022
+    },
+    {
+      "epoch": 3.4391353811149035,
+      "grad_norm": 1.0653672218322754,
+      "learning_rate": 0.0003130830489192264,
+      "loss": 1.2112,
+      "step": 3023
+    },
+    {
+      "epoch": 3.4402730375426622,
+      "grad_norm": 0.8921288251876831,
+      "learning_rate": 0.00031285551763367465,
+      "loss": 1.5596,
+      "step": 3024
+    },
+    {
+      "epoch": 3.441410693970421,
+      "grad_norm": 1.0506072044372559,
+      "learning_rate": 0.0003126279863481229,
+      "loss": 1.9131,
+      "step": 3025
+    },
+    {
+      "epoch": 3.4425483503981797,
+      "grad_norm": 0.9992786645889282,
+      "learning_rate": 0.00031240045506257107,
+      "loss": 2.2033,
+      "step": 3026
+    },
+    {
+      "epoch": 3.4436860068259385,
+      "grad_norm": 1.0742168426513672,
+      "learning_rate": 0.00031217292377701934,
+      "loss": 1.793,
+      "step": 3027
+    },
+    {
+      "epoch": 3.4448236632536973,
+      "grad_norm": 1.2853068113327026,
+      "learning_rate": 0.0003119453924914676,
+      "loss": 2.5013,
+      "step": 3028
+    },
+    {
+      "epoch": 3.445961319681456,
+      "grad_norm": 1.1094430685043335,
+      "learning_rate": 0.0003117178612059158,
+      "loss": 2.518,
+      "step": 3029
+    },
+    {
+      "epoch": 3.4470989761092152,
+      "grad_norm": 1.2212568521499634,
+      "learning_rate": 0.0003114903299203641,
+      "loss": 1.1286,
+      "step": 3030
+    },
+    {
+      "epoch": 3.448236632536974,
+      "grad_norm": 0.8297713398933411,
+      "learning_rate": 0.0003112627986348123,
+      "loss": 2.0985,
+      "step": 3031
+    },
+    {
+      "epoch": 3.4493742889647327,
+      "grad_norm": 1.1644102334976196,
+      "learning_rate": 0.0003110352673492605,
+      "loss": 2.4079,
+      "step": 3032
+    },
+    {
+      "epoch": 3.4505119453924915,
+      "grad_norm": 0.8777130842208862,
+      "learning_rate": 0.00031080773606370876,
+      "loss": 3.0359,
+      "step": 3033
+    },
+    {
+      "epoch": 3.4516496018202503,
+      "grad_norm": 1.0054441690444946,
+      "learning_rate": 0.000310580204778157,
+      "loss": 1.4005,
+      "step": 3034
+    },
+    {
+      "epoch": 3.452787258248009,
+      "grad_norm": 0.7273943424224854,
+      "learning_rate": 0.0003103526734926053,
+      "loss": 0.9642,
+      "step": 3035
+    },
+    {
+      "epoch": 3.453924914675768,
+      "grad_norm": 0.9213597774505615,
+      "learning_rate": 0.00031012514220705344,
+      "loss": 1.603,
+      "step": 3036
+    },
+    {
+      "epoch": 3.4550625711035265,
+      "grad_norm": 0.8535187840461731,
+      "learning_rate": 0.0003098976109215017,
+      "loss": 1.8995,
+      "step": 3037
+    },
+    {
+      "epoch": 3.4562002275312853,
+      "grad_norm": 0.6558015942573547,
+      "learning_rate": 0.00030967007963594997,
+      "loss": 1.7263,
+      "step": 3038
+    },
+    {
+      "epoch": 3.4573378839590445,
+      "grad_norm": 0.8745822310447693,
+      "learning_rate": 0.0003094425483503982,
+      "loss": 2.1463,
+      "step": 3039
+    },
+    {
+      "epoch": 3.4584755403868033,
+      "grad_norm": 0.829898476600647,
+      "learning_rate": 0.00030921501706484644,
+      "loss": 1.1866,
+      "step": 3040
+    },
+    {
+      "epoch": 3.459613196814562,
+      "grad_norm": 0.7422157526016235,
+      "learning_rate": 0.00030898748577929465,
+      "loss": 1.6347,
+      "step": 3041
+    },
+    {
+      "epoch": 3.460750853242321,
+      "grad_norm": 1.017354965209961,
+      "learning_rate": 0.00030875995449374286,
+      "loss": 2.092,
+      "step": 3042
+    },
+    {
+      "epoch": 3.4618885096700796,
+      "grad_norm": 0.9957495331764221,
+      "learning_rate": 0.00030853242320819113,
+      "loss": 1.6618,
+      "step": 3043
+    },
+    {
+      "epoch": 3.4630261660978383,
+      "grad_norm": 0.9194496273994446,
+      "learning_rate": 0.0003083048919226394,
+      "loss": 1.3234,
+      "step": 3044
+    },
+    {
+      "epoch": 3.464163822525597,
+      "grad_norm": 0.8070917129516602,
+      "learning_rate": 0.00030807736063708766,
+      "loss": 1.0518,
+      "step": 3045
+    },
+    {
+      "epoch": 3.4653014789533563,
+      "grad_norm": 0.9527956247329712,
+      "learning_rate": 0.0003078498293515358,
+      "loss": 1.8434,
+      "step": 3046
+    },
+    {
+      "epoch": 3.466439135381115,
+      "grad_norm": 0.9919708967208862,
+      "learning_rate": 0.0003076222980659841,
+      "loss": 1.8574,
+      "step": 3047
+    },
+    {
+      "epoch": 3.467576791808874,
+      "grad_norm": 0.7566163539886475,
+      "learning_rate": 0.00030739476678043234,
+      "loss": 2.2815,
+      "step": 3048
+    },
+    {
+      "epoch": 3.4687144482366326,
+      "grad_norm": 0.9245063066482544,
+      "learning_rate": 0.00030716723549488055,
+      "loss": 1.8061,
+      "step": 3049
+    },
+    {
+      "epoch": 3.4698521046643913,
+      "grad_norm": 0.5763617753982544,
+      "learning_rate": 0.00030693970420932876,
+      "loss": 0.8955,
+      "step": 3050
+    },
+    {
+      "epoch": 3.47098976109215,
+      "grad_norm": 0.6619337797164917,
+      "learning_rate": 0.000306712172923777,
+      "loss": 0.8753,
+      "step": 3051
+    },
+    {
+      "epoch": 3.472127417519909,
+      "grad_norm": 0.9321338534355164,
+      "learning_rate": 0.00030648464163822523,
+      "loss": 1.3895,
+      "step": 3052
+    },
+    {
+      "epoch": 3.473265073947668,
+      "grad_norm": 1.0468645095825195,
+      "learning_rate": 0.0003062571103526735,
+      "loss": 1.9941,
+      "step": 3053
+    },
+    {
+      "epoch": 3.474402730375427,
+      "grad_norm": 1.2254489660263062,
+      "learning_rate": 0.00030602957906712176,
+      "loss": 2.2438,
+      "step": 3054
+    },
+    {
+      "epoch": 3.4755403868031856,
+      "grad_norm": 0.965639591217041,
+      "learning_rate": 0.00030580204778156997,
+      "loss": 1.6008,
+      "step": 3055
+    },
+    {
+      "epoch": 3.4766780432309443,
+      "grad_norm": 0.7972604632377625,
+      "learning_rate": 0.0003055745164960182,
+      "loss": 1.2332,
+      "step": 3056
+    },
+    {
+      "epoch": 3.477815699658703,
+      "grad_norm": 0.8752893209457397,
+      "learning_rate": 0.00030534698521046645,
+      "loss": 1.3047,
+      "step": 3057
+    },
+    {
+      "epoch": 3.478953356086462,
+      "grad_norm": 0.7717936038970947,
+      "learning_rate": 0.0003051194539249147,
+      "loss": 1.4136,
+      "step": 3058
+    },
+    {
+      "epoch": 3.4800910125142206,
+      "grad_norm": 0.8218960165977478,
+      "learning_rate": 0.0003048919226393629,
+      "loss": 1.5268,
+      "step": 3059
+    },
+    {
+      "epoch": 3.4812286689419794,
+      "grad_norm": 0.8942006230354309,
+      "learning_rate": 0.00030466439135381113,
+      "loss": 1.9143,
+      "step": 3060
+    },
+    {
+      "epoch": 3.482366325369738,
+      "grad_norm": 0.8847683668136597,
+      "learning_rate": 0.0003044368600682594,
+      "loss": 2.8478,
+      "step": 3061
+    },
+    {
+      "epoch": 3.4835039817974973,
+      "grad_norm": 0.8269094228744507,
+      "learning_rate": 0.0003042093287827076,
+      "loss": 2.0533,
+      "step": 3062
+    },
+    {
+      "epoch": 3.484641638225256,
+      "grad_norm": 0.7389246225357056,
+      "learning_rate": 0.00030398179749715587,
+      "loss": 1.5518,
+      "step": 3063
+    },
+    {
+      "epoch": 3.485779294653015,
+      "grad_norm": 0.7460439205169678,
+      "learning_rate": 0.00030375426621160413,
+      "loss": 0.9565,
+      "step": 3064
+    },
+    {
+      "epoch": 3.4869169510807736,
+      "grad_norm": 0.8218055367469788,
+      "learning_rate": 0.00030352673492605234,
+      "loss": 1.8419,
+      "step": 3065
+    },
+    {
+      "epoch": 3.4880546075085324,
+      "grad_norm": 1.2303011417388916,
+      "learning_rate": 0.00030329920364050055,
+      "loss": 2.0404,
+      "step": 3066
+    },
+    {
+      "epoch": 3.489192263936291,
+      "grad_norm": 0.7625948190689087,
+      "learning_rate": 0.0003030716723549488,
+      "loss": 1.502,
+      "step": 3067
+    },
+    {
+      "epoch": 3.49032992036405,
+      "grad_norm": 0.9459888339042664,
+      "learning_rate": 0.0003028441410693971,
+      "loss": 1.4141,
+      "step": 3068
+    },
+    {
+      "epoch": 3.491467576791809,
+      "grad_norm": 0.9511308670043945,
+      "learning_rate": 0.0003026166097838453,
+      "loss": 2.9199,
+      "step": 3069
+    },
+    {
+      "epoch": 3.492605233219568,
+      "grad_norm": 0.9939665198326111,
+      "learning_rate": 0.0003023890784982935,
+      "loss": 1.7376,
+      "step": 3070
+    },
+    {
+      "epoch": 3.4937428896473266,
+      "grad_norm": 1.1104254722595215,
+      "learning_rate": 0.00030216154721274176,
+      "loss": 2.1018,
+      "step": 3071
+    },
+    {
+      "epoch": 3.4948805460750854,
+      "grad_norm": 0.8199077248573303,
+      "learning_rate": 0.00030193401592718997,
+      "loss": 2.0834,
+      "step": 3072
+    },
+    {
+      "epoch": 3.496018202502844,
+      "grad_norm": 0.7726981043815613,
+      "learning_rate": 0.00030170648464163824,
+      "loss": 1.2873,
+      "step": 3073
+    },
+    {
+      "epoch": 3.497155858930603,
+      "grad_norm": 1.0992076396942139,
+      "learning_rate": 0.0003014789533560865,
+      "loss": 2.3819,
+      "step": 3074
+    },
+    {
+      "epoch": 3.4982935153583616,
+      "grad_norm": 0.7880586385726929,
+      "learning_rate": 0.0003012514220705347,
+      "loss": 2.0231,
+      "step": 3075
+    },
+    {
+      "epoch": 3.4994311717861204,
+      "grad_norm": 0.5555304884910583,
+      "learning_rate": 0.0003010238907849829,
+      "loss": 1.0044,
+      "step": 3076
+    },
+    {
+      "epoch": 3.500568828213879,
+      "grad_norm": 1.2383909225463867,
+      "learning_rate": 0.0003007963594994312,
+      "loss": 1.2884,
+      "step": 3077
+    },
+    {
+      "epoch": 3.5017064846416384,
+      "grad_norm": 1.214747667312622,
+      "learning_rate": 0.00030056882821387945,
+      "loss": 1.4196,
+      "step": 3078
+    },
+    {
+      "epoch": 3.502844141069397,
+      "grad_norm": 0.7822383046150208,
+      "learning_rate": 0.00030034129692832766,
+      "loss": 1.3068,
+      "step": 3079
+    },
+    {
+      "epoch": 3.503981797497156,
+      "grad_norm": 1.200726866722107,
+      "learning_rate": 0.00030011376564277587,
+      "loss": 2.0791,
+      "step": 3080
+    },
+    {
+      "epoch": 3.5051194539249146,
+      "grad_norm": 0.7341413497924805,
+      "learning_rate": 0.00029988623435722413,
+      "loss": 0.9889,
+      "step": 3081
+    },
+    {
+      "epoch": 3.5062571103526734,
+      "grad_norm": 1.4545503854751587,
+      "learning_rate": 0.00029965870307167234,
+      "loss": 2.6975,
+      "step": 3082
+    },
+    {
+      "epoch": 3.507394766780432,
+      "grad_norm": 0.8442320227622986,
+      "learning_rate": 0.0002994311717861206,
+      "loss": 1.6271,
+      "step": 3083
+    },
+    {
+      "epoch": 3.508532423208191,
+      "grad_norm": 0.7989565134048462,
+      "learning_rate": 0.0002992036405005688,
+      "loss": 0.9025,
+      "step": 3084
+    },
+    {
+      "epoch": 3.50967007963595,
+      "grad_norm": 0.9094103574752808,
+      "learning_rate": 0.0002989761092150171,
+      "loss": 1.4245,
+      "step": 3085
+    },
+    {
+      "epoch": 3.510807736063709,
+      "grad_norm": 1.1836727857589722,
+      "learning_rate": 0.0002987485779294653,
+      "loss": 2.3158,
+      "step": 3086
+    },
+    {
+      "epoch": 3.5119453924914676,
+      "grad_norm": 0.9379547238349915,
+      "learning_rate": 0.00029852104664391355,
+      "loss": 1.3348,
+      "step": 3087
+    },
+    {
+      "epoch": 3.5130830489192264,
+      "grad_norm": 1.4176782369613647,
+      "learning_rate": 0.0002982935153583618,
+      "loss": 2.056,
+      "step": 3088
+    },
+    {
+      "epoch": 3.514220705346985,
+      "grad_norm": 0.8050400018692017,
+      "learning_rate": 0.00029806598407281,
+      "loss": 1.7547,
+      "step": 3089
+    },
+    {
+      "epoch": 3.515358361774744,
+      "grad_norm": 0.6850399971008301,
+      "learning_rate": 0.00029783845278725824,
+      "loss": 0.8674,
+      "step": 3090
+    },
+    {
+      "epoch": 3.5164960182025027,
+      "grad_norm": 1.1448007822036743,
+      "learning_rate": 0.0002976109215017065,
+      "loss": 2.3783,
+      "step": 3091
+    },
+    {
+      "epoch": 3.517633674630262,
+      "grad_norm": 0.645057737827301,
+      "learning_rate": 0.0002973833902161547,
+      "loss": 0.9246,
+      "step": 3092
+    },
+    {
+      "epoch": 3.51877133105802,
+      "grad_norm": 1.0810428857803345,
+      "learning_rate": 0.000297155858930603,
+      "loss": 1.9262,
+      "step": 3093
+    },
+    {
+      "epoch": 3.5199089874857794,
+      "grad_norm": 0.7463341951370239,
+      "learning_rate": 0.0002969283276450512,
+      "loss": 1.8343,
+      "step": 3094
+    },
+    {
+      "epoch": 3.521046643913538,
+      "grad_norm": 1.1941145658493042,
+      "learning_rate": 0.00029670079635949945,
+      "loss": 3.0099,
+      "step": 3095
+    },
+    {
+      "epoch": 3.522184300341297,
+      "grad_norm": 1.1003155708312988,
+      "learning_rate": 0.00029647326507394766,
+      "loss": 1.9862,
+      "step": 3096
+    },
+    {
+      "epoch": 3.5233219567690557,
+      "grad_norm": 0.7779751420021057,
+      "learning_rate": 0.0002962457337883959,
+      "loss": 1.5242,
+      "step": 3097
+    },
+    {
+      "epoch": 3.5244596131968144,
+      "grad_norm": 0.7989513874053955,
+      "learning_rate": 0.0002960182025028442,
+      "loss": 1.2492,
+      "step": 3098
+    },
+    {
+      "epoch": 3.5255972696245736,
+      "grad_norm": 0.9354263544082642,
+      "learning_rate": 0.00029579067121729234,
+      "loss": 1.1991,
+      "step": 3099
+    },
+    {
+      "epoch": 3.526734926052332,
+      "grad_norm": 1.2096282243728638,
+      "learning_rate": 0.0002955631399317406,
+      "loss": 2.2768,
+      "step": 3100
+    },
+    {
+      "epoch": 3.527872582480091,
+      "grad_norm": 1.2972865104675293,
+      "learning_rate": 0.00029533560864618887,
+      "loss": 2.3998,
+      "step": 3101
+    },
+    {
+      "epoch": 3.52901023890785,
+      "grad_norm": 0.9327741861343384,
+      "learning_rate": 0.0002951080773606371,
+      "loss": 2.3759,
+      "step": 3102
+    },
+    {
+      "epoch": 3.5301478953356087,
+      "grad_norm": 1.7685364484786987,
+      "learning_rate": 0.00029488054607508535,
+      "loss": 2.9593,
+      "step": 3103
+    },
+    {
+      "epoch": 3.5312855517633674,
+      "grad_norm": 1.1170707941055298,
+      "learning_rate": 0.00029465301478953356,
+      "loss": 1.7477,
+      "step": 3104
+    },
+    {
+      "epoch": 3.532423208191126,
+      "grad_norm": 0.635349452495575,
+      "learning_rate": 0.0002944254835039818,
+      "loss": 0.4989,
+      "step": 3105
+    },
+    {
+      "epoch": 3.533560864618885,
+      "grad_norm": 1.165709376335144,
+      "learning_rate": 0.00029419795221843003,
+      "loss": 1.5922,
+      "step": 3106
+    },
+    {
+      "epoch": 3.5346985210466437,
+      "grad_norm": 0.8228427171707153,
+      "learning_rate": 0.0002939704209328783,
+      "loss": 1.6231,
+      "step": 3107
+    },
+    {
+      "epoch": 3.535836177474403,
+      "grad_norm": 0.7035273313522339,
+      "learning_rate": 0.00029374288964732656,
+      "loss": 1.2547,
+      "step": 3108
+    },
+    {
+      "epoch": 3.5369738339021617,
+      "grad_norm": 0.9411433935165405,
+      "learning_rate": 0.0002935153583617747,
+      "loss": 1.4946,
+      "step": 3109
+    },
+    {
+      "epoch": 3.5381114903299204,
+      "grad_norm": 0.8050945401191711,
+      "learning_rate": 0.000293287827076223,
+      "loss": 1.2655,
+      "step": 3110
+    },
+    {
+      "epoch": 3.539249146757679,
+      "grad_norm": 0.5025121569633484,
+      "learning_rate": 0.00029306029579067124,
+      "loss": 1.1542,
+      "step": 3111
+    },
+    {
+      "epoch": 3.540386803185438,
+      "grad_norm": 0.6222087740898132,
+      "learning_rate": 0.00029283276450511945,
+      "loss": 0.8863,
+      "step": 3112
+    },
+    {
+      "epoch": 3.5415244596131967,
+      "grad_norm": 1.4298913478851318,
+      "learning_rate": 0.0002926052332195677,
+      "loss": 2.4176,
+      "step": 3113
+    },
+    {
+      "epoch": 3.5426621160409555,
+      "grad_norm": 0.5780576467514038,
+      "learning_rate": 0.0002923777019340159,
+      "loss": 0.5984,
+      "step": 3114
+    },
+    {
+      "epoch": 3.5437997724687147,
+      "grad_norm": 0.6261698007583618,
+      "learning_rate": 0.0002921501706484642,
+      "loss": 0.7561,
+      "step": 3115
+    },
+    {
+      "epoch": 3.544937428896473,
+      "grad_norm": 0.7118838429450989,
+      "learning_rate": 0.0002919226393629124,
+      "loss": 0.7812,
+      "step": 3116
+    },
+    {
+      "epoch": 3.546075085324232,
+      "grad_norm": 0.8261246681213379,
+      "learning_rate": 0.00029169510807736066,
+      "loss": 1.0061,
+      "step": 3117
+    },
+    {
+      "epoch": 3.547212741751991,
+      "grad_norm": 1.191845178604126,
+      "learning_rate": 0.0002914675767918089,
+      "loss": 3.8113,
+      "step": 3118
+    },
+    {
+      "epoch": 3.5483503981797497,
+      "grad_norm": 1.1474705934524536,
+      "learning_rate": 0.0002912400455062571,
+      "loss": 2.7153,
+      "step": 3119
+    },
+    {
+      "epoch": 3.5494880546075085,
+      "grad_norm": 0.9370527267456055,
+      "learning_rate": 0.00029101251422070535,
+      "loss": 1.5725,
+      "step": 3120
+    },
+    {
+      "epoch": 3.5506257110352673,
+      "grad_norm": 0.9595047235488892,
+      "learning_rate": 0.0002907849829351536,
+      "loss": 1.8677,
+      "step": 3121
+    },
+    {
+      "epoch": 3.551763367463026,
+      "grad_norm": 1.2102956771850586,
+      "learning_rate": 0.0002905574516496018,
+      "loss": 1.2953,
+      "step": 3122
+    },
+    {
+      "epoch": 3.5529010238907848,
+      "grad_norm": 0.595016360282898,
+      "learning_rate": 0.00029032992036405003,
+      "loss": 0.6797,
+      "step": 3123
+    },
+    {
+      "epoch": 3.554038680318544,
+      "grad_norm": 0.8877463936805725,
+      "learning_rate": 0.0002901023890784983,
+      "loss": 1.4732,
+      "step": 3124
+    },
+    {
+      "epoch": 3.5551763367463027,
+      "grad_norm": 0.9120120406150818,
+      "learning_rate": 0.00028987485779294656,
+      "loss": 2.0458,
+      "step": 3125
+    },
+    {
+      "epoch": 3.5563139931740615,
+      "grad_norm": 0.845203697681427,
+      "learning_rate": 0.00028964732650739477,
+      "loss": 1.9619,
+      "step": 3126
+    },
+    {
+      "epoch": 3.5574516496018203,
+      "grad_norm": 1.032363772392273,
+      "learning_rate": 0.00028941979522184303,
+      "loss": 1.3608,
+      "step": 3127
+    },
+    {
+      "epoch": 3.558589306029579,
+      "grad_norm": 0.7119994759559631,
+      "learning_rate": 0.00028919226393629124,
+      "loss": 1.3793,
+      "step": 3128
+    },
+    {
+      "epoch": 3.5597269624573378,
+      "grad_norm": 2.150618553161621,
+      "learning_rate": 0.00028896473265073945,
+      "loss": 4.2388,
+      "step": 3129
+    },
+    {
+      "epoch": 3.5608646188850965,
+      "grad_norm": 0.7926281094551086,
+      "learning_rate": 0.0002887372013651877,
+      "loss": 1.7266,
+      "step": 3130
+    },
+    {
+      "epoch": 3.5620022753128557,
+      "grad_norm": 1.1008011102676392,
+      "learning_rate": 0.000288509670079636,
+      "loss": 2.8486,
+      "step": 3131
+    },
+    {
+      "epoch": 3.5631399317406145,
+      "grad_norm": 0.8936535716056824,
+      "learning_rate": 0.0002882821387940842,
+      "loss": 2.1958,
+      "step": 3132
+    },
+    {
+      "epoch": 3.5642775881683733,
+      "grad_norm": 0.7171263694763184,
+      "learning_rate": 0.0002880546075085324,
+      "loss": 0.8248,
+      "step": 3133
+    },
+    {
+      "epoch": 3.565415244596132,
+      "grad_norm": 1.07510244846344,
+      "learning_rate": 0.00028782707622298067,
+      "loss": 1.7625,
+      "step": 3134
+    },
+    {
+      "epoch": 3.5665529010238908,
+      "grad_norm": 0.6037983298301697,
+      "learning_rate": 0.0002875995449374289,
+      "loss": 0.8,
+      "step": 3135
+    },
+    {
+      "epoch": 3.5676905574516495,
+      "grad_norm": 0.737299919128418,
+      "learning_rate": 0.00028737201365187714,
+      "loss": 1.5646,
+      "step": 3136
+    },
+    {
+      "epoch": 3.5688282138794083,
+      "grad_norm": 0.8927704095840454,
+      "learning_rate": 0.0002871444823663254,
+      "loss": 1.8448,
+      "step": 3137
+    },
+    {
+      "epoch": 3.5699658703071675,
+      "grad_norm": 0.7696796655654907,
+      "learning_rate": 0.0002869169510807736,
+      "loss": 1.8304,
+      "step": 3138
+    },
+    {
+      "epoch": 3.571103526734926,
+      "grad_norm": 1.4837048053741455,
+      "learning_rate": 0.0002866894197952218,
+      "loss": 3.0218,
+      "step": 3139
+    },
+    {
+      "epoch": 3.572241183162685,
+      "grad_norm": 0.9347735643386841,
+      "learning_rate": 0.0002864618885096701,
+      "loss": 1.3186,
+      "step": 3140
+    },
+    {
+      "epoch": 3.573378839590444,
+      "grad_norm": 0.8641846179962158,
+      "learning_rate": 0.00028623435722411835,
+      "loss": 1.4623,
+      "step": 3141
+    },
+    {
+      "epoch": 3.5745164960182025,
+      "grad_norm": 1.1673868894577026,
+      "learning_rate": 0.00028600682593856656,
+      "loss": 2.0682,
+      "step": 3142
+    },
+    {
+      "epoch": 3.5756541524459613,
+      "grad_norm": 1.1850996017456055,
+      "learning_rate": 0.00028577929465301477,
+      "loss": 2.0922,
+      "step": 3143
+    },
+    {
+      "epoch": 3.57679180887372,
+      "grad_norm": 0.799368679523468,
+      "learning_rate": 0.00028555176336746304,
+      "loss": 2.0306,
+      "step": 3144
+    },
+    {
+      "epoch": 3.577929465301479,
+      "grad_norm": 0.7206624150276184,
+      "learning_rate": 0.00028532423208191125,
+      "loss": 1.3149,
+      "step": 3145
+    },
+    {
+      "epoch": 3.5790671217292376,
+      "grad_norm": 1.1149860620498657,
+      "learning_rate": 0.0002850967007963595,
+      "loss": 2.1369,
+      "step": 3146
+    },
+    {
+      "epoch": 3.580204778156997,
+      "grad_norm": 1.0406419038772583,
+      "learning_rate": 0.0002848691695108078,
+      "loss": 1.652,
+      "step": 3147
+    },
+    {
+      "epoch": 3.5813424345847555,
+      "grad_norm": 1.1601135730743408,
+      "learning_rate": 0.000284641638225256,
+      "loss": 1.6278,
+      "step": 3148
+    },
+    {
+      "epoch": 3.5824800910125143,
+      "grad_norm": 1.0218907594680786,
+      "learning_rate": 0.0002844141069397042,
+      "loss": 1.2107,
+      "step": 3149
+    },
+    {
+      "epoch": 3.583617747440273,
+      "grad_norm": 1.2365930080413818,
+      "learning_rate": 0.00028418657565415246,
+      "loss": 2.1005,
+      "step": 3150
+    },
+    {
+      "epoch": 3.584755403868032,
+      "grad_norm": 0.7121724486351013,
+      "learning_rate": 0.0002839590443686007,
+      "loss": 1.5354,
+      "step": 3151
+    },
+    {
+      "epoch": 3.5858930602957906,
+      "grad_norm": 0.9739391207695007,
+      "learning_rate": 0.0002837315130830489,
+      "loss": 2.6124,
+      "step": 3152
+    },
+    {
+      "epoch": 3.5870307167235493,
+      "grad_norm": 1.4687827825546265,
+      "learning_rate": 0.00028350398179749714,
+      "loss": 2.4509,
+      "step": 3153
+    },
+    {
+      "epoch": 3.5881683731513085,
+      "grad_norm": 1.106471061706543,
+      "learning_rate": 0.0002832764505119454,
+      "loss": 1.6431,
+      "step": 3154
+    },
+    {
+      "epoch": 3.589306029579067,
+      "grad_norm": 1.566792607307434,
+      "learning_rate": 0.0002830489192263936,
+      "loss": 2.5289,
+      "step": 3155
+    },
+    {
+      "epoch": 3.590443686006826,
+      "grad_norm": 0.8209744095802307,
+      "learning_rate": 0.0002828213879408419,
+      "loss": 1.251,
+      "step": 3156
+    },
+    {
+      "epoch": 3.591581342434585,
+      "grad_norm": 1.5616912841796875,
+      "learning_rate": 0.0002825938566552901,
+      "loss": 2.958,
+      "step": 3157
+    },
+    {
+      "epoch": 3.5927189988623436,
+      "grad_norm": 0.8948552012443542,
+      "learning_rate": 0.00028236632536973835,
+      "loss": 1.6619,
+      "step": 3158
+    },
+    {
+      "epoch": 3.5938566552901023,
+      "grad_norm": 1.0657151937484741,
+      "learning_rate": 0.00028213879408418656,
+      "loss": 1.8547,
+      "step": 3159
+    },
+    {
+      "epoch": 3.594994311717861,
+      "grad_norm": 0.7531653046607971,
+      "learning_rate": 0.00028191126279863483,
+      "loss": 1.5972,
+      "step": 3160
+    },
+    {
+      "epoch": 3.59613196814562,
+      "grad_norm": 0.9963746070861816,
+      "learning_rate": 0.0002816837315130831,
+      "loss": 2.1391,
+      "step": 3161
+    },
+    {
+      "epoch": 3.5972696245733786,
+      "grad_norm": 1.1848362684249878,
+      "learning_rate": 0.00028145620022753125,
+      "loss": 1.8631,
+      "step": 3162
+    },
+    {
+      "epoch": 3.598407281001138,
+      "grad_norm": 1.2183339595794678,
+      "learning_rate": 0.0002812286689419795,
+      "loss": 2.2846,
+      "step": 3163
+    },
+    {
+      "epoch": 3.5995449374288966,
+      "grad_norm": 0.654116153717041,
+      "learning_rate": 0.0002810011376564278,
+      "loss": 1.3402,
+      "step": 3164
+    },
+    {
+      "epoch": 3.6006825938566553,
+      "grad_norm": 1.1731293201446533,
+      "learning_rate": 0.000280773606370876,
+      "loss": 2.3222,
+      "step": 3165
+    },
+    {
+      "epoch": 3.601820250284414,
+      "grad_norm": 0.6737352013587952,
+      "learning_rate": 0.00028054607508532425,
+      "loss": 1.3616,
+      "step": 3166
+    },
+    {
+      "epoch": 3.602957906712173,
+      "grad_norm": 0.6698241233825684,
+      "learning_rate": 0.00028031854379977246,
+      "loss": 1.0889,
+      "step": 3167
+    },
+    {
+      "epoch": 3.6040955631399316,
+      "grad_norm": 1.0666381120681763,
+      "learning_rate": 0.0002800910125142207,
+      "loss": 1.6506,
+      "step": 3168
+    },
+    {
+      "epoch": 3.6052332195676904,
+      "grad_norm": 0.7355189323425293,
+      "learning_rate": 0.00027986348122866893,
+      "loss": 1.2037,
+      "step": 3169
+    },
+    {
+      "epoch": 3.6063708759954496,
+      "grad_norm": 1.5804355144500732,
+      "learning_rate": 0.0002796359499431172,
+      "loss": 3.5022,
+      "step": 3170
+    },
+    {
+      "epoch": 3.6075085324232083,
+      "grad_norm": 0.602844774723053,
+      "learning_rate": 0.00027940841865756546,
+      "loss": 0.6136,
+      "step": 3171
+    },
+    {
+      "epoch": 3.608646188850967,
+      "grad_norm": 0.6072266101837158,
+      "learning_rate": 0.0002791808873720136,
+      "loss": 0.932,
+      "step": 3172
+    },
+    {
+      "epoch": 3.609783845278726,
+      "grad_norm": 0.9740130305290222,
+      "learning_rate": 0.0002789533560864619,
+      "loss": 1.3407,
+      "step": 3173
+    },
+    {
+      "epoch": 3.6109215017064846,
+      "grad_norm": 0.99945068359375,
+      "learning_rate": 0.00027872582480091015,
+      "loss": 3.0923,
+      "step": 3174
+    },
+    {
+      "epoch": 3.6120591581342434,
+      "grad_norm": 0.8721227645874023,
+      "learning_rate": 0.00027849829351535836,
+      "loss": 1.1302,
+      "step": 3175
+    },
+    {
+      "epoch": 3.613196814562002,
+      "grad_norm": 0.9758340716362,
+      "learning_rate": 0.0002782707622298066,
+      "loss": 1.2336,
+      "step": 3176
+    },
+    {
+      "epoch": 3.6143344709897613,
+      "grad_norm": 0.9425876140594482,
+      "learning_rate": 0.00027804323094425483,
+      "loss": 1.7407,
+      "step": 3177
+    },
+    {
+      "epoch": 3.6154721274175197,
+      "grad_norm": 1.0528191328048706,
+      "learning_rate": 0.0002778156996587031,
+      "loss": 1.7902,
+      "step": 3178
+    },
+    {
+      "epoch": 3.616609783845279,
+      "grad_norm": 0.9743272662162781,
+      "learning_rate": 0.0002775881683731513,
+      "loss": 1.5341,
+      "step": 3179
+    },
+    {
+      "epoch": 3.6177474402730376,
+      "grad_norm": 1.2372887134552002,
+      "learning_rate": 0.00027736063708759957,
+      "loss": 2.7279,
+      "step": 3180
+    },
+    {
+      "epoch": 3.6188850967007964,
+      "grad_norm": 1.1999444961547852,
+      "learning_rate": 0.00027713310580204783,
+      "loss": 2.0523,
+      "step": 3181
+    },
+    {
+      "epoch": 3.620022753128555,
+      "grad_norm": 0.903556227684021,
+      "learning_rate": 0.000276905574516496,
+      "loss": 1.3633,
+      "step": 3182
+    },
+    {
+      "epoch": 3.621160409556314,
+      "grad_norm": 0.7182222008705139,
+      "learning_rate": 0.00027667804323094425,
+      "loss": 1.5335,
+      "step": 3183
+    },
+    {
+      "epoch": 3.6222980659840727,
+      "grad_norm": 0.9772674441337585,
+      "learning_rate": 0.0002764505119453925,
+      "loss": 2.0408,
+      "step": 3184
+    },
+    {
+      "epoch": 3.6234357224118314,
+      "grad_norm": 0.6779370903968811,
+      "learning_rate": 0.0002762229806598407,
+      "loss": 0.963,
+      "step": 3185
+    },
+    {
+      "epoch": 3.6245733788395906,
+      "grad_norm": 1.3481802940368652,
+      "learning_rate": 0.00027599544937428894,
+      "loss": 1.9255,
+      "step": 3186
+    },
+    {
+      "epoch": 3.6257110352673494,
+      "grad_norm": 0.7263208031654358,
+      "learning_rate": 0.0002757679180887372,
+      "loss": 1.388,
+      "step": 3187
+    },
+    {
+      "epoch": 3.626848691695108,
+      "grad_norm": 0.8536869287490845,
+      "learning_rate": 0.00027554038680318546,
+      "loss": 1.5152,
+      "step": 3188
+    },
+    {
+      "epoch": 3.627986348122867,
+      "grad_norm": 0.7139433026313782,
+      "learning_rate": 0.0002753128555176337,
+      "loss": 1.4358,
+      "step": 3189
+    },
+    {
+      "epoch": 3.6291240045506257,
+      "grad_norm": 1.031976580619812,
+      "learning_rate": 0.00027508532423208194,
+      "loss": 2.6185,
+      "step": 3190
+    },
+    {
+      "epoch": 3.6302616609783844,
+      "grad_norm": 0.670224130153656,
+      "learning_rate": 0.00027485779294653015,
+      "loss": 1.2865,
+      "step": 3191
+    },
+    {
+      "epoch": 3.631399317406143,
+      "grad_norm": 0.8890179991722107,
+      "learning_rate": 0.00027463026166097836,
+      "loss": 1.5146,
+      "step": 3192
+    },
+    {
+      "epoch": 3.6325369738339024,
+      "grad_norm": 1.02622652053833,
+      "learning_rate": 0.0002744027303754266,
+      "loss": 1.8741,
+      "step": 3193
+    },
+    {
+      "epoch": 3.6336746302616607,
+      "grad_norm": 0.7721883654594421,
+      "learning_rate": 0.0002741751990898749,
+      "loss": 1.5905,
+      "step": 3194
+    },
+    {
+      "epoch": 3.63481228668942,
+      "grad_norm": 1.2086652517318726,
+      "learning_rate": 0.0002739476678043231,
+      "loss": 2.4846,
+      "step": 3195
+    },
+    {
+      "epoch": 3.6359499431171787,
+      "grad_norm": 1.0432493686676025,
+      "learning_rate": 0.0002737201365187713,
+      "loss": 1.6268,
+      "step": 3196
+    },
+    {
+      "epoch": 3.6370875995449374,
+      "grad_norm": 0.746393084526062,
+      "learning_rate": 0.00027349260523321957,
+      "loss": 1.5132,
+      "step": 3197
+    },
+    {
+      "epoch": 3.638225255972696,
+      "grad_norm": 0.9152166843414307,
+      "learning_rate": 0.00027326507394766783,
+      "loss": 2.676,
+      "step": 3198
+    },
+    {
+      "epoch": 3.639362912400455,
+      "grad_norm": 0.8825878500938416,
+      "learning_rate": 0.00027303754266211604,
+      "loss": 1.5822,
+      "step": 3199
+    },
+    {
+      "epoch": 3.640500568828214,
+      "grad_norm": 1.0070838928222656,
+      "learning_rate": 0.0002728100113765643,
+      "loss": 2.4247,
+      "step": 3200
+    },
+    {
+      "epoch": 3.6416382252559725,
+      "grad_norm": 0.7467294931411743,
+      "learning_rate": 0.0002725824800910125,
+      "loss": 1.0876,
+      "step": 3201
+    },
+    {
+      "epoch": 3.6427758816837317,
+      "grad_norm": 0.6660275459289551,
+      "learning_rate": 0.00027235494880546073,
+      "loss": 1.7321,
+      "step": 3202
+    },
+    {
+      "epoch": 3.6439135381114904,
+      "grad_norm": 0.9477649927139282,
+      "learning_rate": 0.000272127417519909,
+      "loss": 2.3247,
+      "step": 3203
+    },
+    {
+      "epoch": 3.645051194539249,
+      "grad_norm": 0.8924974203109741,
+      "learning_rate": 0.00027189988623435726,
+      "loss": 2.3003,
+      "step": 3204
+    },
+    {
+      "epoch": 3.646188850967008,
+      "grad_norm": 1.107689619064331,
+      "learning_rate": 0.00027167235494880547,
+      "loss": 2.079,
+      "step": 3205
+    },
+    {
+      "epoch": 3.6473265073947667,
+      "grad_norm": 0.7451488971710205,
+      "learning_rate": 0.0002714448236632537,
+      "loss": 1.4819,
+      "step": 3206
+    },
+    {
+      "epoch": 3.6484641638225255,
+      "grad_norm": 0.881750226020813,
+      "learning_rate": 0.00027121729237770194,
+      "loss": 1.4403,
+      "step": 3207
+    },
+    {
+      "epoch": 3.6496018202502842,
+      "grad_norm": 1.023970365524292,
+      "learning_rate": 0.0002709897610921502,
+      "loss": 1.5968,
+      "step": 3208
+    },
+    {
+      "epoch": 3.6507394766780434,
+      "grad_norm": 0.9700276255607605,
+      "learning_rate": 0.0002707622298065984,
+      "loss": 2.5977,
+      "step": 3209
+    },
+    {
+      "epoch": 3.651877133105802,
+      "grad_norm": 1.010783076286316,
+      "learning_rate": 0.0002705346985210467,
+      "loss": 2.2983,
+      "step": 3210
+    },
+    {
+      "epoch": 3.653014789533561,
+      "grad_norm": 1.0232635736465454,
+      "learning_rate": 0.0002703071672354949,
+      "loss": 1.461,
+      "step": 3211
+    },
+    {
+      "epoch": 3.6541524459613197,
+      "grad_norm": 0.6770385503768921,
+      "learning_rate": 0.0002700796359499431,
+      "loss": 0.8051,
+      "step": 3212
+    },
+    {
+      "epoch": 3.6552901023890785,
+      "grad_norm": 0.5668506026268005,
+      "learning_rate": 0.00026985210466439136,
+      "loss": 1.0955,
+      "step": 3213
+    },
+    {
+      "epoch": 3.6564277588168372,
+      "grad_norm": 1.4066870212554932,
+      "learning_rate": 0.0002696245733788396,
+      "loss": 3.6724,
+      "step": 3214
+    },
+    {
+      "epoch": 3.657565415244596,
+      "grad_norm": 1.153571605682373,
+      "learning_rate": 0.00026939704209328784,
+      "loss": 1.989,
+      "step": 3215
+    },
+    {
+      "epoch": 3.658703071672355,
+      "grad_norm": 0.728625476360321,
+      "learning_rate": 0.00026916951080773605,
+      "loss": 0.9875,
+      "step": 3216
+    },
+    {
+      "epoch": 3.6598407281001135,
+      "grad_norm": 0.9735997915267944,
+      "learning_rate": 0.0002689419795221843,
+      "loss": 2.0278,
+      "step": 3217
+    },
+    {
+      "epoch": 3.6609783845278727,
+      "grad_norm": 0.7140381932258606,
+      "learning_rate": 0.0002687144482366326,
+      "loss": 1.2533,
+      "step": 3218
+    },
+    {
+      "epoch": 3.6621160409556315,
+      "grad_norm": 0.6878055930137634,
+      "learning_rate": 0.0002684869169510808,
+      "loss": 1.1472,
+      "step": 3219
+    },
+    {
+      "epoch": 3.6632536973833902,
+      "grad_norm": 0.9914765954017639,
+      "learning_rate": 0.00026825938566552905,
+      "loss": 2.0081,
+      "step": 3220
+    },
+    {
+      "epoch": 3.664391353811149,
+      "grad_norm": 1.0459471940994263,
+      "learning_rate": 0.00026803185437997726,
+      "loss": 2.5561,
+      "step": 3221
+    },
+    {
+      "epoch": 3.6655290102389078,
+      "grad_norm": 0.8058758974075317,
+      "learning_rate": 0.00026780432309442547,
+      "loss": 1.2284,
+      "step": 3222
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 1.2225570678710938,
+      "learning_rate": 0.00026757679180887373,
+      "loss": 1.9648,
+      "step": 3223
+    },
+    {
+      "epoch": 3.6678043230944253,
+      "grad_norm": 1.3064903020858765,
+      "learning_rate": 0.000267349260523322,
+      "loss": 2.1073,
+      "step": 3224
+    },
+    {
+      "epoch": 3.6689419795221845,
+      "grad_norm": 0.7548709511756897,
+      "learning_rate": 0.00026712172923777015,
+      "loss": 1.5539,
+      "step": 3225
+    },
+    {
+      "epoch": 3.6700796359499432,
+      "grad_norm": 1.034284234046936,
+      "learning_rate": 0.0002668941979522184,
+      "loss": 2.125,
+      "step": 3226
+    },
+    {
+      "epoch": 3.671217292377702,
+      "grad_norm": 1.494303822517395,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 2.775,
+      "step": 3227
+    },
+    {
+      "epoch": 3.6723549488054608,
+      "grad_norm": 1.1169606447219849,
+      "learning_rate": 0.00026643913538111494,
+      "loss": 1.798,
+      "step": 3228
+    },
+    {
+      "epoch": 3.6734926052332195,
+      "grad_norm": 0.8246114253997803,
+      "learning_rate": 0.00026621160409556315,
+      "loss": 2.3232,
+      "step": 3229
+    },
+    {
+      "epoch": 3.6746302616609783,
+      "grad_norm": 1.181199312210083,
+      "learning_rate": 0.00026598407281001136,
+      "loss": 2.0188,
+      "step": 3230
+    },
+    {
+      "epoch": 3.675767918088737,
+      "grad_norm": 0.7052603960037231,
+      "learning_rate": 0.00026575654152445963,
+      "loss": 0.8771,
+      "step": 3231
+    },
+    {
+      "epoch": 3.6769055745164962,
+      "grad_norm": 0.5976938605308533,
+      "learning_rate": 0.00026552901023890784,
+      "loss": 0.9885,
+      "step": 3232
+    },
+    {
+      "epoch": 3.6780432309442546,
+      "grad_norm": 0.6874876022338867,
+      "learning_rate": 0.0002653014789533561,
+      "loss": 1.0909,
+      "step": 3233
+    },
+    {
+      "epoch": 3.6791808873720138,
+      "grad_norm": 0.9620742797851562,
+      "learning_rate": 0.00026507394766780437,
+      "loss": 2.0523,
+      "step": 3234
+    },
+    {
+      "epoch": 3.6803185437997725,
+      "grad_norm": 1.0894056558609009,
+      "learning_rate": 0.0002648464163822525,
+      "loss": 2.1868,
+      "step": 3235
+    },
+    {
+      "epoch": 3.6814562002275313,
+      "grad_norm": 0.8352320194244385,
+      "learning_rate": 0.0002646188850967008,
+      "loss": 1.5531,
+      "step": 3236
+    },
+    {
+      "epoch": 3.68259385665529,
+      "grad_norm": 0.7866209149360657,
+      "learning_rate": 0.00026439135381114905,
+      "loss": 1.2869,
+      "step": 3237
+    },
+    {
+      "epoch": 3.683731513083049,
+      "grad_norm": 0.8318602442741394,
+      "learning_rate": 0.00026416382252559726,
+      "loss": 1.4013,
+      "step": 3238
+    },
+    {
+      "epoch": 3.684869169510808,
+      "grad_norm": 0.6473860144615173,
+      "learning_rate": 0.0002639362912400455,
+      "loss": 1.0885,
+      "step": 3239
+    },
+    {
+      "epoch": 3.6860068259385663,
+      "grad_norm": 0.8395644426345825,
+      "learning_rate": 0.00026370875995449373,
+      "loss": 1.8404,
+      "step": 3240
+    },
+    {
+      "epoch": 3.6871444823663255,
+      "grad_norm": 1.3969433307647705,
+      "learning_rate": 0.000263481228668942,
+      "loss": 2.3692,
+      "step": 3241
+    },
+    {
+      "epoch": 3.6882821387940843,
+      "grad_norm": 0.7492949366569519,
+      "learning_rate": 0.0002632536973833902,
+      "loss": 1.4541,
+      "step": 3242
+    },
+    {
+      "epoch": 3.689419795221843,
+      "grad_norm": 0.7749677300453186,
+      "learning_rate": 0.00026302616609783847,
+      "loss": 1.3882,
+      "step": 3243
+    },
+    {
+      "epoch": 3.690557451649602,
+      "grad_norm": 0.507800817489624,
+      "learning_rate": 0.00026279863481228674,
+      "loss": 0.6856,
+      "step": 3244
+    },
+    {
+      "epoch": 3.6916951080773606,
+      "grad_norm": 0.7074844837188721,
+      "learning_rate": 0.0002625711035267349,
+      "loss": 1.4151,
+      "step": 3245
+    },
+    {
+      "epoch": 3.6928327645051193,
+      "grad_norm": 1.3755199909210205,
+      "learning_rate": 0.00026234357224118316,
+      "loss": 2.833,
+      "step": 3246
+    },
+    {
+      "epoch": 3.693970420932878,
+      "grad_norm": 0.7762560248374939,
+      "learning_rate": 0.0002621160409556314,
+      "loss": 1.0901,
+      "step": 3247
+    },
+    {
+      "epoch": 3.6951080773606373,
+      "grad_norm": 1.1140830516815186,
+      "learning_rate": 0.00026188850967007963,
+      "loss": 3.3667,
+      "step": 3248
+    },
+    {
+      "epoch": 3.696245733788396,
+      "grad_norm": 1.8301548957824707,
+      "learning_rate": 0.0002616609783845279,
+      "loss": 2.7103,
+      "step": 3249
+    },
+    {
+      "epoch": 3.697383390216155,
+      "grad_norm": 0.9659634828567505,
+      "learning_rate": 0.0002614334470989761,
+      "loss": 2.4665,
+      "step": 3250
+    },
+    {
+      "epoch": 3.6985210466439136,
+      "grad_norm": 1.0916699171066284,
+      "learning_rate": 0.00026120591581342437,
+      "loss": 1.5426,
+      "step": 3251
+    },
+    {
+      "epoch": 3.6996587030716723,
+      "grad_norm": 1.2825108766555786,
+      "learning_rate": 0.0002609783845278726,
+      "loss": 1.7881,
+      "step": 3252
+    },
+    {
+      "epoch": 3.700796359499431,
+      "grad_norm": 1.1956276893615723,
+      "learning_rate": 0.00026075085324232084,
+      "loss": 1.8214,
+      "step": 3253
+    },
+    {
+      "epoch": 3.70193401592719,
+      "grad_norm": 1.0599949359893799,
+      "learning_rate": 0.0002605233219567691,
+      "loss": 2.1982,
+      "step": 3254
+    },
+    {
+      "epoch": 3.703071672354949,
+      "grad_norm": 0.9849620461463928,
+      "learning_rate": 0.00026029579067121726,
+      "loss": 1.133,
+      "step": 3255
+    },
+    {
+      "epoch": 3.7042093287827074,
+      "grad_norm": 1.0015318393707275,
+      "learning_rate": 0.0002600682593856655,
+      "loss": 1.0853,
+      "step": 3256
+    },
+    {
+      "epoch": 3.7053469852104666,
+      "grad_norm": 1.1830706596374512,
+      "learning_rate": 0.0002598407281001138,
+      "loss": 2.3646,
+      "step": 3257
+    },
+    {
+      "epoch": 3.7064846416382253,
+      "grad_norm": 0.695976734161377,
+      "learning_rate": 0.000259613196814562,
+      "loss": 1.4282,
+      "step": 3258
+    },
+    {
+      "epoch": 3.707622298065984,
+      "grad_norm": 0.9679135680198669,
+      "learning_rate": 0.0002593856655290102,
+      "loss": 1.3835,
+      "step": 3259
+    },
+    {
+      "epoch": 3.708759954493743,
+      "grad_norm": 1.6105592250823975,
+      "learning_rate": 0.0002591581342434585,
+      "loss": 1.4964,
+      "step": 3260
+    },
+    {
+      "epoch": 3.7098976109215016,
+      "grad_norm": 1.0834482908248901,
+      "learning_rate": 0.00025893060295790674,
+      "loss": 1.6271,
+      "step": 3261
+    },
+    {
+      "epoch": 3.7110352673492604,
+      "grad_norm": 1.945980429649353,
+      "learning_rate": 0.00025870307167235495,
+      "loss": 1.3425,
+      "step": 3262
+    },
+    {
+      "epoch": 3.712172923777019,
+      "grad_norm": 0.9838660359382629,
+      "learning_rate": 0.0002584755403868032,
+      "loss": 1.7534,
+      "step": 3263
+    },
+    {
+      "epoch": 3.7133105802047783,
+      "grad_norm": 0.8386029005050659,
+      "learning_rate": 0.0002582480091012514,
+      "loss": 1.7482,
+      "step": 3264
+    },
+    {
+      "epoch": 3.714448236632537,
+      "grad_norm": 0.7002511620521545,
+      "learning_rate": 0.00025802047781569963,
+      "loss": 1.1395,
+      "step": 3265
+    },
+    {
+      "epoch": 3.715585893060296,
+      "grad_norm": 0.9224847555160522,
+      "learning_rate": 0.0002577929465301479,
+      "loss": 1.4517,
+      "step": 3266
+    },
+    {
+      "epoch": 3.7167235494880546,
+      "grad_norm": 1.0433372259140015,
+      "learning_rate": 0.00025756541524459616,
+      "loss": 2.0467,
+      "step": 3267
+    },
+    {
+      "epoch": 3.7178612059158134,
+      "grad_norm": 1.5209763050079346,
+      "learning_rate": 0.00025733788395904437,
+      "loss": 2.7371,
+      "step": 3268
+    },
+    {
+      "epoch": 3.718998862343572,
+      "grad_norm": 0.7279452085494995,
+      "learning_rate": 0.0002571103526734926,
+      "loss": 1.0362,
+      "step": 3269
+    },
+    {
+      "epoch": 3.720136518771331,
+      "grad_norm": 0.7574169635772705,
+      "learning_rate": 0.00025688282138794084,
+      "loss": 1.2043,
+      "step": 3270
+    },
+    {
+      "epoch": 3.72127417519909,
+      "grad_norm": 1.3098034858703613,
+      "learning_rate": 0.0002566552901023891,
+      "loss": 2.5448,
+      "step": 3271
+    },
+    {
+      "epoch": 3.722411831626849,
+      "grad_norm": 0.8687241673469543,
+      "learning_rate": 0.0002564277588168373,
+      "loss": 1.4687,
+      "step": 3272
+    },
+    {
+      "epoch": 3.7235494880546076,
+      "grad_norm": 0.9659656882286072,
+      "learning_rate": 0.0002562002275312856,
+      "loss": 2.4018,
+      "step": 3273
+    },
+    {
+      "epoch": 3.7246871444823664,
+      "grad_norm": 1.438889503479004,
+      "learning_rate": 0.0002559726962457338,
+      "loss": 3.2661,
+      "step": 3274
+    },
+    {
+      "epoch": 3.725824800910125,
+      "grad_norm": 1.4293714761734009,
+      "learning_rate": 0.000255745164960182,
+      "loss": 3.2473,
+      "step": 3275
+    },
+    {
+      "epoch": 3.726962457337884,
+      "grad_norm": 1.1246088743209839,
+      "learning_rate": 0.00025551763367463026,
+      "loss": 2.274,
+      "step": 3276
+    },
+    {
+      "epoch": 3.7281001137656427,
+      "grad_norm": 0.6158525943756104,
+      "learning_rate": 0.00025529010238907853,
+      "loss": 1.1515,
+      "step": 3277
+    },
+    {
+      "epoch": 3.729237770193402,
+      "grad_norm": 1.0718598365783691,
+      "learning_rate": 0.00025506257110352674,
+      "loss": 1.6976,
+      "step": 3278
+    },
+    {
+      "epoch": 3.73037542662116,
+      "grad_norm": 0.9811798334121704,
+      "learning_rate": 0.00025483503981797495,
+      "loss": 1.6137,
+      "step": 3279
+    },
+    {
+      "epoch": 3.7315130830489194,
+      "grad_norm": 1.0085113048553467,
+      "learning_rate": 0.0002546075085324232,
+      "loss": 2.4258,
+      "step": 3280
+    },
+    {
+      "epoch": 3.732650739476678,
+      "grad_norm": 1.0862224102020264,
+      "learning_rate": 0.0002543799772468715,
+      "loss": 1.6301,
+      "step": 3281
+    },
+    {
+      "epoch": 3.733788395904437,
+      "grad_norm": 1.057241678237915,
+      "learning_rate": 0.0002541524459613197,
+      "loss": 2.4249,
+      "step": 3282
+    },
+    {
+      "epoch": 3.7349260523321957,
+      "grad_norm": 0.9607272148132324,
+      "learning_rate": 0.00025392491467576795,
+      "loss": 2.4097,
+      "step": 3283
+    },
+    {
+      "epoch": 3.7360637087599544,
+      "grad_norm": 0.769008994102478,
+      "learning_rate": 0.00025369738339021616,
+      "loss": 2.0803,
+      "step": 3284
+    },
+    {
+      "epoch": 3.737201365187713,
+      "grad_norm": 1.317842721939087,
+      "learning_rate": 0.00025346985210466437,
+      "loss": 3.1254,
+      "step": 3285
+    },
+    {
+      "epoch": 3.738339021615472,
+      "grad_norm": 0.8753394484519958,
+      "learning_rate": 0.00025324232081911263,
+      "loss": 0.9374,
+      "step": 3286
+    },
+    {
+      "epoch": 3.739476678043231,
+      "grad_norm": 0.7274285554885864,
+      "learning_rate": 0.0002530147895335609,
+      "loss": 1.4164,
+      "step": 3287
+    },
+    {
+      "epoch": 3.74061433447099,
+      "grad_norm": 0.7005650997161865,
+      "learning_rate": 0.0002527872582480091,
+      "loss": 1.3413,
+      "step": 3288
+    },
+    {
+      "epoch": 3.7417519908987487,
+      "grad_norm": 1.0076242685317993,
+      "learning_rate": 0.0002525597269624573,
+      "loss": 2.4045,
+      "step": 3289
+    },
+    {
+      "epoch": 3.7428896473265074,
+      "grad_norm": 1.4078526496887207,
+      "learning_rate": 0.0002523321956769056,
+      "loss": 2.6869,
+      "step": 3290
+    },
+    {
+      "epoch": 3.744027303754266,
+      "grad_norm": 1.1299002170562744,
+      "learning_rate": 0.00025210466439135385,
+      "loss": 2.3044,
+      "step": 3291
+    },
+    {
+      "epoch": 3.745164960182025,
+      "grad_norm": 0.8027879595756531,
+      "learning_rate": 0.00025187713310580206,
+      "loss": 1.8643,
+      "step": 3292
+    },
+    {
+      "epoch": 3.7463026166097837,
+      "grad_norm": 1.083298921585083,
+      "learning_rate": 0.00025164960182025027,
+      "loss": 1.4689,
+      "step": 3293
+    },
+    {
+      "epoch": 3.747440273037543,
+      "grad_norm": 0.9221919178962708,
+      "learning_rate": 0.00025142207053469853,
+      "loss": 1.6231,
+      "step": 3294
+    },
+    {
+      "epoch": 3.748577929465301,
+      "grad_norm": 0.8968214392662048,
+      "learning_rate": 0.00025119453924914674,
+      "loss": 2.3925,
+      "step": 3295
+    },
+    {
+      "epoch": 3.7497155858930604,
+      "grad_norm": 0.8066171407699585,
+      "learning_rate": 0.000250967007963595,
+      "loss": 1.804,
+      "step": 3296
+    },
+    {
+      "epoch": 3.750853242320819,
+      "grad_norm": 1.2196367979049683,
+      "learning_rate": 0.00025073947667804327,
+      "loss": 3.5051,
+      "step": 3297
+    },
+    {
+      "epoch": 3.751990898748578,
+      "grad_norm": 1.0004627704620361,
+      "learning_rate": 0.0002505119453924914,
+      "loss": 1.7358,
+      "step": 3298
+    },
+    {
+      "epoch": 3.7531285551763367,
+      "grad_norm": 1.1003127098083496,
+      "learning_rate": 0.0002502844141069397,
+      "loss": 1.6266,
+      "step": 3299
+    },
+    {
+      "epoch": 3.7542662116040955,
+      "grad_norm": 0.8804592490196228,
+      "learning_rate": 0.00025005688282138795,
+      "loss": 1.6458,
+      "step": 3300
+    },
+    {
+      "epoch": 3.755403868031854,
+      "grad_norm": 0.6198902130126953,
+      "learning_rate": 0.0002498293515358362,
+      "loss": 1.0018,
+      "step": 3301
+    },
+    {
+      "epoch": 3.756541524459613,
+      "grad_norm": 0.9782140851020813,
+      "learning_rate": 0.0002496018202502844,
+      "loss": 2.0916,
+      "step": 3302
+    },
+    {
+      "epoch": 3.757679180887372,
+      "grad_norm": 1.2002933025360107,
+      "learning_rate": 0.00024937428896473264,
+      "loss": 1.7946,
+      "step": 3303
+    },
+    {
+      "epoch": 3.758816837315131,
+      "grad_norm": 0.9678136110305786,
+      "learning_rate": 0.0002491467576791809,
+      "loss": 1.476,
+      "step": 3304
+    },
+    {
+      "epoch": 3.7599544937428897,
+      "grad_norm": 0.7648789286613464,
+      "learning_rate": 0.0002489192263936291,
+      "loss": 1.3207,
+      "step": 3305
+    },
+    {
+      "epoch": 3.7610921501706485,
+      "grad_norm": 1.2675551176071167,
+      "learning_rate": 0.0002486916951080774,
+      "loss": 2.6937,
+      "step": 3306
+    },
+    {
+      "epoch": 3.7622298065984072,
+      "grad_norm": 0.7276213765144348,
+      "learning_rate": 0.0002484641638225256,
+      "loss": 0.9917,
+      "step": 3307
+    },
+    {
+      "epoch": 3.763367463026166,
+      "grad_norm": 1.3405811786651611,
+      "learning_rate": 0.00024823663253697385,
+      "loss": 1.6142,
+      "step": 3308
+    },
+    {
+      "epoch": 3.7645051194539247,
+      "grad_norm": 0.7004644274711609,
+      "learning_rate": 0.00024800910125142206,
+      "loss": 0.9751,
+      "step": 3309
+    },
+    {
+      "epoch": 3.765642775881684,
+      "grad_norm": 0.8266583681106567,
+      "learning_rate": 0.0002477815699658703,
+      "loss": 1.7234,
+      "step": 3310
+    },
+    {
+      "epoch": 3.7667804323094427,
+      "grad_norm": 0.827187180519104,
+      "learning_rate": 0.0002475540386803186,
+      "loss": 1.5155,
+      "step": 3311
+    },
+    {
+      "epoch": 3.7679180887372015,
+      "grad_norm": 0.8548292517662048,
+      "learning_rate": 0.0002473265073947668,
+      "loss": 1.7327,
+      "step": 3312
+    },
+    {
+      "epoch": 3.7690557451649602,
+      "grad_norm": 0.884003221988678,
+      "learning_rate": 0.000247098976109215,
+      "loss": 1.4591,
+      "step": 3313
+    },
+    {
+      "epoch": 3.770193401592719,
+      "grad_norm": 0.7317343950271606,
+      "learning_rate": 0.00024687144482366327,
+      "loss": 1.0964,
+      "step": 3314
+    },
+    {
+      "epoch": 3.7713310580204777,
+      "grad_norm": 0.8710649013519287,
+      "learning_rate": 0.0002466439135381115,
+      "loss": 1.5287,
+      "step": 3315
+    },
+    {
+      "epoch": 3.7724687144482365,
+      "grad_norm": 1.5312626361846924,
+      "learning_rate": 0.00024641638225255974,
+      "loss": 3.5776,
+      "step": 3316
+    },
+    {
+      "epoch": 3.7736063708759957,
+      "grad_norm": 0.6833708882331848,
+      "learning_rate": 0.00024618885096700795,
+      "loss": 1.2004,
+      "step": 3317
+    },
+    {
+      "epoch": 3.774744027303754,
+      "grad_norm": 1.1912914514541626,
+      "learning_rate": 0.0002459613196814562,
+      "loss": 2.4428,
+      "step": 3318
+    },
+    {
+      "epoch": 3.7758816837315132,
+      "grad_norm": 1.0153708457946777,
+      "learning_rate": 0.00024573378839590443,
+      "loss": 1.9598,
+      "step": 3319
+    },
+    {
+      "epoch": 3.777019340159272,
+      "grad_norm": 1.0456105470657349,
+      "learning_rate": 0.0002455062571103527,
+      "loss": 1.7178,
+      "step": 3320
+    },
+    {
+      "epoch": 3.7781569965870307,
+      "grad_norm": 0.9774885177612305,
+      "learning_rate": 0.0002452787258248009,
+      "loss": 1.9783,
+      "step": 3321
+    },
+    {
+      "epoch": 3.7792946530147895,
+      "grad_norm": 1.089418649673462,
+      "learning_rate": 0.00024505119453924917,
+      "loss": 2.1403,
+      "step": 3322
+    },
+    {
+      "epoch": 3.7804323094425483,
+      "grad_norm": 1.1333575248718262,
+      "learning_rate": 0.0002448236632536974,
+      "loss": 1.3416,
+      "step": 3323
+    },
+    {
+      "epoch": 3.781569965870307,
+      "grad_norm": 0.8414488434791565,
+      "learning_rate": 0.00024459613196814564,
+      "loss": 1.5206,
+      "step": 3324
+    },
+    {
+      "epoch": 3.782707622298066,
+      "grad_norm": 1.0938860177993774,
+      "learning_rate": 0.00024436860068259385,
+      "loss": 2.2782,
+      "step": 3325
+    },
+    {
+      "epoch": 3.783845278725825,
+      "grad_norm": 1.1637766361236572,
+      "learning_rate": 0.00024414106939704206,
+      "loss": 1.9136,
+      "step": 3326
+    },
+    {
+      "epoch": 3.7849829351535837,
+      "grad_norm": 0.893345057964325,
+      "learning_rate": 0.00024391353811149032,
+      "loss": 1.1827,
+      "step": 3327
+    },
+    {
+      "epoch": 3.7861205915813425,
+      "grad_norm": 0.8583236336708069,
+      "learning_rate": 0.0002436860068259386,
+      "loss": 1.831,
+      "step": 3328
+    },
+    {
+      "epoch": 3.7872582480091013,
+      "grad_norm": 1.1517513990402222,
+      "learning_rate": 0.0002434584755403868,
+      "loss": 1.955,
+      "step": 3329
+    },
+    {
+      "epoch": 3.78839590443686,
+      "grad_norm": 0.8208685517311096,
+      "learning_rate": 0.00024323094425483506,
+      "loss": 1.3271,
+      "step": 3330
+    },
+    {
+      "epoch": 3.789533560864619,
+      "grad_norm": 0.6080604195594788,
+      "learning_rate": 0.00024300341296928327,
+      "loss": 1.1819,
+      "step": 3331
+    },
+    {
+      "epoch": 3.7906712172923775,
+      "grad_norm": 0.7177555561065674,
+      "learning_rate": 0.0002427758816837315,
+      "loss": 1.8782,
+      "step": 3332
+    },
+    {
+      "epoch": 3.7918088737201368,
+      "grad_norm": 0.7688407301902771,
+      "learning_rate": 0.00024254835039817977,
+      "loss": 2.0169,
+      "step": 3333
+    },
+    {
+      "epoch": 3.792946530147895,
+      "grad_norm": 1.3213694095611572,
+      "learning_rate": 0.00024232081911262798,
+      "loss": 1.6509,
+      "step": 3334
+    },
+    {
+      "epoch": 3.7940841865756543,
+      "grad_norm": 1.1717759370803833,
+      "learning_rate": 0.00024209328782707625,
+      "loss": 2.5551,
+      "step": 3335
+    },
+    {
+      "epoch": 3.795221843003413,
+      "grad_norm": 1.1276628971099854,
+      "learning_rate": 0.00024186575654152446,
+      "loss": 1.9977,
+      "step": 3336
+    },
+    {
+      "epoch": 3.796359499431172,
+      "grad_norm": 0.9729685187339783,
+      "learning_rate": 0.0002416382252559727,
+      "loss": 2.0503,
+      "step": 3337
+    },
+    {
+      "epoch": 3.7974971558589306,
+      "grad_norm": 1.1408803462982178,
+      "learning_rate": 0.00024141069397042093,
+      "loss": 1.7418,
+      "step": 3338
+    },
+    {
+      "epoch": 3.7986348122866893,
+      "grad_norm": 0.7034707069396973,
+      "learning_rate": 0.00024118316268486917,
+      "loss": 1.622,
+      "step": 3339
+    },
+    {
+      "epoch": 3.799772468714448,
+      "grad_norm": 1.2321475744247437,
+      "learning_rate": 0.00024095563139931743,
+      "loss": 2.719,
+      "step": 3340
+    },
+    {
+      "epoch": 3.800910125142207,
+      "grad_norm": 0.9584794640541077,
+      "learning_rate": 0.00024072810011376564,
+      "loss": 1.7564,
+      "step": 3341
+    },
+    {
+      "epoch": 3.802047781569966,
+      "grad_norm": 0.508548378944397,
+      "learning_rate": 0.00024050056882821388,
+      "loss": 0.6892,
+      "step": 3342
+    },
+    {
+      "epoch": 3.803185437997725,
+      "grad_norm": 1.1406333446502686,
+      "learning_rate": 0.00024027303754266212,
+      "loss": 2.4856,
+      "step": 3343
+    },
+    {
+      "epoch": 3.8043230944254836,
+      "grad_norm": 1.3431506156921387,
+      "learning_rate": 0.00024004550625711035,
+      "loss": 2.4605,
+      "step": 3344
+    },
+    {
+      "epoch": 3.8054607508532423,
+      "grad_norm": 1.2452622652053833,
+      "learning_rate": 0.00023981797497155862,
+      "loss": 2.2794,
+      "step": 3345
+    },
+    {
+      "epoch": 3.806598407281001,
+      "grad_norm": 0.8428059816360474,
+      "learning_rate": 0.00023959044368600683,
+      "loss": 1.701,
+      "step": 3346
+    },
+    {
+      "epoch": 3.80773606370876,
+      "grad_norm": 1.1248981952667236,
+      "learning_rate": 0.00023936291240045506,
+      "loss": 2.0512,
+      "step": 3347
+    },
+    {
+      "epoch": 3.8088737201365186,
+      "grad_norm": 0.9891200661659241,
+      "learning_rate": 0.0002391353811149033,
+      "loss": 2.4231,
+      "step": 3348
+    },
+    {
+      "epoch": 3.810011376564278,
+      "grad_norm": 1.2691020965576172,
+      "learning_rate": 0.00023890784982935154,
+      "loss": 1.9103,
+      "step": 3349
+    },
+    {
+      "epoch": 3.8111490329920366,
+      "grad_norm": 1.2526246309280396,
+      "learning_rate": 0.0002386803185437998,
+      "loss": 2.4644,
+      "step": 3350
+    },
+    {
+      "epoch": 3.8122866894197953,
+      "grad_norm": 1.0178083181381226,
+      "learning_rate": 0.000238452787258248,
+      "loss": 1.6648,
+      "step": 3351
+    },
+    {
+      "epoch": 3.813424345847554,
+      "grad_norm": 0.7109336256980896,
+      "learning_rate": 0.00023822525597269625,
+      "loss": 1.1645,
+      "step": 3352
+    },
+    {
+      "epoch": 3.814562002275313,
+      "grad_norm": 1.2275941371917725,
+      "learning_rate": 0.00023799772468714449,
+      "loss": 1.998,
+      "step": 3353
+    },
+    {
+      "epoch": 3.8156996587030716,
+      "grad_norm": 1.1337257623672485,
+      "learning_rate": 0.00023777019340159272,
+      "loss": 3.3256,
+      "step": 3354
+    },
+    {
+      "epoch": 3.8168373151308304,
+      "grad_norm": 1.2316417694091797,
+      "learning_rate": 0.00023754266211604096,
+      "loss": 2.1339,
+      "step": 3355
+    },
+    {
+      "epoch": 3.8179749715585896,
+      "grad_norm": 2.0374646186828613,
+      "learning_rate": 0.0002373151308304892,
+      "loss": 3.4686,
+      "step": 3356
+    },
+    {
+      "epoch": 3.819112627986348,
+      "grad_norm": 1.059798002243042,
+      "learning_rate": 0.00023708759954493743,
+      "loss": 1.6049,
+      "step": 3357
+    },
+    {
+      "epoch": 3.820250284414107,
+      "grad_norm": 0.8509323596954346,
+      "learning_rate": 0.00023686006825938567,
+      "loss": 1.2672,
+      "step": 3358
+    },
+    {
+      "epoch": 3.821387940841866,
+      "grad_norm": 0.603013277053833,
+      "learning_rate": 0.0002366325369738339,
+      "loss": 1.0302,
+      "step": 3359
+    },
+    {
+      "epoch": 3.8225255972696246,
+      "grad_norm": 1.5710378885269165,
+      "learning_rate": 0.00023640500568828215,
+      "loss": 2.2583,
+      "step": 3360
+    },
+    {
+      "epoch": 3.8236632536973834,
+      "grad_norm": 1.1296995878219604,
+      "learning_rate": 0.00023617747440273038,
+      "loss": 2.1769,
+      "step": 3361
+    },
+    {
+      "epoch": 3.824800910125142,
+      "grad_norm": 1.1707825660705566,
+      "learning_rate": 0.00023594994311717862,
+      "loss": 2.6797,
+      "step": 3362
+    },
+    {
+      "epoch": 3.825938566552901,
+      "grad_norm": 1.0534515380859375,
+      "learning_rate": 0.00023572241183162686,
+      "loss": 0.8871,
+      "step": 3363
+    },
+    {
+      "epoch": 3.8270762229806596,
+      "grad_norm": 0.9066932797431946,
+      "learning_rate": 0.0002354948805460751,
+      "loss": 1.6269,
+      "step": 3364
+    },
+    {
+      "epoch": 3.828213879408419,
+      "grad_norm": 0.8703908324241638,
+      "learning_rate": 0.00023526734926052333,
+      "loss": 1.5106,
+      "step": 3365
+    },
+    {
+      "epoch": 3.8293515358361776,
+      "grad_norm": 0.8020843267440796,
+      "learning_rate": 0.00023503981797497157,
+      "loss": 2.0528,
+      "step": 3366
+    },
+    {
+      "epoch": 3.8304891922639364,
+      "grad_norm": 0.9272792935371399,
+      "learning_rate": 0.0002348122866894198,
+      "loss": 1.6559,
+      "step": 3367
+    },
+    {
+      "epoch": 3.831626848691695,
+      "grad_norm": 0.5369601845741272,
+      "learning_rate": 0.00023458475540386804,
+      "loss": 1.0992,
+      "step": 3368
+    },
+    {
+      "epoch": 3.832764505119454,
+      "grad_norm": 0.6304998397827148,
+      "learning_rate": 0.00023435722411831628,
+      "loss": 0.8554,
+      "step": 3369
+    },
+    {
+      "epoch": 3.8339021615472126,
+      "grad_norm": 0.7873075604438782,
+      "learning_rate": 0.0002341296928327645,
+      "loss": 2.2207,
+      "step": 3370
+    },
+    {
+      "epoch": 3.8350398179749714,
+      "grad_norm": 0.6331322193145752,
+      "learning_rate": 0.00023390216154721275,
+      "loss": 1.4673,
+      "step": 3371
+    },
+    {
+      "epoch": 3.8361774744027306,
+      "grad_norm": 0.5962525010108948,
+      "learning_rate": 0.00023367463026166096,
+      "loss": 1.5718,
+      "step": 3372
+    },
+    {
+      "epoch": 3.837315130830489,
+      "grad_norm": 1.0758075714111328,
+      "learning_rate": 0.00023344709897610923,
+      "loss": 2.0164,
+      "step": 3373
+    },
+    {
+      "epoch": 3.838452787258248,
+      "grad_norm": 0.6574288010597229,
+      "learning_rate": 0.00023321956769055746,
+      "loss": 1.0371,
+      "step": 3374
+    },
+    {
+      "epoch": 3.839590443686007,
+      "grad_norm": 0.984384298324585,
+      "learning_rate": 0.00023299203640500567,
+      "loss": 2.0911,
+      "step": 3375
+    },
+    {
+      "epoch": 3.8407281001137656,
+      "grad_norm": 1.0093640089035034,
+      "learning_rate": 0.00023276450511945394,
+      "loss": 1.3854,
+      "step": 3376
+    },
+    {
+      "epoch": 3.8418657565415244,
+      "grad_norm": 1.2866005897521973,
+      "learning_rate": 0.00023253697383390215,
+      "loss": 2.0335,
+      "step": 3377
+    },
+    {
+      "epoch": 3.843003412969283,
+      "grad_norm": 1.504915475845337,
+      "learning_rate": 0.0002323094425483504,
+      "loss": 2.1961,
+      "step": 3378
+    },
+    {
+      "epoch": 3.8441410693970424,
+      "grad_norm": 0.6888614296913147,
+      "learning_rate": 0.00023208191126279865,
+      "loss": 1.6042,
+      "step": 3379
+    },
+    {
+      "epoch": 3.8452787258248007,
+      "grad_norm": 0.6555305123329163,
+      "learning_rate": 0.00023185437997724686,
+      "loss": 0.8984,
+      "step": 3380
+    },
+    {
+      "epoch": 3.84641638225256,
+      "grad_norm": 0.7594724893569946,
+      "learning_rate": 0.00023162684869169512,
+      "loss": 1.2783,
+      "step": 3381
+    },
+    {
+      "epoch": 3.8475540386803186,
+      "grad_norm": 0.9284168481826782,
+      "learning_rate": 0.00023139931740614333,
+      "loss": 1.9291,
+      "step": 3382
+    },
+    {
+      "epoch": 3.8486916951080774,
+      "grad_norm": 0.5923244953155518,
+      "learning_rate": 0.0002311717861205916,
+      "loss": 0.7256,
+      "step": 3383
+    },
+    {
+      "epoch": 3.849829351535836,
+      "grad_norm": 0.9779441952705383,
+      "learning_rate": 0.00023094425483503983,
+      "loss": 1.9968,
+      "step": 3384
+    },
+    {
+      "epoch": 3.850967007963595,
+      "grad_norm": 1.1208325624465942,
+      "learning_rate": 0.00023071672354948804,
+      "loss": 2.5118,
+      "step": 3385
+    },
+    {
+      "epoch": 3.8521046643913537,
+      "grad_norm": 0.8016112446784973,
+      "learning_rate": 0.0002304891922639363,
+      "loss": 0.7837,
+      "step": 3386
+    },
+    {
+      "epoch": 3.8532423208191124,
+      "grad_norm": 0.9915149807929993,
+      "learning_rate": 0.00023026166097838452,
+      "loss": 2.0725,
+      "step": 3387
+    },
+    {
+      "epoch": 3.8543799772468716,
+      "grad_norm": 0.5172927379608154,
+      "learning_rate": 0.00023003412969283278,
+      "loss": 0.6931,
+      "step": 3388
+    },
+    {
+      "epoch": 3.8555176336746304,
+      "grad_norm": 0.6758817434310913,
+      "learning_rate": 0.000229806598407281,
+      "loss": 1.3358,
+      "step": 3389
+    },
+    {
+      "epoch": 3.856655290102389,
+      "grad_norm": 0.9508668780326843,
+      "learning_rate": 0.00022957906712172923,
+      "loss": 2.2465,
+      "step": 3390
+    },
+    {
+      "epoch": 3.857792946530148,
+      "grad_norm": 0.6855764389038086,
+      "learning_rate": 0.0002293515358361775,
+      "loss": 1.5222,
+      "step": 3391
+    },
+    {
+      "epoch": 3.8589306029579067,
+      "grad_norm": 0.8102338314056396,
+      "learning_rate": 0.0002291240045506257,
+      "loss": 1.9551,
+      "step": 3392
+    },
+    {
+      "epoch": 3.8600682593856654,
+      "grad_norm": 0.8812468647956848,
+      "learning_rate": 0.00022889647326507397,
+      "loss": 1.9685,
+      "step": 3393
+    },
+    {
+      "epoch": 3.861205915813424,
+      "grad_norm": 0.6355249881744385,
+      "learning_rate": 0.00022866894197952218,
+      "loss": 1.231,
+      "step": 3394
+    },
+    {
+      "epoch": 3.8623435722411834,
+      "grad_norm": 0.5649857521057129,
+      "learning_rate": 0.0002284414106939704,
+      "loss": 0.5657,
+      "step": 3395
+    },
+    {
+      "epoch": 3.8634812286689417,
+      "grad_norm": 0.7780776619911194,
+      "learning_rate": 0.00022821387940841868,
+      "loss": 0.8215,
+      "step": 3396
+    },
+    {
+      "epoch": 3.864618885096701,
+      "grad_norm": 0.7705486416816711,
+      "learning_rate": 0.0002279863481228669,
+      "loss": 1.7517,
+      "step": 3397
+    },
+    {
+      "epoch": 3.8657565415244597,
+      "grad_norm": 0.7325091361999512,
+      "learning_rate": 0.00022775881683731515,
+      "loss": 1.9095,
+      "step": 3398
+    },
+    {
+      "epoch": 3.8668941979522184,
+      "grad_norm": 0.6553176045417786,
+      "learning_rate": 0.00022753128555176336,
+      "loss": 1.3486,
+      "step": 3399
+    },
+    {
+      "epoch": 3.868031854379977,
+      "grad_norm": 1.1297094821929932,
+      "learning_rate": 0.0002273037542662116,
+      "loss": 1.3066,
+      "step": 3400
+    },
+    {
+      "epoch": 3.869169510807736,
+      "grad_norm": 0.6546614170074463,
+      "learning_rate": 0.00022707622298065986,
+      "loss": 1.5045,
+      "step": 3401
+    },
+    {
+      "epoch": 3.8703071672354947,
+      "grad_norm": 0.6554498076438904,
+      "learning_rate": 0.00022684869169510807,
+      "loss": 1.1808,
+      "step": 3402
+    },
+    {
+      "epoch": 3.8714448236632535,
+      "grad_norm": 0.9071396589279175,
+      "learning_rate": 0.00022662116040955634,
+      "loss": 1.5988,
+      "step": 3403
+    },
+    {
+      "epoch": 3.8725824800910127,
+      "grad_norm": 1.2640100717544556,
+      "learning_rate": 0.00022639362912400455,
+      "loss": 2.3752,
+      "step": 3404
+    },
+    {
+      "epoch": 3.8737201365187715,
+      "grad_norm": 1.221620798110962,
+      "learning_rate": 0.00022616609783845278,
+      "loss": 2.2681,
+      "step": 3405
+    },
+    {
+      "epoch": 3.87485779294653,
+      "grad_norm": 0.8574184775352478,
+      "learning_rate": 0.00022593856655290102,
+      "loss": 1.9581,
+      "step": 3406
+    },
+    {
+      "epoch": 3.875995449374289,
+      "grad_norm": 0.9591223001480103,
+      "learning_rate": 0.00022571103526734926,
+      "loss": 2.0804,
+      "step": 3407
+    },
+    {
+      "epoch": 3.8771331058020477,
+      "grad_norm": 0.7188199758529663,
+      "learning_rate": 0.00022548350398179752,
+      "loss": 1.6862,
+      "step": 3408
+    },
+    {
+      "epoch": 3.8782707622298065,
+      "grad_norm": 1.205315351486206,
+      "learning_rate": 0.00022525597269624573,
+      "loss": 2.6322,
+      "step": 3409
+    },
+    {
+      "epoch": 3.8794084186575652,
+      "grad_norm": 0.6590704321861267,
+      "learning_rate": 0.00022502844141069397,
+      "loss": 0.9434,
+      "step": 3410
+    },
+    {
+      "epoch": 3.8805460750853245,
+      "grad_norm": 1.0419220924377441,
+      "learning_rate": 0.0002248009101251422,
+      "loss": 1.7335,
+      "step": 3411
+    },
+    {
+      "epoch": 3.881683731513083,
+      "grad_norm": 0.9429000616073608,
+      "learning_rate": 0.00022457337883959044,
+      "loss": 1.6675,
+      "step": 3412
+    },
+    {
+      "epoch": 3.882821387940842,
+      "grad_norm": 0.8173443675041199,
+      "learning_rate": 0.0002243458475540387,
+      "loss": 1.5351,
+      "step": 3413
+    },
+    {
+      "epoch": 3.8839590443686007,
+      "grad_norm": 1.2141562700271606,
+      "learning_rate": 0.00022411831626848692,
+      "loss": 2.0883,
+      "step": 3414
+    },
+    {
+      "epoch": 3.8850967007963595,
+      "grad_norm": 0.9572274088859558,
+      "learning_rate": 0.00022389078498293515,
+      "loss": 2.0156,
+      "step": 3415
+    },
+    {
+      "epoch": 3.8862343572241183,
+      "grad_norm": 0.7675147652626038,
+      "learning_rate": 0.0002236632536973834,
+      "loss": 1.4844,
+      "step": 3416
+    },
+    {
+      "epoch": 3.887372013651877,
+      "grad_norm": 0.8228021264076233,
+      "learning_rate": 0.00022343572241183163,
+      "loss": 2.0502,
+      "step": 3417
+    },
+    {
+      "epoch": 3.888509670079636,
+      "grad_norm": 1.1497551202774048,
+      "learning_rate": 0.0002232081911262799,
+      "loss": 1.1353,
+      "step": 3418
+    },
+    {
+      "epoch": 3.8896473265073945,
+      "grad_norm": 0.7159321308135986,
+      "learning_rate": 0.0002229806598407281,
+      "loss": 1.627,
+      "step": 3419
+    },
+    {
+      "epoch": 3.8907849829351537,
+      "grad_norm": 1.681058645248413,
+      "learning_rate": 0.00022275312855517634,
+      "loss": 3.7341,
+      "step": 3420
+    },
+    {
+      "epoch": 3.8919226393629125,
+      "grad_norm": 0.4869871437549591,
+      "learning_rate": 0.00022252559726962457,
+      "loss": 0.7067,
+      "step": 3421
+    },
+    {
+      "epoch": 3.8930602957906713,
+      "grad_norm": 0.6947728395462036,
+      "learning_rate": 0.0002222980659840728,
+      "loss": 1.0103,
+      "step": 3422
+    },
+    {
+      "epoch": 3.89419795221843,
+      "grad_norm": 0.7470018267631531,
+      "learning_rate": 0.00022207053469852108,
+      "loss": 0.5964,
+      "step": 3423
+    },
+    {
+      "epoch": 3.8953356086461888,
+      "grad_norm": 1.2415233850479126,
+      "learning_rate": 0.00022184300341296929,
+      "loss": 2.5444,
+      "step": 3424
+    },
+    {
+      "epoch": 3.8964732650739475,
+      "grad_norm": 0.9338128566741943,
+      "learning_rate": 0.00022161547212741752,
+      "loss": 1.5243,
+      "step": 3425
+    },
+    {
+      "epoch": 3.8976109215017063,
+      "grad_norm": 0.6499525904655457,
+      "learning_rate": 0.00022138794084186576,
+      "loss": 1.009,
+      "step": 3426
+    },
+    {
+      "epoch": 3.8987485779294655,
+      "grad_norm": 1.1649621725082397,
+      "learning_rate": 0.000221160409556314,
+      "loss": 2.3154,
+      "step": 3427
+    },
+    {
+      "epoch": 3.8998862343572243,
+      "grad_norm": 1.090055227279663,
+      "learning_rate": 0.00022093287827076223,
+      "loss": 2.3174,
+      "step": 3428
+    },
+    {
+      "epoch": 3.901023890784983,
+      "grad_norm": 0.9271589517593384,
+      "learning_rate": 0.00022070534698521047,
+      "loss": 1.8594,
+      "step": 3429
+    },
+    {
+      "epoch": 3.9021615472127418,
+      "grad_norm": 0.8594130277633667,
+      "learning_rate": 0.0002204778156996587,
+      "loss": 1.9265,
+      "step": 3430
+    },
+    {
+      "epoch": 3.9032992036405005,
+      "grad_norm": 0.860653817653656,
+      "learning_rate": 0.00022025028441410694,
+      "loss": 1.9662,
+      "step": 3431
+    },
+    {
+      "epoch": 3.9044368600682593,
+      "grad_norm": 0.8960944414138794,
+      "learning_rate": 0.00022002275312855518,
+      "loss": 1.6532,
+      "step": 3432
+    },
+    {
+      "epoch": 3.905574516496018,
+      "grad_norm": 0.753968358039856,
+      "learning_rate": 0.00021979522184300342,
+      "loss": 0.8833,
+      "step": 3433
+    },
+    {
+      "epoch": 3.9067121729237773,
+      "grad_norm": 0.7025309205055237,
+      "learning_rate": 0.00021956769055745166,
+      "loss": 0.9119,
+      "step": 3434
+    },
+    {
+      "epoch": 3.9078498293515356,
+      "grad_norm": 0.9567241668701172,
+      "learning_rate": 0.0002193401592718999,
+      "loss": 1.3435,
+      "step": 3435
+    },
+    {
+      "epoch": 3.908987485779295,
+      "grad_norm": 0.8972539901733398,
+      "learning_rate": 0.00021911262798634813,
+      "loss": 1.5827,
+      "step": 3436
+    },
+    {
+      "epoch": 3.9101251422070535,
+      "grad_norm": 0.8140578269958496,
+      "learning_rate": 0.00021888509670079637,
+      "loss": 1.5984,
+      "step": 3437
+    },
+    {
+      "epoch": 3.9112627986348123,
+      "grad_norm": 1.117913007736206,
+      "learning_rate": 0.0002186575654152446,
+      "loss": 2.3435,
+      "step": 3438
+    },
+    {
+      "epoch": 3.912400455062571,
+      "grad_norm": 1.821190357208252,
+      "learning_rate": 0.00021843003412969284,
+      "loss": 3.6124,
+      "step": 3439
+    },
+    {
+      "epoch": 3.91353811149033,
+      "grad_norm": 1.1013044118881226,
+      "learning_rate": 0.00021820250284414108,
+      "loss": 2.9663,
+      "step": 3440
+    },
+    {
+      "epoch": 3.9146757679180886,
+      "grad_norm": 0.9850253462791443,
+      "learning_rate": 0.00021797497155858931,
+      "loss": 1.6718,
+      "step": 3441
+    },
+    {
+      "epoch": 3.9158134243458473,
+      "grad_norm": 0.7211486101150513,
+      "learning_rate": 0.00021774744027303755,
+      "loss": 1.7512,
+      "step": 3442
+    },
+    {
+      "epoch": 3.9169510807736065,
+      "grad_norm": 1.0833756923675537,
+      "learning_rate": 0.0002175199089874858,
+      "loss": 2.52,
+      "step": 3443
+    },
+    {
+      "epoch": 3.9180887372013653,
+      "grad_norm": 1.1811580657958984,
+      "learning_rate": 0.00021729237770193403,
+      "loss": 2.2193,
+      "step": 3444
+    },
+    {
+      "epoch": 3.919226393629124,
+      "grad_norm": 1.0001695156097412,
+      "learning_rate": 0.00021706484641638224,
+      "loss": 1.7069,
+      "step": 3445
+    },
+    {
+      "epoch": 3.920364050056883,
+      "grad_norm": 0.64976966381073,
+      "learning_rate": 0.0002168373151308305,
+      "loss": 1.7047,
+      "step": 3446
+    },
+    {
+      "epoch": 3.9215017064846416,
+      "grad_norm": 0.8505579233169556,
+      "learning_rate": 0.00021660978384527874,
+      "loss": 1.1988,
+      "step": 3447
+    },
+    {
+      "epoch": 3.9226393629124003,
+      "grad_norm": 1.036294937133789,
+      "learning_rate": 0.00021638225255972697,
+      "loss": 1.8824,
+      "step": 3448
+    },
+    {
+      "epoch": 3.923777019340159,
+      "grad_norm": 1.0031839609146118,
+      "learning_rate": 0.0002161547212741752,
+      "loss": 1.9741,
+      "step": 3449
+    },
+    {
+      "epoch": 3.9249146757679183,
+      "grad_norm": 0.8414771556854248,
+      "learning_rate": 0.00021592718998862342,
+      "loss": 1.6352,
+      "step": 3450
+    },
+    {
+      "epoch": 3.926052332195677,
+      "grad_norm": 1.0658117532730103,
+      "learning_rate": 0.00021569965870307168,
+      "loss": 1.2177,
+      "step": 3451
+    },
+    {
+      "epoch": 3.927189988623436,
+      "grad_norm": 1.3510658740997314,
+      "learning_rate": 0.00021547212741751992,
+      "loss": 2.6798,
+      "step": 3452
+    },
+    {
+      "epoch": 3.9283276450511946,
+      "grad_norm": 1.097466230392456,
+      "learning_rate": 0.00021524459613196816,
+      "loss": 1.955,
+      "step": 3453
+    },
+    {
+      "epoch": 3.9294653014789533,
+      "grad_norm": 0.5628653168678284,
+      "learning_rate": 0.0002150170648464164,
+      "loss": 1.148,
+      "step": 3454
+    },
+    {
+      "epoch": 3.930602957906712,
+      "grad_norm": 0.7025729417800903,
+      "learning_rate": 0.0002147895335608646,
+      "loss": 1.5446,
+      "step": 3455
+    },
+    {
+      "epoch": 3.931740614334471,
+      "grad_norm": 1.8733407258987427,
+      "learning_rate": 0.00021456200227531287,
+      "loss": 4.242,
+      "step": 3456
+    },
+    {
+      "epoch": 3.93287827076223,
+      "grad_norm": 1.2189491987228394,
+      "learning_rate": 0.0002143344709897611,
+      "loss": 2.2309,
+      "step": 3457
+    },
+    {
+      "epoch": 3.9340159271899884,
+      "grad_norm": 1.0556656122207642,
+      "learning_rate": 0.00021410693970420934,
+      "loss": 1.535,
+      "step": 3458
+    },
+    {
+      "epoch": 3.9351535836177476,
+      "grad_norm": 1.2942211627960205,
+      "learning_rate": 0.00021387940841865758,
+      "loss": 1.7564,
+      "step": 3459
+    },
+    {
+      "epoch": 3.9362912400455063,
+      "grad_norm": 1.3906549215316772,
+      "learning_rate": 0.0002136518771331058,
+      "loss": 2.4698,
+      "step": 3460
+    },
+    {
+      "epoch": 3.937428896473265,
+      "grad_norm": 1.5894346237182617,
+      "learning_rate": 0.00021342434584755405,
+      "loss": 3.9154,
+      "step": 3461
+    },
+    {
+      "epoch": 3.938566552901024,
+      "grad_norm": 1.111507534980774,
+      "learning_rate": 0.00021319681456200226,
+      "loss": 2.2502,
+      "step": 3462
+    },
+    {
+      "epoch": 3.9397042093287826,
+      "grad_norm": 0.9626011848449707,
+      "learning_rate": 0.00021296928327645053,
+      "loss": 2.6883,
+      "step": 3463
+    },
+    {
+      "epoch": 3.9408418657565414,
+      "grad_norm": 0.8860620856285095,
+      "learning_rate": 0.00021274175199089877,
+      "loss": 1.697,
+      "step": 3464
+    },
+    {
+      "epoch": 3.9419795221843,
+      "grad_norm": 0.9714303612709045,
+      "learning_rate": 0.00021251422070534698,
+      "loss": 3.0761,
+      "step": 3465
+    },
+    {
+      "epoch": 3.9431171786120593,
+      "grad_norm": 0.7767373323440552,
+      "learning_rate": 0.00021228668941979524,
+      "loss": 2.555,
+      "step": 3466
+    },
+    {
+      "epoch": 3.944254835039818,
+      "grad_norm": 1.177393913269043,
+      "learning_rate": 0.00021205915813424345,
+      "loss": 2.1534,
+      "step": 3467
+    },
+    {
+      "epoch": 3.945392491467577,
+      "grad_norm": 0.7708033919334412,
+      "learning_rate": 0.0002118316268486917,
+      "loss": 1.6853,
+      "step": 3468
+    },
+    {
+      "epoch": 3.9465301478953356,
+      "grad_norm": 1.034865140914917,
+      "learning_rate": 0.00021160409556313995,
+      "loss": 1.7026,
+      "step": 3469
+    },
+    {
+      "epoch": 3.9476678043230944,
+      "grad_norm": 0.769185483455658,
+      "learning_rate": 0.00021137656427758816,
+      "loss": 1.1644,
+      "step": 3470
+    },
+    {
+      "epoch": 3.948805460750853,
+      "grad_norm": 1.4568527936935425,
+      "learning_rate": 0.00021114903299203642,
+      "loss": 2.0562,
+      "step": 3471
+    },
+    {
+      "epoch": 3.949943117178612,
+      "grad_norm": 0.8477489352226257,
+      "learning_rate": 0.00021092150170648463,
+      "loss": 1.6948,
+      "step": 3472
+    },
+    {
+      "epoch": 3.951080773606371,
+      "grad_norm": 0.6970076560974121,
+      "learning_rate": 0.00021069397042093287,
+      "loss": 1.3396,
+      "step": 3473
+    },
+    {
+      "epoch": 3.9522184300341294,
+      "grad_norm": 0.5849756598472595,
+      "learning_rate": 0.00021046643913538114,
+      "loss": 1.2495,
+      "step": 3474
+    },
+    {
+      "epoch": 3.9533560864618886,
+      "grad_norm": 0.9044439196586609,
+      "learning_rate": 0.00021023890784982934,
+      "loss": 2.0495,
+      "step": 3475
+    },
+    {
+      "epoch": 3.9544937428896474,
+      "grad_norm": 1.0596073865890503,
+      "learning_rate": 0.0002100113765642776,
+      "loss": 2.21,
+      "step": 3476
+    },
+    {
+      "epoch": 3.955631399317406,
+      "grad_norm": 1.0529900789260864,
+      "learning_rate": 0.00020978384527872582,
+      "loss": 2.0394,
+      "step": 3477
+    },
+    {
+      "epoch": 3.956769055745165,
+      "grad_norm": 1.1782349348068237,
+      "learning_rate": 0.00020955631399317406,
+      "loss": 2.4599,
+      "step": 3478
+    },
+    {
+      "epoch": 3.9579067121729237,
+      "grad_norm": 0.6079378128051758,
+      "learning_rate": 0.0002093287827076223,
+      "loss": 1.3727,
+      "step": 3479
+    },
+    {
+      "epoch": 3.9590443686006824,
+      "grad_norm": 0.9474525451660156,
+      "learning_rate": 0.00020910125142207053,
+      "loss": 2.3577,
+      "step": 3480
+    },
+    {
+      "epoch": 3.960182025028441,
+      "grad_norm": 0.8396434783935547,
+      "learning_rate": 0.0002088737201365188,
+      "loss": 1.6286,
+      "step": 3481
+    },
+    {
+      "epoch": 3.9613196814562004,
+      "grad_norm": 0.784811794757843,
+      "learning_rate": 0.000208646188850967,
+      "loss": 1.852,
+      "step": 3482
+    },
+    {
+      "epoch": 3.962457337883959,
+      "grad_norm": 1.2220088243484497,
+      "learning_rate": 0.00020841865756541524,
+      "loss": 2.4076,
+      "step": 3483
+    },
+    {
+      "epoch": 3.963594994311718,
+      "grad_norm": 0.8224531412124634,
+      "learning_rate": 0.00020819112627986348,
+      "loss": 2.0785,
+      "step": 3484
+    },
+    {
+      "epoch": 3.9647326507394767,
+      "grad_norm": 1.1083797216415405,
+      "learning_rate": 0.00020796359499431171,
+      "loss": 2.125,
+      "step": 3485
+    },
+    {
+      "epoch": 3.9658703071672354,
+      "grad_norm": 0.8573135137557983,
+      "learning_rate": 0.00020773606370875998,
+      "loss": 1.063,
+      "step": 3486
+    },
+    {
+      "epoch": 3.967007963594994,
+      "grad_norm": 1.380513072013855,
+      "learning_rate": 0.0002075085324232082,
+      "loss": 3.7957,
+      "step": 3487
+    },
+    {
+      "epoch": 3.968145620022753,
+      "grad_norm": 0.7844088673591614,
+      "learning_rate": 0.00020728100113765643,
+      "loss": 1.1714,
+      "step": 3488
+    },
+    {
+      "epoch": 3.969283276450512,
+      "grad_norm": 0.8056669235229492,
+      "learning_rate": 0.00020705346985210466,
+      "loss": 1.4436,
+      "step": 3489
+    },
+    {
+      "epoch": 3.970420932878271,
+      "grad_norm": 1.3273180723190308,
+      "learning_rate": 0.0002068259385665529,
+      "loss": 3.3308,
+      "step": 3490
+    },
+    {
+      "epoch": 3.9715585893060297,
+      "grad_norm": 0.7416298389434814,
+      "learning_rate": 0.00020659840728100116,
+      "loss": 1.7255,
+      "step": 3491
+    },
+    {
+      "epoch": 3.9726962457337884,
+      "grad_norm": 1.28645658493042,
+      "learning_rate": 0.00020637087599544937,
+      "loss": 2.2977,
+      "step": 3492
+    },
+    {
+      "epoch": 3.973833902161547,
+      "grad_norm": 0.637898862361908,
+      "learning_rate": 0.0002061433447098976,
+      "loss": 1.4102,
+      "step": 3493
+    },
+    {
+      "epoch": 3.974971558589306,
+      "grad_norm": 0.5494396686553955,
+      "learning_rate": 0.00020591581342434585,
+      "loss": 1.001,
+      "step": 3494
+    },
+    {
+      "epoch": 3.9761092150170647,
+      "grad_norm": 1.295616626739502,
+      "learning_rate": 0.00020568828213879408,
+      "loss": 3.9937,
+      "step": 3495
+    },
+    {
+      "epoch": 3.977246871444824,
+      "grad_norm": 1.2073827981948853,
+      "learning_rate": 0.00020546075085324232,
+      "loss": 2.9927,
+      "step": 3496
+    },
+    {
+      "epoch": 3.9783845278725822,
+      "grad_norm": 0.5181747674942017,
+      "learning_rate": 0.00020523321956769056,
+      "loss": 0.6257,
+      "step": 3497
+    },
+    {
+      "epoch": 3.9795221843003414,
+      "grad_norm": 1.6364872455596924,
+      "learning_rate": 0.0002050056882821388,
+      "loss": 3.7021,
+      "step": 3498
+    },
+    {
+      "epoch": 3.9806598407281,
+      "grad_norm": 0.7388929724693298,
+      "learning_rate": 0.00020477815699658703,
+      "loss": 1.1443,
+      "step": 3499
+    },
+    {
+      "epoch": 3.981797497155859,
+      "grad_norm": 0.8518296480178833,
+      "learning_rate": 0.00020455062571103527,
+      "loss": 1.7027,
+      "step": 3500
+    },
+    {
+      "epoch": 3.9829351535836177,
+      "grad_norm": 0.885672390460968,
+      "learning_rate": 0.0002043230944254835,
+      "loss": 1.5378,
+      "step": 3501
+    },
+    {
+      "epoch": 3.9840728100113765,
+      "grad_norm": 0.7524988055229187,
+      "learning_rate": 0.00020409556313993174,
+      "loss": 1.4504,
+      "step": 3502
+    },
+    {
+      "epoch": 3.9852104664391352,
+      "grad_norm": 0.8979845643043518,
+      "learning_rate": 0.00020386803185437998,
+      "loss": 1.747,
+      "step": 3503
+    },
+    {
+      "epoch": 3.986348122866894,
+      "grad_norm": 0.9114024639129639,
+      "learning_rate": 0.00020364050056882822,
+      "loss": 2.2745,
+      "step": 3504
+    },
+    {
+      "epoch": 3.987485779294653,
+      "grad_norm": 1.3460856676101685,
+      "learning_rate": 0.00020341296928327645,
+      "loss": 2.4317,
+      "step": 3505
+    },
+    {
+      "epoch": 3.988623435722412,
+      "grad_norm": 0.6706535816192627,
+      "learning_rate": 0.0002031854379977247,
+      "loss": 1.5125,
+      "step": 3506
+    },
+    {
+      "epoch": 3.9897610921501707,
+      "grad_norm": 0.6658063530921936,
+      "learning_rate": 0.00020295790671217293,
+      "loss": 0.6909,
+      "step": 3507
+    },
+    {
+      "epoch": 3.9908987485779295,
+      "grad_norm": 0.7595200538635254,
+      "learning_rate": 0.00020273037542662117,
+      "loss": 1.3346,
+      "step": 3508
+    },
+    {
+      "epoch": 3.9920364050056882,
+      "grad_norm": 1.1703503131866455,
+      "learning_rate": 0.0002025028441410694,
+      "loss": 1.7334,
+      "step": 3509
+    },
+    {
+      "epoch": 3.993174061433447,
+      "grad_norm": 1.095198154449463,
+      "learning_rate": 0.00020227531285551764,
+      "loss": 2.9393,
+      "step": 3510
+    },
+    {
+      "epoch": 3.9943117178612058,
+      "grad_norm": 1.123388409614563,
+      "learning_rate": 0.00020204778156996588,
+      "loss": 1.6582,
+      "step": 3511
+    },
+    {
+      "epoch": 3.995449374288965,
+      "grad_norm": 0.6648460626602173,
+      "learning_rate": 0.00020182025028441411,
+      "loss": 1.0768,
+      "step": 3512
+    },
+    {
+      "epoch": 3.9965870307167233,
+      "grad_norm": 1.1515564918518066,
+      "learning_rate": 0.00020159271899886232,
+      "loss": 3.5262,
+      "step": 3513
+    },
+    {
+      "epoch": 3.9977246871444825,
+      "grad_norm": 0.7331502437591553,
+      "learning_rate": 0.0002013651877133106,
+      "loss": 1.3639,
+      "step": 3514
+    },
+    {
+      "epoch": 3.9988623435722412,
+      "grad_norm": 1.1474095582962036,
+      "learning_rate": 0.00020113765642775882,
+      "loss": 1.7056,
+      "step": 3515
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.1696233749389648,
+      "learning_rate": 0.00020091012514220706,
+      "loss": 2.7192,
+      "step": 3516
+    },
+    {
+      "epoch": 4.0,
+      "eval_f1": 0.8897,
+      "eval_gen_len": 49.6091,
+      "eval_loss": 1.8271052837371826,
+      "eval_precision": 0.8878,
+      "eval_recall": 0.8917,
+      "eval_rouge1": 0.4366,
+      "eval_rouge2": 0.1966,
+      "eval_rougeL": 0.3643,
+      "eval_rougeLsum": 0.4041,
+      "eval_runtime": 28.1666,
+      "eval_samples_per_second": 3.905,
+      "eval_steps_per_second": 0.497,
+      "step": 3516
+    },
+    {
+      "epoch": 4.001137656427759,
+      "grad_norm": 0.8035992383956909,
+      "learning_rate": 0.0002006825938566553,
+      "loss": 2.5708,
+      "step": 3517
+    },
+    {
+      "epoch": 4.0022753128555175,
+      "grad_norm": 0.5220395922660828,
+      "learning_rate": 0.0002004550625711035,
+      "loss": 0.9217,
+      "step": 3518
+    },
+    {
+      "epoch": 4.003412969283277,
+      "grad_norm": 0.8917536735534668,
+      "learning_rate": 0.00020022753128555177,
+      "loss": 1.7236,
+      "step": 3519
+    },
+    {
+      "epoch": 4.004550625711035,
+      "grad_norm": 1.0690449476242065,
+      "learning_rate": 0.0002,
+      "loss": 2.2547,
+      "step": 3520
+    },
+    {
+      "epoch": 4.005688282138794,
+      "grad_norm": 1.0998704433441162,
+      "learning_rate": 0.00019977246871444825,
+      "loss": 1.9161,
+      "step": 3521
+    },
+    {
+      "epoch": 4.006825938566553,
+      "grad_norm": 0.6901816129684448,
+      "learning_rate": 0.00019954493742889648,
+      "loss": 0.828,
+      "step": 3522
+    },
+    {
+      "epoch": 4.007963594994312,
+      "grad_norm": 1.0238609313964844,
+      "learning_rate": 0.0001993174061433447,
+      "loss": 1.0532,
+      "step": 3523
+    },
+    {
+      "epoch": 4.009101251422071,
+      "grad_norm": 0.8400004506111145,
+      "learning_rate": 0.00019908987485779296,
+      "loss": 1.5558,
+      "step": 3524
+    },
+    {
+      "epoch": 4.010238907849829,
+      "grad_norm": 0.8628343343734741,
+      "learning_rate": 0.0001988623435722412,
+      "loss": 1.6871,
+      "step": 3525
+    },
+    {
+      "epoch": 4.0113765642775885,
+      "grad_norm": 1.2406960725784302,
+      "learning_rate": 0.00019863481228668943,
+      "loss": 1.9634,
+      "step": 3526
+    },
+    {
+      "epoch": 4.012514220705347,
+      "grad_norm": 1.2915693521499634,
+      "learning_rate": 0.00019840728100113767,
+      "loss": 2.4605,
+      "step": 3527
+    },
+    {
+      "epoch": 4.013651877133106,
+      "grad_norm": 1.227972149848938,
+      "learning_rate": 0.00019817974971558588,
+      "loss": 2.6666,
+      "step": 3528
+    },
+    {
+      "epoch": 4.014789533560864,
+      "grad_norm": 1.1289844512939453,
+      "learning_rate": 0.00019795221843003414,
+      "loss": 1.9807,
+      "step": 3529
+    },
+    {
+      "epoch": 4.0159271899886235,
+      "grad_norm": 0.8085461258888245,
+      "learning_rate": 0.00019772468714448235,
+      "loss": 0.9486,
+      "step": 3530
+    },
+    {
+      "epoch": 4.017064846416382,
+      "grad_norm": 0.6080414652824402,
+      "learning_rate": 0.00019749715585893062,
+      "loss": 0.995,
+      "step": 3531
+    },
+    {
+      "epoch": 4.018202502844141,
+      "grad_norm": 0.7787399291992188,
+      "learning_rate": 0.00019726962457337885,
+      "loss": 1.6868,
+      "step": 3532
+    },
+    {
+      "epoch": 4.0193401592719,
+      "grad_norm": 0.7362167239189148,
+      "learning_rate": 0.00019704209328782706,
+      "loss": 1.096,
+      "step": 3533
+    },
+    {
+      "epoch": 4.020477815699659,
+      "grad_norm": 1.1093764305114746,
+      "learning_rate": 0.00019681456200227533,
+      "loss": 1.16,
+      "step": 3534
+    },
+    {
+      "epoch": 4.021615472127418,
+      "grad_norm": 1.1228350400924683,
+      "learning_rate": 0.00019658703071672354,
+      "loss": 2.6241,
+      "step": 3535
+    },
+    {
+      "epoch": 4.022753128555176,
+      "grad_norm": 0.7320429086685181,
+      "learning_rate": 0.0001963594994311718,
+      "loss": 1.7036,
+      "step": 3536
+    },
+    {
+      "epoch": 4.023890784982935,
+      "grad_norm": 0.8620187640190125,
+      "learning_rate": 0.00019613196814562004,
+      "loss": 1.7516,
+      "step": 3537
+    },
+    {
+      "epoch": 4.025028441410694,
+      "grad_norm": 0.8271589875221252,
+      "learning_rate": 0.00019590443686006825,
+      "loss": 1.4492,
+      "step": 3538
+    },
+    {
+      "epoch": 4.026166097838453,
+      "grad_norm": 1.1758748292922974,
+      "learning_rate": 0.0001956769055745165,
+      "loss": 2.0319,
+      "step": 3539
+    },
+    {
+      "epoch": 4.027303754266212,
+      "grad_norm": 1.373871088027954,
+      "learning_rate": 0.00019544937428896472,
+      "loss": 1.6359,
+      "step": 3540
+    },
+    {
+      "epoch": 4.02844141069397,
+      "grad_norm": 0.7867767214775085,
+      "learning_rate": 0.00019522184300341299,
+      "loss": 1.3931,
+      "step": 3541
+    },
+    {
+      "epoch": 4.0295790671217295,
+      "grad_norm": 0.9068583846092224,
+      "learning_rate": 0.00019499431171786122,
+      "loss": 1.5971,
+      "step": 3542
+    },
+    {
+      "epoch": 4.030716723549488,
+      "grad_norm": 1.1074446439743042,
+      "learning_rate": 0.00019476678043230943,
+      "loss": 2.0918,
+      "step": 3543
+    },
+    {
+      "epoch": 4.031854379977247,
+      "grad_norm": 0.8541181087493896,
+      "learning_rate": 0.0001945392491467577,
+      "loss": 2.3143,
+      "step": 3544
+    },
+    {
+      "epoch": 4.032992036405005,
+      "grad_norm": 1.084362268447876,
+      "learning_rate": 0.0001943117178612059,
+      "loss": 1.7493,
+      "step": 3545
+    },
+    {
+      "epoch": 4.034129692832765,
+      "grad_norm": 1.041742205619812,
+      "learning_rate": 0.00019408418657565417,
+      "loss": 1.8627,
+      "step": 3546
+    },
+    {
+      "epoch": 4.035267349260523,
+      "grad_norm": 0.7245301604270935,
+      "learning_rate": 0.00019385665529010238,
+      "loss": 0.726,
+      "step": 3547
+    },
+    {
+      "epoch": 4.036405005688282,
+      "grad_norm": 0.8224193453788757,
+      "learning_rate": 0.00019362912400455062,
+      "loss": 1.5722,
+      "step": 3548
+    },
+    {
+      "epoch": 4.037542662116041,
+      "grad_norm": 1.1694352626800537,
+      "learning_rate": 0.00019340159271899888,
+      "loss": 1.0027,
+      "step": 3549
+    },
+    {
+      "epoch": 4.0386803185438,
+      "grad_norm": 0.9688643217086792,
+      "learning_rate": 0.0001931740614334471,
+      "loss": 2.2008,
+      "step": 3550
+    },
+    {
+      "epoch": 4.039817974971559,
+      "grad_norm": 0.6902188062667847,
+      "learning_rate": 0.00019294653014789536,
+      "loss": 1.4584,
+      "step": 3551
+    },
+    {
+      "epoch": 4.040955631399317,
+      "grad_norm": 1.1102920770645142,
+      "learning_rate": 0.00019271899886234357,
+      "loss": 2.3507,
+      "step": 3552
+    },
+    {
+      "epoch": 4.042093287827076,
+      "grad_norm": 0.5606656074523926,
+      "learning_rate": 0.0001924914675767918,
+      "loss": 1.1701,
+      "step": 3553
+    },
+    {
+      "epoch": 4.043230944254835,
+      "grad_norm": 0.5422776937484741,
+      "learning_rate": 0.00019226393629124007,
+      "loss": 1.1612,
+      "step": 3554
+    },
+    {
+      "epoch": 4.044368600682594,
+      "grad_norm": 1.0809518098831177,
+      "learning_rate": 0.00019203640500568828,
+      "loss": 2.5527,
+      "step": 3555
+    },
+    {
+      "epoch": 4.045506257110353,
+      "grad_norm": 0.9332743287086487,
+      "learning_rate": 0.00019180887372013654,
+      "loss": 1.4009,
+      "step": 3556
+    },
+    {
+      "epoch": 4.046643913538111,
+      "grad_norm": 0.721682608127594,
+      "learning_rate": 0.00019158134243458475,
+      "loss": 0.9377,
+      "step": 3557
+    },
+    {
+      "epoch": 4.047781569965871,
+      "grad_norm": 0.8674675226211548,
+      "learning_rate": 0.000191353811149033,
+      "loss": 1.783,
+      "step": 3558
+    },
+    {
+      "epoch": 4.048919226393629,
+      "grad_norm": 0.8048381209373474,
+      "learning_rate": 0.00019112627986348125,
+      "loss": 1.9681,
+      "step": 3559
+    },
+    {
+      "epoch": 4.050056882821388,
+      "grad_norm": 1.0859109163284302,
+      "learning_rate": 0.00019089874857792946,
+      "loss": 1.4804,
+      "step": 3560
+    },
+    {
+      "epoch": 4.051194539249146,
+      "grad_norm": 0.8473076820373535,
+      "learning_rate": 0.00019067121729237773,
+      "loss": 1.2142,
+      "step": 3561
+    },
+    {
+      "epoch": 4.052332195676906,
+      "grad_norm": 0.7596136331558228,
+      "learning_rate": 0.00019044368600682594,
+      "loss": 1.2756,
+      "step": 3562
+    },
+    {
+      "epoch": 4.053469852104665,
+      "grad_norm": 0.7931225895881653,
+      "learning_rate": 0.00019021615472127417,
+      "loss": 1.4067,
+      "step": 3563
+    },
+    {
+      "epoch": 4.054607508532423,
+      "grad_norm": 0.8623807430267334,
+      "learning_rate": 0.0001899886234357224,
+      "loss": 1.6287,
+      "step": 3564
+    },
+    {
+      "epoch": 4.055745164960182,
+      "grad_norm": 1.0621230602264404,
+      "learning_rate": 0.00018976109215017065,
+      "loss": 2.8103,
+      "step": 3565
+    },
+    {
+      "epoch": 4.056882821387941,
+      "grad_norm": 0.8181642293930054,
+      "learning_rate": 0.0001895335608646189,
+      "loss": 1.1884,
+      "step": 3566
+    },
+    {
+      "epoch": 4.0580204778157,
+      "grad_norm": 0.9738771319389343,
+      "learning_rate": 0.00018930602957906712,
+      "loss": 1.9755,
+      "step": 3567
+    },
+    {
+      "epoch": 4.059158134243458,
+      "grad_norm": 0.9224938154220581,
+      "learning_rate": 0.00018907849829351536,
+      "loss": 1.288,
+      "step": 3568
+    },
+    {
+      "epoch": 4.060295790671217,
+      "grad_norm": 0.8679994940757751,
+      "learning_rate": 0.0001888509670079636,
+      "loss": 1.4742,
+      "step": 3569
+    },
+    {
+      "epoch": 4.061433447098976,
+      "grad_norm": 1.6250495910644531,
+      "learning_rate": 0.00018862343572241183,
+      "loss": 3.5226,
+      "step": 3570
+    },
+    {
+      "epoch": 4.062571103526735,
+      "grad_norm": NaN,
+      "learning_rate": 0.00018862343572241183,
+      "loss": 1.1473,
+      "step": 3571
+    },
+    {
+      "epoch": 4.063708759954494,
+      "grad_norm": 1.0141960382461548,
+      "learning_rate": 0.0001883959044368601,
+      "loss": 2.936,
+      "step": 3572
+    },
+    {
+      "epoch": 4.064846416382252,
+      "grad_norm": 1.2739934921264648,
+      "learning_rate": 0.0001881683731513083,
+      "loss": 2.1047,
+      "step": 3573
+    },
+    {
+      "epoch": 4.065984072810012,
+      "grad_norm": 1.6481419801712036,
+      "learning_rate": 0.00018794084186575654,
+      "loss": 3.6404,
+      "step": 3574
+    },
+    {
+      "epoch": 4.06712172923777,
+      "grad_norm": 0.9871388673782349,
+      "learning_rate": 0.00018771331058020478,
+      "loss": 2.5576,
+      "step": 3575
+    },
+    {
+      "epoch": 4.068259385665529,
+      "grad_norm": 0.8460843563079834,
+      "learning_rate": 0.00018748577929465302,
+      "loss": 2.5027,
+      "step": 3576
+    },
+    {
+      "epoch": 4.0693970420932875,
+      "grad_norm": 0.9054187536239624,
+      "learning_rate": 0.00018725824800910125,
+      "loss": 1.5227,
+      "step": 3577
+    },
+    {
+      "epoch": 4.070534698521047,
+      "grad_norm": 1.1341551542282104,
+      "learning_rate": 0.0001870307167235495,
+      "loss": 1.8531,
+      "step": 3578
+    },
+    {
+      "epoch": 4.071672354948806,
+      "grad_norm": 0.7718223929405212,
+      "learning_rate": 0.00018680318543799773,
+      "loss": 1.1577,
+      "step": 3579
+    },
+    {
+      "epoch": 4.072810011376564,
+      "grad_norm": 0.7591809034347534,
+      "learning_rate": 0.00018657565415244596,
+      "loss": 1.069,
+      "step": 3580
+    },
+    {
+      "epoch": 4.073947667804323,
+      "grad_norm": 1.5402214527130127,
+      "learning_rate": 0.0001863481228668942,
+      "loss": 3.2458,
+      "step": 3581
+    },
+    {
+      "epoch": 4.075085324232082,
+      "grad_norm": 1.6219172477722168,
+      "learning_rate": 0.00018612059158134244,
+      "loss": 2.8352,
+      "step": 3582
+    },
+    {
+      "epoch": 4.076222980659841,
+      "grad_norm": 0.7609260082244873,
+      "learning_rate": 0.00018589306029579068,
+      "loss": 1.5297,
+      "step": 3583
+    },
+    {
+      "epoch": 4.077360637087599,
+      "grad_norm": 0.9463198184967041,
+      "learning_rate": 0.0001856655290102389,
+      "loss": 1.4272,
+      "step": 3584
+    },
+    {
+      "epoch": 4.078498293515358,
+      "grad_norm": 1.0422461032867432,
+      "learning_rate": 0.00018543799772468715,
+      "loss": 2.3036,
+      "step": 3585
+    },
+    {
+      "epoch": 4.079635949943118,
+      "grad_norm": 1.6263346672058105,
+      "learning_rate": 0.0001852104664391354,
+      "loss": 2.6899,
+      "step": 3586
+    },
+    {
+      "epoch": 4.080773606370876,
+      "grad_norm": 0.8726171851158142,
+      "learning_rate": 0.0001849829351535836,
+      "loss": 1.9618,
+      "step": 3587
+    },
+    {
+      "epoch": 4.081911262798635,
+      "grad_norm": 1.0334999561309814,
+      "learning_rate": 0.00018475540386803186,
+      "loss": 1.9655,
+      "step": 3588
+    },
+    {
+      "epoch": 4.0830489192263935,
+      "grad_norm": 1.2761929035186768,
+      "learning_rate": 0.0001845278725824801,
+      "loss": 1.1134,
+      "step": 3589
+    },
+    {
+      "epoch": 4.084186575654153,
+      "grad_norm": 1.0597755908966064,
+      "learning_rate": 0.00018430034129692833,
+      "loss": 1.6925,
+      "step": 3590
+    },
+    {
+      "epoch": 4.085324232081911,
+      "grad_norm": 1.0300536155700684,
+      "learning_rate": 0.00018407281001137657,
+      "loss": 1.877,
+      "step": 3591
+    },
+    {
+      "epoch": 4.08646188850967,
+      "grad_norm": 1.1007896661758423,
+      "learning_rate": 0.00018384527872582478,
+      "loss": 1.8541,
+      "step": 3592
+    },
+    {
+      "epoch": 4.0875995449374285,
+      "grad_norm": 0.9704376459121704,
+      "learning_rate": 0.00018361774744027305,
+      "loss": 1.1683,
+      "step": 3593
+    },
+    {
+      "epoch": 4.088737201365188,
+      "grad_norm": 1.4303745031356812,
+      "learning_rate": 0.00018339021615472128,
+      "loss": 1.9568,
+      "step": 3594
+    },
+    {
+      "epoch": 4.089874857792947,
+      "grad_norm": 0.7820960879325867,
+      "learning_rate": 0.00018316268486916952,
+      "loss": 1.3428,
+      "step": 3595
+    },
+    {
+      "epoch": 4.091012514220705,
+      "grad_norm": 0.9371392726898193,
+      "learning_rate": 0.00018293515358361776,
+      "loss": 1.4801,
+      "step": 3596
+    },
+    {
+      "epoch": 4.092150170648464,
+      "grad_norm": 0.9383748173713684,
+      "learning_rate": 0.00018270762229806597,
+      "loss": 1.8923,
+      "step": 3597
+    },
+    {
+      "epoch": 4.093287827076223,
+      "grad_norm": 1.0831904411315918,
+      "learning_rate": 0.00018248009101251423,
+      "loss": 2.4795,
+      "step": 3598
+    },
+    {
+      "epoch": 4.094425483503982,
+      "grad_norm": 0.872491180896759,
+      "learning_rate": 0.00018225255972696247,
+      "loss": 1.6777,
+      "step": 3599
+    },
+    {
+      "epoch": 4.09556313993174,
+      "grad_norm": 1.4505406618118286,
+      "learning_rate": 0.0001820250284414107,
+      "loss": 2.7823,
+      "step": 3600
+    },
+    {
+      "epoch": 4.0967007963594995,
+      "grad_norm": 1.2872380018234253,
+      "learning_rate": 0.00018179749715585894,
+      "loss": 2.3393,
+      "step": 3601
+    },
+    {
+      "epoch": 4.097838452787259,
+      "grad_norm": 1.019313097000122,
+      "learning_rate": 0.00018156996587030715,
+      "loss": 2.0507,
+      "step": 3602
+    },
+    {
+      "epoch": 4.098976109215017,
+      "grad_norm": 1.1505863666534424,
+      "learning_rate": 0.00018134243458475542,
+      "loss": 1.7178,
+      "step": 3603
+    },
+    {
+      "epoch": 4.100113765642776,
+      "grad_norm": 0.6773290038108826,
+      "learning_rate": 0.00018111490329920363,
+      "loss": 1.6069,
+      "step": 3604
+    },
+    {
+      "epoch": 4.1012514220705345,
+      "grad_norm": 0.7799268960952759,
+      "learning_rate": 0.0001808873720136519,
+      "loss": 1.6404,
+      "step": 3605
+    },
+    {
+      "epoch": 4.102389078498294,
+      "grad_norm": 1.3413171768188477,
+      "learning_rate": 0.00018065984072810013,
+      "loss": 4.3731,
+      "step": 3606
+    },
+    {
+      "epoch": 4.103526734926052,
+      "grad_norm": 1.068517804145813,
+      "learning_rate": 0.00018043230944254834,
+      "loss": 1.7495,
+      "step": 3607
+    },
+    {
+      "epoch": 4.104664391353811,
+      "grad_norm": 0.895536482334137,
+      "learning_rate": 0.0001802047781569966,
+      "loss": 2.339,
+      "step": 3608
+    },
+    {
+      "epoch": 4.1058020477815695,
+      "grad_norm": 1.09737229347229,
+      "learning_rate": 0.0001799772468714448,
+      "loss": 1.8625,
+      "step": 3609
+    },
+    {
+      "epoch": 4.106939704209329,
+      "grad_norm": 1.0622013807296753,
+      "learning_rate": 0.00017974971558589307,
+      "loss": 1.309,
+      "step": 3610
+    },
+    {
+      "epoch": 4.108077360637088,
+      "grad_norm": 1.0724921226501465,
+      "learning_rate": 0.0001795221843003413,
+      "loss": 2.1258,
+      "step": 3611
+    },
+    {
+      "epoch": 4.109215017064846,
+      "grad_norm": 0.6296405792236328,
+      "learning_rate": 0.00017929465301478952,
+      "loss": 1.1405,
+      "step": 3612
+    },
+    {
+      "epoch": 4.1103526734926055,
+      "grad_norm": 0.6233872771263123,
+      "learning_rate": 0.00017906712172923779,
+      "loss": 1.4618,
+      "step": 3613
+    },
+    {
+      "epoch": 4.111490329920364,
+      "grad_norm": 1.0284563302993774,
+      "learning_rate": 0.000178839590443686,
+      "loss": 2.1088,
+      "step": 3614
+    },
+    {
+      "epoch": 4.112627986348123,
+      "grad_norm": 0.5234109163284302,
+      "learning_rate": 0.00017861205915813426,
+      "loss": 0.6546,
+      "step": 3615
+    },
+    {
+      "epoch": 4.113765642775881,
+      "grad_norm": 1.1184611320495605,
+      "learning_rate": 0.0001783845278725825,
+      "loss": 2.2601,
+      "step": 3616
+    },
+    {
+      "epoch": 4.1149032992036405,
+      "grad_norm": 0.5588635802268982,
+      "learning_rate": 0.0001781569965870307,
+      "loss": 1.2117,
+      "step": 3617
+    },
+    {
+      "epoch": 4.1160409556314,
+      "grad_norm": 1.029264211654663,
+      "learning_rate": 0.00017792946530147897,
+      "loss": 2.4449,
+      "step": 3618
+    },
+    {
+      "epoch": 4.117178612059158,
+      "grad_norm": 0.6462013125419617,
+      "learning_rate": 0.00017770193401592718,
+      "loss": 1.0404,
+      "step": 3619
+    },
+    {
+      "epoch": 4.118316268486917,
+      "grad_norm": 0.8388407826423645,
+      "learning_rate": 0.00017747440273037544,
+      "loss": 1.1833,
+      "step": 3620
+    },
+    {
+      "epoch": 4.1194539249146755,
+      "grad_norm": 0.6785479187965393,
+      "learning_rate": 0.00017724687144482365,
+      "loss": 1.404,
+      "step": 3621
+    },
+    {
+      "epoch": 4.120591581342435,
+      "grad_norm": 0.742090106010437,
+      "learning_rate": 0.0001770193401592719,
+      "loss": 1.844,
+      "step": 3622
+    },
+    {
+      "epoch": 4.121729237770193,
+      "grad_norm": 0.5642427206039429,
+      "learning_rate": 0.00017679180887372016,
+      "loss": 0.8566,
+      "step": 3623
+    },
+    {
+      "epoch": 4.122866894197952,
+      "grad_norm": 1.279248595237732,
+      "learning_rate": 0.00017656427758816837,
+      "loss": 2.091,
+      "step": 3624
+    },
+    {
+      "epoch": 4.1240045506257115,
+      "grad_norm": 0.8221580386161804,
+      "learning_rate": 0.00017633674630261663,
+      "loss": 1.0874,
+      "step": 3625
+    },
+    {
+      "epoch": 4.12514220705347,
+      "grad_norm": 1.105421543121338,
+      "learning_rate": 0.00017610921501706484,
+      "loss": 1.8684,
+      "step": 3626
+    },
+    {
+      "epoch": 4.126279863481229,
+      "grad_norm": 0.6208410859107971,
+      "learning_rate": 0.00017588168373151308,
+      "loss": 1.1682,
+      "step": 3627
+    },
+    {
+      "epoch": 4.127417519908987,
+      "grad_norm": 1.1576206684112549,
+      "learning_rate": 0.00017565415244596134,
+      "loss": 2.1256,
+      "step": 3628
+    },
+    {
+      "epoch": 4.1285551763367465,
+      "grad_norm": 0.9814143180847168,
+      "learning_rate": 0.00017542662116040955,
+      "loss": 2.8486,
+      "step": 3629
+    },
+    {
+      "epoch": 4.129692832764505,
+      "grad_norm": 1.1456234455108643,
+      "learning_rate": 0.00017519908987485781,
+      "loss": 1.3626,
+      "step": 3630
+    },
+    {
+      "epoch": 4.130830489192264,
+      "grad_norm": 0.8578894734382629,
+      "learning_rate": 0.00017497155858930602,
+      "loss": 1.7545,
+      "step": 3631
+    },
+    {
+      "epoch": 4.131968145620022,
+      "grad_norm": 0.7834432721138,
+      "learning_rate": 0.00017474402730375426,
+      "loss": 1.7061,
+      "step": 3632
+    },
+    {
+      "epoch": 4.1331058020477816,
+      "grad_norm": 1.3007863759994507,
+      "learning_rate": 0.00017451649601820253,
+      "loss": 1.5035,
+      "step": 3633
+    },
+    {
+      "epoch": 4.134243458475541,
+      "grad_norm": 1.0654963254928589,
+      "learning_rate": 0.00017428896473265074,
+      "loss": 2.8058,
+      "step": 3634
+    },
+    {
+      "epoch": 4.135381114903299,
+      "grad_norm": 1.042611002922058,
+      "learning_rate": 0.000174061433447099,
+      "loss": 1.5158,
+      "step": 3635
+    },
+    {
+      "epoch": 4.136518771331058,
+      "grad_norm": 0.8858871459960938,
+      "learning_rate": 0.0001738339021615472,
+      "loss": 2.2742,
+      "step": 3636
+    },
+    {
+      "epoch": 4.137656427758817,
+      "grad_norm": 1.099360466003418,
+      "learning_rate": 0.00017360637087599545,
+      "loss": 2.7292,
+      "step": 3637
+    },
+    {
+      "epoch": 4.138794084186576,
+      "grad_norm": 0.520845890045166,
+      "learning_rate": 0.00017337883959044368,
+      "loss": 0.512,
+      "step": 3638
+    },
+    {
+      "epoch": 4.139931740614334,
+      "grad_norm": 1.025590181350708,
+      "learning_rate": 0.00017315130830489192,
+      "loss": 2.4849,
+      "step": 3639
+    },
+    {
+      "epoch": 4.141069397042093,
+      "grad_norm": 2.0368833541870117,
+      "learning_rate": 0.00017292377701934018,
+      "loss": 2.6333,
+      "step": 3640
+    },
+    {
+      "epoch": 4.1422070534698525,
+      "grad_norm": 0.952337920665741,
+      "learning_rate": 0.0001726962457337884,
+      "loss": 0.9957,
+      "step": 3641
+    },
+    {
+      "epoch": 4.143344709897611,
+      "grad_norm": 0.782004177570343,
+      "learning_rate": 0.00017246871444823663,
+      "loss": 1.4577,
+      "step": 3642
+    },
+    {
+      "epoch": 4.14448236632537,
+      "grad_norm": 0.8281341195106506,
+      "learning_rate": 0.00017224118316268487,
+      "loss": 1.6821,
+      "step": 3643
+    },
+    {
+      "epoch": 4.145620022753128,
+      "grad_norm": 0.8365357518196106,
+      "learning_rate": 0.0001720136518771331,
+      "loss": 2.0037,
+      "step": 3644
+    },
+    {
+      "epoch": 4.146757679180888,
+      "grad_norm": 0.7121614813804626,
+      "learning_rate": 0.00017178612059158137,
+      "loss": 1.2755,
+      "step": 3645
+    },
+    {
+      "epoch": 4.147895335608646,
+      "grad_norm": 0.890727162361145,
+      "learning_rate": 0.00017155858930602958,
+      "loss": 1.4112,
+      "step": 3646
+    },
+    {
+      "epoch": 4.149032992036405,
+      "grad_norm": 0.9248006343841553,
+      "learning_rate": 0.00017133105802047782,
+      "loss": 1.4232,
+      "step": 3647
+    },
+    {
+      "epoch": 4.150170648464163,
+      "grad_norm": 0.775360107421875,
+      "learning_rate": 0.00017110352673492605,
+      "loss": 1.675,
+      "step": 3648
+    },
+    {
+      "epoch": 4.151308304891923,
+      "grad_norm": 0.8243028521537781,
+      "learning_rate": 0.0001708759954493743,
+      "loss": 1.8722,
+      "step": 3649
+    },
+    {
+      "epoch": 4.152445961319682,
+      "grad_norm": 1.020837426185608,
+      "learning_rate": 0.00017064846416382255,
+      "loss": 2.4388,
+      "step": 3650
+    },
+    {
+      "epoch": 4.15358361774744,
+      "grad_norm": 1.433733344078064,
+      "learning_rate": 0.00017042093287827076,
+      "loss": 3.276,
+      "step": 3651
+    },
+    {
+      "epoch": 4.154721274175199,
+      "grad_norm": 0.7089780569076538,
+      "learning_rate": 0.000170193401592719,
+      "loss": 1.5784,
+      "step": 3652
+    },
+    {
+      "epoch": 4.155858930602958,
+      "grad_norm": 1.084784984588623,
+      "learning_rate": 0.00016996587030716724,
+      "loss": 1.4955,
+      "step": 3653
+    },
+    {
+      "epoch": 4.156996587030717,
+      "grad_norm": 0.6325297951698303,
+      "learning_rate": 0.00016973833902161548,
+      "loss": 0.6437,
+      "step": 3654
+    },
+    {
+      "epoch": 4.158134243458475,
+      "grad_norm": 0.7360290884971619,
+      "learning_rate": 0.0001695108077360637,
+      "loss": 0.833,
+      "step": 3655
+    },
+    {
+      "epoch": 4.159271899886234,
+      "grad_norm": 0.983505129814148,
+      "learning_rate": 0.00016928327645051195,
+      "loss": 2.5266,
+      "step": 3656
+    },
+    {
+      "epoch": 4.160409556313994,
+      "grad_norm": 1.1645041704177856,
+      "learning_rate": 0.00016905574516496019,
+      "loss": 1.9561,
+      "step": 3657
+    },
+    {
+      "epoch": 4.161547212741752,
+      "grad_norm": 0.8973751068115234,
+      "learning_rate": 0.00016882821387940842,
+      "loss": 0.9499,
+      "step": 3658
+    },
+    {
+      "epoch": 4.162684869169511,
+      "grad_norm": 0.748742401599884,
+      "learning_rate": 0.00016860068259385666,
+      "loss": 1.542,
+      "step": 3659
+    },
+    {
+      "epoch": 4.163822525597269,
+      "grad_norm": 0.6661348342895508,
+      "learning_rate": 0.0001683731513083049,
+      "loss": 1.4095,
+      "step": 3660
+    },
+    {
+      "epoch": 4.164960182025029,
+      "grad_norm": 1.0042301416397095,
+      "learning_rate": 0.00016814562002275313,
+      "loss": 1.6135,
+      "step": 3661
+    },
+    {
+      "epoch": 4.166097838452787,
+      "grad_norm": 1.056707739830017,
+      "learning_rate": 0.00016791808873720137,
+      "loss": 1.8746,
+      "step": 3662
+    },
+    {
+      "epoch": 4.167235494880546,
+      "grad_norm": 1.1747101545333862,
+      "learning_rate": 0.0001676905574516496,
+      "loss": 1.8581,
+      "step": 3663
+    },
+    {
+      "epoch": 4.168373151308305,
+      "grad_norm": 0.9841439723968506,
+      "learning_rate": 0.00016746302616609785,
+      "loss": 2.1848,
+      "step": 3664
+    },
+    {
+      "epoch": 4.169510807736064,
+      "grad_norm": 1.0959367752075195,
+      "learning_rate": 0.00016723549488054606,
+      "loss": 2.9336,
+      "step": 3665
+    },
+    {
+      "epoch": 4.170648464163823,
+      "grad_norm": 1.0343743562698364,
+      "learning_rate": 0.00016700796359499432,
+      "loss": 1.8865,
+      "step": 3666
+    },
+    {
+      "epoch": 4.171786120591581,
+      "grad_norm": 1.1250258684158325,
+      "learning_rate": 0.00016678043230944256,
+      "loss": 2.206,
+      "step": 3667
+    },
+    {
+      "epoch": 4.17292377701934,
+      "grad_norm": 0.8602643609046936,
+      "learning_rate": 0.0001665529010238908,
+      "loss": 1.6804,
+      "step": 3668
+    },
+    {
+      "epoch": 4.174061433447099,
+      "grad_norm": 1.1515132188796997,
+      "learning_rate": 0.00016632536973833903,
+      "loss": 2.0477,
+      "step": 3669
+    },
+    {
+      "epoch": 4.175199089874858,
+      "grad_norm": 1.0483062267303467,
+      "learning_rate": 0.00016609783845278724,
+      "loss": 1.4297,
+      "step": 3670
+    },
+    {
+      "epoch": 4.176336746302616,
+      "grad_norm": 1.0967140197753906,
+      "learning_rate": 0.0001658703071672355,
+      "loss": 2.4236,
+      "step": 3671
+    },
+    {
+      "epoch": 4.177474402730375,
+      "grad_norm": 0.9350886940956116,
+      "learning_rate": 0.00016564277588168371,
+      "loss": 2.0317,
+      "step": 3672
+    },
+    {
+      "epoch": 4.178612059158135,
+      "grad_norm": 0.6748828291893005,
+      "learning_rate": 0.00016541524459613198,
+      "loss": 1.6643,
+      "step": 3673
+    },
+    {
+      "epoch": 4.179749715585893,
+      "grad_norm": 1.0818042755126953,
+      "learning_rate": 0.00016518771331058022,
+      "loss": 1.7651,
+      "step": 3674
+    },
+    {
+      "epoch": 4.180887372013652,
+      "grad_norm": 0.8341732621192932,
+      "learning_rate": 0.00016496018202502842,
+      "loss": 1.7182,
+      "step": 3675
+    },
+    {
+      "epoch": 4.18202502844141,
+      "grad_norm": 0.7781969904899597,
+      "learning_rate": 0.0001647326507394767,
+      "loss": 1.5752,
+      "step": 3676
+    },
+    {
+      "epoch": 4.18316268486917,
+      "grad_norm": 0.9930599927902222,
+      "learning_rate": 0.0001645051194539249,
+      "loss": 1.7508,
+      "step": 3677
+    },
+    {
+      "epoch": 4.184300341296928,
+      "grad_norm": 1.0659433603286743,
+      "learning_rate": 0.00016427758816837316,
+      "loss": 2.5191,
+      "step": 3678
+    },
+    {
+      "epoch": 4.185437997724687,
+      "grad_norm": 0.7574253082275391,
+      "learning_rate": 0.0001640500568828214,
+      "loss": 1.2169,
+      "step": 3679
+    },
+    {
+      "epoch": 4.186575654152446,
+      "grad_norm": 0.7488005757331848,
+      "learning_rate": 0.0001638225255972696,
+      "loss": 1.2127,
+      "step": 3680
+    },
+    {
+      "epoch": 4.187713310580205,
+      "grad_norm": 0.8474146127700806,
+      "learning_rate": 0.00016359499431171787,
+      "loss": 1.2748,
+      "step": 3681
+    },
+    {
+      "epoch": 4.188850967007964,
+      "grad_norm": 0.9467219710350037,
+      "learning_rate": 0.00016336746302616608,
+      "loss": 2.0191,
+      "step": 3682
+    },
+    {
+      "epoch": 4.189988623435722,
+      "grad_norm": 0.7899359464645386,
+      "learning_rate": 0.00016313993174061435,
+      "loss": 0.7179,
+      "step": 3683
+    },
+    {
+      "epoch": 4.191126279863481,
+      "grad_norm": 1.5700042247772217,
+      "learning_rate": 0.00016291240045506259,
+      "loss": 2.2769,
+      "step": 3684
+    },
+    {
+      "epoch": 4.19226393629124,
+      "grad_norm": 0.9096818566322327,
+      "learning_rate": 0.0001626848691695108,
+      "loss": 1.828,
+      "step": 3685
+    },
+    {
+      "epoch": 4.193401592718999,
+      "grad_norm": 0.9237043857574463,
+      "learning_rate": 0.00016245733788395906,
+      "loss": 1.4392,
+      "step": 3686
+    },
+    {
+      "epoch": 4.194539249146757,
+      "grad_norm": 1.1367732286453247,
+      "learning_rate": 0.00016222980659840727,
+      "loss": 3.0341,
+      "step": 3687
+    },
+    {
+      "epoch": 4.1956769055745164,
+      "grad_norm": 0.9193979501724243,
+      "learning_rate": 0.00016200227531285553,
+      "loss": 1.32,
+      "step": 3688
+    },
+    {
+      "epoch": 4.196814562002276,
+      "grad_norm": 1.5365267992019653,
+      "learning_rate": 0.00016177474402730374,
+      "loss": 3.1212,
+      "step": 3689
+    },
+    {
+      "epoch": 4.197952218430034,
+      "grad_norm": 0.8413816094398499,
+      "learning_rate": 0.00016154721274175198,
+      "loss": 1.7901,
+      "step": 3690
+    },
+    {
+      "epoch": 4.199089874857793,
+      "grad_norm": 0.6427741050720215,
+      "learning_rate": 0.00016131968145620024,
+      "loss": 0.5028,
+      "step": 3691
+    },
+    {
+      "epoch": 4.2002275312855515,
+      "grad_norm": 0.9684869647026062,
+      "learning_rate": 0.00016109215017064845,
+      "loss": 2.0259,
+      "step": 3692
+    },
+    {
+      "epoch": 4.201365187713311,
+      "grad_norm": 0.6879338622093201,
+      "learning_rate": 0.00016086461888509672,
+      "loss": 1.3484,
+      "step": 3693
+    },
+    {
+      "epoch": 4.202502844141069,
+      "grad_norm": 0.9934858679771423,
+      "learning_rate": 0.00016063708759954493,
+      "loss": 1.5266,
+      "step": 3694
+    },
+    {
+      "epoch": 4.203640500568828,
+      "grad_norm": 1.18158757686615,
+      "learning_rate": 0.00016040955631399316,
+      "loss": 2.0914,
+      "step": 3695
+    },
+    {
+      "epoch": 4.204778156996587,
+      "grad_norm": 0.776360809803009,
+      "learning_rate": 0.00016018202502844143,
+      "loss": 1.8111,
+      "step": 3696
+    },
+    {
+      "epoch": 4.205915813424346,
+      "grad_norm": 0.5315882563591003,
+      "learning_rate": 0.00015995449374288964,
+      "loss": 0.8625,
+      "step": 3697
+    },
+    {
+      "epoch": 4.207053469852105,
+      "grad_norm": 0.7266376614570618,
+      "learning_rate": 0.0001597269624573379,
+      "loss": 0.9552,
+      "step": 3698
+    },
+    {
+      "epoch": 4.208191126279863,
+      "grad_norm": 0.8702644109725952,
+      "learning_rate": 0.0001594994311717861,
+      "loss": 1.4781,
+      "step": 3699
+    },
+    {
+      "epoch": 4.2093287827076225,
+      "grad_norm": 0.6491544246673584,
+      "learning_rate": 0.00015927189988623435,
+      "loss": 0.9899,
+      "step": 3700
+    },
+    {
+      "epoch": 4.210466439135381,
+      "grad_norm": 1.0166810750961304,
+      "learning_rate": 0.00015904436860068261,
+      "loss": 1.7634,
+      "step": 3701
+    },
+    {
+      "epoch": 4.21160409556314,
+      "grad_norm": 1.268738031387329,
+      "learning_rate": 0.00015881683731513082,
+      "loss": 2.6249,
+      "step": 3702
+    },
+    {
+      "epoch": 4.212741751990899,
+      "grad_norm": 1.180303692817688,
+      "learning_rate": 0.0001585893060295791,
+      "loss": 3.388,
+      "step": 3703
+    },
+    {
+      "epoch": 4.2138794084186575,
+      "grad_norm": 0.736415684223175,
+      "learning_rate": 0.0001583617747440273,
+      "loss": 1.1843,
+      "step": 3704
+    },
+    {
+      "epoch": 4.215017064846417,
+      "grad_norm": 1.1527396440505981,
+      "learning_rate": 0.00015813424345847553,
+      "loss": 2.0684,
+      "step": 3705
+    },
+    {
+      "epoch": 4.216154721274175,
+      "grad_norm": 1.2436633110046387,
+      "learning_rate": 0.00015790671217292377,
+      "loss": 3.2018,
+      "step": 3706
+    },
+    {
+      "epoch": 4.217292377701934,
+      "grad_norm": 0.5855286717414856,
+      "learning_rate": 0.000157679180887372,
+      "loss": 1.0727,
+      "step": 3707
+    },
+    {
+      "epoch": 4.2184300341296925,
+      "grad_norm": 1.469887375831604,
+      "learning_rate": 0.00015745164960182027,
+      "loss": 3.1109,
+      "step": 3708
+    },
+    {
+      "epoch": 4.219567690557452,
+      "grad_norm": 0.6294847130775452,
+      "learning_rate": 0.00015722411831626848,
+      "loss": 1.3746,
+      "step": 3709
+    },
+    {
+      "epoch": 4.22070534698521,
+      "grad_norm": 0.7737325429916382,
+      "learning_rate": 0.00015699658703071672,
+      "loss": 1.4918,
+      "step": 3710
+    },
+    {
+      "epoch": 4.221843003412969,
+      "grad_norm": 0.7594680786132812,
+      "learning_rate": 0.00015676905574516496,
+      "loss": 2.4034,
+      "step": 3711
+    },
+    {
+      "epoch": 4.2229806598407285,
+      "grad_norm": 0.9010804295539856,
+      "learning_rate": 0.0001565415244596132,
+      "loss": 1.1525,
+      "step": 3712
+    },
+    {
+      "epoch": 4.224118316268487,
+      "grad_norm": 1.6272693872451782,
+      "learning_rate": 0.00015631399317406146,
+      "loss": 3.6835,
+      "step": 3713
+    },
+    {
+      "epoch": 4.225255972696246,
+      "grad_norm": 0.7173957228660583,
+      "learning_rate": 0.00015608646188850967,
+      "loss": 1.3503,
+      "step": 3714
+    },
+    {
+      "epoch": 4.226393629124004,
+      "grad_norm": 0.8401118516921997,
+      "learning_rate": 0.0001558589306029579,
+      "loss": 1.1993,
+      "step": 3715
+    },
+    {
+      "epoch": 4.2275312855517635,
+      "grad_norm": 0.8311299681663513,
+      "learning_rate": 0.00015563139931740614,
+      "loss": 1.3891,
+      "step": 3716
+    },
+    {
+      "epoch": 4.228668941979522,
+      "grad_norm": 2.7499144077301025,
+      "learning_rate": 0.00015540386803185438,
+      "loss": 4.453,
+      "step": 3717
+    },
+    {
+      "epoch": 4.229806598407281,
+      "grad_norm": 0.736031174659729,
+      "learning_rate": 0.00015517633674630264,
+      "loss": 1.815,
+      "step": 3718
+    },
+    {
+      "epoch": 4.23094425483504,
+      "grad_norm": 0.7462246417999268,
+      "learning_rate": 0.00015494880546075085,
+      "loss": 1.0605,
+      "step": 3719
+    },
+    {
+      "epoch": 4.2320819112627985,
+      "grad_norm": 1.0690699815750122,
+      "learning_rate": 0.0001547212741751991,
+      "loss": 1.1685,
+      "step": 3720
+    },
+    {
+      "epoch": 4.233219567690558,
+      "grad_norm": 0.781626284122467,
+      "learning_rate": 0.00015449374288964733,
+      "loss": 0.8943,
+      "step": 3721
+    },
+    {
+      "epoch": 4.234357224118316,
+      "grad_norm": 0.9446958899497986,
+      "learning_rate": 0.00015426621160409556,
+      "loss": 1.8685,
+      "step": 3722
+    },
+    {
+      "epoch": 4.235494880546075,
+      "grad_norm": 1.0105640888214111,
+      "learning_rate": 0.00015403868031854383,
+      "loss": 2.961,
+      "step": 3723
+    },
+    {
+      "epoch": 4.236632536973834,
+      "grad_norm": 0.6021206974983215,
+      "learning_rate": 0.00015381114903299204,
+      "loss": 1.5929,
+      "step": 3724
+    },
+    {
+      "epoch": 4.237770193401593,
+      "grad_norm": 0.7067481279373169,
+      "learning_rate": 0.00015358361774744027,
+      "loss": 0.7998,
+      "step": 3725
+    },
+    {
+      "epoch": 4.238907849829351,
+      "grad_norm": 0.968774676322937,
+      "learning_rate": 0.0001533560864618885,
+      "loss": 1.5134,
+      "step": 3726
+    },
+    {
+      "epoch": 4.24004550625711,
+      "grad_norm": 0.7576475739479065,
+      "learning_rate": 0.00015312855517633675,
+      "loss": 1.3597,
+      "step": 3727
+    },
+    {
+      "epoch": 4.2411831626848695,
+      "grad_norm": 0.8560457229614258,
+      "learning_rate": 0.00015290102389078499,
+      "loss": 1.7398,
+      "step": 3728
+    },
+    {
+      "epoch": 4.242320819112628,
+      "grad_norm": 1.2698020935058594,
+      "learning_rate": 0.00015267349260523322,
+      "loss": 1.4396,
+      "step": 3729
+    },
+    {
+      "epoch": 4.243458475540387,
+      "grad_norm": 0.9575531482696533,
+      "learning_rate": 0.00015244596131968146,
+      "loss": 1.3904,
+      "step": 3730
+    },
+    {
+      "epoch": 4.244596131968145,
+      "grad_norm": 1.038975715637207,
+      "learning_rate": 0.0001522184300341297,
+      "loss": 1.6743,
+      "step": 3731
+    },
+    {
+      "epoch": 4.2457337883959045,
+      "grad_norm": 1.3596773147583008,
+      "learning_rate": 0.00015199089874857793,
+      "loss": 3.709,
+      "step": 3732
+    },
+    {
+      "epoch": 4.246871444823663,
+      "grad_norm": 0.5811980962753296,
+      "learning_rate": 0.00015176336746302617,
+      "loss": 1.1902,
+      "step": 3733
+    },
+    {
+      "epoch": 4.248009101251422,
+      "grad_norm": 0.7279687523841858,
+      "learning_rate": 0.0001515358361774744,
+      "loss": 1.28,
+      "step": 3734
+    },
+    {
+      "epoch": 4.249146757679181,
+      "grad_norm": 0.8053296208381653,
+      "learning_rate": 0.00015130830489192264,
+      "loss": 1.5716,
+      "step": 3735
+    },
+    {
+      "epoch": 4.25028441410694,
+      "grad_norm": 1.2876355648040771,
+      "learning_rate": 0.00015108077360637088,
+      "loss": 2.0535,
+      "step": 3736
+    },
+    {
+      "epoch": 4.251422070534699,
+      "grad_norm": 1.4081701040267944,
+      "learning_rate": 0.00015085324232081912,
+      "loss": 2.8273,
+      "step": 3737
+    },
+    {
+      "epoch": 4.252559726962457,
+      "grad_norm": 1.2183709144592285,
+      "learning_rate": 0.00015062571103526736,
+      "loss": 2.7009,
+      "step": 3738
+    },
+    {
+      "epoch": 4.253697383390216,
+      "grad_norm": 1.22077476978302,
+      "learning_rate": 0.0001503981797497156,
+      "loss": 2.3349,
+      "step": 3739
+    },
+    {
+      "epoch": 4.254835039817975,
+      "grad_norm": 0.965623140335083,
+      "learning_rate": 0.00015017064846416383,
+      "loss": 2.194,
+      "step": 3740
+    },
+    {
+      "epoch": 4.255972696245734,
+      "grad_norm": 0.9789025187492371,
+      "learning_rate": 0.00014994311717861207,
+      "loss": 1.3343,
+      "step": 3741
+    },
+    {
+      "epoch": 4.257110352673493,
+      "grad_norm": 0.7638621926307678,
+      "learning_rate": 0.0001497155858930603,
+      "loss": 1.2936,
+      "step": 3742
+    },
+    {
+      "epoch": 4.258248009101251,
+      "grad_norm": 0.8102353811264038,
+      "learning_rate": 0.00014948805460750854,
+      "loss": 1.6045,
+      "step": 3743
+    },
+    {
+      "epoch": 4.2593856655290105,
+      "grad_norm": 1.2423808574676514,
+      "learning_rate": 0.00014926052332195678,
+      "loss": 2.631,
+      "step": 3744
+    },
+    {
+      "epoch": 4.260523321956769,
+      "grad_norm": 1.3112363815307617,
+      "learning_rate": 0.000149032992036405,
+      "loss": 2.9128,
+      "step": 3745
+    },
+    {
+      "epoch": 4.261660978384528,
+      "grad_norm": 1.219910740852356,
+      "learning_rate": 0.00014880546075085325,
+      "loss": 2.0505,
+      "step": 3746
+    },
+    {
+      "epoch": 4.262798634812286,
+      "grad_norm": 0.9391239881515503,
+      "learning_rate": 0.0001485779294653015,
+      "loss": 1.2113,
+      "step": 3747
+    },
+    {
+      "epoch": 4.263936291240046,
+      "grad_norm": 1.231005072593689,
+      "learning_rate": 0.00014835039817974973,
+      "loss": 1.5756,
+      "step": 3748
+    },
+    {
+      "epoch": 4.265073947667805,
+      "grad_norm": 1.1906815767288208,
+      "learning_rate": 0.00014812286689419796,
+      "loss": 2.0473,
+      "step": 3749
+    },
+    {
+      "epoch": 4.266211604095563,
+      "grad_norm": 0.9939584136009216,
+      "learning_rate": 0.00014789533560864617,
+      "loss": 1.4565,
+      "step": 3750
+    },
+    {
+      "epoch": 4.267349260523322,
+      "grad_norm": 0.7417742609977722,
+      "learning_rate": 0.00014766780432309444,
+      "loss": 1.7072,
+      "step": 3751
+    },
+    {
+      "epoch": 4.268486916951081,
+      "grad_norm": 0.9343417882919312,
+      "learning_rate": 0.00014744027303754267,
+      "loss": 1.6886,
+      "step": 3752
+    },
+    {
+      "epoch": 4.26962457337884,
+      "grad_norm": 1.7899564504623413,
+      "learning_rate": 0.0001472127417519909,
+      "loss": 3.9775,
+      "step": 3753
+    },
+    {
+      "epoch": 4.270762229806598,
+      "grad_norm": 0.759251058101654,
+      "learning_rate": 0.00014698521046643915,
+      "loss": 1.273,
+      "step": 3754
+    },
+    {
+      "epoch": 4.271899886234357,
+      "grad_norm": 1.0627682209014893,
+      "learning_rate": 0.00014675767918088736,
+      "loss": 1.5295,
+      "step": 3755
+    },
+    {
+      "epoch": 4.273037542662116,
+      "grad_norm": 0.836782693862915,
+      "learning_rate": 0.00014653014789533562,
+      "loss": 1.7053,
+      "step": 3756
+    },
+    {
+      "epoch": 4.274175199089875,
+      "grad_norm": 0.8767544031143188,
+      "learning_rate": 0.00014630261660978386,
+      "loss": 1.7385,
+      "step": 3757
+    },
+    {
+      "epoch": 4.275312855517634,
+      "grad_norm": 1.1457442045211792,
+      "learning_rate": 0.0001460750853242321,
+      "loss": 2.2233,
+      "step": 3758
+    },
+    {
+      "epoch": 4.276450511945392,
+      "grad_norm": 1.1655712127685547,
+      "learning_rate": 0.00014584755403868033,
+      "loss": 1.6773,
+      "step": 3759
+    },
+    {
+      "epoch": 4.277588168373152,
+      "grad_norm": 0.9990490674972534,
+      "learning_rate": 0.00014562002275312854,
+      "loss": 1.7472,
+      "step": 3760
+    },
+    {
+      "epoch": 4.27872582480091,
+      "grad_norm": 1.2598174810409546,
+      "learning_rate": 0.0001453924914675768,
+      "loss": 2.318,
+      "step": 3761
+    },
+    {
+      "epoch": 4.279863481228669,
+      "grad_norm": 1.3087960481643677,
+      "learning_rate": 0.00014516496018202502,
+      "loss": 2.8482,
+      "step": 3762
+    },
+    {
+      "epoch": 4.281001137656427,
+      "grad_norm": 1.1006321907043457,
+      "learning_rate": 0.00014493742889647328,
+      "loss": 2.0884,
+      "step": 3763
+    },
+    {
+      "epoch": 4.282138794084187,
+      "grad_norm": 0.648513674736023,
+      "learning_rate": 0.00014470989761092152,
+      "loss": 1.5729,
+      "step": 3764
+    },
+    {
+      "epoch": 4.283276450511945,
+      "grad_norm": 0.9822235107421875,
+      "learning_rate": 0.00014448236632536973,
+      "loss": 1.5298,
+      "step": 3765
+    },
+    {
+      "epoch": 4.284414106939704,
+      "grad_norm": 0.9084975123405457,
+      "learning_rate": 0.000144254835039818,
+      "loss": 2.1044,
+      "step": 3766
+    },
+    {
+      "epoch": 4.285551763367463,
+      "grad_norm": 1.9358890056610107,
+      "learning_rate": 0.0001440273037542662,
+      "loss": 3.062,
+      "step": 3767
+    },
+    {
+      "epoch": 4.286689419795222,
+      "grad_norm": 0.8079110980033875,
+      "learning_rate": 0.00014379977246871444,
+      "loss": 1.7288,
+      "step": 3768
+    },
+    {
+      "epoch": 4.287827076222981,
+      "grad_norm": 0.6145922541618347,
+      "learning_rate": 0.0001435722411831627,
+      "loss": 1.0093,
+      "step": 3769
+    },
+    {
+      "epoch": 4.288964732650739,
+      "grad_norm": 0.9286468625068665,
+      "learning_rate": 0.0001433447098976109,
+      "loss": 1.4367,
+      "step": 3770
+    },
+    {
+      "epoch": 4.290102389078498,
+      "grad_norm": 0.7882009148597717,
+      "learning_rate": 0.00014311717861205918,
+      "loss": 1.2809,
+      "step": 3771
+    },
+    {
+      "epoch": 4.291240045506257,
+      "grad_norm": 1.0285731554031372,
+      "learning_rate": 0.00014288964732650739,
+      "loss": 2.5163,
+      "step": 3772
+    },
+    {
+      "epoch": 4.292377701934016,
+      "grad_norm": 1.0870704650878906,
+      "learning_rate": 0.00014266211604095562,
+      "loss": 1.9514,
+      "step": 3773
+    },
+    {
+      "epoch": 4.293515358361775,
+      "grad_norm": 0.9731798768043518,
+      "learning_rate": 0.0001424345847554039,
+      "loss": 1.2031,
+      "step": 3774
+    },
+    {
+      "epoch": 4.294653014789533,
+      "grad_norm": 0.6875901222229004,
+      "learning_rate": 0.0001422070534698521,
+      "loss": 0.9267,
+      "step": 3775
+    },
+    {
+      "epoch": 4.295790671217293,
+      "grad_norm": 0.746791422367096,
+      "learning_rate": 0.00014197952218430036,
+      "loss": 1.5462,
+      "step": 3776
+    },
+    {
+      "epoch": 4.296928327645051,
+      "grad_norm": 0.7123384475708008,
+      "learning_rate": 0.00014175199089874857,
+      "loss": 1.3801,
+      "step": 3777
+    },
+    {
+      "epoch": 4.29806598407281,
+      "grad_norm": 0.554326593875885,
+      "learning_rate": 0.0001415244596131968,
+      "loss": 0.6866,
+      "step": 3778
+    },
+    {
+      "epoch": 4.2992036405005685,
+      "grad_norm": 0.9146575331687927,
+      "learning_rate": 0.00014129692832764505,
+      "loss": 1.5966,
+      "step": 3779
+    },
+    {
+      "epoch": 4.300341296928328,
+      "grad_norm": 0.6338298916816711,
+      "learning_rate": 0.00014106939704209328,
+      "loss": 1.0133,
+      "step": 3780
+    },
+    {
+      "epoch": 4.301478953356087,
+      "grad_norm": 0.6664391756057739,
+      "learning_rate": 0.00014084186575654155,
+      "loss": 1.1051,
+      "step": 3781
+    },
+    {
+      "epoch": 4.302616609783845,
+      "grad_norm": 0.8975256085395813,
+      "learning_rate": 0.00014061433447098976,
+      "loss": 1.273,
+      "step": 3782
+    },
+    {
+      "epoch": 4.303754266211604,
+      "grad_norm": 1.1492512226104736,
+      "learning_rate": 0.000140386803185438,
+      "loss": 2.3168,
+      "step": 3783
+    },
+    {
+      "epoch": 4.304891922639363,
+      "grad_norm": 1.3415923118591309,
+      "learning_rate": 0.00014015927189988623,
+      "loss": 2.4386,
+      "step": 3784
+    },
+    {
+      "epoch": 4.306029579067122,
+      "grad_norm": 1.2044072151184082,
+      "learning_rate": 0.00013993174061433447,
+      "loss": 3.6453,
+      "step": 3785
+    },
+    {
+      "epoch": 4.30716723549488,
+      "grad_norm": 0.7362768650054932,
+      "learning_rate": 0.00013970420932878273,
+      "loss": 1.5149,
+      "step": 3786
+    },
+    {
+      "epoch": 4.308304891922639,
+      "grad_norm": 0.8360229134559631,
+      "learning_rate": 0.00013947667804323094,
+      "loss": 2.0411,
+      "step": 3787
+    },
+    {
+      "epoch": 4.309442548350399,
+      "grad_norm": 0.8958834409713745,
+      "learning_rate": 0.00013924914675767918,
+      "loss": 1.5186,
+      "step": 3788
+    },
+    {
+      "epoch": 4.310580204778157,
+      "grad_norm": 0.7361008524894714,
+      "learning_rate": 0.00013902161547212741,
+      "loss": 2.1691,
+      "step": 3789
+    },
+    {
+      "epoch": 4.311717861205916,
+      "grad_norm": 0.9051206111907959,
+      "learning_rate": 0.00013879408418657565,
+      "loss": 1.7265,
+      "step": 3790
+    },
+    {
+      "epoch": 4.3128555176336745,
+      "grad_norm": 1.077871561050415,
+      "learning_rate": 0.00013856655290102392,
+      "loss": 3.1665,
+      "step": 3791
+    },
+    {
+      "epoch": 4.313993174061434,
+      "grad_norm": 0.9962257742881775,
+      "learning_rate": 0.00013833902161547213,
+      "loss": 0.8876,
+      "step": 3792
+    },
+    {
+      "epoch": 4.315130830489192,
+      "grad_norm": 0.7965003848075867,
+      "learning_rate": 0.00013811149032992036,
+      "loss": 0.8467,
+      "step": 3793
+    },
+    {
+      "epoch": 4.316268486916951,
+      "grad_norm": 1.0827739238739014,
+      "learning_rate": 0.0001378839590443686,
+      "loss": 2.0453,
+      "step": 3794
+    },
+    {
+      "epoch": 4.3174061433447095,
+      "grad_norm": 1.0761308670043945,
+      "learning_rate": 0.00013765642775881684,
+      "loss": 2.2351,
+      "step": 3795
+    },
+    {
+      "epoch": 4.318543799772469,
+      "grad_norm": 0.7478647232055664,
+      "learning_rate": 0.00013742889647326507,
+      "loss": 1.662,
+      "step": 3796
+    },
+    {
+      "epoch": 4.319681456200228,
+      "grad_norm": 1.300010085105896,
+      "learning_rate": 0.0001372013651877133,
+      "loss": 2.644,
+      "step": 3797
+    },
+    {
+      "epoch": 4.320819112627986,
+      "grad_norm": 1.0593494176864624,
+      "learning_rate": 0.00013697383390216155,
+      "loss": 2.2775,
+      "step": 3798
+    },
+    {
+      "epoch": 4.321956769055745,
+      "grad_norm": 1.0908184051513672,
+      "learning_rate": 0.00013674630261660978,
+      "loss": 1.6699,
+      "step": 3799
+    },
+    {
+      "epoch": 4.323094425483504,
+      "grad_norm": 1.0574840307235718,
+      "learning_rate": 0.00013651877133105802,
+      "loss": 2.6921,
+      "step": 3800
+    },
+    {
+      "epoch": 4.324232081911263,
+      "grad_norm": 0.6852555871009827,
+      "learning_rate": 0.00013629124004550626,
+      "loss": 1.0971,
+      "step": 3801
+    },
+    {
+      "epoch": 4.325369738339021,
+      "grad_norm": 1.1705669164657593,
+      "learning_rate": 0.0001360637087599545,
+      "loss": 2.0903,
+      "step": 3802
+    },
+    {
+      "epoch": 4.3265073947667805,
+      "grad_norm": 0.9083330035209656,
+      "learning_rate": 0.00013583617747440273,
+      "loss": 2.8104,
+      "step": 3803
+    },
+    {
+      "epoch": 4.327645051194539,
+      "grad_norm": 1.045404076576233,
+      "learning_rate": 0.00013560864618885097,
+      "loss": 1.8789,
+      "step": 3804
+    },
+    {
+      "epoch": 4.328782707622298,
+      "grad_norm": 1.4782066345214844,
+      "learning_rate": 0.0001353811149032992,
+      "loss": 3.033,
+      "step": 3805
+    },
+    {
+      "epoch": 4.329920364050057,
+      "grad_norm": 0.8685378432273865,
+      "learning_rate": 0.00013515358361774744,
+      "loss": 1.8752,
+      "step": 3806
+    },
+    {
+      "epoch": 4.3310580204778155,
+      "grad_norm": 0.6213506460189819,
+      "learning_rate": 0.00013492605233219568,
+      "loss": 1.4381,
+      "step": 3807
+    },
+    {
+      "epoch": 4.332195676905575,
+      "grad_norm": 1.271410346031189,
+      "learning_rate": 0.00013469852104664392,
+      "loss": 2.2921,
+      "step": 3808
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 1.084229826927185,
+      "learning_rate": 0.00013447098976109215,
+      "loss": 2.1447,
+      "step": 3809
+    },
+    {
+      "epoch": 4.334470989761092,
+      "grad_norm": 1.2566276788711548,
+      "learning_rate": 0.0001342434584755404,
+      "loss": 2.3514,
+      "step": 3810
+    },
+    {
+      "epoch": 4.335608646188851,
+      "grad_norm": 0.6356386542320251,
+      "learning_rate": 0.00013401592718998863,
+      "loss": 0.815,
+      "step": 3811
+    },
+    {
+      "epoch": 4.33674630261661,
+      "grad_norm": 0.9741693139076233,
+      "learning_rate": 0.00013378839590443687,
+      "loss": 2.4285,
+      "step": 3812
+    },
+    {
+      "epoch": 4.337883959044369,
+      "grad_norm": 0.9355404376983643,
+      "learning_rate": 0.00013356086461888508,
+      "loss": 2.015,
+      "step": 3813
+    },
+    {
+      "epoch": 4.339021615472127,
+      "grad_norm": 0.9928659796714783,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.7225,
+      "step": 3814
+    },
+    {
+      "epoch": 4.3401592718998865,
+      "grad_norm": 0.7346398830413818,
+      "learning_rate": 0.00013310580204778158,
+      "loss": 0.8244,
+      "step": 3815
+    },
+    {
+      "epoch": 4.341296928327645,
+      "grad_norm": 0.8423483967781067,
+      "learning_rate": 0.00013287827076222981,
+      "loss": 1.5708,
+      "step": 3816
+    },
+    {
+      "epoch": 4.342434584755404,
+      "grad_norm": 1.3044344186782837,
+      "learning_rate": 0.00013265073947667805,
+      "loss": 2.0567,
+      "step": 3817
+    },
+    {
+      "epoch": 4.343572241183162,
+      "grad_norm": 0.7337809801101685,
+      "learning_rate": 0.00013242320819112626,
+      "loss": 0.9073,
+      "step": 3818
+    },
+    {
+      "epoch": 4.3447098976109215,
+      "grad_norm": 1.2432464361190796,
+      "learning_rate": 0.00013219567690557452,
+      "loss": 2.3843,
+      "step": 3819
+    },
+    {
+      "epoch": 4.345847554038681,
+      "grad_norm": 0.8972126245498657,
+      "learning_rate": 0.00013196814562002276,
+      "loss": 1.5984,
+      "step": 3820
+    },
+    {
+      "epoch": 4.346985210466439,
+      "grad_norm": 0.7023994326591492,
+      "learning_rate": 0.000131740614334471,
+      "loss": 1.7012,
+      "step": 3821
+    },
+    {
+      "epoch": 4.348122866894198,
+      "grad_norm": 0.857025146484375,
+      "learning_rate": 0.00013151308304891924,
+      "loss": 1.8499,
+      "step": 3822
+    },
+    {
+      "epoch": 4.349260523321957,
+      "grad_norm": 0.8133417963981628,
+      "learning_rate": 0.00013128555176336745,
+      "loss": 1.8394,
+      "step": 3823
+    },
+    {
+      "epoch": 4.350398179749716,
+      "grad_norm": 1.043709635734558,
+      "learning_rate": 0.0001310580204778157,
+      "loss": 1.6459,
+      "step": 3824
+    },
+    {
+      "epoch": 4.351535836177474,
+      "grad_norm": 0.9391498565673828,
+      "learning_rate": 0.00013083048919226395,
+      "loss": 1.4526,
+      "step": 3825
+    },
+    {
+      "epoch": 4.352673492605233,
+      "grad_norm": 1.0450690984725952,
+      "learning_rate": 0.00013060295790671218,
+      "loss": 1.5307,
+      "step": 3826
+    },
+    {
+      "epoch": 4.3538111490329925,
+      "grad_norm": 1.0160549879074097,
+      "learning_rate": 0.00013037542662116042,
+      "loss": 2.3228,
+      "step": 3827
+    },
+    {
+      "epoch": 4.354948805460751,
+      "grad_norm": 1.861527919769287,
+      "learning_rate": 0.00013014789533560863,
+      "loss": 2.6769,
+      "step": 3828
+    },
+    {
+      "epoch": 4.35608646188851,
+      "grad_norm": 0.8127307295799255,
+      "learning_rate": 0.0001299203640500569,
+      "loss": 1.3464,
+      "step": 3829
+    },
+    {
+      "epoch": 4.357224118316268,
+      "grad_norm": 0.8189122080802917,
+      "learning_rate": 0.0001296928327645051,
+      "loss": 1.2768,
+      "step": 3830
+    },
+    {
+      "epoch": 4.3583617747440275,
+      "grad_norm": 0.8693903088569641,
+      "learning_rate": 0.00012946530147895337,
+      "loss": 2.2173,
+      "step": 3831
+    },
+    {
+      "epoch": 4.359499431171786,
+      "grad_norm": 1.17019522190094,
+      "learning_rate": 0.0001292377701934016,
+      "loss": 2.2724,
+      "step": 3832
+    },
+    {
+      "epoch": 4.360637087599545,
+      "grad_norm": 0.7961968183517456,
+      "learning_rate": 0.00012901023890784982,
+      "loss": 0.8773,
+      "step": 3833
+    },
+    {
+      "epoch": 4.361774744027303,
+      "grad_norm": 0.9709134697914124,
+      "learning_rate": 0.00012878270762229808,
+      "loss": 1.8771,
+      "step": 3834
+    },
+    {
+      "epoch": 4.362912400455063,
+      "grad_norm": 1.0111744403839111,
+      "learning_rate": 0.0001285551763367463,
+      "loss": 1.7679,
+      "step": 3835
+    },
+    {
+      "epoch": 4.364050056882822,
+      "grad_norm": 1.0463035106658936,
+      "learning_rate": 0.00012832764505119455,
+      "loss": 1.8787,
+      "step": 3836
+    },
+    {
+      "epoch": 4.36518771331058,
+      "grad_norm": 0.9706122875213623,
+      "learning_rate": 0.0001281001137656428,
+      "loss": 1.8892,
+      "step": 3837
+    },
+    {
+      "epoch": 4.366325369738339,
+      "grad_norm": 1.487461805343628,
+      "learning_rate": 0.000127872582480091,
+      "loss": 3.0566,
+      "step": 3838
+    },
+    {
+      "epoch": 4.367463026166098,
+      "grad_norm": 0.578607976436615,
+      "learning_rate": 0.00012764505119453926,
+      "loss": 1.1034,
+      "step": 3839
+    },
+    {
+      "epoch": 4.368600682593857,
+      "grad_norm": 0.8049798607826233,
+      "learning_rate": 0.00012741751990898747,
+      "loss": 1.4093,
+      "step": 3840
+    },
+    {
+      "epoch": 4.369738339021615,
+      "grad_norm": 0.9491667747497559,
+      "learning_rate": 0.00012718998862343574,
+      "loss": 1.2235,
+      "step": 3841
+    },
+    {
+      "epoch": 4.370875995449374,
+      "grad_norm": 0.6280384063720703,
+      "learning_rate": 0.00012696245733788398,
+      "loss": 1.5624,
+      "step": 3842
+    },
+    {
+      "epoch": 4.372013651877133,
+      "grad_norm": 1.2607163190841675,
+      "learning_rate": 0.00012673492605233219,
+      "loss": 2.3174,
+      "step": 3843
+    },
+    {
+      "epoch": 4.373151308304892,
+      "grad_norm": 0.9251709580421448,
+      "learning_rate": 0.00012650739476678045,
+      "loss": 1.8042,
+      "step": 3844
+    },
+    {
+      "epoch": 4.374288964732651,
+      "grad_norm": 1.0482347011566162,
+      "learning_rate": 0.00012627986348122866,
+      "loss": 1.7866,
+      "step": 3845
+    },
+    {
+      "epoch": 4.375426621160409,
+      "grad_norm": 1.0927660465240479,
+      "learning_rate": 0.00012605233219567692,
+      "loss": 2.0026,
+      "step": 3846
+    },
+    {
+      "epoch": 4.376564277588169,
+      "grad_norm": 0.9109655618667603,
+      "learning_rate": 0.00012582480091012513,
+      "loss": 1.6597,
+      "step": 3847
+    },
+    {
+      "epoch": 4.377701934015927,
+      "grad_norm": 1.1354659795761108,
+      "learning_rate": 0.00012559726962457337,
+      "loss": 2.0034,
+      "step": 3848
+    },
+    {
+      "epoch": 4.378839590443686,
+      "grad_norm": 0.624963641166687,
+      "learning_rate": 0.00012536973833902163,
+      "loss": 1.1846,
+      "step": 3849
+    },
+    {
+      "epoch": 4.379977246871444,
+      "grad_norm": 0.7862791419029236,
+      "learning_rate": 0.00012514220705346984,
+      "loss": 1.0518,
+      "step": 3850
+    },
+    {
+      "epoch": 4.381114903299204,
+      "grad_norm": 1.0999188423156738,
+      "learning_rate": 0.0001249146757679181,
+      "loss": 3.3598,
+      "step": 3851
+    },
+    {
+      "epoch": 4.382252559726963,
+      "grad_norm": 1.0978202819824219,
+      "learning_rate": 0.00012468714448236632,
+      "loss": 1.4192,
+      "step": 3852
+    },
+    {
+      "epoch": 4.383390216154721,
+      "grad_norm": 0.8978061079978943,
+      "learning_rate": 0.00012445961319681456,
+      "loss": 1.5036,
+      "step": 3853
+    },
+    {
+      "epoch": 4.38452787258248,
+      "grad_norm": 0.7772718071937561,
+      "learning_rate": 0.0001242320819112628,
+      "loss": 1.9559,
+      "step": 3854
+    },
+    {
+      "epoch": 4.385665529010239,
+      "grad_norm": 0.6393569111824036,
+      "learning_rate": 0.00012400455062571103,
+      "loss": 1.2368,
+      "step": 3855
+    },
+    {
+      "epoch": 4.386803185437998,
+      "grad_norm": 1.3224518299102783,
+      "learning_rate": 0.0001237770193401593,
+      "loss": 2.4815,
+      "step": 3856
+    },
+    {
+      "epoch": 4.387940841865756,
+      "grad_norm": 0.8100547194480896,
+      "learning_rate": 0.0001235494880546075,
+      "loss": 2.0509,
+      "step": 3857
+    },
+    {
+      "epoch": 4.389078498293515,
+      "grad_norm": 1.1932857036590576,
+      "learning_rate": 0.00012332195676905574,
+      "loss": 2.3144,
+      "step": 3858
+    },
+    {
+      "epoch": 4.390216154721275,
+      "grad_norm": 1.004632592201233,
+      "learning_rate": 0.00012309442548350398,
+      "loss": 1.9511,
+      "step": 3859
+    },
+    {
+      "epoch": 4.391353811149033,
+      "grad_norm": 0.73143070936203,
+      "learning_rate": 0.00012286689419795221,
+      "loss": 1.5956,
+      "step": 3860
+    },
+    {
+      "epoch": 4.392491467576792,
+      "grad_norm": 0.8859034180641174,
+      "learning_rate": 0.00012263936291240045,
+      "loss": 1.6946,
+      "step": 3861
+    },
+    {
+      "epoch": 4.39362912400455,
+      "grad_norm": 1.6607998609542847,
+      "learning_rate": 0.0001224118316268487,
+      "loss": 3.6334,
+      "step": 3862
+    },
+    {
+      "epoch": 4.39476678043231,
+      "grad_norm": 0.5776720643043518,
+      "learning_rate": 0.00012218430034129693,
+      "loss": 0.8281,
+      "step": 3863
+    },
+    {
+      "epoch": 4.395904436860068,
+      "grad_norm": 0.8924266695976257,
+      "learning_rate": 0.00012195676905574516,
+      "loss": 1.7758,
+      "step": 3864
+    },
+    {
+      "epoch": 4.397042093287827,
+      "grad_norm": 0.5048301815986633,
+      "learning_rate": 0.0001217292377701934,
+      "loss": 0.5899,
+      "step": 3865
+    },
+    {
+      "epoch": 4.398179749715586,
+      "grad_norm": 0.9036781191825867,
+      "learning_rate": 0.00012150170648464164,
+      "loss": 1.7569,
+      "step": 3866
+    },
+    {
+      "epoch": 4.399317406143345,
+      "grad_norm": 1.0674494504928589,
+      "learning_rate": 0.00012127417519908989,
+      "loss": 2.3646,
+      "step": 3867
+    },
+    {
+      "epoch": 4.400455062571104,
+      "grad_norm": 1.6074708700180054,
+      "learning_rate": 0.00012104664391353812,
+      "loss": 2.0718,
+      "step": 3868
+    },
+    {
+      "epoch": 4.401592718998862,
+      "grad_norm": 1.289448618888855,
+      "learning_rate": 0.00012081911262798635,
+      "loss": 2.4961,
+      "step": 3869
+    },
+    {
+      "epoch": 4.402730375426621,
+      "grad_norm": 1.5797637701034546,
+      "learning_rate": 0.00012059158134243458,
+      "loss": 2.0542,
+      "step": 3870
+    },
+    {
+      "epoch": 4.40386803185438,
+      "grad_norm": 0.8559139370918274,
+      "learning_rate": 0.00012036405005688282,
+      "loss": 1.8854,
+      "step": 3871
+    },
+    {
+      "epoch": 4.405005688282139,
+      "grad_norm": 0.749015748500824,
+      "learning_rate": 0.00012013651877133106,
+      "loss": 1.3702,
+      "step": 3872
+    },
+    {
+      "epoch": 4.406143344709897,
+      "grad_norm": 0.95188969373703,
+      "learning_rate": 0.00011990898748577931,
+      "loss": 1.6222,
+      "step": 3873
+    },
+    {
+      "epoch": 4.407281001137656,
+      "grad_norm": 0.8665216565132141,
+      "learning_rate": 0.00011968145620022753,
+      "loss": 1.9693,
+      "step": 3874
+    },
+    {
+      "epoch": 4.408418657565416,
+      "grad_norm": 0.7548373937606812,
+      "learning_rate": 0.00011945392491467577,
+      "loss": 2.2095,
+      "step": 3875
+    },
+    {
+      "epoch": 4.409556313993174,
+      "grad_norm": 0.7718438506126404,
+      "learning_rate": 0.000119226393629124,
+      "loss": 1.0631,
+      "step": 3876
+    },
+    {
+      "epoch": 4.410693970420933,
+      "grad_norm": 0.7795398235321045,
+      "learning_rate": 0.00011899886234357224,
+      "loss": 1.5037,
+      "step": 3877
+    },
+    {
+      "epoch": 4.4118316268486915,
+      "grad_norm": 0.9634506702423096,
+      "learning_rate": 0.00011877133105802048,
+      "loss": 1.2737,
+      "step": 3878
+    },
+    {
+      "epoch": 4.412969283276451,
+      "grad_norm": 1.0782710313796997,
+      "learning_rate": 0.00011854379977246872,
+      "loss": 2.0432,
+      "step": 3879
+    },
+    {
+      "epoch": 4.414106939704209,
+      "grad_norm": 0.8432207107543945,
+      "learning_rate": 0.00011831626848691695,
+      "loss": 1.4583,
+      "step": 3880
+    },
+    {
+      "epoch": 4.415244596131968,
+      "grad_norm": 1.0028852224349976,
+      "learning_rate": 0.00011808873720136519,
+      "loss": 1.3358,
+      "step": 3881
+    },
+    {
+      "epoch": 4.4163822525597265,
+      "grad_norm": 0.9383002519607544,
+      "learning_rate": 0.00011786120591581343,
+      "loss": 1.9518,
+      "step": 3882
+    },
+    {
+      "epoch": 4.417519908987486,
+      "grad_norm": 0.7450307607650757,
+      "learning_rate": 0.00011763367463026167,
+      "loss": 0.7314,
+      "step": 3883
+    },
+    {
+      "epoch": 4.418657565415245,
+      "grad_norm": 0.8783142566680908,
+      "learning_rate": 0.0001174061433447099,
+      "loss": 1.8254,
+      "step": 3884
+    },
+    {
+      "epoch": 4.419795221843003,
+      "grad_norm": 0.7310676574707031,
+      "learning_rate": 0.00011717861205915814,
+      "loss": 1.7363,
+      "step": 3885
+    },
+    {
+      "epoch": 4.420932878270762,
+      "grad_norm": 0.4681646525859833,
+      "learning_rate": 0.00011695108077360638,
+      "loss": 1.0317,
+      "step": 3886
+    },
+    {
+      "epoch": 4.422070534698521,
+      "grad_norm": 0.9998228549957275,
+      "learning_rate": 0.00011672354948805461,
+      "loss": 1.5986,
+      "step": 3887
+    },
+    {
+      "epoch": 4.42320819112628,
+      "grad_norm": 0.6129392981529236,
+      "learning_rate": 0.00011649601820250284,
+      "loss": 0.952,
+      "step": 3888
+    },
+    {
+      "epoch": 4.424345847554038,
+      "grad_norm": 0.8074744343757629,
+      "learning_rate": 0.00011626848691695107,
+      "loss": 1.1051,
+      "step": 3889
+    },
+    {
+      "epoch": 4.4254835039817975,
+      "grad_norm": 0.9240376949310303,
+      "learning_rate": 0.00011604095563139932,
+      "loss": 1.1525,
+      "step": 3890
+    },
+    {
+      "epoch": 4.426621160409557,
+      "grad_norm": 0.9832790493965149,
+      "learning_rate": 0.00011581342434584756,
+      "loss": 1.538,
+      "step": 3891
+    },
+    {
+      "epoch": 4.427758816837315,
+      "grad_norm": 1.5793112516403198,
+      "learning_rate": 0.0001155858930602958,
+      "loss": 2.8963,
+      "step": 3892
+    },
+    {
+      "epoch": 4.428896473265074,
+      "grad_norm": 0.6335533261299133,
+      "learning_rate": 0.00011535836177474402,
+      "loss": 0.7007,
+      "step": 3893
+    },
+    {
+      "epoch": 4.4300341296928325,
+      "grad_norm": 0.8912391066551208,
+      "learning_rate": 0.00011513083048919226,
+      "loss": 1.6569,
+      "step": 3894
+    },
+    {
+      "epoch": 4.431171786120592,
+      "grad_norm": 1.0746057033538818,
+      "learning_rate": 0.0001149032992036405,
+      "loss": 1.9166,
+      "step": 3895
+    },
+    {
+      "epoch": 4.43230944254835,
+      "grad_norm": 0.8969407677650452,
+      "learning_rate": 0.00011467576791808875,
+      "loss": 1.823,
+      "step": 3896
+    },
+    {
+      "epoch": 4.433447098976109,
+      "grad_norm": 1.103092074394226,
+      "learning_rate": 0.00011444823663253698,
+      "loss": 2.2962,
+      "step": 3897
+    },
+    {
+      "epoch": 4.434584755403868,
+      "grad_norm": 0.9207330942153931,
+      "learning_rate": 0.0001142207053469852,
+      "loss": 1.9893,
+      "step": 3898
+    },
+    {
+      "epoch": 4.435722411831627,
+      "grad_norm": 0.7149428725242615,
+      "learning_rate": 0.00011399317406143344,
+      "loss": 1.9576,
+      "step": 3899
+    },
+    {
+      "epoch": 4.436860068259386,
+      "grad_norm": 1.1259676218032837,
+      "learning_rate": 0.00011376564277588168,
+      "loss": 1.7732,
+      "step": 3900
+    },
+    {
+      "epoch": 4.437997724687144,
+      "grad_norm": 1.0312747955322266,
+      "learning_rate": 0.00011353811149032993,
+      "loss": 2.2845,
+      "step": 3901
+    },
+    {
+      "epoch": 4.4391353811149035,
+      "grad_norm": 1.1275362968444824,
+      "learning_rate": 0.00011331058020477817,
+      "loss": 2.7354,
+      "step": 3902
+    },
+    {
+      "epoch": 4.440273037542662,
+      "grad_norm": 0.8789055943489075,
+      "learning_rate": 0.00011308304891922639,
+      "loss": 2.0587,
+      "step": 3903
+    },
+    {
+      "epoch": 4.441410693970421,
+      "grad_norm": 2.0354743003845215,
+      "learning_rate": 0.00011285551763367463,
+      "loss": 2.5204,
+      "step": 3904
+    },
+    {
+      "epoch": 4.44254835039818,
+      "grad_norm": 1.1655513048171997,
+      "learning_rate": 0.00011262798634812287,
+      "loss": 1.7625,
+      "step": 3905
+    },
+    {
+      "epoch": 4.4436860068259385,
+      "grad_norm": 0.7838327288627625,
+      "learning_rate": 0.0001124004550625711,
+      "loss": 1.7538,
+      "step": 3906
+    },
+    {
+      "epoch": 4.444823663253698,
+      "grad_norm": 1.042612075805664,
+      "learning_rate": 0.00011217292377701935,
+      "loss": 1.9573,
+      "step": 3907
+    },
+    {
+      "epoch": 4.445961319681456,
+      "grad_norm": 0.8811992406845093,
+      "learning_rate": 0.00011194539249146758,
+      "loss": 1.237,
+      "step": 3908
+    },
+    {
+      "epoch": 4.447098976109215,
+      "grad_norm": 0.8946145176887512,
+      "learning_rate": 0.00011171786120591581,
+      "loss": 1.7659,
+      "step": 3909
+    },
+    {
+      "epoch": 4.4482366325369735,
+      "grad_norm": 0.6377555727958679,
+      "learning_rate": 0.00011149032992036405,
+      "loss": 1.1502,
+      "step": 3910
+    },
+    {
+      "epoch": 4.449374288964733,
+      "grad_norm": 1.0547488927841187,
+      "learning_rate": 0.00011126279863481229,
+      "loss": 1.3714,
+      "step": 3911
+    },
+    {
+      "epoch": 4.450511945392491,
+      "grad_norm": 0.7710375785827637,
+      "learning_rate": 0.00011103526734926054,
+      "loss": 1.2559,
+      "step": 3912
+    },
+    {
+      "epoch": 4.45164960182025,
+      "grad_norm": 0.5346420407295227,
+      "learning_rate": 0.00011080773606370876,
+      "loss": 0.5416,
+      "step": 3913
+    },
+    {
+      "epoch": 4.4527872582480095,
+      "grad_norm": 0.8821654915809631,
+      "learning_rate": 0.000110580204778157,
+      "loss": 1.1061,
+      "step": 3914
+    },
+    {
+      "epoch": 4.453924914675768,
+      "grad_norm": 2.041219472885132,
+      "learning_rate": 0.00011035267349260524,
+      "loss": 3.7444,
+      "step": 3915
+    },
+    {
+      "epoch": 4.455062571103527,
+      "grad_norm": 0.9593986868858337,
+      "learning_rate": 0.00011012514220705347,
+      "loss": 1.6756,
+      "step": 3916
+    },
+    {
+      "epoch": 4.456200227531285,
+      "grad_norm": 0.8779603838920593,
+      "learning_rate": 0.00010989761092150171,
+      "loss": 2.0283,
+      "step": 3917
+    },
+    {
+      "epoch": 4.4573378839590445,
+      "grad_norm": 1.146952748298645,
+      "learning_rate": 0.00010967007963594995,
+      "loss": 2.2362,
+      "step": 3918
+    },
+    {
+      "epoch": 4.458475540386803,
+      "grad_norm": 0.8655639290809631,
+      "learning_rate": 0.00010944254835039818,
+      "loss": 1.9992,
+      "step": 3919
+    },
+    {
+      "epoch": 4.459613196814562,
+      "grad_norm": 1.1940038204193115,
+      "learning_rate": 0.00010921501706484642,
+      "loss": 1.7449,
+      "step": 3920
+    },
+    {
+      "epoch": 4.460750853242321,
+      "grad_norm": 0.7749353647232056,
+      "learning_rate": 0.00010898748577929466,
+      "loss": 1.1885,
+      "step": 3921
+    },
+    {
+      "epoch": 4.4618885096700796,
+      "grad_norm": 0.8721646666526794,
+      "learning_rate": 0.0001087599544937429,
+      "loss": 1.8104,
+      "step": 3922
+    },
+    {
+      "epoch": 4.463026166097839,
+      "grad_norm": 0.6543747782707214,
+      "learning_rate": 0.00010853242320819112,
+      "loss": 1.2385,
+      "step": 3923
+    },
+    {
+      "epoch": 4.464163822525597,
+      "grad_norm": 0.9697179198265076,
+      "learning_rate": 0.00010830489192263937,
+      "loss": 1.6976,
+      "step": 3924
+    },
+    {
+      "epoch": 4.465301478953356,
+      "grad_norm": 0.9224042892456055,
+      "learning_rate": 0.0001080773606370876,
+      "loss": 1.214,
+      "step": 3925
+    },
+    {
+      "epoch": 4.466439135381115,
+      "grad_norm": 0.9176324605941772,
+      "learning_rate": 0.00010784982935153584,
+      "loss": 1.2962,
+      "step": 3926
+    },
+    {
+      "epoch": 4.467576791808874,
+      "grad_norm": 0.9372009038925171,
+      "learning_rate": 0.00010762229806598408,
+      "loss": 1.9831,
+      "step": 3927
+    },
+    {
+      "epoch": 4.468714448236632,
+      "grad_norm": 1.1387556791305542,
+      "learning_rate": 0.0001073947667804323,
+      "loss": 1.9173,
+      "step": 3928
+    },
+    {
+      "epoch": 4.469852104664391,
+      "grad_norm": 1.1000502109527588,
+      "learning_rate": 0.00010716723549488055,
+      "loss": 1.9951,
+      "step": 3929
+    },
+    {
+      "epoch": 4.4709897610921505,
+      "grad_norm": 0.8819757699966431,
+      "learning_rate": 0.00010693970420932879,
+      "loss": 1.3953,
+      "step": 3930
+    },
+    {
+      "epoch": 4.472127417519909,
+      "grad_norm": 0.9281446933746338,
+      "learning_rate": 0.00010671217292377703,
+      "loss": 1.3608,
+      "step": 3931
+    },
+    {
+      "epoch": 4.473265073947668,
+      "grad_norm": 1.0403672456741333,
+      "learning_rate": 0.00010648464163822526,
+      "loss": 2.4958,
+      "step": 3932
+    },
+    {
+      "epoch": 4.474402730375426,
+      "grad_norm": 0.6432840824127197,
+      "learning_rate": 0.00010625711035267349,
+      "loss": 1.0788,
+      "step": 3933
+    },
+    {
+      "epoch": 4.4755403868031856,
+      "grad_norm": 0.8006333112716675,
+      "learning_rate": 0.00010602957906712172,
+      "loss": 1.5623,
+      "step": 3934
+    },
+    {
+      "epoch": 4.476678043230944,
+      "grad_norm": 0.8013043403625488,
+      "learning_rate": 0.00010580204778156998,
+      "loss": 1.5541,
+      "step": 3935
+    },
+    {
+      "epoch": 4.477815699658703,
+      "grad_norm": 1.071979284286499,
+      "learning_rate": 0.00010557451649601821,
+      "loss": 3.0613,
+      "step": 3936
+    },
+    {
+      "epoch": 4.478953356086462,
+      "grad_norm": 0.6907356381416321,
+      "learning_rate": 0.00010534698521046644,
+      "loss": 1.8117,
+      "step": 3937
+    },
+    {
+      "epoch": 4.480091012514221,
+      "grad_norm": 0.7573530077934265,
+      "learning_rate": 0.00010511945392491467,
+      "loss": 1.4187,
+      "step": 3938
+    },
+    {
+      "epoch": 4.48122866894198,
+      "grad_norm": 0.7063055038452148,
+      "learning_rate": 0.00010489192263936291,
+      "loss": 0.7842,
+      "step": 3939
+    },
+    {
+      "epoch": 4.482366325369738,
+      "grad_norm": 0.9287145733833313,
+      "learning_rate": 0.00010466439135381115,
+      "loss": 1.7274,
+      "step": 3940
+    },
+    {
+      "epoch": 4.483503981797497,
+      "grad_norm": 0.7556503415107727,
+      "learning_rate": 0.0001044368600682594,
+      "loss": 1.4371,
+      "step": 3941
+    },
+    {
+      "epoch": 4.484641638225256,
+      "grad_norm": 1.12169349193573,
+      "learning_rate": 0.00010420932878270762,
+      "loss": 1.4631,
+      "step": 3942
+    },
+    {
+      "epoch": 4.485779294653015,
+      "grad_norm": 1.066739797592163,
+      "learning_rate": 0.00010398179749715586,
+      "loss": 2.0575,
+      "step": 3943
+    },
+    {
+      "epoch": 4.486916951080774,
+      "grad_norm": 0.6979597210884094,
+      "learning_rate": 0.0001037542662116041,
+      "loss": 1.2034,
+      "step": 3944
+    },
+    {
+      "epoch": 4.488054607508532,
+      "grad_norm": 0.9167845249176025,
+      "learning_rate": 0.00010352673492605233,
+      "loss": 1.3631,
+      "step": 3945
+    },
+    {
+      "epoch": 4.489192263936292,
+      "grad_norm": 1.0211360454559326,
+      "learning_rate": 0.00010329920364050058,
+      "loss": 2.8476,
+      "step": 3946
+    },
+    {
+      "epoch": 4.49032992036405,
+      "grad_norm": 1.0867899656295776,
+      "learning_rate": 0.0001030716723549488,
+      "loss": 2.0269,
+      "step": 3947
+    },
+    {
+      "epoch": 4.491467576791809,
+      "grad_norm": 0.963070809841156,
+      "learning_rate": 0.00010284414106939704,
+      "loss": 1.9714,
+      "step": 3948
+    },
+    {
+      "epoch": 4.492605233219567,
+      "grad_norm": 1.0459845066070557,
+      "learning_rate": 0.00010261660978384528,
+      "loss": 1.8135,
+      "step": 3949
+    },
+    {
+      "epoch": 4.493742889647327,
+      "grad_norm": 1.2111842632293701,
+      "learning_rate": 0.00010238907849829352,
+      "loss": 2.337,
+      "step": 3950
+    },
+    {
+      "epoch": 4.494880546075085,
+      "grad_norm": 0.9683245420455933,
+      "learning_rate": 0.00010216154721274175,
+      "loss": 1.2889,
+      "step": 3951
+    },
+    {
+      "epoch": 4.496018202502844,
+      "grad_norm": 0.9261424541473389,
+      "learning_rate": 0.00010193401592718999,
+      "loss": 1.5083,
+      "step": 3952
+    },
+    {
+      "epoch": 4.497155858930603,
+      "grad_norm": 0.6647104620933533,
+      "learning_rate": 0.00010170648464163823,
+      "loss": 1.3438,
+      "step": 3953
+    },
+    {
+      "epoch": 4.498293515358362,
+      "grad_norm": 0.7070716023445129,
+      "learning_rate": 0.00010147895335608646,
+      "loss": 1.2005,
+      "step": 3954
+    },
+    {
+      "epoch": 4.499431171786121,
+      "grad_norm": 0.6603410840034485,
+      "learning_rate": 0.0001012514220705347,
+      "loss": 0.869,
+      "step": 3955
+    },
+    {
+      "epoch": 4.500568828213879,
+      "grad_norm": 1.0688494443893433,
+      "learning_rate": 0.00010102389078498294,
+      "loss": 1.8926,
+      "step": 3956
+    },
+    {
+      "epoch": 4.501706484641638,
+      "grad_norm": 0.7224915027618408,
+      "learning_rate": 0.00010079635949943116,
+      "loss": 1.9618,
+      "step": 3957
+    },
+    {
+      "epoch": 4.502844141069397,
+      "grad_norm": 1.0792325735092163,
+      "learning_rate": 0.00010056882821387941,
+      "loss": 1.9267,
+      "step": 3958
+    },
+    {
+      "epoch": 4.503981797497156,
+      "grad_norm": 0.7390668392181396,
+      "learning_rate": 0.00010034129692832765,
+      "loss": 1.4238,
+      "step": 3959
+    },
+    {
+      "epoch": 4.505119453924914,
+      "grad_norm": 1.3947380781173706,
+      "learning_rate": 0.00010011376564277589,
+      "loss": 2.3192,
+      "step": 3960
+    },
+    {
+      "epoch": 4.506257110352673,
+      "grad_norm": 0.9918806552886963,
+      "learning_rate": 9.988623435722412e-05,
+      "loss": 1.6336,
+      "step": 3961
+    },
+    {
+      "epoch": 4.507394766780433,
+      "grad_norm": 1.1548150777816772,
+      "learning_rate": 9.965870307167235e-05,
+      "loss": 1.2732,
+      "step": 3962
+    },
+    {
+      "epoch": 4.508532423208191,
+      "grad_norm": 1.110249400138855,
+      "learning_rate": 9.94311717861206e-05,
+      "loss": 1.8542,
+      "step": 3963
+    },
+    {
+      "epoch": 4.50967007963595,
+      "grad_norm": 0.9127480387687683,
+      "learning_rate": 9.920364050056883e-05,
+      "loss": 1.477,
+      "step": 3964
+    },
+    {
+      "epoch": 4.510807736063708,
+      "grad_norm": 1.161213755607605,
+      "learning_rate": 9.897610921501707e-05,
+      "loss": 1.8094,
+      "step": 3965
+    },
+    {
+      "epoch": 4.511945392491468,
+      "grad_norm": 1.106597661972046,
+      "learning_rate": 9.874857792946531e-05,
+      "loss": 1.9764,
+      "step": 3966
+    },
+    {
+      "epoch": 4.513083048919226,
+      "grad_norm": 0.7442317008972168,
+      "learning_rate": 9.852104664391353e-05,
+      "loss": 1.4526,
+      "step": 3967
+    },
+    {
+      "epoch": 4.514220705346985,
+      "grad_norm": 0.7829383015632629,
+      "learning_rate": 9.829351535836177e-05,
+      "loss": 2.0345,
+      "step": 3968
+    },
+    {
+      "epoch": 4.515358361774744,
+      "grad_norm": 1.1125015020370483,
+      "learning_rate": 9.806598407281002e-05,
+      "loss": 2.6904,
+      "step": 3969
+    },
+    {
+      "epoch": 4.516496018202503,
+      "grad_norm": 0.7292245626449585,
+      "learning_rate": 9.783845278725826e-05,
+      "loss": 1.323,
+      "step": 3970
+    },
+    {
+      "epoch": 4.517633674630262,
+      "grad_norm": 1.535749912261963,
+      "learning_rate": 9.761092150170649e-05,
+      "loss": 3.5423,
+      "step": 3971
+    },
+    {
+      "epoch": 4.51877133105802,
+      "grad_norm": 0.9439961314201355,
+      "learning_rate": 9.738339021615472e-05,
+      "loss": 3.0287,
+      "step": 3972
+    },
+    {
+      "epoch": 4.519908987485779,
+      "grad_norm": 0.8544148206710815,
+      "learning_rate": 9.715585893060295e-05,
+      "loss": 1.7422,
+      "step": 3973
+    },
+    {
+      "epoch": 4.521046643913538,
+      "grad_norm": 0.7354198694229126,
+      "learning_rate": 9.692832764505119e-05,
+      "loss": 1.4524,
+      "step": 3974
+    },
+    {
+      "epoch": 4.522184300341297,
+      "grad_norm": 0.7296625375747681,
+      "learning_rate": 9.670079635949944e-05,
+      "loss": 1.3034,
+      "step": 3975
+    },
+    {
+      "epoch": 4.523321956769056,
+      "grad_norm": 1.0089813470840454,
+      "learning_rate": 9.647326507394768e-05,
+      "loss": 1.5669,
+      "step": 3976
+    },
+    {
+      "epoch": 4.5244596131968144,
+      "grad_norm": 1.005476474761963,
+      "learning_rate": 9.62457337883959e-05,
+      "loss": 1.4951,
+      "step": 3977
+    },
+    {
+      "epoch": 4.525597269624574,
+      "grad_norm": 0.7817225456237793,
+      "learning_rate": 9.601820250284414e-05,
+      "loss": 1.6755,
+      "step": 3978
+    },
+    {
+      "epoch": 4.526734926052332,
+      "grad_norm": 0.814610481262207,
+      "learning_rate": 9.579067121729238e-05,
+      "loss": 1.0051,
+      "step": 3979
+    },
+    {
+      "epoch": 4.527872582480091,
+      "grad_norm": 0.869544267654419,
+      "learning_rate": 9.556313993174063e-05,
+      "loss": 1.5352,
+      "step": 3980
+    },
+    {
+      "epoch": 4.5290102389078495,
+      "grad_norm": 0.805802047252655,
+      "learning_rate": 9.533560864618886e-05,
+      "loss": 1.9839,
+      "step": 3981
+    },
+    {
+      "epoch": 4.530147895335609,
+      "grad_norm": 0.7213151454925537,
+      "learning_rate": 9.510807736063709e-05,
+      "loss": 2.1419,
+      "step": 3982
+    },
+    {
+      "epoch": 4.531285551763368,
+      "grad_norm": 0.743634045124054,
+      "learning_rate": 9.488054607508532e-05,
+      "loss": 1.2667,
+      "step": 3983
+    },
+    {
+      "epoch": 4.532423208191126,
+      "grad_norm": 0.8204901218414307,
+      "learning_rate": 9.465301478953356e-05,
+      "loss": 1.5102,
+      "step": 3984
+    },
+    {
+      "epoch": 4.533560864618885,
+      "grad_norm": 0.902324378490448,
+      "learning_rate": 9.44254835039818e-05,
+      "loss": 1.2913,
+      "step": 3985
+    },
+    {
+      "epoch": 4.534698521046644,
+      "grad_norm": 1.2690314054489136,
+      "learning_rate": 9.419795221843005e-05,
+      "loss": 3.7403,
+      "step": 3986
+    },
+    {
+      "epoch": 4.535836177474403,
+      "grad_norm": 0.6763675212860107,
+      "learning_rate": 9.397042093287827e-05,
+      "loss": 0.8402,
+      "step": 3987
+    },
+    {
+      "epoch": 4.536973833902161,
+      "grad_norm": 0.6812355518341064,
+      "learning_rate": 9.374288964732651e-05,
+      "loss": 0.8881,
+      "step": 3988
+    },
+    {
+      "epoch": 4.5381114903299204,
+      "grad_norm": 0.7160151600837708,
+      "learning_rate": 9.351535836177475e-05,
+      "loss": 1.3187,
+      "step": 3989
+    },
+    {
+      "epoch": 4.53924914675768,
+      "grad_norm": 0.6161930561065674,
+      "learning_rate": 9.328782707622298e-05,
+      "loss": 1.0509,
+      "step": 3990
+    },
+    {
+      "epoch": 4.540386803185438,
+      "grad_norm": 0.9480435848236084,
+      "learning_rate": 9.306029579067122e-05,
+      "loss": 1.898,
+      "step": 3991
+    },
+    {
+      "epoch": 4.541524459613197,
+      "grad_norm": 1.0816932916641235,
+      "learning_rate": 9.283276450511946e-05,
+      "loss": 2.1839,
+      "step": 3992
+    },
+    {
+      "epoch": 4.5426621160409555,
+      "grad_norm": 0.9521051049232483,
+      "learning_rate": 9.26052332195677e-05,
+      "loss": 1.5058,
+      "step": 3993
+    },
+    {
+      "epoch": 4.543799772468715,
+      "grad_norm": 0.5779545903205872,
+      "learning_rate": 9.237770193401593e-05,
+      "loss": 1.2215,
+      "step": 3994
+    },
+    {
+      "epoch": 4.544937428896473,
+      "grad_norm": 0.7004631161689758,
+      "learning_rate": 9.215017064846417e-05,
+      "loss": 0.8524,
+      "step": 3995
+    },
+    {
+      "epoch": 4.546075085324232,
+      "grad_norm": 1.5310842990875244,
+      "learning_rate": 9.192263936291239e-05,
+      "loss": 2.2935,
+      "step": 3996
+    },
+    {
+      "epoch": 4.5472127417519905,
+      "grad_norm": 1.0182299613952637,
+      "learning_rate": 9.169510807736064e-05,
+      "loss": 1.4897,
+      "step": 3997
+    },
+    {
+      "epoch": 4.54835039817975,
+      "grad_norm": 1.2902427911758423,
+      "learning_rate": 9.146757679180888e-05,
+      "loss": 2.0195,
+      "step": 3998
+    },
+    {
+      "epoch": 4.549488054607508,
+      "grad_norm": 0.9700065851211548,
+      "learning_rate": 9.124004550625712e-05,
+      "loss": 1.4082,
+      "step": 3999
+    },
+    {
+      "epoch": 4.550625711035267,
+      "grad_norm": 1.0845290422439575,
+      "learning_rate": 9.101251422070535e-05,
+      "loss": 2.3722,
+      "step": 4000
+    },
+    {
+      "epoch": 4.5517633674630265,
+      "grad_norm": 1.14189612865448,
+      "learning_rate": 9.078498293515358e-05,
+      "loss": 2.0645,
+      "step": 4001
+    },
+    {
+      "epoch": 4.552901023890785,
+      "grad_norm": 1.1628812551498413,
+      "learning_rate": 9.055745164960181e-05,
+      "loss": 2.7279,
+      "step": 4002
+    },
+    {
+      "epoch": 4.554038680318544,
+      "grad_norm": 0.8541833758354187,
+      "learning_rate": 9.032992036405006e-05,
+      "loss": 1.8064,
+      "step": 4003
+    },
+    {
+      "epoch": 4.555176336746302,
+      "grad_norm": 0.7190375924110413,
+      "learning_rate": 9.01023890784983e-05,
+      "loss": 1.7207,
+      "step": 4004
+    },
+    {
+      "epoch": 4.5563139931740615,
+      "grad_norm": 0.6545979976654053,
+      "learning_rate": 8.987485779294654e-05,
+      "loss": 1.4735,
+      "step": 4005
+    },
+    {
+      "epoch": 4.55745164960182,
+      "grad_norm": 1.12898588180542,
+      "learning_rate": 8.964732650739476e-05,
+      "loss": 1.8204,
+      "step": 4006
+    },
+    {
+      "epoch": 4.558589306029579,
+      "grad_norm": 1.265575885772705,
+      "learning_rate": 8.9419795221843e-05,
+      "loss": 2.1016,
+      "step": 4007
+    },
+    {
+      "epoch": 4.559726962457338,
+      "grad_norm": 1.3039684295654297,
+      "learning_rate": 8.919226393629125e-05,
+      "loss": 2.1831,
+      "step": 4008
+    },
+    {
+      "epoch": 4.5608646188850965,
+      "grad_norm": 0.6052323579788208,
+      "learning_rate": 8.896473265073949e-05,
+      "loss": 1.5035,
+      "step": 4009
+    },
+    {
+      "epoch": 4.562002275312856,
+      "grad_norm": 1.2771997451782227,
+      "learning_rate": 8.873720136518772e-05,
+      "loss": 2.3026,
+      "step": 4010
+    },
+    {
+      "epoch": 4.563139931740614,
+      "grad_norm": 1.577376365661621,
+      "learning_rate": 8.850967007963595e-05,
+      "loss": 2.7743,
+      "step": 4011
+    },
+    {
+      "epoch": 4.564277588168373,
+      "grad_norm": 0.78837651014328,
+      "learning_rate": 8.828213879408418e-05,
+      "loss": 1.4329,
+      "step": 4012
+    },
+    {
+      "epoch": 4.565415244596132,
+      "grad_norm": 1.392586350440979,
+      "learning_rate": 8.805460750853242e-05,
+      "loss": 1.9271,
+      "step": 4013
+    },
+    {
+      "epoch": 4.566552901023891,
+      "grad_norm": 0.6456694602966309,
+      "learning_rate": 8.782707622298067e-05,
+      "loss": 1.1636,
+      "step": 4014
+    },
+    {
+      "epoch": 4.56769055745165,
+      "grad_norm": 0.8244945406913757,
+      "learning_rate": 8.759954493742891e-05,
+      "loss": 1.3006,
+      "step": 4015
+    },
+    {
+      "epoch": 4.568828213879408,
+      "grad_norm": 1.0723069906234741,
+      "learning_rate": 8.737201365187713e-05,
+      "loss": 1.835,
+      "step": 4016
+    },
+    {
+      "epoch": 4.5699658703071675,
+      "grad_norm": 1.0027074813842773,
+      "learning_rate": 8.714448236632537e-05,
+      "loss": 1.806,
+      "step": 4017
+    },
+    {
+      "epoch": 4.571103526734926,
+      "grad_norm": 1.1087925434112549,
+      "learning_rate": 8.69169510807736e-05,
+      "loss": 1.7219,
+      "step": 4018
+    },
+    {
+      "epoch": 4.572241183162685,
+      "grad_norm": 0.5315782427787781,
+      "learning_rate": 8.668941979522184e-05,
+      "loss": 0.856,
+      "step": 4019
+    },
+    {
+      "epoch": 4.573378839590443,
+      "grad_norm": 0.8060416579246521,
+      "learning_rate": 8.646188850967009e-05,
+      "loss": 1.3456,
+      "step": 4020
+    },
+    {
+      "epoch": 4.5745164960182025,
+      "grad_norm": 0.8221978545188904,
+      "learning_rate": 8.623435722411832e-05,
+      "loss": 1.7365,
+      "step": 4021
+    },
+    {
+      "epoch": 4.575654152445962,
+      "grad_norm": 0.8591867685317993,
+      "learning_rate": 8.600682593856655e-05,
+      "loss": 1.5206,
+      "step": 4022
+    },
+    {
+      "epoch": 4.57679180887372,
+      "grad_norm": 0.694990336894989,
+      "learning_rate": 8.577929465301479e-05,
+      "loss": 0.6024,
+      "step": 4023
+    },
+    {
+      "epoch": 4.577929465301479,
+      "grad_norm": 1.0141761302947998,
+      "learning_rate": 8.555176336746303e-05,
+      "loss": 1.8285,
+      "step": 4024
+    },
+    {
+      "epoch": 4.579067121729238,
+      "grad_norm": 0.9870526194572449,
+      "learning_rate": 8.532423208191128e-05,
+      "loss": 2.191,
+      "step": 4025
+    },
+    {
+      "epoch": 4.580204778156997,
+      "grad_norm": 1.3243286609649658,
+      "learning_rate": 8.50967007963595e-05,
+      "loss": 2.1655,
+      "step": 4026
+    },
+    {
+      "epoch": 4.581342434584755,
+      "grad_norm": 0.9310430884361267,
+      "learning_rate": 8.486916951080774e-05,
+      "loss": 1.5505,
+      "step": 4027
+    },
+    {
+      "epoch": 4.582480091012514,
+      "grad_norm": 0.8326849937438965,
+      "learning_rate": 8.464163822525597e-05,
+      "loss": 1.2742,
+      "step": 4028
+    },
+    {
+      "epoch": 4.5836177474402735,
+      "grad_norm": 1.1547021865844727,
+      "learning_rate": 8.441410693970421e-05,
+      "loss": 1.895,
+      "step": 4029
+    },
+    {
+      "epoch": 4.584755403868032,
+      "grad_norm": 4.82136869430542,
+      "learning_rate": 8.418657565415245e-05,
+      "loss": 1.7264,
+      "step": 4030
+    },
+    {
+      "epoch": 4.585893060295791,
+      "grad_norm": 2.2376370429992676,
+      "learning_rate": 8.395904436860069e-05,
+      "loss": 3.246,
+      "step": 4031
+    },
+    {
+      "epoch": 4.587030716723549,
+      "grad_norm": 1.1774529218673706,
+      "learning_rate": 8.373151308304892e-05,
+      "loss": 2.3837,
+      "step": 4032
+    },
+    {
+      "epoch": 4.5881683731513085,
+      "grad_norm": 0.8420902490615845,
+      "learning_rate": 8.350398179749716e-05,
+      "loss": 1.535,
+      "step": 4033
+    },
+    {
+      "epoch": 4.589306029579067,
+      "grad_norm": 1.1322485208511353,
+      "learning_rate": 8.32764505119454e-05,
+      "loss": 3.0796,
+      "step": 4034
+    },
+    {
+      "epoch": 4.590443686006826,
+      "grad_norm": 0.7916383743286133,
+      "learning_rate": 8.304891922639362e-05,
+      "loss": 1.7278,
+      "step": 4035
+    },
+    {
+      "epoch": 4.591581342434584,
+      "grad_norm": 0.9946612119674683,
+      "learning_rate": 8.282138794084186e-05,
+      "loss": 1.3464,
+      "step": 4036
+    },
+    {
+      "epoch": 4.592718998862344,
+      "grad_norm": 0.9056809544563293,
+      "learning_rate": 8.259385665529011e-05,
+      "loss": 1.2729,
+      "step": 4037
+    },
+    {
+      "epoch": 4.593856655290102,
+      "grad_norm": 1.1132405996322632,
+      "learning_rate": 8.236632536973834e-05,
+      "loss": 1.6577,
+      "step": 4038
+    },
+    {
+      "epoch": 4.594994311717861,
+      "grad_norm": 1.0267456769943237,
+      "learning_rate": 8.213879408418658e-05,
+      "loss": 2.0504,
+      "step": 4039
+    },
+    {
+      "epoch": 4.59613196814562,
+      "grad_norm": 0.6499386429786682,
+      "learning_rate": 8.19112627986348e-05,
+      "loss": 0.6439,
+      "step": 4040
+    },
+    {
+      "epoch": 4.597269624573379,
+      "grad_norm": 0.9048442244529724,
+      "learning_rate": 8.168373151308304e-05,
+      "loss": 1.5015,
+      "step": 4041
+    },
+    {
+      "epoch": 4.598407281001138,
+      "grad_norm": 1.0397703647613525,
+      "learning_rate": 8.145620022753129e-05,
+      "loss": 1.5219,
+      "step": 4042
+    },
+    {
+      "epoch": 4.599544937428896,
+      "grad_norm": 0.7324240803718567,
+      "learning_rate": 8.122866894197953e-05,
+      "loss": 1.4209,
+      "step": 4043
+    },
+    {
+      "epoch": 4.600682593856655,
+      "grad_norm": 0.9085008502006531,
+      "learning_rate": 8.100113765642777e-05,
+      "loss": 1.7762,
+      "step": 4044
+    },
+    {
+      "epoch": 4.601820250284414,
+      "grad_norm": 1.1294033527374268,
+      "learning_rate": 8.077360637087599e-05,
+      "loss": 2.8995,
+      "step": 4045
+    },
+    {
+      "epoch": 4.602957906712173,
+      "grad_norm": 0.8178098201751709,
+      "learning_rate": 8.054607508532423e-05,
+      "loss": 2.0034,
+      "step": 4046
+    },
+    {
+      "epoch": 4.604095563139932,
+      "grad_norm": 1.7326245307922363,
+      "learning_rate": 8.031854379977246e-05,
+      "loss": 2.8204,
+      "step": 4047
+    },
+    {
+      "epoch": 4.60523321956769,
+      "grad_norm": 1.0977067947387695,
+      "learning_rate": 8.009101251422071e-05,
+      "loss": 1.5176,
+      "step": 4048
+    },
+    {
+      "epoch": 4.60637087599545,
+      "grad_norm": 0.7887221574783325,
+      "learning_rate": 7.986348122866895e-05,
+      "loss": 1.7636,
+      "step": 4049
+    },
+    {
+      "epoch": 4.607508532423208,
+      "grad_norm": 1.0273473262786865,
+      "learning_rate": 7.963594994311717e-05,
+      "loss": 1.5681,
+      "step": 4050
+    },
+    {
+      "epoch": 4.608646188850967,
+      "grad_norm": 0.9911322593688965,
+      "learning_rate": 7.940841865756541e-05,
+      "loss": 1.8449,
+      "step": 4051
+    },
+    {
+      "epoch": 4.609783845278725,
+      "grad_norm": 0.6486635804176331,
+      "learning_rate": 7.918088737201365e-05,
+      "loss": 1.4309,
+      "step": 4052
+    },
+    {
+      "epoch": 4.610921501706485,
+      "grad_norm": 1.0114645957946777,
+      "learning_rate": 7.895335608646189e-05,
+      "loss": 2.1855,
+      "step": 4053
+    },
+    {
+      "epoch": 4.612059158134244,
+      "grad_norm": 1.1461702585220337,
+      "learning_rate": 7.872582480091014e-05,
+      "loss": 2.1323,
+      "step": 4054
+    },
+    {
+      "epoch": 4.613196814562002,
+      "grad_norm": 0.8668122291564941,
+      "learning_rate": 7.849829351535836e-05,
+      "loss": 1.418,
+      "step": 4055
+    },
+    {
+      "epoch": 4.614334470989761,
+      "grad_norm": 0.5558659434318542,
+      "learning_rate": 7.82707622298066e-05,
+      "loss": 0.7871,
+      "step": 4056
+    },
+    {
+      "epoch": 4.61547212741752,
+      "grad_norm": 1.008002758026123,
+      "learning_rate": 7.804323094425483e-05,
+      "loss": 1.9257,
+      "step": 4057
+    },
+    {
+      "epoch": 4.616609783845279,
+      "grad_norm": 1.01564621925354,
+      "learning_rate": 7.781569965870307e-05,
+      "loss": 2.2797,
+      "step": 4058
+    },
+    {
+      "epoch": 4.617747440273037,
+      "grad_norm": 0.8626942038536072,
+      "learning_rate": 7.758816837315132e-05,
+      "loss": 1.7334,
+      "step": 4059
+    },
+    {
+      "epoch": 4.618885096700796,
+      "grad_norm": 0.9426528811454773,
+      "learning_rate": 7.736063708759954e-05,
+      "loss": 2.4934,
+      "step": 4060
+    },
+    {
+      "epoch": 4.620022753128556,
+      "grad_norm": 0.8705387711524963,
+      "learning_rate": 7.713310580204778e-05,
+      "loss": 1.5291,
+      "step": 4061
+    },
+    {
+      "epoch": 4.621160409556314,
+      "grad_norm": 0.6598314046859741,
+      "learning_rate": 7.690557451649602e-05,
+      "loss": 1.3372,
+      "step": 4062
+    },
+    {
+      "epoch": 4.622298065984073,
+      "grad_norm": 1.0638949871063232,
+      "learning_rate": 7.667804323094426e-05,
+      "loss": 2.271,
+      "step": 4063
+    },
+    {
+      "epoch": 4.623435722411831,
+      "grad_norm": 1.2114907503128052,
+      "learning_rate": 7.645051194539249e-05,
+      "loss": 1.9116,
+      "step": 4064
+    },
+    {
+      "epoch": 4.624573378839591,
+      "grad_norm": 0.896415650844574,
+      "learning_rate": 7.622298065984073e-05,
+      "loss": 1.5814,
+      "step": 4065
+    },
+    {
+      "epoch": 4.625711035267349,
+      "grad_norm": 0.9475829005241394,
+      "learning_rate": 7.599544937428897e-05,
+      "loss": 1.3713,
+      "step": 4066
+    },
+    {
+      "epoch": 4.626848691695108,
+      "grad_norm": 1.0576077699661255,
+      "learning_rate": 7.57679180887372e-05,
+      "loss": 1.3842,
+      "step": 4067
+    },
+    {
+      "epoch": 4.627986348122867,
+      "grad_norm": 1.035212755203247,
+      "learning_rate": 7.554038680318544e-05,
+      "loss": 1.5382,
+      "step": 4068
+    },
+    {
+      "epoch": 4.629124004550626,
+      "grad_norm": 1.0625888109207153,
+      "learning_rate": 7.531285551763368e-05,
+      "loss": 1.9357,
+      "step": 4069
+    },
+    {
+      "epoch": 4.630261660978385,
+      "grad_norm": 0.7255961894989014,
+      "learning_rate": 7.508532423208191e-05,
+      "loss": 1.8392,
+      "step": 4070
+    },
+    {
+      "epoch": 4.631399317406143,
+      "grad_norm": 0.6510151624679565,
+      "learning_rate": 7.485779294653015e-05,
+      "loss": 1.6608,
+      "step": 4071
+    },
+    {
+      "epoch": 4.632536973833902,
+      "grad_norm": 0.7432663440704346,
+      "learning_rate": 7.463026166097839e-05,
+      "loss": 1.4151,
+      "step": 4072
+    },
+    {
+      "epoch": 4.633674630261661,
+      "grad_norm": 1.0191010236740112,
+      "learning_rate": 7.440273037542663e-05,
+      "loss": 2.6943,
+      "step": 4073
+    },
+    {
+      "epoch": 4.63481228668942,
+      "grad_norm": 0.7630909085273743,
+      "learning_rate": 7.417519908987486e-05,
+      "loss": 1.6592,
+      "step": 4074
+    },
+    {
+      "epoch": 4.635949943117178,
+      "grad_norm": 0.8889945149421692,
+      "learning_rate": 7.394766780432309e-05,
+      "loss": 1.8455,
+      "step": 4075
+    },
+    {
+      "epoch": 4.637087599544937,
+      "grad_norm": 1.1400561332702637,
+      "learning_rate": 7.372013651877134e-05,
+      "loss": 2.2122,
+      "step": 4076
+    },
+    {
+      "epoch": 4.638225255972696,
+      "grad_norm": 0.7421183586120605,
+      "learning_rate": 7.349260523321957e-05,
+      "loss": 1.1382,
+      "step": 4077
+    },
+    {
+      "epoch": 4.639362912400455,
+      "grad_norm": 0.8412690162658691,
+      "learning_rate": 7.326507394766781e-05,
+      "loss": 1.8067,
+      "step": 4078
+    },
+    {
+      "epoch": 4.640500568828214,
+      "grad_norm": 0.6977559924125671,
+      "learning_rate": 7.303754266211605e-05,
+      "loss": 0.9484,
+      "step": 4079
+    },
+    {
+      "epoch": 4.6416382252559725,
+      "grad_norm": 1.04801607131958,
+      "learning_rate": 7.281001137656427e-05,
+      "loss": 3.1128,
+      "step": 4080
+    },
+    {
+      "epoch": 4.642775881683732,
+      "grad_norm": 0.740193247795105,
+      "learning_rate": 7.258248009101251e-05,
+      "loss": 1.061,
+      "step": 4081
+    },
+    {
+      "epoch": 4.64391353811149,
+      "grad_norm": 0.5949002504348755,
+      "learning_rate": 7.235494880546076e-05,
+      "loss": 0.9576,
+      "step": 4082
+    },
+    {
+      "epoch": 4.645051194539249,
+      "grad_norm": 0.8039756417274475,
+      "learning_rate": 7.2127417519909e-05,
+      "loss": 1.3286,
+      "step": 4083
+    },
+    {
+      "epoch": 4.6461888509670075,
+      "grad_norm": 1.1032826900482178,
+      "learning_rate": 7.189988623435722e-05,
+      "loss": 2.7052,
+      "step": 4084
+    },
+    {
+      "epoch": 4.647326507394767,
+      "grad_norm": 1.0323725938796997,
+      "learning_rate": 7.167235494880546e-05,
+      "loss": 1.7848,
+      "step": 4085
+    },
+    {
+      "epoch": 4.648464163822526,
+      "grad_norm": 0.5838832259178162,
+      "learning_rate": 7.144482366325369e-05,
+      "loss": 1.4307,
+      "step": 4086
+    },
+    {
+      "epoch": 4.649601820250284,
+      "grad_norm": 1.3185147047042847,
+      "learning_rate": 7.121729237770194e-05,
+      "loss": 1.7205,
+      "step": 4087
+    },
+    {
+      "epoch": 4.650739476678043,
+      "grad_norm": 0.5053433179855347,
+      "learning_rate": 7.098976109215018e-05,
+      "loss": 0.4623,
+      "step": 4088
+    },
+    {
+      "epoch": 4.651877133105802,
+      "grad_norm": 0.9480729699134827,
+      "learning_rate": 7.07622298065984e-05,
+      "loss": 1.9962,
+      "step": 4089
+    },
+    {
+      "epoch": 4.653014789533561,
+      "grad_norm": 0.9530803561210632,
+      "learning_rate": 7.053469852104664e-05,
+      "loss": 2.4195,
+      "step": 4090
+    },
+    {
+      "epoch": 4.654152445961319,
+      "grad_norm": 0.7480978965759277,
+      "learning_rate": 7.030716723549488e-05,
+      "loss": 0.7856,
+      "step": 4091
+    },
+    {
+      "epoch": 4.6552901023890785,
+      "grad_norm": 0.8531357049942017,
+      "learning_rate": 7.007963594994311e-05,
+      "loss": 1.9666,
+      "step": 4092
+    },
+    {
+      "epoch": 4.656427758816838,
+      "grad_norm": 0.8261699080467224,
+      "learning_rate": 6.985210466439137e-05,
+      "loss": 1.2341,
+      "step": 4093
+    },
+    {
+      "epoch": 4.657565415244596,
+      "grad_norm": 1.0157686471939087,
+      "learning_rate": 6.962457337883959e-05,
+      "loss": 1.1924,
+      "step": 4094
+    },
+    {
+      "epoch": 4.658703071672355,
+      "grad_norm": 0.8260225057601929,
+      "learning_rate": 6.939704209328783e-05,
+      "loss": 1.092,
+      "step": 4095
+    },
+    {
+      "epoch": 4.6598407281001135,
+      "grad_norm": 0.836290717124939,
+      "learning_rate": 6.916951080773606e-05,
+      "loss": 0.9278,
+      "step": 4096
+    },
+    {
+      "epoch": 4.660978384527873,
+      "grad_norm": 1.124277949333191,
+      "learning_rate": 6.89419795221843e-05,
+      "loss": 1.9577,
+      "step": 4097
+    },
+    {
+      "epoch": 4.662116040955631,
+      "grad_norm": 1.066027045249939,
+      "learning_rate": 6.871444823663254e-05,
+      "loss": 2.6585,
+      "step": 4098
+    },
+    {
+      "epoch": 4.66325369738339,
+      "grad_norm": 0.5382254123687744,
+      "learning_rate": 6.848691695108077e-05,
+      "loss": 1.025,
+      "step": 4099
+    },
+    {
+      "epoch": 4.664391353811149,
+      "grad_norm": 0.6736378073692322,
+      "learning_rate": 6.825938566552901e-05,
+      "loss": 1.2211,
+      "step": 4100
+    },
+    {
+      "epoch": 4.665529010238908,
+      "grad_norm": 1.0047065019607544,
+      "learning_rate": 6.803185437997725e-05,
+      "loss": 1.528,
+      "step": 4101
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 1.521615743637085,
+      "learning_rate": 6.780432309442548e-05,
+      "loss": 3.3803,
+      "step": 4102
+    },
+    {
+      "epoch": 4.667804323094425,
+      "grad_norm": 0.9080568552017212,
+      "learning_rate": 6.757679180887372e-05,
+      "loss": 1.778,
+      "step": 4103
+    },
+    {
+      "epoch": 4.6689419795221845,
+      "grad_norm": 0.7825086116790771,
+      "learning_rate": 6.734926052332196e-05,
+      "loss": 1.6855,
+      "step": 4104
+    },
+    {
+      "epoch": 4.670079635949943,
+      "grad_norm": 0.905820906162262,
+      "learning_rate": 6.71217292377702e-05,
+      "loss": 2.29,
+      "step": 4105
+    },
+    {
+      "epoch": 4.671217292377702,
+      "grad_norm": 0.7313637137413025,
+      "learning_rate": 6.689419795221843e-05,
+      "loss": 1.6036,
+      "step": 4106
+    },
+    {
+      "epoch": 4.672354948805461,
+      "grad_norm": 0.683791995048523,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.4341,
+      "step": 4107
+    },
+    {
+      "epoch": 4.6734926052332195,
+      "grad_norm": 1.0041923522949219,
+      "learning_rate": 6.643913538111491e-05,
+      "loss": 1.4274,
+      "step": 4108
+    },
+    {
+      "epoch": 4.674630261660979,
+      "grad_norm": 0.662774920463562,
+      "learning_rate": 6.621160409556313e-05,
+      "loss": 1.4739,
+      "step": 4109
+    },
+    {
+      "epoch": 4.675767918088737,
+      "grad_norm": 1.1887277364730835,
+      "learning_rate": 6.598407281001138e-05,
+      "loss": 1.7699,
+      "step": 4110
+    },
+    {
+      "epoch": 4.676905574516496,
+      "grad_norm": 1.0310107469558716,
+      "learning_rate": 6.575654152445962e-05,
+      "loss": 1.565,
+      "step": 4111
+    },
+    {
+      "epoch": 4.678043230944255,
+      "grad_norm": 0.9942086338996887,
+      "learning_rate": 6.552901023890785e-05,
+      "loss": 1.6227,
+      "step": 4112
+    },
+    {
+      "epoch": 4.679180887372014,
+      "grad_norm": 1.4443100690841675,
+      "learning_rate": 6.530147895335609e-05,
+      "loss": 3.6737,
+      "step": 4113
+    },
+    {
+      "epoch": 4.680318543799773,
+      "grad_norm": 0.9084301590919495,
+      "learning_rate": 6.507394766780432e-05,
+      "loss": 1.8005,
+      "step": 4114
+    },
+    {
+      "epoch": 4.681456200227531,
+      "grad_norm": 1.0044715404510498,
+      "learning_rate": 6.484641638225255e-05,
+      "loss": 2.3198,
+      "step": 4115
+    },
+    {
+      "epoch": 4.6825938566552905,
+      "grad_norm": 0.6797364950180054,
+      "learning_rate": 6.46188850967008e-05,
+      "loss": 1.0637,
+      "step": 4116
+    },
+    {
+      "epoch": 4.683731513083049,
+      "grad_norm": 1.2862342596054077,
+      "learning_rate": 6.439135381114904e-05,
+      "loss": 2.6791,
+      "step": 4117
+    },
+    {
+      "epoch": 4.684869169510808,
+      "grad_norm": 0.9734237790107727,
+      "learning_rate": 6.416382252559728e-05,
+      "loss": 1.7895,
+      "step": 4118
+    },
+    {
+      "epoch": 4.686006825938566,
+      "grad_norm": 1.044211506843567,
+      "learning_rate": 6.39362912400455e-05,
+      "loss": 3.2597,
+      "step": 4119
+    },
+    {
+      "epoch": 4.6871444823663255,
+      "grad_norm": 1.082377552986145,
+      "learning_rate": 6.370875995449374e-05,
+      "loss": 1.8831,
+      "step": 4120
+    },
+    {
+      "epoch": 4.688282138794084,
+      "grad_norm": 1.1628836393356323,
+      "learning_rate": 6.348122866894199e-05,
+      "loss": 2.3543,
+      "step": 4121
+    },
+    {
+      "epoch": 4.689419795221843,
+      "grad_norm": 0.9119383692741394,
+      "learning_rate": 6.325369738339022e-05,
+      "loss": 1.5563,
+      "step": 4122
+    },
+    {
+      "epoch": 4.690557451649601,
+      "grad_norm": 1.1674100160598755,
+      "learning_rate": 6.302616609783846e-05,
+      "loss": 1.7306,
+      "step": 4123
+    },
+    {
+      "epoch": 4.691695108077361,
+      "grad_norm": 0.7952550053596497,
+      "learning_rate": 6.279863481228669e-05,
+      "loss": 1.5032,
+      "step": 4124
+    },
+    {
+      "epoch": 4.69283276450512,
+      "grad_norm": 0.9031153917312622,
+      "learning_rate": 6.257110352673492e-05,
+      "loss": 2.8563,
+      "step": 4125
+    },
+    {
+      "epoch": 4.693970420932878,
+      "grad_norm": 1.1440151929855347,
+      "learning_rate": 6.234357224118316e-05,
+      "loss": 1.1994,
+      "step": 4126
+    },
+    {
+      "epoch": 4.695108077360637,
+      "grad_norm": 0.8565515279769897,
+      "learning_rate": 6.21160409556314e-05,
+      "loss": 0.904,
+      "step": 4127
+    },
+    {
+      "epoch": 4.696245733788396,
+      "grad_norm": 1.0120102167129517,
+      "learning_rate": 6.188850967007965e-05,
+      "loss": 1.6285,
+      "step": 4128
+    },
+    {
+      "epoch": 4.697383390216155,
+      "grad_norm": 0.95466148853302,
+      "learning_rate": 6.166097838452787e-05,
+      "loss": 1.4088,
+      "step": 4129
+    },
+    {
+      "epoch": 4.698521046643913,
+      "grad_norm": 0.9938770532608032,
+      "learning_rate": 6.143344709897611e-05,
+      "loss": 1.9033,
+      "step": 4130
+    },
+    {
+      "epoch": 4.699658703071672,
+      "grad_norm": 0.8543733954429626,
+      "learning_rate": 6.120591581342434e-05,
+      "loss": 2.0826,
+      "step": 4131
+    },
+    {
+      "epoch": 4.7007963594994315,
+      "grad_norm": 1.0578027963638306,
+      "learning_rate": 6.097838452787258e-05,
+      "loss": 1.5184,
+      "step": 4132
+    },
+    {
+      "epoch": 4.70193401592719,
+      "grad_norm": 1.0492652654647827,
+      "learning_rate": 6.075085324232082e-05,
+      "loss": 2.2705,
+      "step": 4133
+    },
+    {
+      "epoch": 4.703071672354949,
+      "grad_norm": 0.7921819686889648,
+      "learning_rate": 6.052332195676906e-05,
+      "loss": 1.2564,
+      "step": 4134
+    },
+    {
+      "epoch": 4.704209328782707,
+      "grad_norm": 0.9057871103286743,
+      "learning_rate": 6.029579067121729e-05,
+      "loss": 1.3608,
+      "step": 4135
+    },
+    {
+      "epoch": 4.705346985210467,
+      "grad_norm": 1.0830860137939453,
+      "learning_rate": 6.006825938566553e-05,
+      "loss": 2.1617,
+      "step": 4136
+    },
+    {
+      "epoch": 4.706484641638225,
+      "grad_norm": 0.9428957104682922,
+      "learning_rate": 5.9840728100113766e-05,
+      "loss": 2.4494,
+      "step": 4137
+    },
+    {
+      "epoch": 4.707622298065984,
+      "grad_norm": 0.7517849206924438,
+      "learning_rate": 5.9613196814562e-05,
+      "loss": 1.2648,
+      "step": 4138
+    },
+    {
+      "epoch": 4.708759954493743,
+      "grad_norm": 1.1423892974853516,
+      "learning_rate": 5.938566552901024e-05,
+      "loss": 1.8682,
+      "step": 4139
+    },
+    {
+      "epoch": 4.709897610921502,
+      "grad_norm": 0.8202763795852661,
+      "learning_rate": 5.915813424345848e-05,
+      "loss": 2.3286,
+      "step": 4140
+    },
+    {
+      "epoch": 4.711035267349261,
+      "grad_norm": 0.7963849306106567,
+      "learning_rate": 5.8930602957906714e-05,
+      "loss": 1.5263,
+      "step": 4141
+    },
+    {
+      "epoch": 4.712172923777019,
+      "grad_norm": 0.8499445915222168,
+      "learning_rate": 5.870307167235495e-05,
+      "loss": 1.4715,
+      "step": 4142
+    },
+    {
+      "epoch": 4.713310580204778,
+      "grad_norm": 0.9363263249397278,
+      "learning_rate": 5.847554038680319e-05,
+      "loss": 2.0218,
+      "step": 4143
+    },
+    {
+      "epoch": 4.714448236632537,
+      "grad_norm": 0.9191851615905762,
+      "learning_rate": 5.824800910125142e-05,
+      "loss": 2.695,
+      "step": 4144
+    },
+    {
+      "epoch": 4.715585893060296,
+      "grad_norm": 1.0178710222244263,
+      "learning_rate": 5.802047781569966e-05,
+      "loss": 2.0507,
+      "step": 4145
+    },
+    {
+      "epoch": 4.716723549488055,
+      "grad_norm": 0.9625295400619507,
+      "learning_rate": 5.77929465301479e-05,
+      "loss": 1.7354,
+      "step": 4146
+    },
+    {
+      "epoch": 4.717861205915813,
+      "grad_norm": 1.0893375873565674,
+      "learning_rate": 5.756541524459613e-05,
+      "loss": 2.4767,
+      "step": 4147
+    },
+    {
+      "epoch": 4.718998862343573,
+      "grad_norm": 0.6518070101737976,
+      "learning_rate": 5.733788395904437e-05,
+      "loss": 1.2152,
+      "step": 4148
+    },
+    {
+      "epoch": 4.720136518771331,
+      "grad_norm": 1.2176522016525269,
+      "learning_rate": 5.71103526734926e-05,
+      "loss": 1.7852,
+      "step": 4149
+    },
+    {
+      "epoch": 4.72127417519909,
+      "grad_norm": 1.0597165822982788,
+      "learning_rate": 5.688282138794084e-05,
+      "loss": 2.6517,
+      "step": 4150
+    },
+    {
+      "epoch": 4.722411831626848,
+      "grad_norm": 0.9489892721176147,
+      "learning_rate": 5.6655290102389084e-05,
+      "loss": 2.0657,
+      "step": 4151
+    },
+    {
+      "epoch": 4.723549488054608,
+      "grad_norm": 0.9729322195053101,
+      "learning_rate": 5.6427758816837314e-05,
+      "loss": 1.7927,
+      "step": 4152
+    },
+    {
+      "epoch": 4.724687144482367,
+      "grad_norm": 0.832109808921814,
+      "learning_rate": 5.620022753128555e-05,
+      "loss": 1.1887,
+      "step": 4153
+    },
+    {
+      "epoch": 4.725824800910125,
+      "grad_norm": 1.0068272352218628,
+      "learning_rate": 5.597269624573379e-05,
+      "loss": 1.3375,
+      "step": 4154
+    },
+    {
+      "epoch": 4.726962457337884,
+      "grad_norm": 0.7367716431617737,
+      "learning_rate": 5.5745164960182025e-05,
+      "loss": 1.3822,
+      "step": 4155
+    },
+    {
+      "epoch": 4.728100113765643,
+      "grad_norm": 1.1299673318862915,
+      "learning_rate": 5.551763367463027e-05,
+      "loss": 2.0459,
+      "step": 4156
+    },
+    {
+      "epoch": 4.729237770193402,
+      "grad_norm": 0.8980764150619507,
+      "learning_rate": 5.52901023890785e-05,
+      "loss": 1.6,
+      "step": 4157
+    },
+    {
+      "epoch": 4.73037542662116,
+      "grad_norm": 1.1862162351608276,
+      "learning_rate": 5.5062571103526736e-05,
+      "loss": 2.6499,
+      "step": 4158
+    },
+    {
+      "epoch": 4.731513083048919,
+      "grad_norm": 0.6605071425437927,
+      "learning_rate": 5.483503981797497e-05,
+      "loss": 0.782,
+      "step": 4159
+    },
+    {
+      "epoch": 4.732650739476678,
+      "grad_norm": 1.0915684700012207,
+      "learning_rate": 5.460750853242321e-05,
+      "loss": 2.1564,
+      "step": 4160
+    },
+    {
+      "epoch": 4.733788395904437,
+      "grad_norm": 0.7124598026275635,
+      "learning_rate": 5.437997724687145e-05,
+      "loss": 1.3858,
+      "step": 4161
+    },
+    {
+      "epoch": 4.734926052332195,
+      "grad_norm": 0.8076978325843811,
+      "learning_rate": 5.4152445961319684e-05,
+      "loss": 1.6923,
+      "step": 4162
+    },
+    {
+      "epoch": 4.736063708759954,
+      "grad_norm": 1.0069531202316284,
+      "learning_rate": 5.392491467576792e-05,
+      "loss": 2.5924,
+      "step": 4163
+    },
+    {
+      "epoch": 4.737201365187714,
+      "grad_norm": 0.7821694016456604,
+      "learning_rate": 5.369738339021615e-05,
+      "loss": 1.5226,
+      "step": 4164
+    },
+    {
+      "epoch": 4.738339021615472,
+      "grad_norm": 1.0969029664993286,
+      "learning_rate": 5.3469852104664395e-05,
+      "loss": 1.791,
+      "step": 4165
+    },
+    {
+      "epoch": 4.739476678043231,
+      "grad_norm": 0.6663936972618103,
+      "learning_rate": 5.324232081911263e-05,
+      "loss": 1.298,
+      "step": 4166
+    },
+    {
+      "epoch": 4.7406143344709895,
+      "grad_norm": 0.6106514930725098,
+      "learning_rate": 5.301478953356086e-05,
+      "loss": 1.0094,
+      "step": 4167
+    },
+    {
+      "epoch": 4.741751990898749,
+      "grad_norm": 0.8213837742805481,
+      "learning_rate": 5.2787258248009106e-05,
+      "loss": 1.5493,
+      "step": 4168
+    },
+    {
+      "epoch": 4.742889647326507,
+      "grad_norm": 1.1655770540237427,
+      "learning_rate": 5.2559726962457336e-05,
+      "loss": 2.0271,
+      "step": 4169
+    },
+    {
+      "epoch": 4.744027303754266,
+      "grad_norm": 0.9248787760734558,
+      "learning_rate": 5.233219567690557e-05,
+      "loss": 2.3724,
+      "step": 4170
+    },
+    {
+      "epoch": 4.745164960182025,
+      "grad_norm": 0.7159304022789001,
+      "learning_rate": 5.210466439135381e-05,
+      "loss": 1.3412,
+      "step": 4171
+    },
+    {
+      "epoch": 4.746302616609784,
+      "grad_norm": 0.773129403591156,
+      "learning_rate": 5.187713310580205e-05,
+      "loss": 2.2651,
+      "step": 4172
+    },
+    {
+      "epoch": 4.747440273037543,
+      "grad_norm": 1.6007509231567383,
+      "learning_rate": 5.164960182025029e-05,
+      "loss": 2.185,
+      "step": 4173
+    },
+    {
+      "epoch": 4.748577929465301,
+      "grad_norm": 0.9146437048912048,
+      "learning_rate": 5.142207053469852e-05,
+      "loss": 1.8578,
+      "step": 4174
+    },
+    {
+      "epoch": 4.74971558589306,
+      "grad_norm": 1.0452239513397217,
+      "learning_rate": 5.119453924914676e-05,
+      "loss": 2.7301,
+      "step": 4175
+    },
+    {
+      "epoch": 4.750853242320819,
+      "grad_norm": 1.1946684122085571,
+      "learning_rate": 5.0967007963594995e-05,
+      "loss": 2.3748,
+      "step": 4176
+    },
+    {
+      "epoch": 4.751990898748578,
+      "grad_norm": 1.1966066360473633,
+      "learning_rate": 5.073947667804323e-05,
+      "loss": 1.9953,
+      "step": 4177
+    },
+    {
+      "epoch": 4.753128555176337,
+      "grad_norm": 0.8742679357528687,
+      "learning_rate": 5.051194539249147e-05,
+      "loss": 1.7049,
+      "step": 4178
+    },
+    {
+      "epoch": 4.7542662116040955,
+      "grad_norm": 0.7945109605789185,
+      "learning_rate": 5.0284414106939706e-05,
+      "loss": 1.4822,
+      "step": 4179
+    },
+    {
+      "epoch": 4.755403868031855,
+      "grad_norm": 0.7654008269309998,
+      "learning_rate": 5.005688282138794e-05,
+      "loss": 1.0475,
+      "step": 4180
+    },
+    {
+      "epoch": 4.756541524459613,
+      "grad_norm": 1.0608311891555786,
+      "learning_rate": 4.982935153583617e-05,
+      "loss": 1.5947,
+      "step": 4181
+    },
+    {
+      "epoch": 4.757679180887372,
+      "grad_norm": 1.2881845235824585,
+      "learning_rate": 4.960182025028442e-05,
+      "loss": 3.6129,
+      "step": 4182
+    },
+    {
+      "epoch": 4.7588168373151305,
+      "grad_norm": 1.1011303663253784,
+      "learning_rate": 4.9374288964732654e-05,
+      "loss": 1.1276,
+      "step": 4183
+    },
+    {
+      "epoch": 4.75995449374289,
+      "grad_norm": 1.142633318901062,
+      "learning_rate": 4.9146757679180884e-05,
+      "loss": 2.4236,
+      "step": 4184
+    },
+    {
+      "epoch": 4.761092150170649,
+      "grad_norm": 0.940455436706543,
+      "learning_rate": 4.891922639362913e-05,
+      "loss": 2.3076,
+      "step": 4185
+    },
+    {
+      "epoch": 4.762229806598407,
+      "grad_norm": 0.8096868991851807,
+      "learning_rate": 4.869169510807736e-05,
+      "loss": 0.983,
+      "step": 4186
+    },
+    {
+      "epoch": 4.763367463026166,
+      "grad_norm": 1.020517349243164,
+      "learning_rate": 4.8464163822525595e-05,
+      "loss": 2.3447,
+      "step": 4187
+    },
+    {
+      "epoch": 4.764505119453925,
+      "grad_norm": 1.09331214427948,
+      "learning_rate": 4.823663253697384e-05,
+      "loss": 2.6737,
+      "step": 4188
+    },
+    {
+      "epoch": 4.765642775881684,
+      "grad_norm": 1.5871385335922241,
+      "learning_rate": 4.800910125142207e-05,
+      "loss": 3.2077,
+      "step": 4189
+    },
+    {
+      "epoch": 4.766780432309442,
+      "grad_norm": 0.6802913546562195,
+      "learning_rate": 4.778156996587031e-05,
+      "loss": 1.2683,
+      "step": 4190
+    },
+    {
+      "epoch": 4.7679180887372015,
+      "grad_norm": 0.8317682147026062,
+      "learning_rate": 4.755403868031854e-05,
+      "loss": 1.9439,
+      "step": 4191
+    },
+    {
+      "epoch": 4.769055745164961,
+      "grad_norm": 0.8209235668182373,
+      "learning_rate": 4.732650739476678e-05,
+      "loss": 1.3119,
+      "step": 4192
+    },
+    {
+      "epoch": 4.770193401592719,
+      "grad_norm": 1.44149649143219,
+      "learning_rate": 4.7098976109215024e-05,
+      "loss": 2.4981,
+      "step": 4193
+    },
+    {
+      "epoch": 4.771331058020478,
+      "grad_norm": 1.0103590488433838,
+      "learning_rate": 4.6871444823663254e-05,
+      "loss": 1.6604,
+      "step": 4194
+    },
+    {
+      "epoch": 4.7724687144482365,
+      "grad_norm": 1.4098118543624878,
+      "learning_rate": 4.664391353811149e-05,
+      "loss": 3.3837,
+      "step": 4195
+    },
+    {
+      "epoch": 4.773606370875996,
+      "grad_norm": 0.829317569732666,
+      "learning_rate": 4.641638225255973e-05,
+      "loss": 1.5868,
+      "step": 4196
+    },
+    {
+      "epoch": 4.774744027303754,
+      "grad_norm": 0.9138002991676331,
+      "learning_rate": 4.6188850967007965e-05,
+      "loss": 2.3734,
+      "step": 4197
+    },
+    {
+      "epoch": 4.775881683731513,
+      "grad_norm": 0.6845062971115112,
+      "learning_rate": 4.5961319681456195e-05,
+      "loss": 1.1712,
+      "step": 4198
+    },
+    {
+      "epoch": 4.7770193401592715,
+      "grad_norm": 0.592144787311554,
+      "learning_rate": 4.573378839590444e-05,
+      "loss": 1.0386,
+      "step": 4199
+    },
+    {
+      "epoch": 4.778156996587031,
+      "grad_norm": 1.318644404411316,
+      "learning_rate": 4.5506257110352676e-05,
+      "loss": 2.859,
+      "step": 4200
+    },
+    {
+      "epoch": 4.779294653014789,
+      "grad_norm": 1.0490177869796753,
+      "learning_rate": 4.5278725824800906e-05,
+      "loss": 2.2225,
+      "step": 4201
+    },
+    {
+      "epoch": 4.780432309442548,
+      "grad_norm": 0.9637972712516785,
+      "learning_rate": 4.505119453924915e-05,
+      "loss": 1.6452,
+      "step": 4202
+    },
+    {
+      "epoch": 4.7815699658703075,
+      "grad_norm": 1.3395143747329712,
+      "learning_rate": 4.482366325369738e-05,
+      "loss": 2.982,
+      "step": 4203
+    },
+    {
+      "epoch": 4.782707622298066,
+      "grad_norm": 0.9925627708435059,
+      "learning_rate": 4.4596131968145624e-05,
+      "loss": 2.2197,
+      "step": 4204
+    },
+    {
+      "epoch": 4.783845278725825,
+      "grad_norm": 0.715560257434845,
+      "learning_rate": 4.436860068259386e-05,
+      "loss": 1.6095,
+      "step": 4205
+    },
+    {
+      "epoch": 4.784982935153583,
+      "grad_norm": 0.7981997132301331,
+      "learning_rate": 4.414106939704209e-05,
+      "loss": 1.859,
+      "step": 4206
+    },
+    {
+      "epoch": 4.7861205915813425,
+      "grad_norm": 0.894432783126831,
+      "learning_rate": 4.3913538111490335e-05,
+      "loss": 1.4185,
+      "step": 4207
+    },
+    {
+      "epoch": 4.787258248009101,
+      "grad_norm": 1.0646681785583496,
+      "learning_rate": 4.3686006825938565e-05,
+      "loss": 2.6534,
+      "step": 4208
+    },
+    {
+      "epoch": 4.78839590443686,
+      "grad_norm": 1.0470125675201416,
+      "learning_rate": 4.34584755403868e-05,
+      "loss": 2.4396,
+      "step": 4209
+    },
+    {
+      "epoch": 4.789533560864619,
+      "grad_norm": 1.2576868534088135,
+      "learning_rate": 4.3230944254835046e-05,
+      "loss": 1.9555,
+      "step": 4210
+    },
+    {
+      "epoch": 4.7906712172923775,
+      "grad_norm": 0.7670148611068726,
+      "learning_rate": 4.3003412969283276e-05,
+      "loss": 1.1409,
+      "step": 4211
+    },
+    {
+      "epoch": 4.791808873720137,
+      "grad_norm": 0.9157593250274658,
+      "learning_rate": 4.277588168373151e-05,
+      "loss": 1.9202,
+      "step": 4212
+    },
+    {
+      "epoch": 4.792946530147895,
+      "grad_norm": 0.4483858048915863,
+      "learning_rate": 4.254835039817975e-05,
+      "loss": 0.7922,
+      "step": 4213
+    },
+    {
+      "epoch": 4.794084186575654,
+      "grad_norm": 1.031166911125183,
+      "learning_rate": 4.232081911262799e-05,
+      "loss": 2.3445,
+      "step": 4214
+    },
+    {
+      "epoch": 4.795221843003413,
+      "grad_norm": 1.042548418045044,
+      "learning_rate": 4.2093287827076224e-05,
+      "loss": 1.7547,
+      "step": 4215
+    },
+    {
+      "epoch": 4.796359499431172,
+      "grad_norm": 1.197379231452942,
+      "learning_rate": 4.186575654152446e-05,
+      "loss": 1.7015,
+      "step": 4216
+    },
+    {
+      "epoch": 4.797497155858931,
+      "grad_norm": 0.9254120588302612,
+      "learning_rate": 4.16382252559727e-05,
+      "loss": 1.1992,
+      "step": 4217
+    },
+    {
+      "epoch": 4.798634812286689,
+      "grad_norm": 0.6915335655212402,
+      "learning_rate": 4.141069397042093e-05,
+      "loss": 0.8966,
+      "step": 4218
+    },
+    {
+      "epoch": 4.7997724687144485,
+      "grad_norm": 1.2596569061279297,
+      "learning_rate": 4.118316268486917e-05,
+      "loss": 1.7151,
+      "step": 4219
+    },
+    {
+      "epoch": 4.800910125142207,
+      "grad_norm": 0.873573362827301,
+      "learning_rate": 4.09556313993174e-05,
+      "loss": 2.261,
+      "step": 4220
+    },
+    {
+      "epoch": 4.802047781569966,
+      "grad_norm": 0.6213791370391846,
+      "learning_rate": 4.0728100113765646e-05,
+      "loss": 1.1226,
+      "step": 4221
+    },
+    {
+      "epoch": 4.803185437997724,
+      "grad_norm": 0.9243321418762207,
+      "learning_rate": 4.050056882821388e-05,
+      "loss": 2.3551,
+      "step": 4222
+    },
+    {
+      "epoch": 4.8043230944254836,
+      "grad_norm": 1.7408545017242432,
+      "learning_rate": 4.0273037542662113e-05,
+      "loss": 3.208,
+      "step": 4223
+    },
+    {
+      "epoch": 4.805460750853243,
+      "grad_norm": 0.5592201352119446,
+      "learning_rate": 4.004550625711036e-05,
+      "loss": 0.7587,
+      "step": 4224
+    },
+    {
+      "epoch": 4.806598407281001,
+      "grad_norm": 0.7832121849060059,
+      "learning_rate": 3.981797497155859e-05,
+      "loss": 1.2192,
+      "step": 4225
+    },
+    {
+      "epoch": 4.80773606370876,
+      "grad_norm": 0.9754132628440857,
+      "learning_rate": 3.9590443686006824e-05,
+      "loss": 1.2067,
+      "step": 4226
+    },
+    {
+      "epoch": 4.808873720136519,
+      "grad_norm": 0.9849193096160889,
+      "learning_rate": 3.936291240045507e-05,
+      "loss": 2.3535,
+      "step": 4227
+    },
+    {
+      "epoch": 4.810011376564278,
+      "grad_norm": 0.9449824690818787,
+      "learning_rate": 3.91353811149033e-05,
+      "loss": 1.7503,
+      "step": 4228
+    },
+    {
+      "epoch": 4.811149032992036,
+      "grad_norm": 1.1584771871566772,
+      "learning_rate": 3.8907849829351535e-05,
+      "loss": 2.9749,
+      "step": 4229
+    },
+    {
+      "epoch": 4.812286689419795,
+      "grad_norm": 0.9928334355354309,
+      "learning_rate": 3.868031854379977e-05,
+      "loss": 2.3772,
+      "step": 4230
+    },
+    {
+      "epoch": 4.8134243458475545,
+      "grad_norm": 1.044175386428833,
+      "learning_rate": 3.845278725824801e-05,
+      "loss": 2.2084,
+      "step": 4231
+    },
+    {
+      "epoch": 4.814562002275313,
+      "grad_norm": 1.501725673675537,
+      "learning_rate": 3.8225255972696246e-05,
+      "loss": 3.0545,
+      "step": 4232
+    },
+    {
+      "epoch": 4.815699658703072,
+      "grad_norm": 0.6129480004310608,
+      "learning_rate": 3.799772468714448e-05,
+      "loss": 1.1711,
+      "step": 4233
+    },
+    {
+      "epoch": 4.81683731513083,
+      "grad_norm": 0.6920023560523987,
+      "learning_rate": 3.777019340159272e-05,
+      "loss": 1.006,
+      "step": 4234
+    },
+    {
+      "epoch": 4.81797497155859,
+      "grad_norm": 1.1781179904937744,
+      "learning_rate": 3.754266211604096e-05,
+      "loss": 2.6292,
+      "step": 4235
+    },
+    {
+      "epoch": 4.819112627986348,
+      "grad_norm": 0.9663881659507751,
+      "learning_rate": 3.7315130830489194e-05,
+      "loss": 1.689,
+      "step": 4236
+    },
+    {
+      "epoch": 4.820250284414107,
+      "grad_norm": 1.1518983840942383,
+      "learning_rate": 3.708759954493743e-05,
+      "loss": 2.804,
+      "step": 4237
+    },
+    {
+      "epoch": 4.821387940841865,
+      "grad_norm": 1.6885265111923218,
+      "learning_rate": 3.686006825938567e-05,
+      "loss": 2.9798,
+      "step": 4238
+    },
+    {
+      "epoch": 4.822525597269625,
+      "grad_norm": 1.2374640703201294,
+      "learning_rate": 3.6632536973833905e-05,
+      "loss": 1.4952,
+      "step": 4239
+    },
+    {
+      "epoch": 4.823663253697383,
+      "grad_norm": 0.8825571537017822,
+      "learning_rate": 3.6405005688282136e-05,
+      "loss": 1.5783,
+      "step": 4240
+    },
+    {
+      "epoch": 4.824800910125142,
+      "grad_norm": 1.4535282850265503,
+      "learning_rate": 3.617747440273038e-05,
+      "loss": 2.3041,
+      "step": 4241
+    },
+    {
+      "epoch": 4.825938566552901,
+      "grad_norm": 0.9175541400909424,
+      "learning_rate": 3.594994311717861e-05,
+      "loss": 1.5203,
+      "step": 4242
+    },
+    {
+      "epoch": 4.82707622298066,
+      "grad_norm": 0.984664797782898,
+      "learning_rate": 3.5722411831626847e-05,
+      "loss": 2.5147,
+      "step": 4243
+    },
+    {
+      "epoch": 4.828213879408419,
+      "grad_norm": 0.7770801782608032,
+      "learning_rate": 3.549488054607509e-05,
+      "loss": 2.0094,
+      "step": 4244
+    },
+    {
+      "epoch": 4.829351535836177,
+      "grad_norm": 0.9367466568946838,
+      "learning_rate": 3.526734926052332e-05,
+      "loss": 2.4352,
+      "step": 4245
+    },
+    {
+      "epoch": 4.830489192263936,
+      "grad_norm": 0.7870509028434753,
+      "learning_rate": 3.503981797497156e-05,
+      "loss": 2.0384,
+      "step": 4246
+    },
+    {
+      "epoch": 4.831626848691695,
+      "grad_norm": 0.8838803172111511,
+      "learning_rate": 3.4812286689419794e-05,
+      "loss": 1.4871,
+      "step": 4247
+    },
+    {
+      "epoch": 4.832764505119454,
+      "grad_norm": 0.8027423024177551,
+      "learning_rate": 3.458475540386803e-05,
+      "loss": 1.8367,
+      "step": 4248
+    },
+    {
+      "epoch": 4.833902161547213,
+      "grad_norm": 0.9481743574142456,
+      "learning_rate": 3.435722411831627e-05,
+      "loss": 0.9802,
+      "step": 4249
+    },
+    {
+      "epoch": 4.835039817974971,
+      "grad_norm": 0.6448602676391602,
+      "learning_rate": 3.4129692832764505e-05,
+      "loss": 1.3029,
+      "step": 4250
+    },
+    {
+      "epoch": 4.836177474402731,
+      "grad_norm": 0.8399932980537415,
+      "learning_rate": 3.390216154721274e-05,
+      "loss": 2.0227,
+      "step": 4251
+    },
+    {
+      "epoch": 4.837315130830489,
+      "grad_norm": 1.314458966255188,
+      "learning_rate": 3.367463026166098e-05,
+      "loss": 2.1751,
+      "step": 4252
+    },
+    {
+      "epoch": 4.838452787258248,
+      "grad_norm": 0.5397687554359436,
+      "learning_rate": 3.3447098976109216e-05,
+      "loss": 1.035,
+      "step": 4253
+    },
+    {
+      "epoch": 4.839590443686006,
+      "grad_norm": 0.7466420531272888,
+      "learning_rate": 3.3219567690557453e-05,
+      "loss": 1.8664,
+      "step": 4254
+    },
+    {
+      "epoch": 4.840728100113766,
+      "grad_norm": 1.525064468383789,
+      "learning_rate": 3.299203640500569e-05,
+      "loss": 1.8065,
+      "step": 4255
+    },
+    {
+      "epoch": 4.841865756541525,
+      "grad_norm": 1.5727421045303345,
+      "learning_rate": 3.276450511945393e-05,
+      "loss": 1.9101,
+      "step": 4256
+    },
+    {
+      "epoch": 4.843003412969283,
+      "grad_norm": 0.8878625631332397,
+      "learning_rate": 3.253697383390216e-05,
+      "loss": 2.3298,
+      "step": 4257
+    },
+    {
+      "epoch": 4.844141069397042,
+      "grad_norm": 0.9675851464271545,
+      "learning_rate": 3.23094425483504e-05,
+      "loss": 1.6869,
+      "step": 4258
+    },
+    {
+      "epoch": 4.845278725824801,
+      "grad_norm": 1.704845666885376,
+      "learning_rate": 3.208191126279864e-05,
+      "loss": 3.9697,
+      "step": 4259
+    },
+    {
+      "epoch": 4.84641638225256,
+      "grad_norm": 0.6775122284889221,
+      "learning_rate": 3.185437997724687e-05,
+      "loss": 1.1285,
+      "step": 4260
+    },
+    {
+      "epoch": 4.847554038680318,
+      "grad_norm": 0.7678422331809998,
+      "learning_rate": 3.162684869169511e-05,
+      "loss": 1.6568,
+      "step": 4261
+    },
+    {
+      "epoch": 4.848691695108077,
+      "grad_norm": 0.8857121467590332,
+      "learning_rate": 3.139931740614334e-05,
+      "loss": 1.3846,
+      "step": 4262
+    },
+    {
+      "epoch": 4.849829351535837,
+      "grad_norm": 0.8548673987388611,
+      "learning_rate": 3.117178612059158e-05,
+      "loss": 1.6368,
+      "step": 4263
+    },
+    {
+      "epoch": 4.850967007963595,
+      "grad_norm": 0.9651059508323669,
+      "learning_rate": 3.094425483503982e-05,
+      "loss": 2.2593,
+      "step": 4264
+    },
+    {
+      "epoch": 4.852104664391354,
+      "grad_norm": 0.593999445438385,
+      "learning_rate": 3.0716723549488054e-05,
+      "loss": 0.7888,
+      "step": 4265
+    },
+    {
+      "epoch": 4.853242320819112,
+      "grad_norm": 1.1292147636413574,
+      "learning_rate": 3.048919226393629e-05,
+      "loss": 1.7061,
+      "step": 4266
+    },
+    {
+      "epoch": 4.854379977246872,
+      "grad_norm": 1.285089135169983,
+      "learning_rate": 3.026166097838453e-05,
+      "loss": 2.3986,
+      "step": 4267
+    },
+    {
+      "epoch": 4.85551763367463,
+      "grad_norm": 0.9155679941177368,
+      "learning_rate": 3.0034129692832765e-05,
+      "loss": 1.9914,
+      "step": 4268
+    },
+    {
+      "epoch": 4.856655290102389,
+      "grad_norm": 1.3086957931518555,
+      "learning_rate": 2.9806598407281e-05,
+      "loss": 2.3898,
+      "step": 4269
+    },
+    {
+      "epoch": 4.857792946530148,
+      "grad_norm": 1.1196303367614746,
+      "learning_rate": 2.957906712172924e-05,
+      "loss": 3.3304,
+      "step": 4270
+    },
+    {
+      "epoch": 4.858930602957907,
+      "grad_norm": 1.0274990797042847,
+      "learning_rate": 2.9351535836177476e-05,
+      "loss": 1.8834,
+      "step": 4271
+    },
+    {
+      "epoch": 4.860068259385666,
+      "grad_norm": 0.9945041537284851,
+      "learning_rate": 2.912400455062571e-05,
+      "loss": 2.645,
+      "step": 4272
+    },
+    {
+      "epoch": 4.861205915813424,
+      "grad_norm": 0.88654625415802,
+      "learning_rate": 2.889647326507395e-05,
+      "loss": 1.7848,
+      "step": 4273
+    },
+    {
+      "epoch": 4.862343572241183,
+      "grad_norm": 0.6680393815040588,
+      "learning_rate": 2.8668941979522186e-05,
+      "loss": 1.1586,
+      "step": 4274
+    },
+    {
+      "epoch": 4.863481228668942,
+      "grad_norm": 1.0366647243499756,
+      "learning_rate": 2.844141069397042e-05,
+      "loss": 1.2493,
+      "step": 4275
+    },
+    {
+      "epoch": 4.864618885096701,
+      "grad_norm": 1.4107547998428345,
+      "learning_rate": 2.8213879408418657e-05,
+      "loss": 3.2325,
+      "step": 4276
+    },
+    {
+      "epoch": 4.865756541524459,
+      "grad_norm": 0.627693772315979,
+      "learning_rate": 2.7986348122866894e-05,
+      "loss": 0.5508,
+      "step": 4277
+    },
+    {
+      "epoch": 4.8668941979522184,
+      "grad_norm": 0.7929319739341736,
+      "learning_rate": 2.7758816837315134e-05,
+      "loss": 1.5375,
+      "step": 4278
+    },
+    {
+      "epoch": 4.868031854379977,
+      "grad_norm": 1.2130954265594482,
+      "learning_rate": 2.7531285551763368e-05,
+      "loss": 2.0214,
+      "step": 4279
+    },
+    {
+      "epoch": 4.869169510807736,
+      "grad_norm": 0.9208192229270935,
+      "learning_rate": 2.7303754266211605e-05,
+      "loss": 1.7519,
+      "step": 4280
+    },
+    {
+      "epoch": 4.870307167235495,
+      "grad_norm": 1.1531082391738892,
+      "learning_rate": 2.7076222980659842e-05,
+      "loss": 2.057,
+      "step": 4281
+    },
+    {
+      "epoch": 4.8714448236632535,
+      "grad_norm": 1.157972812652588,
+      "learning_rate": 2.6848691695108076e-05,
+      "loss": 2.9342,
+      "step": 4282
+    },
+    {
+      "epoch": 4.872582480091013,
+      "grad_norm": 0.9264168739318848,
+      "learning_rate": 2.6621160409556316e-05,
+      "loss": 1.2696,
+      "step": 4283
+    },
+    {
+      "epoch": 4.873720136518771,
+      "grad_norm": 1.2955060005187988,
+      "learning_rate": 2.6393629124004553e-05,
+      "loss": 3.8855,
+      "step": 4284
+    },
+    {
+      "epoch": 4.87485779294653,
+      "grad_norm": 0.8939307332038879,
+      "learning_rate": 2.6166097838452787e-05,
+      "loss": 1.0266,
+      "step": 4285
+    },
+    {
+      "epoch": 4.8759954493742885,
+      "grad_norm": 0.6583961844444275,
+      "learning_rate": 2.5938566552901024e-05,
+      "loss": 1.4206,
+      "step": 4286
+    },
+    {
+      "epoch": 4.877133105802048,
+      "grad_norm": 0.642296552658081,
+      "learning_rate": 2.571103526734926e-05,
+      "loss": 0.9931,
+      "step": 4287
+    },
+    {
+      "epoch": 4.878270762229807,
+      "grad_norm": 0.9714062809944153,
+      "learning_rate": 2.5483503981797498e-05,
+      "loss": 1.8689,
+      "step": 4288
+    },
+    {
+      "epoch": 4.879408418657565,
+      "grad_norm": 1.0246009826660156,
+      "learning_rate": 2.5255972696245735e-05,
+      "loss": 2.1948,
+      "step": 4289
+    },
+    {
+      "epoch": 4.8805460750853245,
+      "grad_norm": 0.6404163837432861,
+      "learning_rate": 2.502844141069397e-05,
+      "loss": 0.7135,
+      "step": 4290
+    },
+    {
+      "epoch": 4.881683731513083,
+      "grad_norm": 0.6843558549880981,
+      "learning_rate": 2.480091012514221e-05,
+      "loss": 1.1422,
+      "step": 4291
+    },
+    {
+      "epoch": 4.882821387940842,
+      "grad_norm": 0.5428027510643005,
+      "learning_rate": 2.4573378839590442e-05,
+      "loss": 1.0285,
+      "step": 4292
+    },
+    {
+      "epoch": 4.8839590443686,
+      "grad_norm": 1.4773718118667603,
+      "learning_rate": 2.434584755403868e-05,
+      "loss": 1.9624,
+      "step": 4293
+    },
+    {
+      "epoch": 4.8850967007963595,
+      "grad_norm": 1.062320590019226,
+      "learning_rate": 2.411831626848692e-05,
+      "loss": 2.0422,
+      "step": 4294
+    },
+    {
+      "epoch": 4.886234357224119,
+      "grad_norm": 0.9401637315750122,
+      "learning_rate": 2.3890784982935157e-05,
+      "loss": 2.0883,
+      "step": 4295
+    },
+    {
+      "epoch": 4.887372013651877,
+      "grad_norm": 0.9944307804107666,
+      "learning_rate": 2.366325369738339e-05,
+      "loss": 2.0369,
+      "step": 4296
+    },
+    {
+      "epoch": 4.888509670079636,
+      "grad_norm": 1.2259140014648438,
+      "learning_rate": 2.3435722411831627e-05,
+      "loss": 2.3347,
+      "step": 4297
+    },
+    {
+      "epoch": 4.8896473265073945,
+      "grad_norm": 0.7835159301757812,
+      "learning_rate": 2.3208191126279864e-05,
+      "loss": 1.4377,
+      "step": 4298
+    },
+    {
+      "epoch": 4.890784982935154,
+      "grad_norm": 1.1913021802902222,
+      "learning_rate": 2.2980659840728098e-05,
+      "loss": 1.6073,
+      "step": 4299
+    },
+    {
+      "epoch": 4.891922639362912,
+      "grad_norm": 0.95366370677948,
+      "learning_rate": 2.2753128555176338e-05,
+      "loss": 1.8692,
+      "step": 4300
+    },
+    {
+      "epoch": 4.893060295790671,
+      "grad_norm": 0.8128577470779419,
+      "learning_rate": 2.2525597269624575e-05,
+      "loss": 1.3592,
+      "step": 4301
+    },
+    {
+      "epoch": 4.8941979522184305,
+      "grad_norm": 1.3554686307907104,
+      "learning_rate": 2.2298065984072812e-05,
+      "loss": 2.6948,
+      "step": 4302
+    },
+    {
+      "epoch": 4.895335608646189,
+      "grad_norm": 0.8826941847801208,
+      "learning_rate": 2.2070534698521046e-05,
+      "loss": 1.4915,
+      "step": 4303
+    },
+    {
+      "epoch": 4.896473265073948,
+      "grad_norm": 1.1675195693969727,
+      "learning_rate": 2.1843003412969283e-05,
+      "loss": 1.9757,
+      "step": 4304
+    },
+    {
+      "epoch": 4.897610921501706,
+      "grad_norm": 0.7045602798461914,
+      "learning_rate": 2.1615472127417523e-05,
+      "loss": 1.5306,
+      "step": 4305
+    },
+    {
+      "epoch": 4.8987485779294655,
+      "grad_norm": 1.0397857427597046,
+      "learning_rate": 2.1387940841865757e-05,
+      "loss": 1.9285,
+      "step": 4306
+    },
+    {
+      "epoch": 4.899886234357224,
+      "grad_norm": 0.5089128613471985,
+      "learning_rate": 2.1160409556313994e-05,
+      "loss": 0.8025,
+      "step": 4307
+    },
+    {
+      "epoch": 4.901023890784983,
+      "grad_norm": 0.6715580821037292,
+      "learning_rate": 2.093287827076223e-05,
+      "loss": 1.3201,
+      "step": 4308
+    },
+    {
+      "epoch": 4.902161547212742,
+      "grad_norm": 1.0820637941360474,
+      "learning_rate": 2.0705346985210464e-05,
+      "loss": 1.5423,
+      "step": 4309
+    },
+    {
+      "epoch": 4.9032992036405005,
+      "grad_norm": 1.0669132471084595,
+      "learning_rate": 2.04778156996587e-05,
+      "loss": 1.2283,
+      "step": 4310
+    },
+    {
+      "epoch": 4.90443686006826,
+      "grad_norm": 0.9476402401924133,
+      "learning_rate": 2.025028441410694e-05,
+      "loss": 1.5601,
+      "step": 4311
+    },
+    {
+      "epoch": 4.905574516496018,
+      "grad_norm": 0.8250419497489929,
+      "learning_rate": 2.002275312855518e-05,
+      "loss": 1.2876,
+      "step": 4312
+    },
+    {
+      "epoch": 4.906712172923777,
+      "grad_norm": 0.8450213074684143,
+      "learning_rate": 1.9795221843003412e-05,
+      "loss": 1.3217,
+      "step": 4313
+    },
+    {
+      "epoch": 4.907849829351536,
+      "grad_norm": 0.6894076466560364,
+      "learning_rate": 1.956769055745165e-05,
+      "loss": 0.8804,
+      "step": 4314
+    },
+    {
+      "epoch": 4.908987485779295,
+      "grad_norm": 0.8378640413284302,
+      "learning_rate": 1.9340159271899886e-05,
+      "loss": 1.5368,
+      "step": 4315
+    },
+    {
+      "epoch": 4.910125142207053,
+      "grad_norm": 0.8414782285690308,
+      "learning_rate": 1.9112627986348123e-05,
+      "loss": 2.0629,
+      "step": 4316
+    },
+    {
+      "epoch": 4.911262798634812,
+      "grad_norm": 0.7455840706825256,
+      "learning_rate": 1.888509670079636e-05,
+      "loss": 2.0014,
+      "step": 4317
+    },
+    {
+      "epoch": 4.912400455062571,
+      "grad_norm": 0.9223942756652832,
+      "learning_rate": 1.8657565415244597e-05,
+      "loss": 2.2794,
+      "step": 4318
+    },
+    {
+      "epoch": 4.91353811149033,
+      "grad_norm": 0.8256419897079468,
+      "learning_rate": 1.8430034129692834e-05,
+      "loss": 0.9545,
+      "step": 4319
+    },
+    {
+      "epoch": 4.914675767918089,
+      "grad_norm": 0.9742276668548584,
+      "learning_rate": 1.8202502844141068e-05,
+      "loss": 1.5479,
+      "step": 4320
+    },
+    {
+      "epoch": 4.915813424345847,
+      "grad_norm": 0.6871644854545593,
+      "learning_rate": 1.7974971558589305e-05,
+      "loss": 1.025,
+      "step": 4321
+    },
+    {
+      "epoch": 4.9169510807736065,
+      "grad_norm": 0.43966150283813477,
+      "learning_rate": 1.7747440273037545e-05,
+      "loss": 0.4528,
+      "step": 4322
+    },
+    {
+      "epoch": 4.918088737201365,
+      "grad_norm": 0.7050820589065552,
+      "learning_rate": 1.751990898748578e-05,
+      "loss": 1.4931,
+      "step": 4323
+    },
+    {
+      "epoch": 4.919226393629124,
+      "grad_norm": 0.9527431130409241,
+      "learning_rate": 1.7292377701934016e-05,
+      "loss": 1.7784,
+      "step": 4324
+    },
+    {
+      "epoch": 4.920364050056882,
+      "grad_norm": 0.806577742099762,
+      "learning_rate": 1.7064846416382253e-05,
+      "loss": 1.7357,
+      "step": 4325
+    },
+    {
+      "epoch": 4.921501706484642,
+      "grad_norm": 0.8829598426818848,
+      "learning_rate": 1.683731513083049e-05,
+      "loss": 0.9894,
+      "step": 4326
+    },
+    {
+      "epoch": 4.922639362912401,
+      "grad_norm": 0.9835923910140991,
+      "learning_rate": 1.6609783845278727e-05,
+      "loss": 2.0375,
+      "step": 4327
+    },
+    {
+      "epoch": 4.923777019340159,
+      "grad_norm": 0.6324472427368164,
+      "learning_rate": 1.6382252559726964e-05,
+      "loss": 1.0425,
+      "step": 4328
+    },
+    {
+      "epoch": 4.924914675767918,
+      "grad_norm": 1.1519763469696045,
+      "learning_rate": 1.61547212741752e-05,
+      "loss": 1.2597,
+      "step": 4329
+    },
+    {
+      "epoch": 4.926052332195677,
+      "grad_norm": 0.9644741415977478,
+      "learning_rate": 1.5927189988623434e-05,
+      "loss": 1.327,
+      "step": 4330
+    },
+    {
+      "epoch": 4.927189988623436,
+      "grad_norm": 1.3414863348007202,
+      "learning_rate": 1.569965870307167e-05,
+      "loss": 1.7992,
+      "step": 4331
+    },
+    {
+      "epoch": 4.928327645051194,
+      "grad_norm": 1.1579735279083252,
+      "learning_rate": 1.547212741751991e-05,
+      "loss": 2.0279,
+      "step": 4332
+    },
+    {
+      "epoch": 4.929465301478953,
+      "grad_norm": 0.6621242761611938,
+      "learning_rate": 1.5244596131968145e-05,
+      "loss": 1.217,
+      "step": 4333
+    },
+    {
+      "epoch": 4.9306029579067125,
+      "grad_norm": 1.626521348953247,
+      "learning_rate": 1.5017064846416382e-05,
+      "loss": 1.5727,
+      "step": 4334
+    },
+    {
+      "epoch": 4.931740614334471,
+      "grad_norm": 1.1708447933197021,
+      "learning_rate": 1.478953356086462e-05,
+      "loss": 1.681,
+      "step": 4335
+    },
+    {
+      "epoch": 4.93287827076223,
+      "grad_norm": 0.654159426689148,
+      "learning_rate": 1.4562002275312855e-05,
+      "loss": 1.478,
+      "step": 4336
+    },
+    {
+      "epoch": 4.934015927189988,
+      "grad_norm": 0.9996544718742371,
+      "learning_rate": 1.4334470989761093e-05,
+      "loss": 2.0139,
+      "step": 4337
+    },
+    {
+      "epoch": 4.935153583617748,
+      "grad_norm": 0.6572261452674866,
+      "learning_rate": 1.4106939704209329e-05,
+      "loss": 1.5581,
+      "step": 4338
+    },
+    {
+      "epoch": 4.936291240045506,
+      "grad_norm": 0.7390673160552979,
+      "learning_rate": 1.3879408418657567e-05,
+      "loss": 1.218,
+      "step": 4339
+    },
+    {
+      "epoch": 4.937428896473265,
+      "grad_norm": 0.8499306440353394,
+      "learning_rate": 1.3651877133105803e-05,
+      "loss": 2.0041,
+      "step": 4340
+    },
+    {
+      "epoch": 4.938566552901024,
+      "grad_norm": 0.7346335053443909,
+      "learning_rate": 1.3424345847554038e-05,
+      "loss": 1.2743,
+      "step": 4341
+    },
+    {
+      "epoch": 4.939704209328783,
+      "grad_norm": 0.9406227469444275,
+      "learning_rate": 1.3196814562002277e-05,
+      "loss": 1.9248,
+      "step": 4342
+    },
+    {
+      "epoch": 4.940841865756542,
+      "grad_norm": 0.9315640330314636,
+      "learning_rate": 1.2969283276450512e-05,
+      "loss": 1.8989,
+      "step": 4343
+    },
+    {
+      "epoch": 4.9419795221843,
+      "grad_norm": 0.8979254961013794,
+      "learning_rate": 1.2741751990898749e-05,
+      "loss": 1.645,
+      "step": 4344
+    },
+    {
+      "epoch": 4.943117178612059,
+      "grad_norm": 1.1764620542526245,
+      "learning_rate": 1.2514220705346986e-05,
+      "loss": 2.31,
+      "step": 4345
+    },
+    {
+      "epoch": 4.944254835039818,
+      "grad_norm": 0.8805619478225708,
+      "learning_rate": 1.2286689419795221e-05,
+      "loss": 2.1559,
+      "step": 4346
+    },
+    {
+      "epoch": 4.945392491467577,
+      "grad_norm": 2.1326658725738525,
+      "learning_rate": 1.205915813424346e-05,
+      "loss": 4.3989,
+      "step": 4347
+    },
+    {
+      "epoch": 4.946530147895336,
+      "grad_norm": 0.6682014465332031,
+      "learning_rate": 1.1831626848691695e-05,
+      "loss": 1.1768,
+      "step": 4348
+    },
+    {
+      "epoch": 4.947667804323094,
+      "grad_norm": 0.696792721748352,
+      "learning_rate": 1.1604095563139932e-05,
+      "loss": 0.8609,
+      "step": 4349
+    },
+    {
+      "epoch": 4.948805460750854,
+      "grad_norm": 1.4207789897918701,
+      "learning_rate": 1.1376564277588169e-05,
+      "loss": 2.1457,
+      "step": 4350
+    },
+    {
+      "epoch": 4.949943117178612,
+      "grad_norm": 1.3908188343048096,
+      "learning_rate": 1.1149032992036406e-05,
+      "loss": 1.7755,
+      "step": 4351
+    },
+    {
+      "epoch": 4.951080773606371,
+      "grad_norm": 1.002634882926941,
+      "learning_rate": 1.0921501706484641e-05,
+      "loss": 2.3076,
+      "step": 4352
+    },
+    {
+      "epoch": 4.952218430034129,
+      "grad_norm": 1.1484334468841553,
+      "learning_rate": 1.0693970420932878e-05,
+      "loss": 2.1246,
+      "step": 4353
+    },
+    {
+      "epoch": 4.953356086461889,
+      "grad_norm": 1.0049216747283936,
+      "learning_rate": 1.0466439135381115e-05,
+      "loss": 2.0208,
+      "step": 4354
+    },
+    {
+      "epoch": 4.954493742889647,
+      "grad_norm": 0.7231261134147644,
+      "learning_rate": 1.023890784982935e-05,
+      "loss": 0.9476,
+      "step": 4355
+    },
+    {
+      "epoch": 4.955631399317406,
+      "grad_norm": 0.9415931105613708,
+      "learning_rate": 1.001137656427759e-05,
+      "loss": 1.935,
+      "step": 4356
+    },
+    {
+      "epoch": 4.9567690557451645,
+      "grad_norm": 1.4136885404586792,
+      "learning_rate": 9.783845278725825e-06,
+      "loss": 2.5718,
+      "step": 4357
+    },
+    {
+      "epoch": 4.957906712172924,
+      "grad_norm": 0.6967998147010803,
+      "learning_rate": 9.556313993174062e-06,
+      "loss": 0.6904,
+      "step": 4358
+    },
+    {
+      "epoch": 4.959044368600683,
+      "grad_norm": 0.5910859107971191,
+      "learning_rate": 9.328782707622299e-06,
+      "loss": 0.8882,
+      "step": 4359
+    },
+    {
+      "epoch": 4.960182025028441,
+      "grad_norm": 0.9837796688079834,
+      "learning_rate": 9.101251422070534e-06,
+      "loss": 0.7616,
+      "step": 4360
+    },
+    {
+      "epoch": 4.9613196814562,
+      "grad_norm": 0.983348548412323,
+      "learning_rate": 8.873720136518773e-06,
+      "loss": 1.538,
+      "step": 4361
+    },
+    {
+      "epoch": 4.962457337883959,
+      "grad_norm": 0.8036980628967285,
+      "learning_rate": 8.646188850967008e-06,
+      "loss": 1.886,
+      "step": 4362
+    },
+    {
+      "epoch": 4.963594994311718,
+      "grad_norm": 1.007638692855835,
+      "learning_rate": 8.418657565415245e-06,
+      "loss": 2.004,
+      "step": 4363
+    },
+    {
+      "epoch": 4.964732650739476,
+      "grad_norm": 0.9947067499160767,
+      "learning_rate": 8.191126279863482e-06,
+      "loss": 1.8577,
+      "step": 4364
+    },
+    {
+      "epoch": 4.965870307167235,
+      "grad_norm": 1.1273043155670166,
+      "learning_rate": 7.963594994311717e-06,
+      "loss": 1.8335,
+      "step": 4365
+    },
+    {
+      "epoch": 4.967007963594995,
+      "grad_norm": 1.045556664466858,
+      "learning_rate": 7.736063708759956e-06,
+      "loss": 1.9307,
+      "step": 4366
+    },
+    {
+      "epoch": 4.968145620022753,
+      "grad_norm": 1.058106541633606,
+      "learning_rate": 7.508532423208191e-06,
+      "loss": 1.9808,
+      "step": 4367
+    },
+    {
+      "epoch": 4.969283276450512,
+      "grad_norm": 0.9825536608695984,
+      "learning_rate": 7.281001137656427e-06,
+      "loss": 1.0482,
+      "step": 4368
+    },
+    {
+      "epoch": 4.9704209328782705,
+      "grad_norm": 1.0794082880020142,
+      "learning_rate": 7.053469852104664e-06,
+      "loss": 1.8367,
+      "step": 4369
+    },
+    {
+      "epoch": 4.97155858930603,
+      "grad_norm": 1.2737934589385986,
+      "learning_rate": 6.825938566552901e-06,
+      "loss": 2.6611,
+      "step": 4370
+    },
+    {
+      "epoch": 4.972696245733788,
+      "grad_norm": 1.13683021068573,
+      "learning_rate": 6.598407281001138e-06,
+      "loss": 1.9974,
+      "step": 4371
+    },
+    {
+      "epoch": 4.973833902161547,
+      "grad_norm": 1.0530338287353516,
+      "learning_rate": 6.370875995449374e-06,
+      "loss": 1.9576,
+      "step": 4372
+    },
+    {
+      "epoch": 4.974971558589306,
+      "grad_norm": 1.3305796384811401,
+      "learning_rate": 6.1433447098976105e-06,
+      "loss": 3.5862,
+      "step": 4373
+    },
+    {
+      "epoch": 4.976109215017065,
+      "grad_norm": 1.4939963817596436,
+      "learning_rate": 5.9158134243458475e-06,
+      "loss": 2.9359,
+      "step": 4374
+    },
+    {
+      "epoch": 4.977246871444824,
+      "grad_norm": 1.0199118852615356,
+      "learning_rate": 5.6882821387940845e-06,
+      "loss": 1.8068,
+      "step": 4375
+    },
+    {
+      "epoch": 4.978384527872582,
+      "grad_norm": 1.2236491441726685,
+      "learning_rate": 5.460750853242321e-06,
+      "loss": 2.3517,
+      "step": 4376
+    },
+    {
+      "epoch": 4.979522184300341,
+      "grad_norm": 0.7806452512741089,
+      "learning_rate": 5.233219567690558e-06,
+      "loss": 1.238,
+      "step": 4377
+    },
+    {
+      "epoch": 4.9806598407281,
+      "grad_norm": 1.1601351499557495,
+      "learning_rate": 5.005688282138795e-06,
+      "loss": 1.9707,
+      "step": 4378
+    },
+    {
+      "epoch": 4.981797497155859,
+      "grad_norm": 1.6477166414260864,
+      "learning_rate": 4.778156996587031e-06,
+      "loss": 3.1835,
+      "step": 4379
+    },
+    {
+      "epoch": 4.982935153583618,
+      "grad_norm": 0.8821132779121399,
+      "learning_rate": 4.550625711035267e-06,
+      "loss": 1.4507,
+      "step": 4380
+    },
+    {
+      "epoch": 4.9840728100113765,
+      "grad_norm": 0.922042965888977,
+      "learning_rate": 4.323094425483504e-06,
+      "loss": 1.9507,
+      "step": 4381
+    },
+    {
+      "epoch": 4.985210466439136,
+      "grad_norm": 0.9775927662849426,
+      "learning_rate": 4.095563139931741e-06,
+      "loss": 1.4578,
+      "step": 4382
+    },
+    {
+      "epoch": 4.986348122866894,
+      "grad_norm": 1.6098992824554443,
+      "learning_rate": 3.868031854379978e-06,
+      "loss": 2.923,
+      "step": 4383
+    },
+    {
+      "epoch": 4.987485779294653,
+      "grad_norm": 0.938714325428009,
+      "learning_rate": 3.6405005688282136e-06,
+      "loss": 1.6787,
+      "step": 4384
+    },
+    {
+      "epoch": 4.9886234357224115,
+      "grad_norm": 0.8386014103889465,
+      "learning_rate": 3.4129692832764506e-06,
+      "loss": 2.1241,
+      "step": 4385
+    },
+    {
+      "epoch": 4.989761092150171,
+      "grad_norm": 1.0702085494995117,
+      "learning_rate": 3.185437997724687e-06,
+      "loss": 2.7462,
+      "step": 4386
+    },
+    {
+      "epoch": 4.99089874857793,
+      "grad_norm": 0.7873029708862305,
+      "learning_rate": 2.9579067121729238e-06,
+      "loss": 1.3806,
+      "step": 4387
+    },
+    {
+      "epoch": 4.992036405005688,
+      "grad_norm": 0.7936583757400513,
+      "learning_rate": 2.7303754266211603e-06,
+      "loss": 1.4501,
+      "step": 4388
+    },
+    {
+      "epoch": 4.993174061433447,
+      "grad_norm": 1.116550087928772,
+      "learning_rate": 2.5028441410693973e-06,
+      "loss": 3.2445,
+      "step": 4389
+    },
+    {
+      "epoch": 4.994311717861206,
+      "grad_norm": 0.6965498328208923,
+      "learning_rate": 2.2753128555176335e-06,
+      "loss": 1.3281,
+      "step": 4390
+    },
+    {
+      "epoch": 4.995449374288965,
+      "grad_norm": 1.0656460523605347,
+      "learning_rate": 2.0477815699658705e-06,
+      "loss": 2.3715,
+      "step": 4391
+    },
+    {
+      "epoch": 4.996587030716723,
+      "grad_norm": 1.0514278411865234,
+      "learning_rate": 1.8202502844141068e-06,
+      "loss": 1.2846,
+      "step": 4392
+    },
+    {
+      "epoch": 4.9977246871444825,
+      "grad_norm": 1.341645359992981,
+      "learning_rate": 1.5927189988623436e-06,
+      "loss": 1.663,
+      "step": 4393
+    },
+    {
+      "epoch": 4.998862343572241,
+      "grad_norm": 0.9820802807807922,
+      "learning_rate": 1.3651877133105802e-06,
+      "loss": 2.5961,
+      "step": 4394
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.8708580136299133,
+      "learning_rate": 1.1376564277588167e-06,
+      "loss": 1.7838,
+      "step": 4395
+    },
+    {
+      "epoch": 5.0,
+      "eval_f1": 0.8906,
+      "eval_gen_len": 49.5727,
+      "eval_loss": 1.8275777101516724,
+      "eval_precision": 0.8894,
+      "eval_recall": 0.8921,
+      "eval_rouge1": 0.4407,
+      "eval_rouge2": 0.1997,
+      "eval_rougeL": 0.3672,
+      "eval_rougeLsum": 0.4075,
+      "eval_runtime": 28.8614,
+      "eval_samples_per_second": 3.811,
+      "eval_steps_per_second": 0.485,
+      "step": 4395
+    },
+    {
+      "epoch": 5.0,
+      "step": 4395,
+      "total_flos": 4146070071306240.0,
+      "train_loss": 1.9703198982044563,
+      "train_runtime": 950.7032,
+      "train_samples_per_second": 4.623,
+      "train_steps_per_second": 4.623
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4395,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4146070071306240.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}