diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9995621077215007,
+  "eval_steps": 500,
+  "global_step": 1712,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011677127426653043,
+      "grad_norm": 0.30564462406845966,
+      "learning_rate": 1.1627906976744187e-07,
+      "loss": 0.7789,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023354254853306085,
+      "grad_norm": 0.32833540717743703,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 0.8214,
+      "step": 2
+    },
+    {
+      "epoch": 0.003503138227995913,
+      "grad_norm": 0.31266936077720214,
+      "learning_rate": 3.488372093023256e-07,
+      "loss": 0.8189,
+      "step": 3
+    },
+    {
+      "epoch": 0.004670850970661217,
+      "grad_norm": 0.3438620091878376,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.8277,
+      "step": 4
+    },
+    {
+      "epoch": 0.005838563713326522,
+      "grad_norm": 0.3280532190217796,
+      "learning_rate": 5.813953488372094e-07,
+      "loss": 0.8116,
+      "step": 5
+    },
+    {
+      "epoch": 0.007006276455991826,
+      "grad_norm": 0.3553360356646081,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 0.7978,
+      "step": 6
+    },
+    {
+      "epoch": 0.00817398919865713,
+      "grad_norm": 0.32450691964765016,
+      "learning_rate": 8.139534883720931e-07,
+      "loss": 0.8095,
+      "step": 7
+    },
+    {
+      "epoch": 0.009341701941322434,
+      "grad_norm": 0.3015688178468145,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.7871,
+      "step": 8
+    },
+    {
+      "epoch": 0.01050941468398774,
+      "grad_norm": 0.3258273578018379,
+      "learning_rate": 1.0465116279069768e-06,
+      "loss": 0.829,
+      "step": 9
+    },
+    {
+      "epoch": 0.011677127426653043,
+      "grad_norm": 0.2855652037368924,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.784,
+      "step": 10
+    },
+    {
+      "epoch": 0.012844840169318347,
+      "grad_norm": 0.25832037727345997,
+      "learning_rate": 1.2790697674418605e-06,
+      "loss": 0.7616,
+      "step": 11
+    },
+    {
+      "epoch": 0.014012552911983653,
+      "grad_norm": 0.25081924619751567,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.7577,
+      "step": 12
+    },
+    {
+      "epoch": 0.015180265654648957,
+      "grad_norm": 0.23839912562689336,
+      "learning_rate": 1.5116279069767443e-06,
+      "loss": 0.7826,
+      "step": 13
+    },
+    {
+      "epoch": 0.01634797839731426,
+      "grad_norm": 0.20039163059899556,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.8334,
+      "step": 14
+    },
+    {
+      "epoch": 0.017515691139979566,
+      "grad_norm": 0.17960170364046102,
+      "learning_rate": 1.7441860465116282e-06,
+      "loss": 0.7525,
+      "step": 15
+    },
+    {
+      "epoch": 0.018683403882644868,
+      "grad_norm": 0.1734714720130586,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.7721,
+      "step": 16
+    },
+    {
+      "epoch": 0.019851116625310174,
+      "grad_norm": 0.16729359608175276,
+      "learning_rate": 1.976744186046512e-06,
+      "loss": 0.7291,
+      "step": 17
+    },
+    {
+      "epoch": 0.02101882936797548,
+      "grad_norm": 0.16247774534275922,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.7535,
+      "step": 18
+    },
+    {
+      "epoch": 0.02218654211064078,
+      "grad_norm": 0.16934346273234752,
+      "learning_rate": 2.2093023255813954e-06,
+      "loss": 0.7083,
+      "step": 19
+    },
+    {
+      "epoch": 0.023354254853306087,
+      "grad_norm": 0.23921133752125487,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.7557,
+      "step": 20
+    },
+    {
+      "epoch": 0.024521967595971392,
+      "grad_norm": 0.25717306590036343,
+      "learning_rate": 2.4418604651162793e-06,
+      "loss": 0.7408,
+      "step": 21
+    },
+    {
+      "epoch": 0.025689680338636695,
+      "grad_norm": 0.23928803494898124,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.7502,
+      "step": 22
+    },
+    {
+      "epoch": 0.026857393081302,
+      "grad_norm": 0.1914043010138273,
+      "learning_rate": 2.674418604651163e-06,
+      "loss": 0.7056,
+      "step": 23
+    },
+    {
+      "epoch": 0.028025105823967306,
+      "grad_norm": 0.16772002698185992,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.7265,
+      "step": 24
+    },
+    {
+      "epoch": 0.029192818566632608,
+      "grad_norm": 0.143058823913016,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 0.6843,
+      "step": 25
+    },
+    {
+      "epoch": 0.030360531309297913,
+      "grad_norm": 0.1549196627019143,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.7458,
+      "step": 26
+    },
+    {
+      "epoch": 0.031528244051963215,
+      "grad_norm": 0.15318812110518076,
+      "learning_rate": 3.1395348837209307e-06,
+      "loss": 0.6964,
+      "step": 27
+    },
+    {
+      "epoch": 0.03269595679462852,
+      "grad_norm": 0.16399738338386272,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.7299,
+      "step": 28
+    },
+    {
+      "epoch": 0.033863669537293827,
+      "grad_norm": 0.16915517016357967,
+      "learning_rate": 3.372093023255814e-06,
+      "loss": 0.6873,
+      "step": 29
+    },
+    {
+      "epoch": 0.03503138227995913,
+      "grad_norm": 0.1632982937823785,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.661,
+      "step": 30
+    },
+    {
+      "epoch": 0.03619909502262444,
+      "grad_norm": 0.14916630228641206,
+      "learning_rate": 3.6046511627906977e-06,
+      "loss": 0.652,
+      "step": 31
+    },
+    {
+      "epoch": 0.037366807765289736,
+      "grad_norm": 0.1428655701120444,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.6855,
+      "step": 32
+    },
+    {
+      "epoch": 0.03853452050795504,
+      "grad_norm": 0.12995605431744386,
+      "learning_rate": 3.837209302325582e-06,
+      "loss": 0.6721,
+      "step": 33
+    },
+    {
+      "epoch": 0.03970223325062035,
+      "grad_norm": 0.134016447795876,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.6353,
+      "step": 34
+    },
+    {
+      "epoch": 0.04086994599328565,
+      "grad_norm": 0.12410669363513019,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 0.6775,
+      "step": 35
+    },
+    {
+      "epoch": 0.04203765873595096,
+      "grad_norm": 0.12620284226721423,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.6385,
+      "step": 36
+    },
+    {
+      "epoch": 0.04320537147861626,
+      "grad_norm": 0.12124188259218677,
+      "learning_rate": 4.302325581395349e-06,
+      "loss": 0.6376,
+      "step": 37
+    },
+    {
+      "epoch": 0.04437308422128156,
+      "grad_norm": 0.12316921627782162,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.6565,
+      "step": 38
+    },
+    {
+      "epoch": 0.04554079696394687,
+      "grad_norm": 0.12711614520871609,
+      "learning_rate": 4.5348837209302326e-06,
+      "loss": 0.6736,
+      "step": 39
+    },
+    {
+      "epoch": 0.046708509706612174,
+      "grad_norm": 0.12498276444104262,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.7088,
+      "step": 40
+    },
+    {
+      "epoch": 0.04787622244927748,
+      "grad_norm": 0.11560834235731308,
+      "learning_rate": 4.767441860465117e-06,
+      "loss": 0.6676,
+      "step": 41
+    },
+    {
+      "epoch": 0.049043935191942785,
+      "grad_norm": 0.10654434189128426,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.6268,
+      "step": 42
+    },
+    {
+      "epoch": 0.050211647934608084,
+      "grad_norm": 0.11883606837428971,
+      "learning_rate": 5e-06,
+      "loss": 0.6597,
+      "step": 43
+    },
+    {
+      "epoch": 0.05137936067727339,
+      "grad_norm": 0.1114144504504184,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.6257,
+      "step": 44
+    },
+    {
+      "epoch": 0.052547073419938695,
+      "grad_norm": 0.11260561068540641,
+      "learning_rate": 5.232558139534885e-06,
+      "loss": 0.6573,
+      "step": 45
+    },
+    {
+      "epoch": 0.053714786162604,
+      "grad_norm": 0.10509147817504694,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.6026,
+      "step": 46
+    },
+    {
+      "epoch": 0.054882498905269306,
+      "grad_norm": 0.10522690062903195,
+      "learning_rate": 5.465116279069767e-06,
+      "loss": 0.6038,
+      "step": 47
+    },
+    {
+      "epoch": 0.05605021164793461,
+      "grad_norm": 0.10764356942965987,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.6487,
+      "step": 48
+    },
+    {
+      "epoch": 0.05721792439059991,
+      "grad_norm": 0.11231343319128377,
+      "learning_rate": 5.697674418604652e-06,
+      "loss": 0.6714,
+      "step": 49
+    },
+    {
+      "epoch": 0.058385637133265215,
+      "grad_norm": 0.10372259404529237,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.5893,
+      "step": 50
+    },
+    {
+      "epoch": 0.05955334987593052,
+      "grad_norm": 0.10166950941063504,
+      "learning_rate": 5.930232558139536e-06,
+      "loss": 0.6024,
+      "step": 51
+    },
+    {
+      "epoch": 0.06072106261859583,
+      "grad_norm": 0.10193864871361023,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.6066,
+      "step": 52
+    },
+    {
+      "epoch": 0.06188877536126113,
+      "grad_norm": 0.10317965605538539,
+      "learning_rate": 6.162790697674419e-06,
+      "loss": 0.6131,
+      "step": 53
+    },
+    {
+      "epoch": 0.06305648810392643,
+      "grad_norm": 0.10941239434280968,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.6236,
+      "step": 54
+    },
+    {
+      "epoch": 0.06422420084659174,
+      "grad_norm": 0.09975106071253902,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 0.5777,
+      "step": 55
+    },
+    {
+      "epoch": 0.06539191358925704,
+      "grad_norm": 0.10389556834832771,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.6096,
+      "step": 56
+    },
+    {
+      "epoch": 0.06655962633192235,
+      "grad_norm": 0.10560366404930781,
+      "learning_rate": 6.627906976744186e-06,
+      "loss": 0.6409,
+      "step": 57
+    },
+    {
+      "epoch": 0.06772733907458765,
+      "grad_norm": 0.11681843674452609,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.6201,
+      "step": 58
+    },
+    {
+      "epoch": 0.06889505181725296,
+      "grad_norm": 0.10585504126991926,
+      "learning_rate": 6.86046511627907e-06,
+      "loss": 0.6284,
+      "step": 59
+    },
+    {
+      "epoch": 0.07006276455991826,
+      "grad_norm": 0.1032381740905318,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.591,
+      "step": 60
+    },
+    {
+      "epoch": 0.07123047730258357,
+      "grad_norm": 0.10850984084343966,
+      "learning_rate": 7.0930232558139545e-06,
+      "loss": 0.5918,
+      "step": 61
+    },
+    {
+      "epoch": 0.07239819004524888,
+      "grad_norm": 0.1037896269857189,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.6177,
+      "step": 62
+    },
+    {
+      "epoch": 0.07356590278791417,
+      "grad_norm": 0.10569396486443379,
+      "learning_rate": 7.325581395348837e-06,
+      "loss": 0.6054,
+      "step": 63
+    },
+    {
+      "epoch": 0.07473361553057947,
+      "grad_norm": 0.10275839245475688,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.6456,
+      "step": 64
+    },
+    {
+      "epoch": 0.07590132827324478,
+      "grad_norm": 0.11474187417659724,
+      "learning_rate": 7.5581395348837215e-06,
+      "loss": 0.5847,
+      "step": 65
+    },
+    {
+      "epoch": 0.07706904101591008,
+      "grad_norm": 0.10441478686516256,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.5814,
+      "step": 66
+    },
+    {
+      "epoch": 0.07823675375857539,
+      "grad_norm": 0.10112190058952489,
+      "learning_rate": 7.790697674418605e-06,
+      "loss": 0.6262,
+      "step": 67
+    },
+    {
+      "epoch": 0.0794044665012407,
+      "grad_norm": 0.09739217624425252,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.5865,
+      "step": 68
+    },
+    {
+      "epoch": 0.080572179243906,
+      "grad_norm": 0.11964031763510963,
+      "learning_rate": 8.023255813953488e-06,
+      "loss": 0.6556,
+      "step": 69
+    },
+    {
+      "epoch": 0.0817398919865713,
+      "grad_norm": 0.10217521138884825,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.6072,
+      "step": 70
+    },
+    {
+      "epoch": 0.08290760472923661,
+      "grad_norm": 0.09602073381876323,
+      "learning_rate": 8.255813953488374e-06,
+      "loss": 0.5741,
+      "step": 71
+    },
+    {
+      "epoch": 0.08407531747190192,
+      "grad_norm": 0.10774601692683897,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.6023,
+      "step": 72
+    },
+    {
+      "epoch": 0.08524303021456722,
+      "grad_norm": 0.10503799831204726,
+      "learning_rate": 8.488372093023256e-06,
+      "loss": 0.5711,
+      "step": 73
+    },
+    {
+      "epoch": 0.08641074295723251,
+      "grad_norm": 0.10116435608645177,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.5793,
+      "step": 74
+    },
+    {
+      "epoch": 0.08757845569989782,
+      "grad_norm": 0.10415600732331867,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 0.5981,
+      "step": 75
+    },
+    {
+      "epoch": 0.08874616844256313,
+      "grad_norm": 0.10905014609493073,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.6114,
+      "step": 76
+    },
+    {
+      "epoch": 0.08991388118522843,
+      "grad_norm": 0.10231591309268155,
+      "learning_rate": 8.953488372093024e-06,
+      "loss": 0.5968,
+      "step": 77
+    },
+    {
+      "epoch": 0.09108159392789374,
+      "grad_norm": 0.10772740662845653,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.5904,
+      "step": 78
+    },
+    {
+      "epoch": 0.09224930667055904,
+      "grad_norm": 0.11613080779076006,
+      "learning_rate": 9.186046511627908e-06,
+      "loss": 0.6312,
+      "step": 79
+    },
+    {
+      "epoch": 0.09341701941322435,
+      "grad_norm": 0.10555992507957129,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.5812,
+      "step": 80
+    },
+    {
+      "epoch": 0.09458473215588965,
+      "grad_norm": 0.11436251664825016,
+      "learning_rate": 9.418604651162791e-06,
+      "loss": 0.616,
+      "step": 81
+    },
+    {
+      "epoch": 0.09575244489855496,
+      "grad_norm": 0.10973879500489721,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.5859,
+      "step": 82
+    },
+    {
+      "epoch": 0.09692015764122026,
+      "grad_norm": 0.10480095235076241,
+      "learning_rate": 9.651162790697676e-06,
+      "loss": 0.6065,
+      "step": 83
+    },
+    {
+      "epoch": 0.09808787038388557,
+      "grad_norm": 0.10525862586860382,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.5755,
+      "step": 84
+    },
+    {
+      "epoch": 0.09925558312655088,
+      "grad_norm": 0.11396935995146483,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 0.5912,
+      "step": 85
+    },
+    {
+      "epoch": 0.10042329586921617,
+      "grad_norm": 0.10434454454033827,
+      "learning_rate": 1e-05,
+      "loss": 0.5916,
+      "step": 86
+    },
+    {
+      "epoch": 0.10159100861188147,
+      "grad_norm": 0.10973805447790101,
+      "learning_rate": 1.0116279069767443e-05,
+      "loss": 0.6364,
+      "step": 87
+    },
+    {
+      "epoch": 0.10275872135454678,
+      "grad_norm": 0.10608098146069943,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 0.5854,
+      "step": 88
+    },
+    {
+      "epoch": 0.10392643409721208,
+      "grad_norm": 0.11196306241994648,
+      "learning_rate": 1.0348837209302327e-05,
+      "loss": 0.643,
+      "step": 89
+    },
+    {
+      "epoch": 0.10509414683987739,
+      "grad_norm": 0.10877344323591649,
+      "learning_rate": 1.046511627906977e-05,
+      "loss": 0.6185,
+      "step": 90
+    },
+    {
+      "epoch": 0.1062618595825427,
+      "grad_norm": 0.10768290489094381,
+      "learning_rate": 1.058139534883721e-05,
+      "loss": 0.6067,
+      "step": 91
+    },
+    {
+      "epoch": 0.107429572325208,
+      "grad_norm": 0.10872599119841052,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 0.5719,
+      "step": 92
+    },
+    {
+      "epoch": 0.1085972850678733,
+      "grad_norm": 0.11252737355361118,
+      "learning_rate": 1.0813953488372094e-05,
+      "loss": 0.5806,
+      "step": 93
+    },
+    {
+      "epoch": 0.10976499781053861,
+      "grad_norm": 0.10808602695875576,
+      "learning_rate": 1.0930232558139535e-05,
+      "loss": 0.6048,
+      "step": 94
+    },
+    {
+      "epoch": 0.11093271055320392,
+      "grad_norm": 0.10699068463024544,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 0.5772,
+      "step": 95
+    },
+    {
+      "epoch": 0.11210042329586922,
+      "grad_norm": 0.11568124535087292,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 0.604,
+      "step": 96
+    },
+    {
+      "epoch": 0.11326813603853451,
+      "grad_norm": 0.1138019166447473,
+      "learning_rate": 1.1279069767441861e-05,
+      "loss": 0.6094,
+      "step": 97
+    },
+    {
+      "epoch": 0.11443584878119982,
+      "grad_norm": 0.11001239735145404,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 0.6037,
+      "step": 98
+    },
+    {
+      "epoch": 0.11560356152386513,
+      "grad_norm": 0.10240225550519928,
+      "learning_rate": 1.1511627906976746e-05,
+      "loss": 0.5822,
+      "step": 99
+    },
+    {
+      "epoch": 0.11677127426653043,
+      "grad_norm": 0.11125282535515749,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.5612,
+      "step": 100
+    },
+    {
+      "epoch": 0.11793898700919574,
+      "grad_norm": 0.10334910049359279,
+      "learning_rate": 1.174418604651163e-05,
+      "loss": 0.5733,
+      "step": 101
+    },
+    {
+      "epoch": 0.11910669975186104,
+      "grad_norm": 0.1106538692610509,
+      "learning_rate": 1.1860465116279072e-05,
+      "loss": 0.578,
+      "step": 102
+    },
+    {
+      "epoch": 0.12027441249452635,
+      "grad_norm": 0.109982250805842,
+      "learning_rate": 1.1976744186046511e-05,
+      "loss": 0.5609,
+      "step": 103
+    },
+    {
+      "epoch": 0.12144212523719165,
+      "grad_norm": 0.11201483656298437,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 0.557,
+      "step": 104
+    },
+    {
+      "epoch": 0.12260983797985696,
+      "grad_norm": 0.11586845140534063,
+      "learning_rate": 1.2209302325581397e-05,
+      "loss": 0.6105,
+      "step": 105
+    },
+    {
+      "epoch": 0.12377755072252226,
+      "grad_norm": 0.10861531177190674,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 0.5498,
+      "step": 106
+    },
+    {
+      "epoch": 0.12494526346518757,
+      "grad_norm": 0.11689355857574382,
+      "learning_rate": 1.244186046511628e-05,
+      "loss": 0.5739,
+      "step": 107
+    },
+    {
+      "epoch": 0.12611297620785286,
+      "grad_norm": 0.10958605556349243,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 0.5781,
+      "step": 108
+    },
+    {
+      "epoch": 0.12728068895051817,
+      "grad_norm": 0.11490281561278697,
+      "learning_rate": 1.2674418604651164e-05,
+      "loss": 0.5985,
+      "step": 109
+    },
+    {
+      "epoch": 0.12844840169318347,
+      "grad_norm": 0.11252917714187195,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 0.5831,
+      "step": 110
+    },
+    {
+      "epoch": 0.12961611443584878,
+      "grad_norm": 0.10632268862482831,
+      "learning_rate": 1.2906976744186049e-05,
+      "loss": 0.5688,
+      "step": 111
+    },
+    {
+      "epoch": 0.13078382717851408,
+      "grad_norm": 0.11322847083844126,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 0.5746,
+      "step": 112
+    },
+    {
+      "epoch": 0.1319515399211794,
+      "grad_norm": 0.1085155206736773,
+      "learning_rate": 1.313953488372093e-05,
+      "loss": 0.5879,
+      "step": 113
+    },
+    {
+      "epoch": 0.1331192526638447,
+      "grad_norm": 0.10353255995650625,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 0.5861,
+      "step": 114
+    },
+    {
+      "epoch": 0.13428696540651,
+      "grad_norm": 0.11691679243373448,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 0.6003,
+      "step": 115
+    },
+    {
+      "epoch": 0.1354546781491753,
+      "grad_norm": 0.10622719934159856,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 0.5442,
+      "step": 116
+    },
+    {
+      "epoch": 0.1366223908918406,
+      "grad_norm": 0.11353339666494162,
+      "learning_rate": 1.36046511627907e-05,
+      "loss": 0.6002,
+      "step": 117
+    },
+    {
+      "epoch": 0.13779010363450592,
+      "grad_norm": 0.1132531379443118,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 0.5964,
+      "step": 118
+    },
+    {
+      "epoch": 0.13895781637717122,
+      "grad_norm": 0.10729109513336833,
+      "learning_rate": 1.3837209302325583e-05,
+      "loss": 0.5631,
+      "step": 119
+    },
+    {
+      "epoch": 0.14012552911983653,
+      "grad_norm": 0.11625992760611893,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 0.5629,
+      "step": 120
+    },
+    {
+      "epoch": 0.14129324186250183,
+      "grad_norm": 0.11996175886502079,
+      "learning_rate": 1.4069767441860466e-05,
+      "loss": 0.6122,
+      "step": 121
+    },
+    {
+      "epoch": 0.14246095460516714,
+      "grad_norm": 0.11117745837967138,
+      "learning_rate": 1.4186046511627909e-05,
+      "loss": 0.5767,
+      "step": 122
+    },
+    {
+      "epoch": 0.14362866734783244,
+      "grad_norm": 0.11349775083476304,
+      "learning_rate": 1.4302325581395352e-05,
+      "loss": 0.5498,
+      "step": 123
+    },
+    {
+      "epoch": 0.14479638009049775,
+      "grad_norm": 0.11035578729101571,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 0.5825,
+      "step": 124
+    },
+    {
+      "epoch": 0.14596409283316303,
+      "grad_norm": 0.11069211832448793,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 0.5581,
+      "step": 125
+    },
+    {
+      "epoch": 0.14713180557582833,
+      "grad_norm": 0.12487620795488663,
+      "learning_rate": 1.4651162790697674e-05,
+      "loss": 0.6035,
+      "step": 126
+    },
+    {
+      "epoch": 0.14829951831849364,
+      "grad_norm": 0.11920735258567208,
+      "learning_rate": 1.4767441860465117e-05,
+      "loss": 0.5831,
+      "step": 127
+    },
+    {
+      "epoch": 0.14946723106115894,
+      "grad_norm": 0.09999991800198216,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 0.5692,
+      "step": 128
+    },
+    {
+      "epoch": 0.15063494380382425,
+      "grad_norm": 0.11809737394191513,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.5965,
+      "step": 129
+    },
+    {
+      "epoch": 0.15180265654648956,
+      "grad_norm": 0.11373229461424854,
+      "learning_rate": 1.5116279069767443e-05,
+      "loss": 0.5808,
+      "step": 130
+    },
+    {
+      "epoch": 0.15297036928915486,
+      "grad_norm": 0.11755585982189523,
+      "learning_rate": 1.5232558139534886e-05,
+      "loss": 0.5529,
+      "step": 131
+    },
+    {
+      "epoch": 0.15413808203182017,
+      "grad_norm": 0.1143749509110744,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 0.5683,
+      "step": 132
+    },
+    {
+      "epoch": 0.15530579477448547,
+      "grad_norm": 0.10617262705915796,
+      "learning_rate": 1.546511627906977e-05,
+      "loss": 0.5596,
+      "step": 133
+    },
+    {
+      "epoch": 0.15647350751715078,
+      "grad_norm": 0.11487677905649916,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 0.5595,
+      "step": 134
+    },
+    {
+      "epoch": 0.15764122025981608,
+      "grad_norm": 0.10593112855506719,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 0.5808,
+      "step": 135
+    },
+    {
+      "epoch": 0.1588089330024814,
+      "grad_norm": 0.11519963367375394,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 0.5907,
+      "step": 136
+    },
+    {
+      "epoch": 0.1599766457451467,
+      "grad_norm": 0.11800612892986873,
+      "learning_rate": 1.5930232558139536e-05,
+      "loss": 0.5801,
+      "step": 137
+    },
+    {
+      "epoch": 0.161144358487812,
+      "grad_norm": 0.10653916315095806,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 0.5429,
+      "step": 138
+    },
+    {
+      "epoch": 0.1623120712304773,
+      "grad_norm": 0.12286810403897815,
+      "learning_rate": 1.616279069767442e-05,
+      "loss": 0.5356,
+      "step": 139
+    },
+    {
+      "epoch": 0.1634797839731426,
+      "grad_norm": 0.11191931847950809,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.5575,
+      "step": 140
+    },
+    {
+      "epoch": 0.16464749671580792,
+      "grad_norm": 0.1166951103523856,
+      "learning_rate": 1.6395348837209303e-05,
+      "loss": 0.5686,
+      "step": 141
+    },
+    {
+      "epoch": 0.16581520945847322,
+      "grad_norm": 0.1121165550696581,
+      "learning_rate": 1.6511627906976747e-05,
+      "loss": 0.578,
+      "step": 142
+    },
+    {
+      "epoch": 0.16698292220113853,
+      "grad_norm": 0.09935706390568835,
+      "learning_rate": 1.6627906976744188e-05,
+      "loss": 0.5738,
+      "step": 143
+    },
+    {
+      "epoch": 0.16815063494380383,
+      "grad_norm": 0.12579530644763948,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 0.5535,
+      "step": 144
+    },
+    {
+      "epoch": 0.16931834768646914,
+      "grad_norm": 0.10944282160312462,
+      "learning_rate": 1.6860465116279073e-05,
+      "loss": 0.5546,
+      "step": 145
+    },
+    {
+      "epoch": 0.17048606042913445,
+      "grad_norm": 0.11257605163324247,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 0.5956,
+      "step": 146
+    },
+    {
+      "epoch": 0.17165377317179975,
+      "grad_norm": 0.11230661159774834,
+      "learning_rate": 1.7093023255813955e-05,
+      "loss": 0.5666,
+      "step": 147
+    },
+    {
+      "epoch": 0.17282148591446503,
+      "grad_norm": 0.11421498469122787,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.5759,
+      "step": 148
+    },
+    {
+      "epoch": 0.17398919865713033,
+      "grad_norm": 0.10913743207065967,
+      "learning_rate": 1.7325581395348837e-05,
+      "loss": 0.5562,
+      "step": 149
+    },
+    {
+      "epoch": 0.17515691139979564,
+      "grad_norm": 0.10926453129073088,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 0.5698,
+      "step": 150
+    },
+    {
+      "epoch": 0.17632462414246095,
+      "grad_norm": 0.12144091036617567,
+      "learning_rate": 1.7558139534883722e-05,
+      "loss": 0.5897,
+      "step": 151
+    },
+    {
+      "epoch": 0.17749233688512625,
+      "grad_norm": 0.11115372786472467,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 0.5788,
+      "step": 152
+    },
+    {
+      "epoch": 0.17866004962779156,
+      "grad_norm": 0.12501837602022495,
+      "learning_rate": 1.7790697674418608e-05,
+      "loss": 0.6025,
+      "step": 153
+    },
+    {
+      "epoch": 0.17982776237045686,
+      "grad_norm": 0.10557510099430313,
+      "learning_rate": 1.790697674418605e-05,
+      "loss": 0.5715,
+      "step": 154
+    },
+    {
+      "epoch": 0.18099547511312217,
+      "grad_norm": 0.10495801312455112,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 0.5672,
+      "step": 155
+    },
+    {
+      "epoch": 0.18216318785578747,
+      "grad_norm": 0.10699126762988374,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 0.5717,
+      "step": 156
+    },
+    {
+      "epoch": 0.18333090059845278,
+      "grad_norm": 0.10733875477946303,
+      "learning_rate": 1.825581395348837e-05,
+      "loss": 0.5782,
+      "step": 157
+    },
+    {
+      "epoch": 0.18449861334111808,
+      "grad_norm": 0.11461042518969218,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 0.5946,
+      "step": 158
+    },
+    {
+      "epoch": 0.1856663260837834,
+      "grad_norm": 0.1182071609356666,
+      "learning_rate": 1.8488372093023256e-05,
+      "loss": 0.5539,
+      "step": 159
+    },
+    {
+      "epoch": 0.1868340388264487,
+      "grad_norm": 0.11565838271541584,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 0.5729,
+      "step": 160
+    },
+    {
+      "epoch": 0.188001751569114,
+      "grad_norm": 0.10947654482446578,
+      "learning_rate": 1.872093023255814e-05,
+      "loss": 0.5833,
+      "step": 161
+    },
+    {
+      "epoch": 0.1891694643117793,
+      "grad_norm": 0.10633133353729567,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 0.5407,
+      "step": 162
+    },
+    {
+      "epoch": 0.1903371770544446,
+      "grad_norm": 0.11714041544407103,
+      "learning_rate": 1.8953488372093027e-05,
+      "loss": 0.5984,
+      "step": 163
+    },
+    {
+      "epoch": 0.19150488979710992,
+      "grad_norm": 0.10764057713012003,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.5681,
+      "step": 164
+    },
+    {
+      "epoch": 0.19267260253977522,
+      "grad_norm": 0.11319428414525053,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 0.6109,
+      "step": 165
+    },
+    {
+      "epoch": 0.19384031528244053,
+      "grad_norm": 0.10734302624506814,
+      "learning_rate": 1.9302325581395353e-05,
+      "loss": 0.5843,
+      "step": 166
+    },
+    {
+      "epoch": 0.19500802802510583,
+      "grad_norm": 0.12052719879146287,
+      "learning_rate": 1.941860465116279e-05,
+      "loss": 0.5898,
+      "step": 167
+    },
+    {
+      "epoch": 0.19617574076777114,
+      "grad_norm": 0.10473976910837862,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 0.5884,
+      "step": 168
+    },
+    {
+      "epoch": 0.19734345351043645,
+      "grad_norm": 0.12348269847643348,
+      "learning_rate": 1.9651162790697676e-05,
+      "loss": 0.5605,
+      "step": 169
+    },
+    {
+      "epoch": 0.19851116625310175,
+      "grad_norm": 0.10883064889144985,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.5916,
+      "step": 170
+    },
+    {
+      "epoch": 0.19967887899576703,
+      "grad_norm": 0.11520640627258975,
+      "learning_rate": 1.988372093023256e-05,
+      "loss": 0.5977,
+      "step": 171
+    },
+    {
+      "epoch": 0.20084659173843233,
+      "grad_norm": 0.11327775743602798,
+      "learning_rate": 2e-05,
+      "loss": 0.5868,
+      "step": 172
+    },
+    {
+      "epoch": 0.20201430448109764,
+      "grad_norm": 0.1137302824544105,
+      "learning_rate": 1.9999979192104532e-05,
+      "loss": 0.5579,
+      "step": 173
+    },
+    {
+      "epoch": 0.20318201722376295,
+      "grad_norm": 0.11295807026074281,
+      "learning_rate": 1.9999916768504724e-05,
+      "loss": 0.5925,
+      "step": 174
+    },
+    {
+      "epoch": 0.20434972996642825,
+      "grad_norm": 0.10785377488578082,
+      "learning_rate": 1.9999812729460353e-05,
+      "loss": 0.5983,
+      "step": 175
+    },
+    {
+      "epoch": 0.20551744270909356,
+      "grad_norm": 0.11723638113975465,
+      "learning_rate": 1.9999667075404385e-05,
+      "loss": 0.5664,
+      "step": 176
+    },
+    {
+      "epoch": 0.20668515545175886,
+      "grad_norm": 0.10695390689877542,
+      "learning_rate": 1.9999479806942976e-05,
+      "loss": 0.5963,
+      "step": 177
+    },
+    {
+      "epoch": 0.20785286819442417,
+      "grad_norm": 0.11800177754670883,
+      "learning_rate": 1.9999250924855458e-05,
+      "loss": 0.578,
+      "step": 178
+    },
+    {
+      "epoch": 0.20902058093708947,
+      "grad_norm": 0.11912320262310404,
+      "learning_rate": 1.9998980430094333e-05,
+      "loss": 0.5749,
+      "step": 179
+    },
+    {
+      "epoch": 0.21018829367975478,
+      "grad_norm": 0.10610763260891623,
+      "learning_rate": 1.9998668323785298e-05,
+      "loss": 0.5423,
+      "step": 180
+    },
+    {
+      "epoch": 0.21135600642242008,
+      "grad_norm": 0.12335270278738927,
+      "learning_rate": 1.9998314607227203e-05,
+      "loss": 0.6018,
+      "step": 181
+    },
+    {
+      "epoch": 0.2125237191650854,
+      "grad_norm": 0.10246185441641448,
+      "learning_rate": 1.9997919281892066e-05,
+      "loss": 0.5373,
+      "step": 182
+    },
+    {
+      "epoch": 0.2136914319077507,
+      "grad_norm": 0.11479865168009573,
+      "learning_rate": 1.999748234942507e-05,
+      "loss": 0.5829,
+      "step": 183
+    },
+    {
+      "epoch": 0.214859144650416,
+      "grad_norm": 0.12046285501999437,
+      "learning_rate": 1.9997003811644534e-05,
+      "loss": 0.5841,
+      "step": 184
+    },
+    {
+      "epoch": 0.2160268573930813,
+      "grad_norm": 0.12683630224767456,
+      "learning_rate": 1.999648367054194e-05,
+      "loss": 0.5579,
+      "step": 185
+    },
+    {
+      "epoch": 0.2171945701357466,
+      "grad_norm": 0.11931662682168222,
+      "learning_rate": 1.9995921928281893e-05,
+      "loss": 0.5791,
+      "step": 186
+    },
+    {
+      "epoch": 0.21836228287841192,
+      "grad_norm": 0.11937256426577998,
+      "learning_rate": 1.9995318587202132e-05,
+      "loss": 0.5805,
+      "step": 187
+    },
+    {
+      "epoch": 0.21952999562107722,
+      "grad_norm": 0.13247451694331203,
+      "learning_rate": 1.99946736498135e-05,
+      "loss": 0.5978,
+      "step": 188
+    },
+    {
+      "epoch": 0.22069770836374253,
+      "grad_norm": 0.09798789179194178,
+      "learning_rate": 1.999398711879996e-05,
+      "loss": 0.5467,
+      "step": 189
+    },
+    {
+      "epoch": 0.22186542110640783,
+      "grad_norm": 0.15692432711442875,
+      "learning_rate": 1.9993258997018564e-05,
+      "loss": 0.5659,
+      "step": 190
+    },
+    {
+      "epoch": 0.22303313384907314,
+      "grad_norm": 0.10574969838966325,
+      "learning_rate": 1.9992489287499454e-05,
+      "loss": 0.5492,
+      "step": 191
+    },
+    {
+      "epoch": 0.22420084659173845,
+      "grad_norm": 0.12497207286074621,
+      "learning_rate": 1.9991677993445832e-05,
+      "loss": 0.5537,
+      "step": 192
+    },
+    {
+      "epoch": 0.22536855933440372,
+      "grad_norm": 0.12144733350351312,
+      "learning_rate": 1.9990825118233958e-05,
+      "loss": 0.5589,
+      "step": 193
+    },
+    {
+      "epoch": 0.22653627207706903,
+      "grad_norm": 0.12261856348321301,
+      "learning_rate": 1.9989930665413148e-05,
+      "loss": 0.5784,
+      "step": 194
+    },
+    {
+      "epoch": 0.22770398481973433,
+      "grad_norm": 0.12086552268607476,
+      "learning_rate": 1.9988994638705734e-05,
+      "loss": 0.5304,
+      "step": 195
+    },
+    {
+      "epoch": 0.22887169756239964,
+      "grad_norm": 0.10646352289246058,
+      "learning_rate": 1.9988017042007066e-05,
+      "loss": 0.5619,
+      "step": 196
+    },
+    {
+      "epoch": 0.23003941030506495,
+      "grad_norm": 0.11632453380800095,
+      "learning_rate": 1.9986997879385488e-05,
+      "loss": 0.5611,
+      "step": 197
+    },
+    {
+      "epoch": 0.23120712304773025,
+      "grad_norm": 0.10851656696993678,
+      "learning_rate": 1.998593715508233e-05,
+      "loss": 0.607,
+      "step": 198
+    },
+    {
+      "epoch": 0.23237483579039556,
+      "grad_norm": 0.10575377583870114,
+      "learning_rate": 1.9984834873511873e-05,
+      "loss": 0.533,
+      "step": 199
+    },
+    {
+      "epoch": 0.23354254853306086,
+      "grad_norm": 0.11360633311836489,
+      "learning_rate": 1.9983691039261358e-05,
+      "loss": 0.5561,
+      "step": 200
+    },
+    {
+      "epoch": 0.23471026127572617,
+      "grad_norm": 0.10651665254006461,
+      "learning_rate": 1.9982505657090936e-05,
+      "loss": 0.5777,
+      "step": 201
+    },
+    {
+      "epoch": 0.23587797401839147,
+      "grad_norm": 0.11968477538578834,
+      "learning_rate": 1.998127873193367e-05,
+      "loss": 0.5683,
+      "step": 202
+    },
+    {
+      "epoch": 0.23704568676105678,
+      "grad_norm": 0.11277577777595266,
+      "learning_rate": 1.9980010268895507e-05,
+      "loss": 0.5802,
+      "step": 203
+    },
+    {
+      "epoch": 0.23821339950372208,
+      "grad_norm": 0.1049714416283311,
+      "learning_rate": 1.9978700273255254e-05,
+      "loss": 0.5746,
+      "step": 204
+    },
+    {
+      "epoch": 0.2393811122463874,
+      "grad_norm": 0.11447880486494638,
+      "learning_rate": 1.997734875046456e-05,
+      "loss": 0.589,
+      "step": 205
+    },
+    {
+      "epoch": 0.2405488249890527,
+      "grad_norm": 0.10224346631999826,
+      "learning_rate": 1.9975955706147902e-05,
+      "loss": 0.5754,
+      "step": 206
+    },
+    {
+      "epoch": 0.241716537731718,
+      "grad_norm": 0.12294767739661006,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.575,
+      "step": 207
+    },
+    {
+      "epoch": 0.2428842504743833,
+      "grad_norm": 0.1044819656584086,
+      "learning_rate": 1.99730450762985e-05,
+      "loss": 0.5482,
+      "step": 208
+    },
+    {
+      "epoch": 0.2440519632170486,
+      "grad_norm": 0.10546156552239568,
+      "learning_rate": 1.9971527502878576e-05,
+      "loss": 0.594,
+      "step": 209
+    },
+    {
+      "epoch": 0.24521967595971392,
+      "grad_norm": 0.11992767807435974,
+      "learning_rate": 1.9969968432158266e-05,
+      "loss": 0.5741,
+      "step": 210
+    },
+    {
+      "epoch": 0.24638738870237922,
+      "grad_norm": 0.1166142439683565,
+      "learning_rate": 1.9968367870625762e-05,
+      "loss": 0.5996,
+      "step": 211
+    },
+    {
+      "epoch": 0.24755510144504453,
+      "grad_norm": 0.11297239366355238,
+      "learning_rate": 1.9966725824941933e-05,
+      "loss": 0.5789,
+      "step": 212
+    },
+    {
+      "epoch": 0.24872281418770983,
+      "grad_norm": 0.11842926758032446,
+      "learning_rate": 1.996504230194028e-05,
+      "loss": 0.5792,
+      "step": 213
+    },
+    {
+      "epoch": 0.24989052693037514,
+      "grad_norm": 0.12084129211926148,
+      "learning_rate": 1.9963317308626916e-05,
+      "loss": 0.5629,
+      "step": 214
+    },
+    {
+      "epoch": 0.2510582396730404,
+      "grad_norm": 0.12141987935232827,
+      "learning_rate": 1.9961550852180535e-05,
+      "loss": 0.5922,
+      "step": 215
+    },
+    {
+      "epoch": 0.2522259524157057,
+      "grad_norm": 0.10769561167853686,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.5367,
+      "step": 216
+    },
+    {
+      "epoch": 0.25339366515837103,
+      "grad_norm": 0.10351008647543627,
+      "learning_rate": 1.9957893579466252e-05,
+      "loss": 0.5434,
+      "step": 217
+    },
+    {
+      "epoch": 0.25456137790103633,
+      "grad_norm": 0.11716413464974285,
+      "learning_rate": 1.995600277841837e-05,
+      "loss": 0.5777,
+      "step": 218
+    },
+    {
+      "epoch": 0.25572909064370164,
+      "grad_norm": 0.11700741821181772,
+      "learning_rate": 1.9954070544677476e-05,
+      "loss": 0.5883,
+      "step": 219
+    },
+    {
+      "epoch": 0.25689680338636695,
+      "grad_norm": 0.11076060293575493,
+      "learning_rate": 1.995209688628471e-05,
+      "loss": 0.5729,
+      "step": 220
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.1125529297134036,
+      "learning_rate": 1.9950081811453598e-05,
+      "loss": 0.5381,
+      "step": 221
+    },
+    {
+      "epoch": 0.25923222887169756,
+      "grad_norm": 0.10404960138609225,
+      "learning_rate": 1.994802532857004e-05,
+      "loss": 0.5762,
+      "step": 222
+    },
+    {
+      "epoch": 0.26039994161436286,
+      "grad_norm": 0.10498282866240406,
+      "learning_rate": 1.9945927446192256e-05,
+      "loss": 0.6024,
+      "step": 223
+    },
+    {
+      "epoch": 0.26156765435702817,
+      "grad_norm": 0.10220926099889541,
+      "learning_rate": 1.9943788173050743e-05,
+      "loss": 0.5827,
+      "step": 224
+    },
+    {
+      "epoch": 0.2627353670996935,
+      "grad_norm": 0.09897782888565257,
+      "learning_rate": 1.994160751804826e-05,
+      "loss": 0.5483,
+      "step": 225
+    },
+    {
+      "epoch": 0.2639030798423588,
+      "grad_norm": 0.10351960698848944,
+      "learning_rate": 1.9939385490259773e-05,
+      "loss": 0.5653,
+      "step": 226
+    },
+    {
+      "epoch": 0.2650707925850241,
+      "grad_norm": 0.10511709587666057,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 0.5493,
+      "step": 227
+    },
+    {
+      "epoch": 0.2662385053276894,
+      "grad_norm": 0.11850231498841193,
+      "learning_rate": 1.99348173534855e-05,
+      "loss": 0.5412,
+      "step": 228
+    },
+    {
+      "epoch": 0.2674062180703547,
+      "grad_norm": 0.1023009334170023,
+      "learning_rate": 1.9932471263510385e-05,
+      "loss": 0.543,
+      "step": 229
+    },
+    {
+      "epoch": 0.26857393081302,
+      "grad_norm": 0.11107875080825416,
+      "learning_rate": 1.9930083838770504e-05,
+      "loss": 0.5599,
+      "step": 230
+    },
+    {
+      "epoch": 0.2697416435556853,
+      "grad_norm": 0.10394299732464432,
+      "learning_rate": 1.9927655089201327e-05,
+      "loss": 0.5539,
+      "step": 231
+    },
+    {
+      "epoch": 0.2709093562983506,
+      "grad_norm": 0.10502870163713204,
+      "learning_rate": 1.992518502491028e-05,
+      "loss": 0.557,
+      "step": 232
+    },
+    {
+      "epoch": 0.2720770690410159,
+      "grad_norm": 0.10072310159616753,
+      "learning_rate": 1.9922673656176733e-05,
+      "loss": 0.5615,
+      "step": 233
+    },
+    {
+      "epoch": 0.2732447817836812,
+      "grad_norm": 0.09980743204384966,
+      "learning_rate": 1.9920120993451947e-05,
+      "loss": 0.5384,
+      "step": 234
+    },
+    {
+      "epoch": 0.27441249452634653,
+      "grad_norm": 0.10616169214774278,
+      "learning_rate": 1.991752704735903e-05,
+      "loss": 0.5875,
+      "step": 235
+    },
+    {
+      "epoch": 0.27558020726901183,
+      "grad_norm": 0.09769309653354638,
+      "learning_rate": 1.991489182869289e-05,
+      "loss": 0.5706,
+      "step": 236
+    },
+    {
+      "epoch": 0.27674792001167714,
+      "grad_norm": 0.09587022258887787,
+      "learning_rate": 1.99122153484202e-05,
+      "loss": 0.5043,
+      "step": 237
+    },
+    {
+      "epoch": 0.27791563275434245,
+      "grad_norm": 0.10242229750172051,
+      "learning_rate": 1.990949761767935e-05,
+      "loss": 0.5612,
+      "step": 238
+    },
+    {
+      "epoch": 0.27908334549700775,
+      "grad_norm": 0.10622302154222721,
+      "learning_rate": 1.9906738647780384e-05,
+      "loss": 0.569,
+      "step": 239
+    },
+    {
+      "epoch": 0.28025105823967306,
+      "grad_norm": 0.09806409896700138,
+      "learning_rate": 1.9903938450204972e-05,
+      "loss": 0.5699,
+      "step": 240
+    },
+    {
+      "epoch": 0.28141877098233836,
+      "grad_norm": 0.11146662712913008,
+      "learning_rate": 1.990109703660637e-05,
+      "loss": 0.5668,
+      "step": 241
+    },
+    {
+      "epoch": 0.28258648372500367,
+      "grad_norm": 0.10549314044212608,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.5556,
+      "step": 242
+    },
+    {
+      "epoch": 0.283754196467669,
+      "grad_norm": 0.10036266870763799,
+      "learning_rate": 1.9895290608810103e-05,
+      "loss": 0.5518,
+      "step": 243
+    },
+    {
+      "epoch": 0.2849219092103343,
+      "grad_norm": 0.11109966556688954,
+      "learning_rate": 1.9892325618776353e-05,
+      "loss": 0.5573,
+      "step": 244
+    },
+    {
+      "epoch": 0.2860896219529996,
+      "grad_norm": 0.11225687090567443,
+      "learning_rate": 1.988931946104712e-05,
+      "loss": 0.5627,
+      "step": 245
+    },
+    {
+      "epoch": 0.2872573346956649,
+      "grad_norm": 0.10050968648973099,
+      "learning_rate": 1.988627214813277e-05,
+      "loss": 0.5663,
+      "step": 246
+    },
+    {
+      "epoch": 0.2884250474383302,
+      "grad_norm": 0.11114818529496283,
+      "learning_rate": 1.9883183692714935e-05,
+      "loss": 0.5679,
+      "step": 247
+    },
+    {
+      "epoch": 0.2895927601809955,
+      "grad_norm": 0.10076119553714703,
+      "learning_rate": 1.9880054107646467e-05,
+      "loss": 0.5612,
+      "step": 248
+    },
+    {
+      "epoch": 0.2907604729236608,
+      "grad_norm": 0.10501901759989775,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.5706,
+      "step": 249
+    },
+    {
+      "epoch": 0.29192818566632606,
+      "grad_norm": 0.10704849716800265,
+      "learning_rate": 1.98736716008248e-05,
+      "loss": 0.5683,
+      "step": 250
+    },
+    {
+      "epoch": 0.29309589840899136,
+      "grad_norm": 0.10568278713866586,
+      "learning_rate": 1.987041870563291e-05,
+      "loss": 0.5674,
+      "step": 251
+    },
+    {
+      "epoch": 0.29426361115165667,
+      "grad_norm": 0.10629280179712927,
+      "learning_rate": 1.986712473391289e-05,
+      "loss": 0.5911,
+      "step": 252
+    },
+    {
+      "epoch": 0.295431323894322,
+      "grad_norm": 0.10418485583941635,
+      "learning_rate": 1.9863789699372863e-05,
+      "loss": 0.5525,
+      "step": 253
+    },
+    {
+      "epoch": 0.2965990366369873,
+      "grad_norm": 0.10021543035790885,
+      "learning_rate": 1.986041361589184e-05,
+      "loss": 0.5592,
+      "step": 254
+    },
+    {
+      "epoch": 0.2977667493796526,
+      "grad_norm": 0.11950645373116194,
+      "learning_rate": 1.985699649751966e-05,
+      "loss": 0.6288,
+      "step": 255
+    },
+    {
+      "epoch": 0.2989344621223179,
+      "grad_norm": 0.10319560335848957,
+      "learning_rate": 1.9853538358476933e-05,
+      "loss": 0.5712,
+      "step": 256
+    },
+    {
+      "epoch": 0.3001021748649832,
+      "grad_norm": 0.11535588869047243,
+      "learning_rate": 1.9850039213154972e-05,
+      "loss": 0.5264,
+      "step": 257
+    },
+    {
+      "epoch": 0.3012698876076485,
+      "grad_norm": 0.10883626910948505,
+      "learning_rate": 1.984649907611575e-05,
+      "loss": 0.5598,
+      "step": 258
+    },
+    {
+      "epoch": 0.3024376003503138,
+      "grad_norm": 0.12337401921698725,
+      "learning_rate": 1.984291796209183e-05,
+      "loss": 0.5877,
+      "step": 259
+    },
+    {
+      "epoch": 0.3036053130929791,
+      "grad_norm": 0.11131298719906058,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.5774,
+      "step": 260
+    },
+    {
+      "epoch": 0.3047730258356444,
+      "grad_norm": 0.11450431966078348,
+      "learning_rate": 1.9835632862872712e-05,
+      "loss": 0.5388,
+      "step": 261
+    },
+    {
+      "epoch": 0.3059407385783097,
+      "grad_norm": 0.10932818433124766,
+      "learning_rate": 1.9831928907995032e-05,
+      "loss": 0.5454,
+      "step": 262
+    },
+    {
+      "epoch": 0.30710845132097503,
+      "grad_norm": 0.10367890290336752,
+      "learning_rate": 1.9828184036767556e-05,
+      "loss": 0.5505,
+      "step": 263
+    },
+    {
+      "epoch": 0.30827616406364033,
+      "grad_norm": 0.11402430570292973,
+      "learning_rate": 1.982439826477487e-05,
+      "loss": 0.533,
+      "step": 264
+    },
+    {
+      "epoch": 0.30944387680630564,
+      "grad_norm": 0.09470007979650907,
+      "learning_rate": 1.9820571607771756e-05,
+      "loss": 0.5308,
+      "step": 265
+    },
+    {
+      "epoch": 0.31061158954897095,
+      "grad_norm": 0.10415876707074928,
+      "learning_rate": 1.9816704081683152e-05,
+      "loss": 0.5395,
+      "step": 266
+    },
+    {
+      "epoch": 0.31177930229163625,
+      "grad_norm": 0.10192960825063536,
+      "learning_rate": 1.9812795702604073e-05,
+      "loss": 0.5433,
+      "step": 267
+    },
+    {
+      "epoch": 0.31294701503430156,
+      "grad_norm": 0.11050212159160243,
+      "learning_rate": 1.980884648679955e-05,
+      "loss": 0.5434,
+      "step": 268
+    },
+    {
+      "epoch": 0.31411472777696686,
+      "grad_norm": 0.10282886857674339,
+      "learning_rate": 1.980485645070455e-05,
+      "loss": 0.539,
+      "step": 269
+    },
+    {
+      "epoch": 0.31528244051963217,
+      "grad_norm": 0.11637309759654478,
+      "learning_rate": 1.9800825610923937e-05,
+      "loss": 0.5705,
+      "step": 270
+    },
+    {
+      "epoch": 0.3164501532622975,
+      "grad_norm": 0.09969507011918745,
+      "learning_rate": 1.9796753984232357e-05,
+      "loss": 0.5593,
+      "step": 271
+    },
+    {
+      "epoch": 0.3176178660049628,
+      "grad_norm": 0.10731993449755584,
+      "learning_rate": 1.9792641587574212e-05,
+      "loss": 0.5738,
+      "step": 272
+    },
+    {
+      "epoch": 0.3187855787476281,
+      "grad_norm": 0.09724737192653705,
+      "learning_rate": 1.9788488438063566e-05,
+      "loss": 0.53,
+      "step": 273
+    },
+    {
+      "epoch": 0.3199532914902934,
+      "grad_norm": 0.09774873965360162,
+      "learning_rate": 1.9784294552984077e-05,
+      "loss": 0.5492,
+      "step": 274
+    },
+    {
+      "epoch": 0.3211210042329587,
+      "grad_norm": 0.10121925341855467,
+      "learning_rate": 1.9780059949788935e-05,
+      "loss": 0.5315,
+      "step": 275
+    },
+    {
+      "epoch": 0.322288716975624,
+      "grad_norm": 0.09752492554276239,
+      "learning_rate": 1.9775784646100768e-05,
+      "loss": 0.561,
+      "step": 276
+    },
+    {
+      "epoch": 0.3234564297182893,
+      "grad_norm": 0.10255621435499539,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.6118,
+      "step": 277
+    },
+    {
+      "epoch": 0.3246241424609546,
+      "grad_norm": 0.10148692341412345,
+      "learning_rate": 1.9767112008582736e-05,
+      "loss": 0.5577,
+      "step": 278
+    },
+    {
+      "epoch": 0.3257918552036199,
+      "grad_norm": 0.10367754133878035,
+      "learning_rate": 1.9762714710844738e-05,
+      "loss": 0.5497,
+      "step": 279
+    },
+    {
+      "epoch": 0.3269595679462852,
+      "grad_norm": 0.10884588000911052,
+      "learning_rate": 1.97582767847973e-05,
+      "loss": 0.5528,
+      "step": 280
+    },
+    {
+      "epoch": 0.32812728068895053,
+      "grad_norm": 0.12372222081528182,
+      "learning_rate": 1.975379824890921e-05,
+      "loss": 0.5716,
+      "step": 281
+    },
+    {
+      "epoch": 0.32929499343161583,
+      "grad_norm": 0.0986910939596042,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.5247,
+      "step": 282
+    },
+    {
+      "epoch": 0.33046270617428114,
+      "grad_norm": 0.1169470941497335,
+      "learning_rate": 1.9744719422331098e-05,
+      "loss": 0.5466,
+      "step": 283
+    },
+    {
+      "epoch": 0.33163041891694645,
+      "grad_norm": 0.09495954087332904,
+      "learning_rate": 1.9740119169423337e-05,
+      "loss": 0.5342,
+      "step": 284
+    },
+    {
+      "epoch": 0.33279813165961175,
+      "grad_norm": 0.11420254007636717,
+      "learning_rate": 1.973547838223927e-05,
+      "loss": 0.5637,
+      "step": 285
+    },
+    {
+      "epoch": 0.33396584440227706,
+      "grad_norm": 0.09886930389119711,
+      "learning_rate": 1.9730797080091904e-05,
+      "loss": 0.5533,
+      "step": 286
+    },
+    {
+      "epoch": 0.33513355714494236,
+      "grad_norm": 0.10469136166126607,
+      "learning_rate": 1.9726075282462847e-05,
+      "loss": 0.5606,
+      "step": 287
+    },
+    {
+      "epoch": 0.33630126988760767,
+      "grad_norm": 0.10853875045500982,
+      "learning_rate": 1.9721313009002228e-05,
+      "loss": 0.576,
+      "step": 288
+    },
+    {
+      "epoch": 0.337468982630273,
+      "grad_norm": 0.09241824611239742,
+      "learning_rate": 1.9716510279528632e-05,
+      "loss": 0.5523,
+      "step": 289
+    },
+    {
+      "epoch": 0.3386366953729383,
+      "grad_norm": 0.104140951888344,
+      "learning_rate": 1.971166711402899e-05,
+      "loss": 0.5675,
+      "step": 290
+    },
+    {
+      "epoch": 0.3398044081156036,
+      "grad_norm": 0.09708792295533575,
+      "learning_rate": 1.9706783532658528e-05,
+      "loss": 0.5243,
+      "step": 291
+    },
+    {
+      "epoch": 0.3409721208582689,
+      "grad_norm": 0.09765720107058758,
+      "learning_rate": 1.9701859555740647e-05,
+      "loss": 0.5406,
+      "step": 292
+    },
+    {
+      "epoch": 0.3421398336009342,
+      "grad_norm": 0.09547556148637255,
+      "learning_rate": 1.969689520376687e-05,
+      "loss": 0.5518,
+      "step": 293
+    },
+    {
+      "epoch": 0.3433075463435995,
+      "grad_norm": 0.09041340612268509,
+      "learning_rate": 1.969189049739674e-05,
+      "loss": 0.5429,
+      "step": 294
+    },
+    {
+      "epoch": 0.34447525908626475,
+      "grad_norm": 0.09169851416597515,
+      "learning_rate": 1.968684545745774e-05,
+      "loss": 0.5304,
+      "step": 295
+    },
+    {
+      "epoch": 0.34564297182893006,
+      "grad_norm": 0.10197431404631534,
+      "learning_rate": 1.9681760104945203e-05,
+      "loss": 0.5751,
+      "step": 296
+    },
+    {
+      "epoch": 0.34681068457159536,
+      "grad_norm": 0.0996352281742712,
+      "learning_rate": 1.9676634461022222e-05,
+      "loss": 0.5179,
+      "step": 297
+    },
+    {
+      "epoch": 0.34797839731426067,
+      "grad_norm": 0.10658803475097145,
+      "learning_rate": 1.9671468547019575e-05,
+      "loss": 0.5443,
+      "step": 298
+    },
+    {
+      "epoch": 0.349146110056926,
+      "grad_norm": 0.10752084207065592,
+      "learning_rate": 1.966626238443561e-05,
+      "loss": 0.546,
+      "step": 299
+    },
+    {
+      "epoch": 0.3503138227995913,
+      "grad_norm": 0.1037735871137313,
+      "learning_rate": 1.9661015994936204e-05,
+      "loss": 0.5596,
+      "step": 300
+    },
+    {
+      "epoch": 0.3514815355422566,
+      "grad_norm": 0.10346435036996218,
+      "learning_rate": 1.9655729400354607e-05,
+      "loss": 0.5468,
+      "step": 301
+    },
+    {
+      "epoch": 0.3526492482849219,
+      "grad_norm": 0.10384738592494464,
+      "learning_rate": 1.96504026226914e-05,
+      "loss": 0.5887,
+      "step": 302
+    },
+    {
+      "epoch": 0.3538169610275872,
+      "grad_norm": 0.09360952637770997,
+      "learning_rate": 1.9645035684114398e-05,
+      "loss": 0.5397,
+      "step": 303
+    },
+    {
+      "epoch": 0.3549846737702525,
+      "grad_norm": 0.10162848536659859,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 0.5664,
+      "step": 304
+    },
+    {
+      "epoch": 0.3561523865129178,
+      "grad_norm": 0.09625335938252519,
+      "learning_rate": 1.963418141372579e-05,
+      "loss": 0.5532,
+      "step": 305
+    },
+    {
+      "epoch": 0.3573200992555831,
+      "grad_norm": 0.10420579569994722,
+      "learning_rate": 1.9628694127085094e-05,
+      "loss": 0.5601,
+      "step": 306
+    },
+    {
+      "epoch": 0.3584878119982484,
+      "grad_norm": 0.10775535579847907,
+      "learning_rate": 1.9623166769872216e-05,
+      "loss": 0.5437,
+      "step": 307
+    },
+    {
+      "epoch": 0.3596555247409137,
+      "grad_norm": 0.09772481948532967,
+      "learning_rate": 1.9617599365089693e-05,
+      "loss": 0.548,
+      "step": 308
+    },
+    {
+      "epoch": 0.36082323748357903,
+      "grad_norm": 0.09485683416200634,
+      "learning_rate": 1.9611991935906727e-05,
+      "loss": 0.5441,
+      "step": 309
+    },
+    {
+      "epoch": 0.36199095022624433,
+      "grad_norm": 0.10230016439870315,
+      "learning_rate": 1.9606344505659072e-05,
+      "loss": 0.5327,
+      "step": 310
+    },
+    {
+      "epoch": 0.36315866296890964,
+      "grad_norm": 0.09410351163307436,
+      "learning_rate": 1.9600657097848957e-05,
+      "loss": 0.5365,
+      "step": 311
+    },
+    {
+      "epoch": 0.36432637571157495,
+      "grad_norm": 0.1047721975865455,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.5155,
+      "step": 312
+    },
+    {
+      "epoch": 0.36549408845424025,
+      "grad_norm": 0.1009114482542302,
+      "learning_rate": 1.9589162444382e-05,
+      "loss": 0.5576,
+      "step": 313
+    },
+    {
+      "epoch": 0.36666180119690556,
+      "grad_norm": 0.09884414658493296,
+      "learning_rate": 1.9583355246561076e-05,
+      "loss": 0.5414,
+      "step": 314
+    },
+    {
+      "epoch": 0.36782951393957086,
+      "grad_norm": 0.09877742406300995,
+      "learning_rate": 1.9577508166849308e-05,
+      "loss": 0.5931,
+      "step": 315
+    },
+    {
+      "epoch": 0.36899722668223617,
+      "grad_norm": 0.10132987122425642,
+      "learning_rate": 1.957162122957978e-05,
+      "loss": 0.5575,
+      "step": 316
+    },
+    {
+      "epoch": 0.3701649394249015,
+      "grad_norm": 0.10777962724030166,
+      "learning_rate": 1.9565694459251457e-05,
+      "loss": 0.5965,
+      "step": 317
+    },
+    {
+      "epoch": 0.3713326521675668,
+      "grad_norm": 0.09382106711820008,
+      "learning_rate": 1.9559727880529057e-05,
+      "loss": 0.526,
+      "step": 318
+    },
+    {
+      "epoch": 0.3725003649102321,
+      "grad_norm": 0.10023736883332071,
+      "learning_rate": 1.955372151824297e-05,
+      "loss": 0.5142,
+      "step": 319
+    },
+    {
+      "epoch": 0.3736680776528974,
+      "grad_norm": 0.10083172228781762,
+      "learning_rate": 1.9547675397389144e-05,
+      "loss": 0.5543,
+      "step": 320
+    },
+    {
+      "epoch": 0.3748357903955627,
+      "grad_norm": 0.1121720464099374,
+      "learning_rate": 1.954158954312899e-05,
+      "loss": 0.5342,
+      "step": 321
+    },
+    {
+      "epoch": 0.376003503138228,
+      "grad_norm": 0.1105734428127215,
+      "learning_rate": 1.9535463980789277e-05,
+      "loss": 0.5659,
+      "step": 322
+    },
+    {
+      "epoch": 0.3771712158808933,
+      "grad_norm": 0.09783175702590335,
+      "learning_rate": 1.9529298735862015e-05,
+      "loss": 0.5346,
+      "step": 323
+    },
+    {
+      "epoch": 0.3783389286235586,
+      "grad_norm": 0.12200018058150439,
+      "learning_rate": 1.9523093834004358e-05,
+      "loss": 0.5849,
+      "step": 324
+    },
+    {
+      "epoch": 0.3795066413662239,
+      "grad_norm": 0.10960456422704468,
+      "learning_rate": 1.951684930103849e-05,
+      "loss": 0.5415,
+      "step": 325
+    },
+    {
+      "epoch": 0.3806743541088892,
+      "grad_norm": 0.09445346701299293,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.5371,
+      "step": 326
+    },
+    {
+      "epoch": 0.38184206685155453,
+      "grad_norm": 0.11802475254609616,
+      "learning_rate": 1.9504241445895434e-05,
+      "loss": 0.5325,
+      "step": 327
+    },
+    {
+      "epoch": 0.38300977959421983,
+      "grad_norm": 0.10677083231020687,
+      "learning_rate": 1.949787817618683e-05,
+      "loss": 0.5446,
+      "step": 328
+    },
+    {
+      "epoch": 0.38417749233688514,
+      "grad_norm": 0.10873152854924739,
+      "learning_rate": 1.9491475380306975e-05,
+      "loss": 0.55,
+      "step": 329
+    },
+    {
+      "epoch": 0.38534520507955045,
+      "grad_norm": 0.09210084333390614,
+      "learning_rate": 1.9485033084901606e-05,
+      "loss": 0.5275,
+      "step": 330
+    },
+    {
+      "epoch": 0.38651291782221575,
+      "grad_norm": 0.09294040828563485,
+      "learning_rate": 1.9478551316780855e-05,
+      "loss": 0.5274,
+      "step": 331
+    },
+    {
+      "epoch": 0.38768063056488106,
+      "grad_norm": 0.10427110053898693,
+      "learning_rate": 1.9472030102919102e-05,
+      "loss": 0.529,
+      "step": 332
+    },
+    {
+      "epoch": 0.38884834330754636,
+      "grad_norm": 0.0958999240207101,
+      "learning_rate": 1.94654694704549e-05,
+      "loss": 0.5461,
+      "step": 333
+    },
+    {
+      "epoch": 0.39001605605021167,
+      "grad_norm": 0.09741301763453894,
+      "learning_rate": 1.945886944669084e-05,
+      "loss": 0.531,
+      "step": 334
+    },
+    {
+      "epoch": 0.391183768792877,
+      "grad_norm": 0.09596347112007247,
+      "learning_rate": 1.9452230059093442e-05,
+      "loss": 0.557,
+      "step": 335
+    },
+    {
+      "epoch": 0.3923514815355423,
+      "grad_norm": 0.0936067431975876,
+      "learning_rate": 1.944555133529304e-05,
+      "loss": 0.5227,
+      "step": 336
+    },
+    {
+      "epoch": 0.3935191942782076,
+      "grad_norm": 0.08832316425828694,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 0.5293,
+      "step": 337
+    },
+    {
+      "epoch": 0.3946869070208729,
+      "grad_norm": 0.09650888791808,
+      "learning_rate": 1.943207599042297e-05,
+      "loss": 0.5416,
+      "step": 338
+    },
+    {
+      "epoch": 0.3958546197635382,
+      "grad_norm": 0.09804772403885612,
+      "learning_rate": 1.9425279425432016e-05,
+      "loss": 0.5571,
+      "step": 339
+    },
+    {
+      "epoch": 0.3970223325062035,
+      "grad_norm": 0.09714595133846421,
+      "learning_rate": 1.941844363639525e-05,
+      "loss": 0.5572,
+      "step": 340
+    },
+    {
+      "epoch": 0.39819004524886875,
+      "grad_norm": 0.09923857360191067,
+      "learning_rate": 1.941156865176035e-05,
+      "loss": 0.572,
+      "step": 341
+    },
+    {
+      "epoch": 0.39935775799153406,
+      "grad_norm": 0.09031999750936985,
+      "learning_rate": 1.9404654500138117e-05,
+      "loss": 0.5211,
+      "step": 342
+    },
+    {
+      "epoch": 0.40052547073419936,
+      "grad_norm": 0.106929417693131,
+      "learning_rate": 1.939770121030233e-05,
+      "loss": 0.5729,
+      "step": 343
+    },
+    {
+      "epoch": 0.40169318347686467,
+      "grad_norm": 0.09641124718489782,
+      "learning_rate": 1.939070881118966e-05,
+      "loss": 0.5143,
+      "step": 344
+    },
+    {
+      "epoch": 0.40286089621953,
+      "grad_norm": 0.10282309622522638,
+      "learning_rate": 1.938367733189953e-05,
+      "loss": 0.5574,
+      "step": 345
+    },
+    {
+      "epoch": 0.4040286089621953,
+      "grad_norm": 0.10252480695651792,
+      "learning_rate": 1.937660680169399e-05,
+      "loss": 0.5509,
+      "step": 346
+    },
+    {
+      "epoch": 0.4051963217048606,
+      "grad_norm": 0.10030808599193926,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.5678,
+      "step": 347
+    },
+    {
+      "epoch": 0.4063640344475259,
+      "grad_norm": 0.11111335707300751,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.5874,
+      "step": 348
+    },
+    {
+      "epoch": 0.4075317471901912,
+      "grad_norm": 0.09836780388962878,
+      "learning_rate": 1.9355161200642485e-05,
+      "loss": 0.5302,
+      "step": 349
+    },
+    {
+      "epoch": 0.4086994599328565,
+      "grad_norm": 0.10137426416205007,
+      "learning_rate": 1.9347934762644326e-05,
+      "loss": 0.5313,
+      "step": 350
+    },
+    {
+      "epoch": 0.4098671726755218,
+      "grad_norm": 0.09635837852994052,
+      "learning_rate": 1.934066942247629e-05,
+      "loss": 0.5656,
+      "step": 351
+    },
+    {
+      "epoch": 0.4110348854181871,
+      "grad_norm": 0.09920524349532533,
+      "learning_rate": 1.9333365210373668e-05,
+      "loss": 0.5392,
+      "step": 352
+    },
+    {
+      "epoch": 0.4122025981608524,
+      "grad_norm": 0.10972436670249126,
+      "learning_rate": 1.9326022156733514e-05,
+      "loss": 0.5423,
+      "step": 353
+    },
+    {
+      "epoch": 0.4133703109035177,
+      "grad_norm": 0.09507783847571877,
+      "learning_rate": 1.9318640292114526e-05,
+      "loss": 0.5328,
+      "step": 354
+    },
+    {
+      "epoch": 0.41453802364618303,
+      "grad_norm": 0.10031806905114238,
+      "learning_rate": 1.9311219647236914e-05,
+      "loss": 0.5125,
+      "step": 355
+    },
+    {
+      "epoch": 0.41570573638884833,
+      "grad_norm": 0.10554156690034494,
+      "learning_rate": 1.9303760252982287e-05,
+      "loss": 0.5095,
+      "step": 356
+    },
+    {
+      "epoch": 0.41687344913151364,
+      "grad_norm": 0.09948270342859926,
+      "learning_rate": 1.9296262140393498e-05,
+      "loss": 0.5385,
+      "step": 357
+    },
+    {
+      "epoch": 0.41804116187417895,
+      "grad_norm": 0.10242340854708372,
+      "learning_rate": 1.9288725340674536e-05,
+      "loss": 0.5441,
+      "step": 358
+    },
+    {
+      "epoch": 0.41920887461684425,
+      "grad_norm": 0.09958346944880063,
+      "learning_rate": 1.928114988519039e-05,
+      "loss": 0.5525,
+      "step": 359
+    },
+    {
+      "epoch": 0.42037658735950956,
+      "grad_norm": 0.09701954311804577,
+      "learning_rate": 1.927353580546692e-05,
+      "loss": 0.523,
+      "step": 360
+    },
+    {
+      "epoch": 0.42154430010217486,
+      "grad_norm": 0.09345741570450718,
+      "learning_rate": 1.9265883133190715e-05,
+      "loss": 0.5521,
+      "step": 361
+    },
+    {
+      "epoch": 0.42271201284484017,
+      "grad_norm": 0.10587998244995132,
+      "learning_rate": 1.925819190020898e-05,
+      "loss": 0.5269,
+      "step": 362
+    },
+    {
+      "epoch": 0.4238797255875055,
+      "grad_norm": 0.10300303713266237,
+      "learning_rate": 1.925046213852939e-05,
+      "loss": 0.5621,
+      "step": 363
+    },
+    {
+      "epoch": 0.4250474383301708,
+      "grad_norm": 0.09643988143943011,
+      "learning_rate": 1.9242693880319962e-05,
+      "loss": 0.5344,
+      "step": 364
+    },
+    {
+      "epoch": 0.4262151510728361,
+      "grad_norm": 0.10277568613430882,
+      "learning_rate": 1.923488715790891e-05,
+      "loss": 0.5487,
+      "step": 365
+    },
+    {
+      "epoch": 0.4273828638155014,
+      "grad_norm": 0.09891093527062521,
+      "learning_rate": 1.922704200378453e-05,
+      "loss": 0.533,
+      "step": 366
+    },
+    {
+      "epoch": 0.4285505765581667,
+      "grad_norm": 0.09098641798535176,
+      "learning_rate": 1.921915845059505e-05,
+      "loss": 0.5403,
+      "step": 367
+    },
+    {
+      "epoch": 0.429718289300832,
+      "grad_norm": 0.09808930264680543,
+      "learning_rate": 1.92112365311485e-05,
+      "loss": 0.5435,
+      "step": 368
+    },
+    {
+      "epoch": 0.4308860020434973,
+      "grad_norm": 0.1022876179069837,
+      "learning_rate": 1.920327627841258e-05,
+      "loss": 0.5447,
+      "step": 369
+    },
+    {
+      "epoch": 0.4320537147861626,
+      "grad_norm": 0.09534187536909185,
+      "learning_rate": 1.919527772551451e-05,
+      "loss": 0.5444,
+      "step": 370
+    },
+    {
+      "epoch": 0.4332214275288279,
+      "grad_norm": 0.10018667790529624,
+      "learning_rate": 1.9187240905740893e-05,
+      "loss": 0.5676,
+      "step": 371
+    },
+    {
+      "epoch": 0.4343891402714932,
+      "grad_norm": 0.1001736325549756,
+      "learning_rate": 1.9179165852537596e-05,
+      "loss": 0.5509,
+      "step": 372
+    },
+    {
+      "epoch": 0.43555685301415853,
+      "grad_norm": 0.10567128205285446,
+      "learning_rate": 1.917105259950959e-05,
+      "loss": 0.5175,
+      "step": 373
+    },
+    {
+      "epoch": 0.43672456575682383,
+      "grad_norm": 0.09476327280175384,
+      "learning_rate": 1.916290118042082e-05,
+      "loss": 0.4734,
+      "step": 374
+    },
+    {
+      "epoch": 0.43789227849948914,
+      "grad_norm": 0.09560073735710471,
+      "learning_rate": 1.9154711629194062e-05,
+      "loss": 0.5174,
+      "step": 375
+    },
+    {
+      "epoch": 0.43905999124215445,
+      "grad_norm": 0.10734423701634045,
+      "learning_rate": 1.914648397991078e-05,
+      "loss": 0.5276,
+      "step": 376
+    },
+    {
+      "epoch": 0.44022770398481975,
+      "grad_norm": 0.09058408377241653,
+      "learning_rate": 1.913821826681099e-05,
+      "loss": 0.4897,
+      "step": 377
+    },
+    {
+      "epoch": 0.44139541672748506,
+      "grad_norm": 0.11029665250724824,
+      "learning_rate": 1.9129914524293102e-05,
+      "loss": 0.5491,
+      "step": 378
+    },
+    {
+      "epoch": 0.44256312947015036,
+      "grad_norm": 0.09837800477306506,
+      "learning_rate": 1.9121572786913808e-05,
+      "loss": 0.5317,
+      "step": 379
+    },
+    {
+      "epoch": 0.44373084221281567,
+      "grad_norm": 0.1051047641816339,
+      "learning_rate": 1.91131930893879e-05,
+      "loss": 0.5457,
+      "step": 380
+    },
+    {
+      "epoch": 0.444898554955481,
+      "grad_norm": 0.10253299845672673,
+      "learning_rate": 1.9104775466588162e-05,
+      "loss": 0.5679,
+      "step": 381
+    },
+    {
+      "epoch": 0.4460662676981463,
+      "grad_norm": 0.09912007250316186,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.5411,
+      "step": 382
+    },
+    {
+      "epoch": 0.4472339804408116,
+      "grad_norm": 0.10080472550647453,
+      "learning_rate": 1.9087826585447264e-05,
+      "loss": 0.532,
+      "step": 383
+    },
+    {
+      "epoch": 0.4484016931834769,
+      "grad_norm": 0.1008564281758169,
+      "learning_rate": 1.9079295397640215e-05,
+      "loss": 0.5379,
+      "step": 384
+    },
+    {
+      "epoch": 0.4495694059261422,
+      "grad_norm": 0.11352819606124316,
+      "learning_rate": 1.9070726425627257e-05,
+      "loss": 0.5591,
+      "step": 385
+    },
+    {
+      "epoch": 0.45073711866880745,
+      "grad_norm": 0.09482521125567568,
+      "learning_rate": 1.9062119705068843e-05,
+      "loss": 0.5788,
+      "step": 386
+    },
+    {
+      "epoch": 0.45190483141147275,
+      "grad_norm": 0.10638489762625036,
+      "learning_rate": 1.9053475271782523e-05,
+      "loss": 0.5441,
+      "step": 387
+    },
+    {
+      "epoch": 0.45307254415413806,
+      "grad_norm": 0.10654019602253878,
+      "learning_rate": 1.9044793161742783e-05,
+      "loss": 0.5715,
+      "step": 388
+    },
+    {
+      "epoch": 0.45424025689680336,
+      "grad_norm": 0.11526826385728377,
+      "learning_rate": 1.9036073411080917e-05,
+      "loss": 0.55,
+      "step": 389
+    },
+    {
+      "epoch": 0.45540796963946867,
+      "grad_norm": 0.09914583611825761,
+      "learning_rate": 1.902731605608486e-05,
+      "loss": 0.5098,
+      "step": 390
+    },
+    {
+      "epoch": 0.456575682382134,
+      "grad_norm": 0.09233790556989405,
+      "learning_rate": 1.901852113319903e-05,
+      "loss": 0.5439,
+      "step": 391
+    },
+    {
+      "epoch": 0.4577433951247993,
+      "grad_norm": 0.11296025898460205,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.5142,
+      "step": 392
+    },
+    {
+      "epoch": 0.4589111078674646,
+      "grad_norm": 0.10262720018297075,
+      "learning_rate": 1.900081873031731e-05,
+      "loss": 0.5463,
+      "step": 393
+    },
+    {
+      "epoch": 0.4600788206101299,
+      "grad_norm": 0.10143075720647761,
+      "learning_rate": 1.899191132399138e-05,
+      "loss": 0.5644,
+      "step": 394
+    },
+    {
+      "epoch": 0.4612465333527952,
+      "grad_norm": 0.10498553422968382,
+      "learning_rate": 1.898296649711527e-05,
+      "loss": 0.5412,
+      "step": 395
+    },
+    {
+      "epoch": 0.4624142460954605,
+      "grad_norm": 0.08810447418635098,
+      "learning_rate": 1.8973984286913584e-05,
+      "loss": 0.4968,
+      "step": 396
+    },
+    {
+      "epoch": 0.4635819588381258,
+      "grad_norm": 0.09490633203114292,
+      "learning_rate": 1.896496473076651e-05,
+      "loss": 0.5195,
+      "step": 397
+    },
+    {
+      "epoch": 0.4647496715807911,
+      "grad_norm": 0.11285167388149728,
+      "learning_rate": 1.895590786620963e-05,
+      "loss": 0.5649,
+      "step": 398
+    },
+    {
+      "epoch": 0.4659173843234564,
+      "grad_norm": 0.08951166314453594,
+      "learning_rate": 1.8946813730933816e-05,
+      "loss": 0.5436,
+      "step": 399
+    },
+    {
+      "epoch": 0.4670850970661217,
+      "grad_norm": 0.09689864742036951,
+      "learning_rate": 1.8937682362785025e-05,
+      "loss": 0.5091,
+      "step": 400
+    },
+    {
+      "epoch": 0.46825280980878703,
+      "grad_norm": 0.09017521933904439,
+      "learning_rate": 1.8928513799764164e-05,
+      "loss": 0.513,
+      "step": 401
+    },
+    {
+      "epoch": 0.46942052255145233,
+      "grad_norm": 0.1026981297608084,
+      "learning_rate": 1.891930808002694e-05,
+      "loss": 0.5548,
+      "step": 402
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.09314647395446474,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.5367,
+      "step": 403
+    },
+    {
+      "epoch": 0.47175594803678295,
+      "grad_norm": 0.09442904199247638,
+      "learning_rate": 1.890078532379919e-05,
+      "loss": 0.5038,
+      "step": 404
+    },
+    {
+      "epoch": 0.47292366077944825,
+      "grad_norm": 0.09816142943702764,
+      "learning_rate": 1.8891468364392577e-05,
+      "loss": 0.5389,
+      "step": 405
+    },
+    {
+      "epoch": 0.47409137352211356,
+      "grad_norm": 0.1009882668250232,
+      "learning_rate": 1.8882114402437105e-05,
+      "loss": 0.5679,
+      "step": 406
+    },
+    {
+      "epoch": 0.47525908626477886,
+      "grad_norm": 0.09624055495473108,
+      "learning_rate": 1.8872723476860033e-05,
+      "loss": 0.5376,
+      "step": 407
+    },
+    {
+      "epoch": 0.47642679900744417,
+      "grad_norm": 0.09420295944960082,
+      "learning_rate": 1.8863295626742438e-05,
+      "loss": 0.5383,
+      "step": 408
+    },
+    {
+      "epoch": 0.4775945117501095,
+      "grad_norm": 0.08884834398323711,
+      "learning_rate": 1.885383089131906e-05,
+      "loss": 0.527,
+      "step": 409
+    },
+    {
+      "epoch": 0.4787622244927748,
+      "grad_norm": 0.09190247718945312,
+      "learning_rate": 1.8844329309978146e-05,
+      "loss": 0.5164,
+      "step": 410
+    },
+    {
+      "epoch": 0.4799299372354401,
+      "grad_norm": 0.09431019707142699,
+      "learning_rate": 1.8834790922261278e-05,
+      "loss": 0.5318,
+      "step": 411
+    },
+    {
+      "epoch": 0.4810976499781054,
+      "grad_norm": 0.09228002068956359,
+      "learning_rate": 1.8825215767863215e-05,
+      "loss": 0.5401,
+      "step": 412
+    },
+    {
+      "epoch": 0.4822653627207707,
+      "grad_norm": 0.0874533376366972,
+      "learning_rate": 1.8815603886631715e-05,
+      "loss": 0.5008,
+      "step": 413
+    },
+    {
+      "epoch": 0.483433075463436,
+      "grad_norm": 0.08686678497381918,
+      "learning_rate": 1.880595531856738e-05,
+      "loss": 0.5307,
+      "step": 414
+    },
+    {
+      "epoch": 0.4846007882061013,
+      "grad_norm": 0.09736343101823981,
+      "learning_rate": 1.8796270103823494e-05,
+      "loss": 0.5346,
+      "step": 415
+    },
+    {
+      "epoch": 0.4857685009487666,
+      "grad_norm": 0.08726657121418524,
+      "learning_rate": 1.8786548282705847e-05,
+      "loss": 0.5426,
+      "step": 416
+    },
+    {
+      "epoch": 0.4869362136914319,
+      "grad_norm": 0.14283649357057415,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.5349,
+      "step": 417
+    },
+    {
+      "epoch": 0.4881039264340972,
+      "grad_norm": 0.10194550819098544,
+      "learning_rate": 1.876699498333393e-05,
+      "loss": 0.5063,
+      "step": 418
+    },
+    {
+      "epoch": 0.48927163917676253,
+      "grad_norm": 0.0975064591161238,
+      "learning_rate": 1.8757163586452266e-05,
+      "loss": 0.562,
+      "step": 419
+    },
+    {
+      "epoch": 0.49043935191942784,
+      "grad_norm": 0.0966455524008552,
+      "learning_rate": 1.8747295745941705e-05,
+      "loss": 0.5524,
+      "step": 420
+    },
+    {
+      "epoch": 0.49160706466209314,
+      "grad_norm": 0.09691607631980698,
+      "learning_rate": 1.8737391502868038e-05,
+      "loss": 0.5557,
+      "step": 421
+    },
+    {
+      "epoch": 0.49277477740475845,
+      "grad_norm": 0.09512275867220013,
+      "learning_rate": 1.8727450898448562e-05,
+      "loss": 0.5108,
+      "step": 422
+    },
+    {
+      "epoch": 0.49394249014742375,
+      "grad_norm": 0.1007754773427735,
+      "learning_rate": 1.8717473974051888e-05,
+      "loss": 0.5201,
+      "step": 423
+    },
+    {
+      "epoch": 0.49511020289008906,
+      "grad_norm": 0.10070987086625786,
+      "learning_rate": 1.8707460771197773e-05,
+      "loss": 0.533,
+      "step": 424
+    },
+    {
+      "epoch": 0.49627791563275436,
+      "grad_norm": 0.10134621779535154,
+      "learning_rate": 1.8697411331556958e-05,
+      "loss": 0.5295,
+      "step": 425
+    },
+    {
+      "epoch": 0.49744562837541967,
+      "grad_norm": 0.09445407679038957,
+      "learning_rate": 1.8687325696950972e-05,
+      "loss": 0.5463,
+      "step": 426
+    },
+    {
+      "epoch": 0.498613341118085,
+      "grad_norm": 0.09476841788833641,
+      "learning_rate": 1.867720390935199e-05,
+      "loss": 0.4944,
+      "step": 427
+    },
+    {
+      "epoch": 0.4997810538607503,
+      "grad_norm": 0.09400445624455463,
+      "learning_rate": 1.8667046010882627e-05,
+      "loss": 0.5144,
+      "step": 428
+    },
+    {
+      "epoch": 0.5009487666034156,
+      "grad_norm": 0.09170189607830459,
+      "learning_rate": 1.865685204381578e-05,
+      "loss": 0.4943,
+      "step": 429
+    },
+    {
+      "epoch": 0.5021164793460808,
+      "grad_norm": 1.4517228497440204,
+      "learning_rate": 1.8646622050574456e-05,
+      "loss": 0.5827,
+      "step": 430
+    },
+    {
+      "epoch": 0.5032841920887462,
+      "grad_norm": 0.11873229683981142,
+      "learning_rate": 1.863635607373157e-05,
+      "loss": 0.5396,
+      "step": 431
+    },
+    {
+      "epoch": 0.5044519048314114,
+      "grad_norm": 0.22844956138454534,
+      "learning_rate": 1.8626054156009807e-05,
+      "loss": 0.5312,
+      "step": 432
+    },
+    {
+      "epoch": 0.5056196175740768,
+      "grad_norm": 0.10716896229268597,
+      "learning_rate": 1.8615716340281408e-05,
+      "loss": 0.5362,
+      "step": 433
+    },
+    {
+      "epoch": 0.5067873303167421,
+      "grad_norm": 0.1770927193530642,
+      "learning_rate": 1.8605342669568013e-05,
+      "loss": 0.5441,
+      "step": 434
+    },
+    {
+      "epoch": 0.5079550430594074,
+      "grad_norm": 0.10935376573205435,
+      "learning_rate": 1.8594933187040468e-05,
+      "loss": 0.5464,
+      "step": 435
+    },
+    {
+      "epoch": 0.5091227558020727,
+      "grad_norm": 0.1185210256945446,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.5206,
+      "step": 436
+    },
+    {
+      "epoch": 0.510290468544738,
+      "grad_norm": 0.11074184943493057,
+      "learning_rate": 1.8574006959971335e-05,
+      "loss": 0.5854,
+      "step": 437
+    },
+    {
+      "epoch": 0.5114581812874033,
+      "grad_norm": 0.11895175299911663,
+      "learning_rate": 1.856349030251589e-05,
+      "loss": 0.5532,
+      "step": 438
+    },
+    {
+      "epoch": 0.5126258940300686,
+      "grad_norm": 0.1110466484181855,
+      "learning_rate": 1.8552938007418237e-05,
+      "loss": 0.5592,
+      "step": 439
+    },
+    {
+      "epoch": 0.5137936067727339,
+      "grad_norm": 0.11067623838660022,
+      "learning_rate": 1.8542350118592585e-05,
+      "loss": 0.542,
+      "step": 440
+    },
+    {
+      "epoch": 0.5149613195153993,
+      "grad_norm": 0.09755819475077775,
+      "learning_rate": 1.8531726680101268e-05,
+      "loss": 0.5243,
+      "step": 441
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.1083993931854557,
+      "learning_rate": 1.8521067736154567e-05,
+      "loss": 0.5265,
+      "step": 442
+    },
+    {
+      "epoch": 0.5172967450007299,
+      "grad_norm": 0.10286941216933793,
+      "learning_rate": 1.851037333111052e-05,
+      "loss": 0.5308,
+      "step": 443
+    },
+    {
+      "epoch": 0.5184644577433951,
+      "grad_norm": 0.11214365794501269,
+      "learning_rate": 1.8499643509474738e-05,
+      "loss": 0.5575,
+      "step": 444
+    },
+    {
+      "epoch": 0.5196321704860605,
+      "grad_norm": 0.10754247898911949,
+      "learning_rate": 1.8488878315900228e-05,
+      "loss": 0.5344,
+      "step": 445
+    },
+    {
+      "epoch": 0.5207998832287257,
+      "grad_norm": 0.0983603339518925,
+      "learning_rate": 1.8478077795187188e-05,
+      "loss": 0.5639,
+      "step": 446
+    },
+    {
+      "epoch": 0.5219675959713911,
+      "grad_norm": 0.10657785813882288,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.5423,
+      "step": 447
+    },
+    {
+      "epoch": 0.5231353087140563,
+      "grad_norm": 0.1074796168475389,
+      "learning_rate": 1.8456370952281243e-05,
+      "loss": 0.5611,
+      "step": 448
+    },
+    {
+      "epoch": 0.5243030214567217,
+      "grad_norm": 0.1007544834069768,
+      "learning_rate": 1.844546472042308e-05,
+      "loss": 0.5371,
+      "step": 449
+    },
+    {
+      "epoch": 0.525470734199387,
+      "grad_norm": 0.10117058382649649,
+      "learning_rate": 1.84345233420955e-05,
+      "loss": 0.5474,
+      "step": 450
+    },
+    {
+      "epoch": 0.5266384469420523,
+      "grad_norm": 0.10294471660659305,
+      "learning_rate": 1.8423546862831917e-05,
+      "loss": 0.5613,
+      "step": 451
+    },
+    {
+      "epoch": 0.5278061596847176,
+      "grad_norm": 0.09654874682031785,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.5078,
+      "step": 452
+    },
+    {
+      "epoch": 0.5289738724273829,
+      "grad_norm": 0.09094513050494832,
+      "learning_rate": 1.8401488784360567e-05,
+      "loss": 0.5577,
+      "step": 453
+    },
+    {
+      "epoch": 0.5301415851700482,
+      "grad_norm": 0.09301606554951869,
+      "learning_rate": 1.8390407276949235e-05,
+      "loss": 0.5363,
+      "step": 454
+    },
+    {
+      "epoch": 0.5313092979127134,
+      "grad_norm": 0.09378085727968428,
+      "learning_rate": 1.8379290852194395e-05,
+      "loss": 0.5324,
+      "step": 455
+    },
+    {
+      "epoch": 0.5324770106553788,
+      "grad_norm": 0.09617249646125954,
+      "learning_rate": 1.836813955635793e-05,
+      "loss": 0.5692,
+      "step": 456
+    },
+    {
+      "epoch": 0.533644723398044,
+      "grad_norm": 0.09469837946398985,
+      "learning_rate": 1.835695343584683e-05,
+      "loss": 0.5206,
+      "step": 457
+    },
+    {
+      "epoch": 0.5348124361407094,
+      "grad_norm": 0.09739149222825773,
+      "learning_rate": 1.834573253721303e-05,
+      "loss": 0.5184,
+      "step": 458
+    },
+    {
+      "epoch": 0.5359801488833746,
+      "grad_norm": 0.09912649676282187,
+      "learning_rate": 1.8334476907153177e-05,
+      "loss": 0.5191,
+      "step": 459
+    },
+    {
+      "epoch": 0.53714786162604,
+      "grad_norm": 0.1054944994692783,
+      "learning_rate": 1.8323186592508474e-05,
+      "loss": 0.572,
+      "step": 460
+    },
+    {
+      "epoch": 0.5383155743687053,
+      "grad_norm": 0.10827683141002112,
+      "learning_rate": 1.8311861640264456e-05,
+      "loss": 0.5585,
+      "step": 461
+    },
+    {
+      "epoch": 0.5394832871113706,
+      "grad_norm": 0.09415594593110363,
+      "learning_rate": 1.8300502097550807e-05,
+      "loss": 0.522,
+      "step": 462
+    },
+    {
+      "epoch": 0.5406509998540359,
+      "grad_norm": 0.1003774850262075,
+      "learning_rate": 1.8289108011641163e-05,
+      "loss": 0.5552,
+      "step": 463
+    },
+    {
+      "epoch": 0.5418187125967012,
+      "grad_norm": 0.09375050663222316,
+      "learning_rate": 1.8277679429952913e-05,
+      "loss": 0.5067,
+      "step": 464
+    },
+    {
+      "epoch": 0.5429864253393665,
+      "grad_norm": 0.0887080483351958,
+      "learning_rate": 1.8266216400047002e-05,
+      "loss": 0.514,
+      "step": 465
+    },
+    {
+      "epoch": 0.5441541380820318,
+      "grad_norm": 0.1058708221076345,
+      "learning_rate": 1.825471896962774e-05,
+      "loss": 0.504,
+      "step": 466
+    },
+    {
+      "epoch": 0.5453218508246971,
+      "grad_norm": 0.09939150210508796,
+      "learning_rate": 1.8243187186542594e-05,
+      "loss": 0.5366,
+      "step": 467
+    },
+    {
+      "epoch": 0.5464895635673624,
+      "grad_norm": 0.09638015080060934,
+      "learning_rate": 1.8231621098781983e-05,
+      "loss": 0.5369,
+      "step": 468
+    },
+    {
+      "epoch": 0.5476572763100277,
+      "grad_norm": 0.09694194215948637,
+      "learning_rate": 1.8220020754479104e-05,
+      "loss": 0.5102,
+      "step": 469
+    },
+    {
+      "epoch": 0.5488249890526931,
+      "grad_norm": 0.09307521727149975,
+      "learning_rate": 1.8208386201909702e-05,
+      "loss": 0.5215,
+      "step": 470
+    },
+    {
+      "epoch": 0.5499927017953583,
+      "grad_norm": 0.10726836181095019,
+      "learning_rate": 1.8196717489491885e-05,
+      "loss": 0.5264,
+      "step": 471
+    },
+    {
+      "epoch": 0.5511604145380237,
+      "grad_norm": 0.10408810449220295,
+      "learning_rate": 1.8185014665785936e-05,
+      "loss": 0.5091,
+      "step": 472
+    },
+    {
+      "epoch": 0.5523281272806889,
+      "grad_norm": 0.10038212650001072,
+      "learning_rate": 1.817327777949407e-05,
+      "loss": 0.5142,
+      "step": 473
+    },
+    {
+      "epoch": 0.5534958400233543,
+      "grad_norm": 0.11293905746818456,
+      "learning_rate": 1.8161506879460274e-05,
+      "loss": 0.5147,
+      "step": 474
+    },
+    {
+      "epoch": 0.5546635527660195,
+      "grad_norm": 0.09469105359194502,
+      "learning_rate": 1.8149702014670074e-05,
+      "loss": 0.4964,
+      "step": 475
+    },
+    {
+      "epoch": 0.5558312655086849,
+      "grad_norm": 0.09748556677890345,
+      "learning_rate": 1.8137863234250346e-05,
+      "loss": 0.5413,
+      "step": 476
+    },
+    {
+      "epoch": 0.5569989782513501,
+      "grad_norm": 0.08767397975967761,
+      "learning_rate": 1.8125990587469125e-05,
+      "loss": 0.4989,
+      "step": 477
+    },
+    {
+      "epoch": 0.5581666909940155,
+      "grad_norm": 0.10123681811904006,
+      "learning_rate": 1.8114084123735355e-05,
+      "loss": 0.5387,
+      "step": 478
+    },
+    {
+      "epoch": 0.5593344037366808,
+      "grad_norm": 0.096357589224639,
+      "learning_rate": 1.8102143892598734e-05,
+      "loss": 0.5405,
+      "step": 479
+    },
+    {
+      "epoch": 0.5605021164793461,
+      "grad_norm": 0.0947964943650361,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.5383,
+      "step": 480
+    },
+    {
+      "epoch": 0.5616698292220114,
+      "grad_norm": 0.09339599342833826,
+      "learning_rate": 1.8078162327018115e-05,
+      "loss": 0.5159,
+      "step": 481
+    },
+    {
+      "epoch": 0.5628375419646767,
+      "grad_norm": 0.08972724567325081,
+      "learning_rate": 1.8066121092375303e-05,
+      "loss": 0.5231,
+      "step": 482
+    },
+    {
+      "epoch": 0.564005254707342,
+      "grad_norm": 0.09559413672630114,
+      "learning_rate": 1.8054046289931585e-05,
+      "loss": 0.5568,
+      "step": 483
+    },
+    {
+      "epoch": 0.5651729674500073,
+      "grad_norm": 0.09242764419426074,
+      "learning_rate": 1.8041937969937206e-05,
+      "loss": 0.5557,
+      "step": 484
+    },
+    {
+      "epoch": 0.5663406801926726,
+      "grad_norm": 0.09152759869334208,
+      "learning_rate": 1.80297961827819e-05,
+      "loss": 0.5076,
+      "step": 485
+    },
+    {
+      "epoch": 0.567508392935338,
+      "grad_norm": 0.09490000738428221,
+      "learning_rate": 1.8017620978994676e-05,
+      "loss": 0.5456,
+      "step": 486
+    },
+    {
+      "epoch": 0.5686761056780032,
+      "grad_norm": 0.09351129725737609,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.5312,
+      "step": 487
+    },
+    {
+      "epoch": 0.5698438184206686,
+      "grad_norm": 0.10013031757811451,
+      "learning_rate": 1.7993170524335614e-05,
+      "loss": 0.5078,
+      "step": 488
+    },
+    {
+      "epoch": 0.5710115311633338,
+      "grad_norm": 0.08904257404941217,
+      "learning_rate": 1.7980895375216282e-05,
+      "loss": 0.5147,
+      "step": 489
+    },
+    {
+      "epoch": 0.5721792439059992,
+      "grad_norm": 0.09366126153376797,
+      "learning_rate": 1.7968587012969606e-05,
+      "loss": 0.5427,
+      "step": 490
+    },
+    {
+      "epoch": 0.5733469566486644,
+      "grad_norm": 0.10672736716298999,
+      "learning_rate": 1.795624548881781e-05,
+      "loss": 0.5315,
+      "step": 491
+    },
+    {
+      "epoch": 0.5745146693913298,
+      "grad_norm": 0.10317266729169318,
+      "learning_rate": 1.7943870854121126e-05,
+      "loss": 0.5007,
+      "step": 492
+    },
+    {
+      "epoch": 0.575682382133995,
+      "grad_norm": 0.08535082873185525,
+      "learning_rate": 1.7931463160377575e-05,
+      "loss": 0.5071,
+      "step": 493
+    },
+    {
+      "epoch": 0.5768500948766604,
+      "grad_norm": 0.09033003687302651,
+      "learning_rate": 1.7919022459222754e-05,
+      "loss": 0.4834,
+      "step": 494
+    },
+    {
+      "epoch": 0.5780178076193256,
+      "grad_norm": 0.08699224352381438,
+      "learning_rate": 1.7906548802429622e-05,
+      "loss": 0.5253,
+      "step": 495
+    },
+    {
+      "epoch": 0.579185520361991,
+      "grad_norm": 0.08880725786849378,
+      "learning_rate": 1.7894042241908293e-05,
+      "loss": 0.5641,
+      "step": 496
+    },
+    {
+      "epoch": 0.5803532331046563,
+      "grad_norm": 0.09658770809796625,
+      "learning_rate": 1.7881502829705808e-05,
+      "loss": 0.5271,
+      "step": 497
+    },
+    {
+      "epoch": 0.5815209458473216,
+      "grad_norm": 0.09206223676651927,
+      "learning_rate": 1.786893061800592e-05,
+      "loss": 0.5343,
+      "step": 498
+    },
+    {
+      "epoch": 0.5826886585899869,
+      "grad_norm": 0.0926677488510619,
+      "learning_rate": 1.785632565912888e-05,
+      "loss": 0.5377,
+      "step": 499
+    },
+    {
+      "epoch": 0.5838563713326521,
+      "grad_norm": 0.10039139304129137,
+      "learning_rate": 1.7843688005531227e-05,
+      "loss": 0.5231,
+      "step": 500
+    },
+    {
+      "epoch": 0.5850240840753175,
+      "grad_norm": 0.09344600327843604,
+      "learning_rate": 1.7831017709805555e-05,
+      "loss": 0.5467,
+      "step": 501
+    },
+    {
+      "epoch": 0.5861917968179827,
+      "grad_norm": 0.09330398511197112,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.5246,
+      "step": 502
+    },
+    {
+      "epoch": 0.5873595095606481,
+      "grad_norm": 0.10263539143517565,
+      "learning_rate": 1.7805579403019523e-05,
+      "loss": 0.5475,
+      "step": 503
+    },
+    {
+      "epoch": 0.5885272223033133,
+      "grad_norm": 0.08822192842682693,
+      "learning_rate": 1.779281149782269e-05,
+      "loss": 0.506,
+      "step": 504
+    },
+    {
+      "epoch": 0.5896949350459787,
+      "grad_norm": 0.10644114881324088,
+      "learning_rate": 1.7780011162224444e-05,
+      "loss": 0.5459,
+      "step": 505
+    },
+    {
+      "epoch": 0.590862647788644,
+      "grad_norm": 0.15571623727419126,
+      "learning_rate": 1.7767178449494403e-05,
+      "loss": 0.5467,
+      "step": 506
+    },
+    {
+      "epoch": 0.5920303605313093,
+      "grad_norm": 0.09657561821693365,
+      "learning_rate": 1.7754313413036907e-05,
+      "loss": 0.5162,
+      "step": 507
+    },
+    {
+      "epoch": 0.5931980732739746,
+      "grad_norm": 0.09477783282539325,
+      "learning_rate": 1.7741416106390828e-05,
+      "loss": 0.5429,
+      "step": 508
+    },
+    {
+      "epoch": 0.5943657860166399,
+      "grad_norm": 0.09051909268624236,
+      "learning_rate": 1.7728486583229323e-05,
+      "loss": 0.5315,
+      "step": 509
+    },
+    {
+      "epoch": 0.5955334987593052,
+      "grad_norm": 0.09397611826617958,
+      "learning_rate": 1.771552489735963e-05,
+      "loss": 0.5147,
+      "step": 510
+    },
+    {
+      "epoch": 0.5967012115019705,
+      "grad_norm": 0.09115347417536236,
+      "learning_rate": 1.770253110272283e-05,
+      "loss": 0.507,
+      "step": 511
+    },
+    {
+      "epoch": 0.5978689242446358,
+      "grad_norm": 0.09539087720964856,
+      "learning_rate": 1.768950525339362e-05,
+      "loss": 0.5448,
+      "step": 512
+    },
+    {
+      "epoch": 0.5990366369873011,
+      "grad_norm": 0.09260392222584905,
+      "learning_rate": 1.7676447403580114e-05,
+      "loss": 0.5189,
+      "step": 513
+    },
+    {
+      "epoch": 0.6002043497299664,
+      "grad_norm": 0.0859039143902055,
+      "learning_rate": 1.7663357607623577e-05,
+      "loss": 0.5009,
+      "step": 514
+    },
+    {
+      "epoch": 0.6013720624726318,
+      "grad_norm": 0.09867677535040326,
+      "learning_rate": 1.7650235919998234e-05,
+      "loss": 0.5247,
+      "step": 515
+    },
+    {
+      "epoch": 0.602539775215297,
+      "grad_norm": 0.08794728192220518,
+      "learning_rate": 1.7637082395311024e-05,
+      "loss": 0.5492,
+      "step": 516
+    },
+    {
+      "epoch": 0.6037074879579624,
+      "grad_norm": 0.09057235682704273,
+      "learning_rate": 1.7623897088301387e-05,
+      "loss": 0.5379,
+      "step": 517
+    },
+    {
+      "epoch": 0.6048752007006276,
+      "grad_norm": 0.09316666978147607,
+      "learning_rate": 1.7610680053841008e-05,
+      "loss": 0.5218,
+      "step": 518
+    },
+    {
+      "epoch": 0.606042913443293,
+      "grad_norm": 0.08596835457007931,
+      "learning_rate": 1.7597431346933633e-05,
+      "loss": 0.505,
+      "step": 519
+    },
+    {
+      "epoch": 0.6072106261859582,
+      "grad_norm": 0.10548783083747705,
+      "learning_rate": 1.75841510227148e-05,
+      "loss": 0.5291,
+      "step": 520
+    },
+    {
+      "epoch": 0.6083783389286236,
+      "grad_norm": 0.09392258411347965,
+      "learning_rate": 1.757083913645163e-05,
+      "loss": 0.5698,
+      "step": 521
+    },
+    {
+      "epoch": 0.6095460516712888,
+      "grad_norm": 0.08477876872877074,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.5095,
+      "step": 522
+    },
+    {
+      "epoch": 0.6107137644139542,
+      "grad_norm": 0.10049415613902883,
+      "learning_rate": 1.7544120899517255e-05,
+      "loss": 0.5322,
+      "step": 523
+    },
+    {
+      "epoch": 0.6118814771566194,
+      "grad_norm": 0.09190026004266191,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.5968,
+      "step": 524
+    },
+    {
+      "epoch": 0.6130491898992848,
+      "grad_norm": 0.10280244814745373,
+      "learning_rate": 1.7517277080890277e-05,
+      "loss": 0.5285,
+      "step": 525
+    },
+    {
+      "epoch": 0.6142169026419501,
+      "grad_norm": 0.10143849463366379,
+      "learning_rate": 1.7503808218001305e-05,
+      "loss": 0.5172,
+      "step": 526
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.10173641097422463,
+      "learning_rate": 1.7490308127420928e-05,
+      "loss": 0.575,
+      "step": 527
+    },
+    {
+      "epoch": 0.6165523281272807,
+      "grad_norm": 0.42473220936132217,
+      "learning_rate": 1.7476776865330847e-05,
+      "loss": 0.5063,
+      "step": 528
+    },
+    {
+      "epoch": 0.617720040869946,
+      "grad_norm": 0.11197522283506745,
+      "learning_rate": 1.7463214488042472e-05,
+      "loss": 0.5252,
+      "step": 529
+    },
+    {
+      "epoch": 0.6188877536126113,
+      "grad_norm": 0.10609447751720964,
+      "learning_rate": 1.7449621051996714e-05,
+      "loss": 0.4841,
+      "step": 530
+    },
+    {
+      "epoch": 0.6200554663552766,
+      "grad_norm": 0.10662071333333323,
+      "learning_rate": 1.743599661376373e-05,
+      "loss": 0.5575,
+      "step": 531
+    },
+    {
+      "epoch": 0.6212231790979419,
+      "grad_norm": 0.1006198057931059,
+      "learning_rate": 1.74223412300427e-05,
+      "loss": 0.5071,
+      "step": 532
+    },
+    {
+      "epoch": 0.6223908918406073,
+      "grad_norm": 0.11383943439861291,
+      "learning_rate": 1.7408654957661585e-05,
+      "loss": 0.5291,
+      "step": 533
+    },
+    {
+      "epoch": 0.6235586045832725,
+      "grad_norm": 0.11906095940245624,
+      "learning_rate": 1.739493785357688e-05,
+      "loss": 0.4878,
+      "step": 534
+    },
+    {
+      "epoch": 0.6247263173259379,
+      "grad_norm": 0.11549552892112545,
+      "learning_rate": 1.738118997487341e-05,
+      "loss": 0.5238,
+      "step": 535
+    },
+    {
+      "epoch": 0.6258940300686031,
+      "grad_norm": 0.11885462736024531,
+      "learning_rate": 1.736741137876405e-05,
+      "loss": 0.4986,
+      "step": 536
+    },
+    {
+      "epoch": 0.6270617428112685,
+      "grad_norm": 0.10626567976512107,
+      "learning_rate": 1.7353602122589528e-05,
+      "loss": 0.5615,
+      "step": 537
+    },
+    {
+      "epoch": 0.6282294555539337,
+      "grad_norm": 0.09276038435512893,
+      "learning_rate": 1.7339762263818144e-05,
+      "loss": 0.499,
+      "step": 538
+    },
+    {
+      "epoch": 0.6293971682965991,
+      "grad_norm": 0.10034518599250988,
+      "learning_rate": 1.7325891860045577e-05,
+      "loss": 0.5047,
+      "step": 539
+    },
+    {
+      "epoch": 0.6305648810392643,
+      "grad_norm": 0.09814536446678611,
+      "learning_rate": 1.7311990968994598e-05,
+      "loss": 0.518,
+      "step": 540
+    },
+    {
+      "epoch": 0.6317325937819297,
+      "grad_norm": 0.1046681915799261,
+      "learning_rate": 1.7298059648514876e-05,
+      "loss": 0.533,
+      "step": 541
+    },
+    {
+      "epoch": 0.632900306524595,
+      "grad_norm": 0.10058162274146695,
+      "learning_rate": 1.7284097956582694e-05,
+      "loss": 0.5141,
+      "step": 542
+    },
+    {
+      "epoch": 0.6340680192672603,
+      "grad_norm": 0.09112734863110987,
+      "learning_rate": 1.727010595130074e-05,
+      "loss": 0.5068,
+      "step": 543
+    },
+    {
+      "epoch": 0.6352357320099256,
+      "grad_norm": 0.10455284790354596,
+      "learning_rate": 1.725608369089785e-05,
+      "loss": 0.522,
+      "step": 544
+    },
+    {
+      "epoch": 0.6364034447525908,
+      "grad_norm": 0.09148938856770766,
+      "learning_rate": 1.724203123372877e-05,
+      "loss": 0.5189,
+      "step": 545
+    },
+    {
+      "epoch": 0.6375711574952562,
+      "grad_norm": 0.10026666613658562,
+      "learning_rate": 1.7227948638273918e-05,
+      "loss": 0.5462,
+      "step": 546
+    },
+    {
+      "epoch": 0.6387388702379214,
+      "grad_norm": 0.09222406896677988,
+      "learning_rate": 1.721383596313912e-05,
+      "loss": 0.5169,
+      "step": 547
+    },
+    {
+      "epoch": 0.6399065829805868,
+      "grad_norm": 0.09310908933089541,
+      "learning_rate": 1.7199693267055392e-05,
+      "loss": 0.5287,
+      "step": 548
+    },
+    {
+      "epoch": 0.641074295723252,
+      "grad_norm": 0.095458614368866,
+      "learning_rate": 1.7185520608878685e-05,
+      "loss": 0.5418,
+      "step": 549
+    },
+    {
+      "epoch": 0.6422420084659174,
+      "grad_norm": 0.0906954275067539,
+      "learning_rate": 1.7171318047589637e-05,
+      "loss": 0.5376,
+      "step": 550
+    },
+    {
+      "epoch": 0.6434097212085826,
+      "grad_norm": 0.08673238986490785,
+      "learning_rate": 1.7157085642293328e-05,
+      "loss": 0.5002,
+      "step": 551
+    },
+    {
+      "epoch": 0.644577433951248,
+      "grad_norm": 0.09404189944392383,
+      "learning_rate": 1.7142823452219036e-05,
+      "loss": 0.5364,
+      "step": 552
+    },
+    {
+      "epoch": 0.6457451466939133,
+      "grad_norm": 0.09626938331900155,
+      "learning_rate": 1.712853153672e-05,
+      "loss": 0.5119,
+      "step": 553
+    },
+    {
+      "epoch": 0.6469128594365786,
+      "grad_norm": 0.09281544794352652,
+      "learning_rate": 1.7114209955273154e-05,
+      "loss": 0.5422,
+      "step": 554
+    },
+    {
+      "epoch": 0.6480805721792439,
+      "grad_norm": 0.10189330950127505,
+      "learning_rate": 1.7099858767478888e-05,
+      "loss": 0.5554,
+      "step": 555
+    },
+    {
+      "epoch": 0.6492482849219092,
+      "grad_norm": 0.09358264568858085,
+      "learning_rate": 1.7085478033060808e-05,
+      "loss": 0.5362,
+      "step": 556
+    },
+    {
+      "epoch": 0.6504159976645745,
+      "grad_norm": 0.08841057495089519,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.4869,
+      "step": 557
+    },
+    {
+      "epoch": 0.6515837104072398,
+      "grad_norm": 0.1031846777597249,
+      "learning_rate": 1.705662816386217e-05,
+      "loss": 0.5172,
+      "step": 558
+    },
+    {
+      "epoch": 0.6527514231499051,
+      "grad_norm": 0.09724358742414549,
+      "learning_rate": 1.7042159149142626e-05,
+      "loss": 0.5082,
+      "step": 559
+    },
+    {
+      "epoch": 0.6539191358925704,
+      "grad_norm": 0.09128619455139929,
+      "learning_rate": 1.7027660827920798e-05,
+      "loss": 0.5107,
+      "step": 560
+    },
+    {
+      "epoch": 0.6550868486352357,
+      "grad_norm": 0.09164788164967556,
+      "learning_rate": 1.7013133260532584e-05,
+      "loss": 0.4992,
+      "step": 561
+    },
+    {
+      "epoch": 0.6562545613779011,
+      "grad_norm": 0.09352237553065389,
+      "learning_rate": 1.699857650743562e-05,
+      "loss": 0.5197,
+      "step": 562
+    },
+    {
+      "epoch": 0.6574222741205663,
+      "grad_norm": 0.09175884616735769,
+      "learning_rate": 1.6983990629208973e-05,
+      "loss": 0.4721,
+      "step": 563
+    },
+    {
+      "epoch": 0.6585899868632317,
+      "grad_norm": 0.09501791325290149,
+      "learning_rate": 1.696937568655294e-05,
+      "loss": 0.5172,
+      "step": 564
+    },
+    {
+      "epoch": 0.6597576996058969,
+      "grad_norm": 0.11030387709371951,
+      "learning_rate": 1.6954731740288744e-05,
+      "loss": 0.539,
+      "step": 565
+    },
+    {
+      "epoch": 0.6609254123485623,
+      "grad_norm": 0.08437933465736824,
+      "learning_rate": 1.6940058851358344e-05,
+      "loss": 0.4969,
+      "step": 566
+    },
+    {
+      "epoch": 0.6620931250912275,
+      "grad_norm": 0.09151047103179918,
+      "learning_rate": 1.692535708082412e-05,
+      "loss": 0.5297,
+      "step": 567
+    },
+    {
+      "epoch": 0.6632608378338929,
+      "grad_norm": 0.10624560596352767,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.5129,
+      "step": 568
+    },
+    {
+      "epoch": 0.6644285505765581,
+      "grad_norm": 0.10333095358925988,
+      "learning_rate": 1.6895867139794455e-05,
+      "loss": 0.5433,
+      "step": 569
+    },
+    {
+      "epoch": 0.6655962633192235,
+      "grad_norm": 0.10297684563382176,
+      "learning_rate": 1.6881079092023742e-05,
+      "loss": 0.5154,
+      "step": 570
+    },
+    {
+      "epoch": 0.6667639760618888,
+      "grad_norm": 0.09512871362532475,
+      "learning_rate": 1.6866262408098134e-05,
+      "loss": 0.5051,
+      "step": 571
+    },
+    {
+      "epoch": 0.6679316888045541,
+      "grad_norm": 0.09342712944817942,
+      "learning_rate": 1.6851417149678442e-05,
+      "loss": 0.4837,
+      "step": 572
+    },
+    {
+      "epoch": 0.6690994015472194,
+      "grad_norm": 0.09250142163593851,
+      "learning_rate": 1.683654337854438e-05,
+      "loss": 0.4959,
+      "step": 573
+    },
+    {
+      "epoch": 0.6702671142898847,
+      "grad_norm": 0.10789071884267618,
+      "learning_rate": 1.682164115659432e-05,
+      "loss": 0.5406,
+      "step": 574
+    },
+    {
+      "epoch": 0.67143482703255,
+      "grad_norm": 0.09167043679757947,
+      "learning_rate": 1.6806710545845036e-05,
+      "loss": 0.5321,
+      "step": 575
+    },
+    {
+      "epoch": 0.6726025397752153,
+      "grad_norm": 0.10161998662632392,
+      "learning_rate": 1.679175160843145e-05,
+      "loss": 0.5527,
+      "step": 576
+    },
+    {
+      "epoch": 0.6737702525178806,
+      "grad_norm": 0.09431769574746432,
+      "learning_rate": 1.6776764406606362e-05,
+      "loss": 0.5432,
+      "step": 577
+    },
+    {
+      "epoch": 0.674937965260546,
+      "grad_norm": 0.0924890085002518,
+      "learning_rate": 1.6761749002740195e-05,
+      "loss": 0.5185,
+      "step": 578
+    },
+    {
+      "epoch": 0.6761056780032112,
+      "grad_norm": 0.09761554426304338,
+      "learning_rate": 1.6746705459320746e-05,
+      "loss": 0.5377,
+      "step": 579
+    },
+    {
+      "epoch": 0.6772733907458766,
+      "grad_norm": 0.10087273149291337,
+      "learning_rate": 1.6731633838952905e-05,
+      "loss": 0.5408,
+      "step": 580
+    },
+    {
+      "epoch": 0.6784411034885418,
+      "grad_norm": 0.09366665120224972,
+      "learning_rate": 1.6716534204358416e-05,
+      "loss": 0.533,
+      "step": 581
+    },
+    {
+      "epoch": 0.6796088162312072,
+      "grad_norm": 0.09795941186131112,
+      "learning_rate": 1.67014066183756e-05,
+      "loss": 0.5418,
+      "step": 582
+    },
+    {
+      "epoch": 0.6807765289738724,
+      "grad_norm": 0.0983414630982643,
+      "learning_rate": 1.6686251143959096e-05,
+      "loss": 0.5311,
+      "step": 583
+    },
+    {
+      "epoch": 0.6819442417165378,
+      "grad_norm": 0.09551613086305391,
+      "learning_rate": 1.6671067844179625e-05,
+      "loss": 0.5225,
+      "step": 584
+    },
+    {
+      "epoch": 0.683111954459203,
+      "grad_norm": 0.08945042186633705,
+      "learning_rate": 1.6655856782223682e-05,
+      "loss": 0.5017,
+      "step": 585
+    },
+    {
+      "epoch": 0.6842796672018684,
+      "grad_norm": 0.08964124612927858,
+      "learning_rate": 1.6640618021393305e-05,
+      "loss": 0.507,
+      "step": 586
+    },
+    {
+      "epoch": 0.6854473799445336,
+      "grad_norm": 0.09271965846916459,
+      "learning_rate": 1.6625351625105796e-05,
+      "loss": 0.5042,
+      "step": 587
+    },
+    {
+      "epoch": 0.686615092687199,
+      "grad_norm": 0.09471448661911139,
+      "learning_rate": 1.6610057656893483e-05,
+      "loss": 0.5132,
+      "step": 588
+    },
+    {
+      "epoch": 0.6877828054298643,
+      "grad_norm": 0.09955924974568266,
+      "learning_rate": 1.6594736180403412e-05,
+      "loss": 0.5438,
+      "step": 589
+    },
+    {
+      "epoch": 0.6889505181725295,
+      "grad_norm": 0.10218813312504628,
+      "learning_rate": 1.657938725939713e-05,
+      "loss": 0.5345,
+      "step": 590
+    },
+    {
+      "epoch": 0.6901182309151949,
+      "grad_norm": 0.09442964280668642,
+      "learning_rate": 1.6564010957750375e-05,
+      "loss": 0.5434,
+      "step": 591
+    },
+    {
+      "epoch": 0.6912859436578601,
+      "grad_norm": 0.09497965214703927,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.5135,
+      "step": 592
+    },
+    {
+      "epoch": 0.6924536564005255,
+      "grad_norm": 0.0904410348230212,
+      "learning_rate": 1.653317646860793e-05,
+      "loss": 0.52,
+      "step": 593
+    },
+    {
+      "epoch": 0.6936213691431907,
+      "grad_norm": 0.09030315927160171,
+      "learning_rate": 1.6517718409432405e-05,
+      "loss": 0.5002,
+      "step": 594
+    },
+    {
+      "epoch": 0.6947890818858561,
+      "grad_norm": 0.09737871871141063,
+      "learning_rate": 1.650223322625621e-05,
+      "loss": 0.5431,
+      "step": 595
+    },
+    {
+      "epoch": 0.6959567946285213,
+      "grad_norm": 0.0892506308950772,
+      "learning_rate": 1.6486720983522156e-05,
+      "loss": 0.5388,
+      "step": 596
+    },
+    {
+      "epoch": 0.6971245073711867,
+      "grad_norm": 0.08548091021198806,
+      "learning_rate": 1.6471181745785673e-05,
+      "loss": 0.5059,
+      "step": 597
+    },
+    {
+      "epoch": 0.698292220113852,
+      "grad_norm": 0.09122715767324607,
+      "learning_rate": 1.645561557771453e-05,
+      "loss": 0.4875,
+      "step": 598
+    },
+    {
+      "epoch": 0.6994599328565173,
+      "grad_norm": 0.08582544336681185,
+      "learning_rate": 1.6440022544088553e-05,
+      "loss": 0.4926,
+      "step": 599
+    },
+    {
+      "epoch": 0.7006276455991826,
+      "grad_norm": 0.08238455935378519,
+      "learning_rate": 1.6424402709799404e-05,
+      "loss": 0.4969,
+      "step": 600
+    },
+    {
+      "epoch": 0.7017953583418479,
+      "grad_norm": 0.09368611772565201,
+      "learning_rate": 1.6408756139850243e-05,
+      "loss": 0.5361,
+      "step": 601
+    },
+    {
+      "epoch": 0.7029630710845132,
+      "grad_norm": 0.09341282156247242,
+      "learning_rate": 1.6393082899355516e-05,
+      "loss": 0.5487,
+      "step": 602
+    },
+    {
+      "epoch": 0.7041307838271785,
+      "grad_norm": 0.09399261389033317,
+      "learning_rate": 1.637738305354065e-05,
+      "loss": 0.5517,
+      "step": 603
+    },
+    {
+      "epoch": 0.7052984965698438,
+      "grad_norm": 0.08968403561556221,
+      "learning_rate": 1.63616566677418e-05,
+      "loss": 0.5302,
+      "step": 604
+    },
+    {
+      "epoch": 0.7064662093125091,
+      "grad_norm": 0.09220097393228797,
+      "learning_rate": 1.634590380740556e-05,
+      "loss": 0.5086,
+      "step": 605
+    },
+    {
+      "epoch": 0.7076339220551744,
+      "grad_norm": 0.089925560372052,
+      "learning_rate": 1.6330124538088705e-05,
+      "loss": 0.5218,
+      "step": 606
+    },
+    {
+      "epoch": 0.7088016347978398,
+      "grad_norm": 0.09119673889996767,
+      "learning_rate": 1.631431892545791e-05,
+      "loss": 0.5303,
+      "step": 607
+    },
+    {
+      "epoch": 0.709969347540505,
+      "grad_norm": 0.08360979291650934,
+      "learning_rate": 1.629848703528949e-05,
+      "loss": 0.4795,
+      "step": 608
+    },
+    {
+      "epoch": 0.7111370602831704,
+      "grad_norm": 0.09017335559301042,
+      "learning_rate": 1.6282628933469097e-05,
+      "loss": 0.5192,
+      "step": 609
+    },
+    {
+      "epoch": 0.7123047730258356,
+      "grad_norm": 0.09221885290202332,
+      "learning_rate": 1.6266744685991487e-05,
+      "loss": 0.4863,
+      "step": 610
+    },
+    {
+      "epoch": 0.713472485768501,
+      "grad_norm": 0.09391154865227909,
+      "learning_rate": 1.6250834358960203e-05,
+      "loss": 0.5071,
+      "step": 611
+    },
+    {
+      "epoch": 0.7146401985111662,
+      "grad_norm": 0.08890469483169615,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.5075,
+      "step": 612
+    },
+    {
+      "epoch": 0.7158079112538316,
+      "grad_norm": 0.08489907360963106,
+      "learning_rate": 1.6218935731193223e-05,
+      "loss": 0.5217,
+      "step": 613
+    },
+    {
+      "epoch": 0.7169756239964968,
+      "grad_norm": 0.08668664760363444,
+      "learning_rate": 1.620294756320619e-05,
+      "loss": 0.4852,
+      "step": 614
+    },
+    {
+      "epoch": 0.7181433367391622,
+      "grad_norm": 0.0966973984335969,
+      "learning_rate": 1.6186933581162254e-05,
+      "loss": 0.5214,
+      "step": 615
+    },
+    {
+      "epoch": 0.7193110494818274,
+      "grad_norm": 0.09519917329133792,
+      "learning_rate": 1.6170893851704875e-05,
+      "loss": 0.5192,
+      "step": 616
+    },
+    {
+      "epoch": 0.7204787622244928,
+      "grad_norm": 0.09107102430888307,
+      "learning_rate": 1.6154828441584655e-05,
+      "loss": 0.534,
+      "step": 617
+    },
+    {
+      "epoch": 0.7216464749671581,
+      "grad_norm": 0.10775428517083704,
+      "learning_rate": 1.613873741765907e-05,
+      "loss": 0.5452,
+      "step": 618
+    },
+    {
+      "epoch": 0.7228141877098234,
+      "grad_norm": 0.09471131995978731,
+      "learning_rate": 1.6122620846892182e-05,
+      "loss": 0.4887,
+      "step": 619
+    },
+    {
+      "epoch": 0.7239819004524887,
+      "grad_norm": 0.1045889578063256,
+      "learning_rate": 1.6106478796354382e-05,
+      "loss": 0.5262,
+      "step": 620
+    },
+    {
+      "epoch": 0.725149613195154,
+      "grad_norm": 0.09175614310120031,
+      "learning_rate": 1.6090311333222088e-05,
+      "loss": 0.5343,
+      "step": 621
+    },
+    {
+      "epoch": 0.7263173259378193,
+      "grad_norm": 0.1005570753015032,
+      "learning_rate": 1.607411852477748e-05,
+      "loss": 0.5044,
+      "step": 622
+    },
+    {
+      "epoch": 0.7274850386804846,
+      "grad_norm": 0.10857733529992022,
+      "learning_rate": 1.60579004384082e-05,
+      "loss": 0.5386,
+      "step": 623
+    },
+    {
+      "epoch": 0.7286527514231499,
+      "grad_norm": 0.08874179555386268,
+      "learning_rate": 1.604165714160711e-05,
+      "loss": 0.4919,
+      "step": 624
+    },
+    {
+      "epoch": 0.7298204641658153,
+      "grad_norm": 0.09348653854346242,
+      "learning_rate": 1.6025388701971963e-05,
+      "loss": 0.5157,
+      "step": 625
+    },
+    {
+      "epoch": 0.7309881769084805,
+      "grad_norm": 0.09860848027634214,
+      "learning_rate": 1.600909518720517e-05,
+      "loss": 0.5209,
+      "step": 626
+    },
+    {
+      "epoch": 0.7321558896511459,
+      "grad_norm": 0.08732147638662077,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.5122,
+      "step": 627
+    },
+    {
+      "epoch": 0.7333236023938111,
+      "grad_norm": 0.09418976054154009,
+      "learning_rate": 1.597643320360769e-05,
+      "loss": 0.5208,
+      "step": 628
+    },
+    {
+      "epoch": 0.7344913151364765,
+      "grad_norm": 0.09874263969316294,
+      "learning_rate": 1.5960064870702434e-05,
+      "loss": 0.5303,
+      "step": 629
+    },
+    {
+      "epoch": 0.7356590278791417,
+      "grad_norm": 0.08985156111029734,
+      "learning_rate": 1.594367173451582e-05,
+      "loss": 0.5235,
+      "step": 630
+    },
+    {
+      "epoch": 0.7368267406218071,
+      "grad_norm": 0.10451566490771613,
+      "learning_rate": 1.5927253863269174e-05,
+      "loss": 0.5264,
+      "step": 631
+    },
+    {
+      "epoch": 0.7379944533644723,
+      "grad_norm": 0.0940739103160632,
+      "learning_rate": 1.5910811325286768e-05,
+      "loss": 0.5137,
+      "step": 632
+    },
+    {
+      "epoch": 0.7391621661071377,
+      "grad_norm": 0.09530428595284947,
+      "learning_rate": 1.5894344188995522e-05,
+      "loss": 0.5187,
+      "step": 633
+    },
+    {
+      "epoch": 0.740329878849803,
+      "grad_norm": 0.09177025050004287,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.4786,
+      "step": 634
+    },
+    {
+      "epoch": 0.7414975915924683,
+      "grad_norm": 0.08932711979618843,
+      "learning_rate": 1.5861336395705765e-05,
+      "loss": 0.5402,
+      "step": 635
+    },
+    {
+      "epoch": 0.7426653043351336,
+      "grad_norm": 0.08446070057890374,
+      "learning_rate": 1.58447958760718e-05,
+      "loss": 0.4842,
+      "step": 636
+    },
+    {
+      "epoch": 0.7438330170777988,
+      "grad_norm": 0.09313694016493819,
+      "learning_rate": 1.5828231032857503e-05,
+      "loss": 0.5235,
+      "step": 637
+    },
+    {
+      "epoch": 0.7450007298204642,
+      "grad_norm": 0.09130161143333161,
+      "learning_rate": 1.581164193499879e-05,
+      "loss": 0.5277,
+      "step": 638
+    },
+    {
+      "epoch": 0.7461684425631294,
+      "grad_norm": 0.08643070441709304,
+      "learning_rate": 1.5795028651532502e-05,
+      "loss": 0.4927,
+      "step": 639
+    },
+    {
+      "epoch": 0.7473361553057948,
+      "grad_norm": 0.09366247346056625,
+      "learning_rate": 1.577839125159613e-05,
+      "loss": 0.5397,
+      "step": 640
+    },
+    {
+      "epoch": 0.74850386804846,
+      "grad_norm": 0.09078511300065908,
+      "learning_rate": 1.576172980442753e-05,
+      "loss": 0.5354,
+      "step": 641
+    },
+    {
+      "epoch": 0.7496715807911254,
+      "grad_norm": 0.094927388750307,
+      "learning_rate": 1.5745044379364637e-05,
+      "loss": 0.5269,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508392935337906,
+      "grad_norm": 0.09141010007874806,
+      "learning_rate": 1.572833504584516e-05,
+      "loss": 0.5512,
+      "step": 643
+    },
+    {
+      "epoch": 0.752007006276456,
+      "grad_norm": 0.1037201991528851,
+      "learning_rate": 1.5711601873406314e-05,
+      "loss": 0.5047,
+      "step": 644
+    },
+    {
+      "epoch": 0.7531747190191213,
+      "grad_norm": 0.08781892207177035,
+      "learning_rate": 1.569484493168452e-05,
+      "loss": 0.5245,
+      "step": 645
+    },
+    {
+      "epoch": 0.7543424317617866,
+      "grad_norm": 0.09072628004970838,
+      "learning_rate": 1.5678064290415122e-05,
+      "loss": 0.5109,
+      "step": 646
+    },
+    {
+      "epoch": 0.7555101445044519,
+      "grad_norm": 0.09838519233908884,
+      "learning_rate": 1.5661260019432077e-05,
+      "loss": 0.5074,
+      "step": 647
+    },
+    {
+      "epoch": 0.7566778572471172,
+      "grad_norm": 0.09576343540002055,
+      "learning_rate": 1.5644432188667695e-05,
+      "loss": 0.5308,
+      "step": 648
+    },
+    {
+      "epoch": 0.7578455699897825,
+      "grad_norm": 0.09269256576988653,
+      "learning_rate": 1.5627580868152316e-05,
+      "loss": 0.5035,
+      "step": 649
+    },
+    {
+      "epoch": 0.7590132827324478,
+      "grad_norm": 0.10979863584510316,
+      "learning_rate": 1.5610706128014054e-05,
+      "loss": 0.5207,
+      "step": 650
+    },
+    {
+      "epoch": 0.7601809954751131,
+      "grad_norm": 0.08759265968043346,
+      "learning_rate": 1.5593808038478465e-05,
+      "loss": 0.4871,
+      "step": 651
+    },
+    {
+      "epoch": 0.7613487082177784,
+      "grad_norm": 0.09296714505768022,
+      "learning_rate": 1.5576886669868297e-05,
+      "loss": 0.5078,
+      "step": 652
+    },
+    {
+      "epoch": 0.7625164209604437,
+      "grad_norm": 0.09280397669805499,
+      "learning_rate": 1.5559942092603152e-05,
+      "loss": 0.5106,
+      "step": 653
+    },
+    {
+      "epoch": 0.7636841337031091,
+      "grad_norm": 0.08943890896744143,
+      "learning_rate": 1.554297437719923e-05,
+      "loss": 0.5309,
+      "step": 654
+    },
+    {
+      "epoch": 0.7648518464457743,
+      "grad_norm": 0.09088271627431413,
+      "learning_rate": 1.5525983594269026e-05,
+      "loss": 0.5067,
+      "step": 655
+    },
+    {
+      "epoch": 0.7660195591884397,
+      "grad_norm": 0.08530195981313075,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.5204,
+      "step": 656
+    },
+    {
+      "epoch": 0.7671872719311049,
+      "grad_norm": 0.08659944828684096,
+      "learning_rate": 1.5491933108759417e-05,
+      "loss": 0.5026,
+      "step": 657
+    },
+    {
+      "epoch": 0.7683549846737703,
+      "grad_norm": 0.08826294699082939,
+      "learning_rate": 1.5474873547883803e-05,
+      "loss": 0.5154,
+      "step": 658
+    },
+    {
+      "epoch": 0.7695226974164355,
+      "grad_norm": 0.09062072607620872,
+      "learning_rate": 1.545779120288889e-05,
+      "loss": 0.5337,
+      "step": 659
+    },
+    {
+      "epoch": 0.7706904101591009,
+      "grad_norm": 0.08108098680171624,
+      "learning_rate": 1.5440686144864207e-05,
+      "loss": 0.4824,
+      "step": 660
+    },
+    {
+      "epoch": 0.7718581229017661,
+      "grad_norm": 0.09826209975284508,
+      "learning_rate": 1.5423558444993808e-05,
+      "loss": 0.5375,
+      "step": 661
+    },
+    {
+      "epoch": 0.7730258356444315,
+      "grad_norm": 0.09460906120111412,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4967,
+      "step": 662
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.08922879715777261,
+      "learning_rate": 1.5389235404922912e-05,
+      "loss": 0.5001,
+      "step": 663
+    },
+    {
+      "epoch": 0.7753612611297621,
+      "grad_norm": 0.09428197176405324,
+      "learning_rate": 1.5372040207560457e-05,
+      "loss": 0.5531,
+      "step": 664
+    },
+    {
+      "epoch": 0.7765289738724274,
+      "grad_norm": 0.09077453542313937,
+      "learning_rate": 1.5354822654027787e-05,
+      "loss": 0.5253,
+      "step": 665
+    },
+    {
+      "epoch": 0.7776966866150927,
+      "grad_norm": 0.09871838104484734,
+      "learning_rate": 1.5337582815977102e-05,
+      "loss": 0.5013,
+      "step": 666
+    },
+    {
+      "epoch": 0.778864399357758,
+      "grad_norm": 0.08954830300297364,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.538,
+      "step": 667
+    },
+    {
+      "epoch": 0.7800321121004233,
+      "grad_norm": 0.08956355728152633,
+      "learning_rate": 1.5303036573393964e-05,
+      "loss": 0.5129,
+      "step": 668
+    },
+    {
+      "epoch": 0.7811998248430886,
+      "grad_norm": 0.09446234198217252,
+      "learning_rate": 1.528573031262842e-05,
+      "loss": 0.5167,
+      "step": 669
+    },
+    {
+      "epoch": 0.782367537585754,
+      "grad_norm": 0.08340356551804785,
+      "learning_rate": 1.5268402054878116e-05,
+      "loss": 0.4784,
+      "step": 670
+    },
+    {
+      "epoch": 0.7835352503284192,
+      "grad_norm": 0.0865915402628961,
+      "learning_rate": 1.5251051872255963e-05,
+      "loss": 0.5494,
+      "step": 671
+    },
+    {
+      "epoch": 0.7847029630710846,
+      "grad_norm": 0.0972844652705535,
+      "learning_rate": 1.5233679836966122e-05,
+      "loss": 0.5236,
+      "step": 672
+    },
+    {
+      "epoch": 0.7858706758137498,
+      "grad_norm": 0.09211851296801833,
+      "learning_rate": 1.5216286021303685e-05,
+      "loss": 0.5103,
+      "step": 673
+    },
+    {
+      "epoch": 0.7870383885564152,
+      "grad_norm": 0.09133465754559285,
+      "learning_rate": 1.5198870497654397e-05,
+      "loss": 0.5135,
+      "step": 674
+    },
+    {
+      "epoch": 0.7882061012990804,
+      "grad_norm": 0.09638151849557211,
+      "learning_rate": 1.5181433338494337e-05,
+      "loss": 0.5312,
+      "step": 675
+    },
+    {
+      "epoch": 0.7893738140417458,
+      "grad_norm": 0.09140424589508486,
+      "learning_rate": 1.5163974616389621e-05,
+      "loss": 0.4968,
+      "step": 676
+    },
+    {
+      "epoch": 0.790541526784411,
+      "grad_norm": 0.08927158638673892,
+      "learning_rate": 1.5146494403996101e-05,
+      "loss": 0.5292,
+      "step": 677
+    },
+    {
+      "epoch": 0.7917092395270764,
+      "grad_norm": 0.10316838805023244,
+      "learning_rate": 1.5128992774059063e-05,
+      "loss": 0.5141,
+      "step": 678
+    },
+    {
+      "epoch": 0.7928769522697416,
+      "grad_norm": 0.09482369050934547,
+      "learning_rate": 1.5111469799412927e-05,
+      "loss": 0.5029,
+      "step": 679
+    },
+    {
+      "epoch": 0.794044665012407,
+      "grad_norm": 0.095015999288536,
+      "learning_rate": 1.5093925552980934e-05,
+      "loss": 0.5388,
+      "step": 680
+    },
+    {
+      "epoch": 0.7952123777550723,
+      "grad_norm": 0.09179816249019541,
+      "learning_rate": 1.5076360107774855e-05,
+      "loss": 0.487,
+      "step": 681
+    },
+    {
+      "epoch": 0.7963800904977375,
+      "grad_norm": 0.1090778248938321,
+      "learning_rate": 1.5058773536894685e-05,
+      "loss": 0.5242,
+      "step": 682
+    },
+    {
+      "epoch": 0.7975478032404029,
+      "grad_norm": 0.08913973508270337,
+      "learning_rate": 1.504116591352832e-05,
+      "loss": 0.4961,
+      "step": 683
+    },
+    {
+      "epoch": 0.7987155159830681,
+      "grad_norm": 1.2499683777590804,
+      "learning_rate": 1.5023537310951284e-05,
+      "loss": 0.5275,
+      "step": 684
+    },
+    {
+      "epoch": 0.7998832287257335,
+      "grad_norm": 1.4127057418026325,
+      "learning_rate": 1.5005887802526398e-05,
+      "loss": 0.7382,
+      "step": 685
+    },
+    {
+      "epoch": 0.8010509414683987,
+      "grad_norm": 1.086764834740458,
+      "learning_rate": 1.4988217461703488e-05,
+      "loss": 0.7481,
+      "step": 686
+    },
+    {
+      "epoch": 0.8022186542110641,
+      "grad_norm": 0.35214333247285007,
+      "learning_rate": 1.4970526362019077e-05,
+      "loss": 0.5911,
+      "step": 687
+    },
+    {
+      "epoch": 0.8033863669537293,
+      "grad_norm": 0.2264472326611188,
+      "learning_rate": 1.4952814577096073e-05,
+      "loss": 0.557,
+      "step": 688
+    },
+    {
+      "epoch": 0.8045540796963947,
+      "grad_norm": 0.18010425439574462,
+      "learning_rate": 1.493508218064347e-05,
+      "loss": 0.5519,
+      "step": 689
+    },
+    {
+      "epoch": 0.80572179243906,
+      "grad_norm": 0.12743279427063928,
+      "learning_rate": 1.491732924645604e-05,
+      "loss": 0.5546,
+      "step": 690
+    },
+    {
+      "epoch": 0.8068895051817253,
+      "grad_norm": 0.1319968533622762,
+      "learning_rate": 1.489955584841402e-05,
+      "loss": 0.558,
+      "step": 691
+    },
+    {
+      "epoch": 0.8080572179243906,
+      "grad_norm": 0.11906975791499813,
+      "learning_rate": 1.4881762060482814e-05,
+      "loss": 0.5268,
+      "step": 692
+    },
+    {
+      "epoch": 0.8092249306670559,
+      "grad_norm": 0.10195435310917256,
+      "learning_rate": 1.4863947956712679e-05,
+      "loss": 0.4768,
+      "step": 693
+    },
+    {
+      "epoch": 0.8103926434097212,
+      "grad_norm": 0.11773143038317281,
+      "learning_rate": 1.4846113611238413e-05,
+      "loss": 0.5121,
+      "step": 694
+    },
+    {
+      "epoch": 0.8115603561523865,
+      "grad_norm": 0.11287021181321717,
+      "learning_rate": 1.4828259098279059e-05,
+      "loss": 0.5284,
+      "step": 695
+    },
+    {
+      "epoch": 0.8127280688950518,
+      "grad_norm": 0.09315270375022444,
+      "learning_rate": 1.4810384492137582e-05,
+      "loss": 0.5281,
+      "step": 696
+    },
+    {
+      "epoch": 0.8138957816377171,
+      "grad_norm": 0.1042211497063363,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.5197,
+      "step": 697
+    },
+    {
+      "epoch": 0.8150634943803824,
+      "grad_norm": 0.10982700226716413,
+      "learning_rate": 1.4774575297937921e-05,
+      "loss": 0.5161,
+      "step": 698
+    },
+    {
+      "epoch": 0.8162312071230478,
+      "grad_norm": 0.09594833445738143,
+      "learning_rate": 1.4756640858902531e-05,
+      "loss": 0.5395,
+      "step": 699
+    },
+    {
+      "epoch": 0.817398919865713,
+      "grad_norm": 0.10772348594508968,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.5099,
+      "step": 700
+    },
+    {
+      "epoch": 0.8185666326083784,
+      "grad_norm": 0.11939656790394067,
+      "learning_rate": 1.4720712670138255e-05,
+      "loss": 0.5104,
+      "step": 701
+    },
+    {
+      "epoch": 0.8197343453510436,
+      "grad_norm": 0.09954420251064322,
+      "learning_rate": 1.4702719069927368e-05,
+      "loss": 0.4964,
+      "step": 702
+    },
+    {
+      "epoch": 0.820902058093709,
+      "grad_norm": 0.10317122864982559,
+      "learning_rate": 1.4684705898979119e-05,
+      "loss": 0.5138,
+      "step": 703
+    },
+    {
+      "epoch": 0.8220697708363742,
+      "grad_norm": 0.0974029935350602,
+      "learning_rate": 1.4666673232256738e-05,
+      "loss": 0.5105,
+      "step": 704
+    },
+    {
+      "epoch": 0.8232374835790396,
+      "grad_norm": 0.09331603348859181,
+      "learning_rate": 1.46486211448046e-05,
+      "loss": 0.5044,
+      "step": 705
+    },
+    {
+      "epoch": 0.8244051963217048,
+      "grad_norm": 0.09850559527347909,
+      "learning_rate": 1.4630549711747888e-05,
+      "loss": 0.5207,
+      "step": 706
+    },
+    {
+      "epoch": 0.8255729090643702,
+      "grad_norm": 0.0974694982215739,
+      "learning_rate": 1.4612459008292307e-05,
+      "loss": 0.518,
+      "step": 707
+    },
+    {
+      "epoch": 0.8267406218070354,
+      "grad_norm": 0.0892555549978633,
+      "learning_rate": 1.4594349109723744e-05,
+      "loss": 0.4973,
+      "step": 708
+    },
+    {
+      "epoch": 0.8279083345497008,
+      "grad_norm": 0.09082452993385741,
+      "learning_rate": 1.4576220091407981e-05,
+      "loss": 0.5356,
+      "step": 709
+    },
+    {
+      "epoch": 0.8290760472923661,
+      "grad_norm": 0.09294161184706613,
+      "learning_rate": 1.4558072028790356e-05,
+      "loss": 0.511,
+      "step": 710
+    },
+    {
+      "epoch": 0.8302437600350314,
+      "grad_norm": 0.09718185218423567,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.5286,
+      "step": 711
+    },
+    {
+      "epoch": 0.8314114727776967,
+      "grad_norm": 0.09402283644678681,
+      "learning_rate": 1.4521719072826858e-05,
+      "loss": 0.5034,
+      "step": 712
+    },
+    {
+      "epoch": 0.832579185520362,
+      "grad_norm": 0.08598696771961127,
+      "learning_rate": 1.4503514330766689e-05,
+      "loss": 0.477,
+      "step": 713
+    },
+    {
+      "epoch": 0.8337468982630273,
+      "grad_norm": 0.09172931553345563,
+      "learning_rate": 1.4485290846975431e-05,
+      "loss": 0.4827,
+      "step": 714
+    },
+    {
+      "epoch": 0.8349146110056926,
+      "grad_norm": 0.09685390842095985,
+      "learning_rate": 1.4467048697291554e-05,
+      "loss": 0.513,
+      "step": 715
+    },
+    {
+      "epoch": 0.8360823237483579,
+      "grad_norm": 0.09951463708629217,
+      "learning_rate": 1.4448787957631209e-05,
+      "loss": 0.5617,
+      "step": 716
+    },
+    {
+      "epoch": 0.8372500364910233,
+      "grad_norm": 0.1018363107419832,
+      "learning_rate": 1.4430508703987914e-05,
+      "loss": 0.5332,
+      "step": 717
+    },
+    {
+      "epoch": 0.8384177492336885,
+      "grad_norm": 0.0984895505963426,
+      "learning_rate": 1.4412211012432213e-05,
+      "loss": 0.5241,
+      "step": 718
+    },
+    {
+      "epoch": 0.8395854619763539,
+      "grad_norm": 0.09504248666419114,
+      "learning_rate": 1.4393894959111415e-05,
+      "loss": 0.5247,
+      "step": 719
+    },
+    {
+      "epoch": 0.8407531747190191,
+      "grad_norm": 0.09268015365876724,
+      "learning_rate": 1.4375560620249209e-05,
+      "loss": 0.521,
+      "step": 720
+    },
+    {
+      "epoch": 0.8419208874616845,
+      "grad_norm": 0.09451953076978113,
+      "learning_rate": 1.4357208072145405e-05,
+      "loss": 0.5468,
+      "step": 721
+    },
+    {
+      "epoch": 0.8430886002043497,
+      "grad_norm": 0.08752822593518737,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.4743,
+      "step": 722
+    },
+    {
+      "epoch": 0.8442563129470151,
+      "grad_norm": 0.08978487814751383,
+      "learning_rate": 1.4320448653790782e-05,
+      "loss": 0.4782,
+      "step": 723
+    },
+    {
+      "epoch": 0.8454240256896803,
+      "grad_norm": 0.09888883302056402,
+      "learning_rate": 1.430204193651719e-05,
+      "loss": 0.5558,
+      "step": 724
+    },
+    {
+      "epoch": 0.8465917384323457,
+      "grad_norm": 0.09075824565993522,
+      "learning_rate": 1.4283617315955815e-05,
+      "loss": 0.535,
+      "step": 725
+    },
+    {
+      "epoch": 0.847759451175011,
+      "grad_norm": 0.09048519617055771,
+      "learning_rate": 1.4265174868782172e-05,
+      "loss": 0.5112,
+      "step": 726
+    },
+    {
+      "epoch": 0.8489271639176762,
+      "grad_norm": 0.08910954736114979,
+      "learning_rate": 1.4246714671745964e-05,
+      "loss": 0.5013,
+      "step": 727
+    },
+    {
+      "epoch": 0.8500948766603416,
+      "grad_norm": 0.09342418531896803,
+      "learning_rate": 1.4228236801670762e-05,
+      "loss": 0.5065,
+      "step": 728
+    },
+    {
+      "epoch": 0.8512625894030068,
+      "grad_norm": 0.08500200165724381,
+      "learning_rate": 1.4209741335453688e-05,
+      "loss": 0.5012,
+      "step": 729
+    },
+    {
+      "epoch": 0.8524303021456722,
+      "grad_norm": 0.08527065877255138,
+      "learning_rate": 1.4191228350065079e-05,
+      "loss": 0.4881,
+      "step": 730
+    },
+    {
+      "epoch": 0.8535980148883374,
+      "grad_norm": 0.09015352187031851,
+      "learning_rate": 1.417269792254819e-05,
+      "loss": 0.5442,
+      "step": 731
+    },
+    {
+      "epoch": 0.8547657276310028,
+      "grad_norm": 0.08796293741170319,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.533,
+      "step": 732
+    },
+    {
+      "epoch": 0.855933440373668,
+      "grad_norm": 0.08859703570286513,
+      "learning_rate": 1.4135585049665207e-05,
+      "loss": 0.4916,
+      "step": 733
+    },
+    {
+      "epoch": 0.8571011531163334,
+      "grad_norm": 0.08545757872199287,
+      "learning_rate": 1.4117002758747268e-05,
+      "loss": 0.5042,
+      "step": 734
+    },
+    {
+      "epoch": 0.8582688658589986,
+      "grad_norm": 0.08749507002070006,
+      "learning_rate": 1.409840333459672e-05,
+      "loss": 0.4739,
+      "step": 735
+    },
+    {
+      "epoch": 0.859436578601664,
+      "grad_norm": 0.0904027767497731,
+      "learning_rate": 1.4079786854616537e-05,
+      "loss": 0.4938,
+      "step": 736
+    },
+    {
+      "epoch": 0.8606042913443293,
+      "grad_norm": 0.0910586616374535,
+      "learning_rate": 1.4061153396280673e-05,
+      "loss": 0.4989,
+      "step": 737
+    },
+    {
+      "epoch": 0.8617720040869946,
+      "grad_norm": 0.09043567675838207,
+      "learning_rate": 1.4042503037133736e-05,
+      "loss": 0.4876,
+      "step": 738
+    },
+    {
+      "epoch": 0.8629397168296599,
+      "grad_norm": 0.08584334578784489,
+      "learning_rate": 1.4023835854790682e-05,
+      "loss": 0.5035,
+      "step": 739
+    },
+    {
+      "epoch": 0.8641074295723252,
+      "grad_norm": 0.09078704012893032,
+      "learning_rate": 1.400515192693645e-05,
+      "loss": 0.4423,
+      "step": 740
+    },
+    {
+      "epoch": 0.8652751423149905,
+      "grad_norm": 0.0918614454505933,
+      "learning_rate": 1.3986451331325697e-05,
+      "loss": 0.4907,
+      "step": 741
+    },
+    {
+      "epoch": 0.8664428550576558,
+      "grad_norm": 0.08735171714060494,
+      "learning_rate": 1.3967734145782425e-05,
+      "loss": 0.5229,
+      "step": 742
+    },
+    {
+      "epoch": 0.8676105678003211,
+      "grad_norm": 0.08607388502136135,
+      "learning_rate": 1.3949000448199686e-05,
+      "loss": 0.511,
+      "step": 743
+    },
+    {
+      "epoch": 0.8687782805429864,
+      "grad_norm": 0.0897865410204032,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.4967,
+      "step": 744
+    },
+    {
+      "epoch": 0.8699459932856517,
+      "grad_norm": 0.0872115148085154,
+      "learning_rate": 1.391148382883124e-05,
+      "loss": 0.4952,
+      "step": 745
+    },
+    {
+      "epoch": 0.8711137060283171,
+      "grad_norm": 0.08738160986088495,
+      "learning_rate": 1.3892701063173917e-05,
+      "loss": 0.5262,
+      "step": 746
+    },
+    {
+      "epoch": 0.8722814187709823,
+      "grad_norm": 0.0866894790564702,
+      "learning_rate": 1.3873902097733231e-05,
+      "loss": 0.5201,
+      "step": 747
+    },
+    {
+      "epoch": 0.8734491315136477,
+      "grad_norm": 0.0847700195746533,
+      "learning_rate": 1.3855087010742563e-05,
+      "loss": 0.5235,
+      "step": 748
+    },
+    {
+      "epoch": 0.8746168442563129,
+      "grad_norm": 0.08901037133981184,
+      "learning_rate": 1.3836255880502389e-05,
+      "loss": 0.5235,
+      "step": 749
+    },
+    {
+      "epoch": 0.8757845569989783,
+      "grad_norm": 0.08996145572264451,
+      "learning_rate": 1.3817408785379944e-05,
+      "loss": 0.4999,
+      "step": 750
+    },
+    {
+      "epoch": 0.8769522697416435,
+      "grad_norm": 0.08571171541871257,
+      "learning_rate": 1.379854580380891e-05,
+      "loss": 0.5048,
+      "step": 751
+    },
+    {
+      "epoch": 0.8781199824843089,
+      "grad_norm": 0.09382132198021056,
+      "learning_rate": 1.3779667014289067e-05,
+      "loss": 0.5056,
+      "step": 752
+    },
+    {
+      "epoch": 0.8792876952269741,
+      "grad_norm": 0.08243502665921021,
+      "learning_rate": 1.3760772495385998e-05,
+      "loss": 0.5035,
+      "step": 753
+    },
+    {
+      "epoch": 0.8804554079696395,
+      "grad_norm": 0.08819243785616018,
+      "learning_rate": 1.3741862325730739e-05,
+      "loss": 0.513,
+      "step": 754
+    },
+    {
+      "epoch": 0.8816231207123048,
+      "grad_norm": 0.0917525409239656,
+      "learning_rate": 1.3722936584019453e-05,
+      "loss": 0.5164,
+      "step": 755
+    },
+    {
+      "epoch": 0.8827908334549701,
+      "grad_norm": 0.08589532071551682,
+      "learning_rate": 1.3703995349013113e-05,
+      "loss": 0.5071,
+      "step": 756
+    },
+    {
+      "epoch": 0.8839585461976354,
+      "grad_norm": 0.08514379673528481,
+      "learning_rate": 1.3685038699537165e-05,
+      "loss": 0.5222,
+      "step": 757
+    },
+    {
+      "epoch": 0.8851262589403007,
+      "grad_norm": 0.08580975655464479,
+      "learning_rate": 1.3666066714481205e-05,
+      "loss": 0.4657,
+      "step": 758
+    },
+    {
+      "epoch": 0.886293971682966,
+      "grad_norm": 0.08374598754744979,
+      "learning_rate": 1.3647079472798652e-05,
+      "loss": 0.4908,
+      "step": 759
+    },
+    {
+      "epoch": 0.8874616844256313,
+      "grad_norm": 0.08415969931911824,
+      "learning_rate": 1.362807705350641e-05,
+      "loss": 0.482,
+      "step": 760
+    },
+    {
+      "epoch": 0.8886293971682966,
+      "grad_norm": 0.08333161579762088,
+      "learning_rate": 1.3609059535684557e-05,
+      "loss": 0.4831,
+      "step": 761
+    },
+    {
+      "epoch": 0.889797109910962,
+      "grad_norm": 0.08718141586375165,
+      "learning_rate": 1.3590026998475987e-05,
+      "loss": 0.4975,
+      "step": 762
+    },
+    {
+      "epoch": 0.8909648226536272,
+      "grad_norm": 0.08477183129792722,
+      "learning_rate": 1.3570979521086118e-05,
+      "loss": 0.4992,
+      "step": 763
+    },
+    {
+      "epoch": 0.8921325353962926,
+      "grad_norm": 0.08886941360065786,
+      "learning_rate": 1.355191718278253e-05,
+      "loss": 0.5153,
+      "step": 764
+    },
+    {
+      "epoch": 0.8933002481389578,
+      "grad_norm": 0.08198000738754052,
+      "learning_rate": 1.3532840062894654e-05,
+      "loss": 0.4856,
+      "step": 765
+    },
+    {
+      "epoch": 0.8944679608816232,
+      "grad_norm": 0.08688188070619318,
+      "learning_rate": 1.3513748240813429e-05,
+      "loss": 0.525,
+      "step": 766
+    },
+    {
+      "epoch": 0.8956356736242884,
+      "grad_norm": 0.08933551275863236,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.5156,
+      "step": 767
+    },
+    {
+      "epoch": 0.8968033863669538,
+      "grad_norm": 0.08904168345314728,
+      "learning_rate": 1.3475520807940303e-05,
+      "loss": 0.4921,
+      "step": 768
+    },
+    {
+      "epoch": 0.897971099109619,
+      "grad_norm": 0.08635737497436785,
+      "learning_rate": 1.345638535623489e-05,
+      "loss": 0.5295,
+      "step": 769
+    },
+    {
+      "epoch": 0.8991388118522844,
+      "grad_norm": 0.09759522269439165,
+      "learning_rate": 1.3437235520508434e-05,
+      "loss": 0.4938,
+      "step": 770
+    },
+    {
+      "epoch": 0.9003065245949496,
+      "grad_norm": 0.093565748137191,
+      "learning_rate": 1.3418071380454501e-05,
+      "loss": 0.4699,
+      "step": 771
+    },
+    {
+      "epoch": 0.9014742373376149,
+      "grad_norm": 0.08664192746902967,
+      "learning_rate": 1.3398893015826166e-05,
+      "loss": 0.4991,
+      "step": 772
+    },
+    {
+      "epoch": 0.9026419500802803,
+      "grad_norm": 0.10162975520202895,
+      "learning_rate": 1.337970050643572e-05,
+      "loss": 0.5503,
+      "step": 773
+    },
+    {
+      "epoch": 0.9038096628229455,
+      "grad_norm": 0.08429035003077145,
+      "learning_rate": 1.3360493932154301e-05,
+      "loss": 0.4644,
+      "step": 774
+    },
+    {
+      "epoch": 0.9049773755656109,
+      "grad_norm": 0.08679002793051512,
+      "learning_rate": 1.3341273372911591e-05,
+      "loss": 0.5201,
+      "step": 775
+    },
+    {
+      "epoch": 0.9061450883082761,
+      "grad_norm": 0.09631702912968285,
+      "learning_rate": 1.3322038908695466e-05,
+      "loss": 0.5122,
+      "step": 776
+    },
+    {
+      "epoch": 0.9073128010509415,
+      "grad_norm": 0.09225052719991648,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.5215,
+      "step": 777
+    },
+    {
+      "epoch": 0.9084805137936067,
+      "grad_norm": 0.0917590742994506,
+      "learning_rate": 1.3283528585583484e-05,
+      "loss": 0.5151,
+      "step": 778
+    },
+    {
+      "epoch": 0.9096482265362721,
+      "grad_norm": 0.09189168646202008,
+      "learning_rate": 1.326425288695138e-05,
+      "loss": 0.5403,
+      "step": 779
+    },
+    {
+      "epoch": 0.9108159392789373,
+      "grad_norm": 0.09010958592883682,
+      "learning_rate": 1.3244963603872707e-05,
+      "loss": 0.5088,
+      "step": 780
+    },
+    {
+      "epoch": 0.9119836520216027,
+      "grad_norm": 0.3408634948069181,
+      "learning_rate": 1.3225660816621342e-05,
+      "loss": 0.5127,
+      "step": 781
+    },
+    {
+      "epoch": 0.913151364764268,
+      "grad_norm": 0.08893967846677533,
+      "learning_rate": 1.3206344605527355e-05,
+      "loss": 0.5356,
+      "step": 782
+    },
+    {
+      "epoch": 0.9143190775069333,
+      "grad_norm": 0.08942385507991694,
+      "learning_rate": 1.3187015050976696e-05,
+      "loss": 0.5005,
+      "step": 783
+    },
+    {
+      "epoch": 0.9154867902495986,
+      "grad_norm": 0.09018750621719458,
+      "learning_rate": 1.3167672233410826e-05,
+      "loss": 0.5207,
+      "step": 784
+    },
+    {
+      "epoch": 0.9166545029922639,
+      "grad_norm": 0.09046429815076398,
+      "learning_rate": 1.3148316233326414e-05,
+      "loss": 0.53,
+      "step": 785
+    },
+    {
+      "epoch": 0.9178222157349292,
+      "grad_norm": 0.0911761330092476,
+      "learning_rate": 1.312894713127499e-05,
+      "loss": 0.4837,
+      "step": 786
+    },
+    {
+      "epoch": 0.9189899284775945,
+      "grad_norm": 0.08461441871387797,
+      "learning_rate": 1.3109565007862597e-05,
+      "loss": 0.5033,
+      "step": 787
+    },
+    {
+      "epoch": 0.9201576412202598,
+      "grad_norm": 0.09222590938550988,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.4776,
+      "step": 788
+    },
+    {
+      "epoch": 0.9213253539629251,
+      "grad_norm": 0.1585097726286984,
+      "learning_rate": 1.3070762019649722e-05,
+      "loss": 0.5193,
+      "step": 789
+    },
+    {
+      "epoch": 0.9224930667055904,
+      "grad_norm": 0.08451451675879941,
+      "learning_rate": 1.3051341316330946e-05,
+      "loss": 0.494,
+      "step": 790
+    },
+    {
+      "epoch": 0.9236607794482558,
+      "grad_norm": 0.08952984905029172,
+      "learning_rate": 1.303190791461394e-05,
+      "loss": 0.5282,
+      "step": 791
+    },
+    {
+      "epoch": 0.924828492190921,
+      "grad_norm": 0.08825561740671896,
+      "learning_rate": 1.3012461895372343e-05,
+      "loss": 0.4953,
+      "step": 792
+    },
+    {
+      "epoch": 0.9259962049335864,
+      "grad_norm": 0.08736348538695109,
+      "learning_rate": 1.2993003339532304e-05,
+      "loss": 0.521,
+      "step": 793
+    },
+    {
+      "epoch": 0.9271639176762516,
+      "grad_norm": 0.08298591370151669,
+      "learning_rate": 1.2973532328072138e-05,
+      "loss": 0.4868,
+      "step": 794
+    },
+    {
+      "epoch": 0.928331630418917,
+      "grad_norm": 0.08341468920438881,
+      "learning_rate": 1.2954048942022002e-05,
+      "loss": 0.4866,
+      "step": 795
+    },
+    {
+      "epoch": 0.9294993431615822,
+      "grad_norm": 0.08944389306269677,
+      "learning_rate": 1.2934553262463548e-05,
+      "loss": 0.491,
+      "step": 796
+    },
+    {
+      "epoch": 0.9306670559042476,
+      "grad_norm": 0.08596137385075837,
+      "learning_rate": 1.2915045370529586e-05,
+      "loss": 0.5035,
+      "step": 797
+    },
+    {
+      "epoch": 0.9318347686469128,
+      "grad_norm": 0.08974862458041069,
+      "learning_rate": 1.2895525347403757e-05,
+      "loss": 0.4949,
+      "step": 798
+    },
+    {
+      "epoch": 0.9330024813895782,
+      "grad_norm": 0.08916752177966264,
+      "learning_rate": 1.2875993274320173e-05,
+      "loss": 0.4877,
+      "step": 799
+    },
+    {
+      "epoch": 0.9341701941322434,
+      "grad_norm": 0.08638017056762691,
+      "learning_rate": 1.285644923256311e-05,
+      "loss": 0.5077,
+      "step": 800
+    },
+    {
+      "epoch": 0.9353379068749088,
+      "grad_norm": 0.09064533601966272,
+      "learning_rate": 1.2836893303466636e-05,
+      "loss": 0.5357,
+      "step": 801
+    },
+    {
+      "epoch": 0.9365056196175741,
+      "grad_norm": 0.09572057834765484,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.5285,
+      "step": 802
+    },
+    {
+      "epoch": 0.9376733323602394,
+      "grad_norm": 0.08607232705847896,
+      "learning_rate": 1.2797746108838777e-05,
+      "loss": 0.4844,
+      "step": 803
+    },
+    {
+      "epoch": 0.9388410451029047,
+      "grad_norm": 0.08943232967642067,
+      "learning_rate": 1.277815500622154e-05,
+      "loss": 0.5194,
+      "step": 804
+    },
+    {
+      "epoch": 0.94000875784557,
+      "grad_norm": 0.08944025844557332,
+      "learning_rate": 1.275855234209251e-05,
+      "loss": 0.535,
+      "step": 805
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.08442694967260898,
+      "learning_rate": 1.2738938198029725e-05,
+      "loss": 0.4617,
+      "step": 806
+    },
+    {
+      "epoch": 0.9423441833309006,
+      "grad_norm": 0.08591729389474993,
+      "learning_rate": 1.2719312655658995e-05,
+      "loss": 0.5035,
+      "step": 807
+    },
+    {
+      "epoch": 0.9435118960735659,
+      "grad_norm": 0.08273958753179908,
+      "learning_rate": 1.269967579665357e-05,
+      "loss": 0.4804,
+      "step": 808
+    },
+    {
+      "epoch": 0.9446796088162313,
+      "grad_norm": 0.09178154659663433,
+      "learning_rate": 1.2680027702733791e-05,
+      "loss": 0.5188,
+      "step": 809
+    },
+    {
+      "epoch": 0.9458473215588965,
+      "grad_norm": 0.0850407201807945,
+      "learning_rate": 1.2660368455666752e-05,
+      "loss": 0.5009,
+      "step": 810
+    },
+    {
+      "epoch": 0.9470150343015619,
+      "grad_norm": 0.0906493955753226,
+      "learning_rate": 1.2640698137265967e-05,
+      "loss": 0.4967,
+      "step": 811
+    },
+    {
+      "epoch": 0.9481827470442271,
+      "grad_norm": 0.08806682720723877,
+      "learning_rate": 1.2621016829391022e-05,
+      "loss": 0.516,
+      "step": 812
+    },
+    {
+      "epoch": 0.9493504597868925,
+      "grad_norm": 0.09088497972958133,
+      "learning_rate": 1.2601324613947239e-05,
+      "loss": 0.5181,
+      "step": 813
+    },
+    {
+      "epoch": 0.9505181725295577,
+      "grad_norm": 0.09030559966729769,
+      "learning_rate": 1.2581621572885324e-05,
+      "loss": 0.5109,
+      "step": 814
+    },
+    {
+      "epoch": 0.9516858852722231,
+      "grad_norm": 0.08606752016403738,
+      "learning_rate": 1.2561907788201042e-05,
+      "loss": 0.4996,
+      "step": 815
+    },
+    {
+      "epoch": 0.9528535980148883,
+      "grad_norm": 0.07880003963682261,
+      "learning_rate": 1.2542183341934873e-05,
+      "loss": 0.4668,
+      "step": 816
+    },
+    {
+      "epoch": 0.9540213107575536,
+      "grad_norm": 0.08817972676348344,
+      "learning_rate": 1.2522448316171651e-05,
+      "loss": 0.4473,
+      "step": 817
+    },
+    {
+      "epoch": 0.955189023500219,
+      "grad_norm": 0.09453976834889187,
+      "learning_rate": 1.2502702793040257e-05,
+      "loss": 0.5126,
+      "step": 818
+    },
+    {
+      "epoch": 0.9563567362428842,
+      "grad_norm": 0.09025415184969021,
+      "learning_rate": 1.2482946854713238e-05,
+      "loss": 0.5221,
+      "step": 819
+    },
+    {
+      "epoch": 0.9575244489855496,
+      "grad_norm": 0.08457799904498127,
+      "learning_rate": 1.24631805834065e-05,
+      "loss": 0.4856,
+      "step": 820
+    },
+    {
+      "epoch": 0.9586921617282148,
+      "grad_norm": 0.08531253536127277,
+      "learning_rate": 1.2443404061378941e-05,
+      "loss": 0.4908,
+      "step": 821
+    },
+    {
+      "epoch": 0.9598598744708802,
+      "grad_norm": 0.08716337921006603,
+      "learning_rate": 1.2423617370932127e-05,
+      "loss": 0.5227,
+      "step": 822
+    },
+    {
+      "epoch": 0.9610275872135454,
+      "grad_norm": 0.08814950873331522,
+      "learning_rate": 1.2403820594409926e-05,
+      "loss": 0.5178,
+      "step": 823
+    },
+    {
+      "epoch": 0.9621952999562108,
+      "grad_norm": 0.08720706085390008,
+      "learning_rate": 1.2384013814198197e-05,
+      "loss": 0.5017,
+      "step": 824
+    },
+    {
+      "epoch": 0.963363012698876,
+      "grad_norm": 0.08607738820485343,
+      "learning_rate": 1.2364197112724421e-05,
+      "loss": 0.5202,
+      "step": 825
+    },
+    {
+      "epoch": 0.9645307254415414,
+      "grad_norm": 0.08802106502487528,
+      "learning_rate": 1.2344370572457368e-05,
+      "loss": 0.4995,
+      "step": 826
+    },
+    {
+      "epoch": 0.9656984381842066,
+      "grad_norm": 0.0824684270189752,
+      "learning_rate": 1.2324534275906747e-05,
+      "loss": 0.455,
+      "step": 827
+    },
+    {
+      "epoch": 0.966866150926872,
+      "grad_norm": 0.08792660038381252,
+      "learning_rate": 1.2304688305622889e-05,
+      "loss": 0.5079,
+      "step": 828
+    },
+    {
+      "epoch": 0.9680338636695373,
+      "grad_norm": 0.08868436456506142,
+      "learning_rate": 1.2284832744196355e-05,
+      "loss": 0.4913,
+      "step": 829
+    },
+    {
+      "epoch": 0.9692015764122026,
+      "grad_norm": 0.08353386099880912,
+      "learning_rate": 1.2264967674257647e-05,
+      "loss": 0.5168,
+      "step": 830
+    },
+    {
+      "epoch": 0.9703692891548679,
+      "grad_norm": 0.09048836477835019,
+      "learning_rate": 1.2245093178476813e-05,
+      "loss": 0.5232,
+      "step": 831
+    },
+    {
+      "epoch": 0.9715370018975332,
+      "grad_norm": 0.0928731930466399,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.5022,
+      "step": 832
+    },
+    {
+      "epoch": 0.9727047146401985,
+      "grad_norm": 0.08470226908546906,
+      "learning_rate": 1.2205316240264812e-05,
+      "loss": 0.4896,
+      "step": 833
+    },
+    {
+      "epoch": 0.9738724273828638,
+      "grad_norm": 0.08209178033230537,
+      "learning_rate": 1.218541396336852e-05,
+      "loss": 0.4836,
+      "step": 834
+    },
+    {
+      "epoch": 0.9750401401255291,
+      "grad_norm": 0.08821733087027568,
+      "learning_rate": 1.2165502591699166e-05,
+      "loss": 0.5208,
+      "step": 835
+    },
+    {
+      "epoch": 0.9762078528681944,
+      "grad_norm": 0.08457877242130701,
+      "learning_rate": 1.2145582208119497e-05,
+      "loss": 0.4972,
+      "step": 836
+    },
+    {
+      "epoch": 0.9773755656108597,
+      "grad_norm": 0.0871570454306533,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.5252,
+      "step": 837
+    },
+    {
+      "epoch": 0.9785432783535251,
+      "grad_norm": 0.08516564972248847,
+      "learning_rate": 1.2105714736867391e-05,
+      "loss": 0.5065,
+      "step": 838
+    },
+    {
+      "epoch": 0.9797109910961903,
+      "grad_norm": 0.08367268707184083,
+      "learning_rate": 1.2085767815106583e-05,
+      "loss": 0.5111,
+      "step": 839
+    },
+    {
+      "epoch": 0.9808787038388557,
+      "grad_norm": 0.08343399149710296,
+      "learning_rate": 1.2065812213258051e-05,
+      "loss": 0.4749,
+      "step": 840
+    },
+    {
+      "epoch": 0.9820464165815209,
+      "grad_norm": 0.08973398731097092,
+      "learning_rate": 1.2045848014368591e-05,
+      "loss": 0.5491,
+      "step": 841
+    },
+    {
+      "epoch": 0.9832141293241863,
+      "grad_norm": 0.08599299831430852,
+      "learning_rate": 1.2025875301520811e-05,
+      "loss": 0.502,
+      "step": 842
+    },
+    {
+      "epoch": 0.9843818420668515,
+      "grad_norm": 0.08318640757764813,
+      "learning_rate": 1.200589415783273e-05,
+      "loss": 0.4916,
+      "step": 843
+    },
+    {
+      "epoch": 0.9855495548095169,
+      "grad_norm": 0.08150977061343474,
+      "learning_rate": 1.1985904666457455e-05,
+      "loss": 0.4851,
+      "step": 844
+    },
+    {
+      "epoch": 0.9867172675521821,
+      "grad_norm": 0.08018931403925618,
+      "learning_rate": 1.196590691058284e-05,
+      "loss": 0.49,
+      "step": 845
+    },
+    {
+      "epoch": 0.9878849802948475,
+      "grad_norm": 0.08425610271600195,
+      "learning_rate": 1.1945900973431127e-05,
+      "loss": 0.4933,
+      "step": 846
+    },
+    {
+      "epoch": 0.9890526930375128,
+      "grad_norm": 0.08613584416372519,
+      "learning_rate": 1.1925886938258606e-05,
+      "loss": 0.48,
+      "step": 847
+    },
+    {
+      "epoch": 0.9902204057801781,
+      "grad_norm": 0.08895467622300753,
+      "learning_rate": 1.1905864888355264e-05,
+      "loss": 0.4977,
+      "step": 848
+    },
+    {
+      "epoch": 0.9913881185228434,
+      "grad_norm": 0.09204308352258604,
+      "learning_rate": 1.1885834907044444e-05,
+      "loss": 0.5373,
+      "step": 849
+    },
+    {
+      "epoch": 0.9925558312655087,
+      "grad_norm": 0.08378311154628072,
+      "learning_rate": 1.1865797077682509e-05,
+      "loss": 0.4821,
+      "step": 850
+    },
+    {
+      "epoch": 0.993723544008174,
+      "grad_norm": 0.08523060710334014,
+      "learning_rate": 1.1845751483658454e-05,
+      "loss": 0.4903,
+      "step": 851
+    },
+    {
+      "epoch": 0.9948912567508393,
+      "grad_norm": 0.08226409678909838,
+      "learning_rate": 1.182569820839362e-05,
+      "loss": 0.4634,
+      "step": 852
+    },
+    {
+      "epoch": 0.9960589694935046,
+      "grad_norm": 0.08920786896214337,
+      "learning_rate": 1.1805637335341285e-05,
+      "loss": 0.5364,
+      "step": 853
+    },
+    {
+      "epoch": 0.99722668223617,
+      "grad_norm": 0.08158275595574004,
+      "learning_rate": 1.1785568947986368e-05,
+      "loss": 0.4998,
+      "step": 854
+    },
+    {
+      "epoch": 0.9983943949788352,
+      "grad_norm": 0.08752745351144177,
+      "learning_rate": 1.1765493129845047e-05,
+      "loss": 0.4985,
+      "step": 855
+    },
+    {
+      "epoch": 0.9995621077215006,
+      "grad_norm": 0.08639331438549855,
+      "learning_rate": 1.1745409964464425e-05,
+      "loss": 0.475,
+      "step": 856
+    },
+    {
+      "epoch": 1.0011677127426653,
+      "grad_norm": 0.10687507579924417,
+      "learning_rate": 1.1725319535422189e-05,
+      "loss": 0.4,
+      "step": 857
+    },
+    {
+      "epoch": 1.0023354254853305,
+      "grad_norm": 0.11329909483779083,
+      "learning_rate": 1.170522192632624e-05,
+      "loss": 0.4037,
+      "step": 858
+    },
+    {
+      "epoch": 1.003503138227996,
+      "grad_norm": 0.08790537333673759,
+      "learning_rate": 1.1685117220814373e-05,
+      "loss": 0.3954,
+      "step": 859
+    },
+    {
+      "epoch": 1.0046708509706612,
+      "grad_norm": 0.09468298173555426,
+      "learning_rate": 1.1665005502553912e-05,
+      "loss": 0.4011,
+      "step": 860
+    },
+    {
+      "epoch": 1.0058385637133265,
+      "grad_norm": 0.09916660077374043,
+      "learning_rate": 1.164488685524136e-05,
+      "loss": 0.3839,
+      "step": 861
+    },
+    {
+      "epoch": 1.0070062764559917,
+      "grad_norm": 0.10650431825855765,
+      "learning_rate": 1.1624761362602061e-05,
+      "loss": 0.3995,
+      "step": 862
+    },
+    {
+      "epoch": 1.0081739891986572,
+      "grad_norm": 0.11564468915026303,
+      "learning_rate": 1.160462910838984e-05,
+      "loss": 0.3933,
+      "step": 863
+    },
+    {
+      "epoch": 1.0093417019413224,
+      "grad_norm": 0.1036613986116272,
+      "learning_rate": 1.1584490176386671e-05,
+      "loss": 0.4016,
+      "step": 864
+    },
+    {
+      "epoch": 1.0105094146839877,
+      "grad_norm": 0.09980107488841199,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.3792,
+      "step": 865
+    },
+    {
+      "epoch": 1.011677127426653,
+      "grad_norm": 0.09971237003592305,
+      "learning_rate": 1.1544192614273957e-05,
+      "loss": 0.3947,
+      "step": 866
+    },
+    {
+      "epoch": 1.0128448401693184,
+      "grad_norm": 0.09343750961048831,
+      "learning_rate": 1.1524034151865904e-05,
+      "loss": 0.374,
+      "step": 867
+    },
+    {
+      "epoch": 1.0140125529119837,
+      "grad_norm": 0.11794863375534724,
+      "learning_rate": 1.1503869347069186e-05,
+      "loss": 0.4023,
+      "step": 868
+    },
+    {
+      "epoch": 1.015180265654649,
+      "grad_norm": 0.09941795063748246,
+      "learning_rate": 1.1483698283801233e-05,
+      "loss": 0.3776,
+      "step": 869
+    },
+    {
+      "epoch": 1.0163479783973142,
+      "grad_norm": 0.10971112599810502,
+      "learning_rate": 1.1463521046005524e-05,
+      "loss": 0.3949,
+      "step": 870
+    },
+    {
+      "epoch": 1.0175156911399796,
+      "grad_norm": 0.10217776725700205,
+      "learning_rate": 1.1443337717651221e-05,
+      "loss": 0.3951,
+      "step": 871
+    },
+    {
+      "epoch": 1.0186834038826449,
+      "grad_norm": 0.0937392057088196,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3762,
+      "step": 872
+    },
+    {
+      "epoch": 1.0198511166253101,
+      "grad_norm": 0.09586457482158174,
+      "learning_rate": 1.1402953125269927e-05,
+      "loss": 0.3913,
+      "step": 873
+    },
+    {
+      "epoch": 1.0210188293679754,
+      "grad_norm": 0.09915260496280234,
+      "learning_rate": 1.1382752029306606e-05,
+      "loss": 0.387,
+      "step": 874
+    },
+    {
+      "epoch": 1.0221865421106409,
+      "grad_norm": 0.09856015609414595,
+      "learning_rate": 1.1362545178911346e-05,
+      "loss": 0.3817,
+      "step": 875
+    },
+    {
+      "epoch": 1.0233542548533061,
+      "grad_norm": 0.09830437451790303,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.3956,
+      "step": 876
+    },
+    {
+      "epoch": 1.0245219675959714,
+      "grad_norm": 0.10212375434668003,
+      "learning_rate": 1.132211455121824e-05,
+      "loss": 0.4095,
+      "step": 877
+    },
+    {
+      "epoch": 1.0256896803386366,
+      "grad_norm": 0.1043852511915365,
+      "learning_rate": 1.1301890942175649e-05,
+      "loss": 0.4239,
+      "step": 878
+    },
+    {
+      "epoch": 1.026857393081302,
+      "grad_norm": 0.09138548878815848,
+      "learning_rate": 1.1281661915210931e-05,
+      "loss": 0.3847,
+      "step": 879
+    },
+    {
+      "epoch": 1.0280251058239673,
+      "grad_norm": 0.11327338693997174,
+      "learning_rate": 1.126142755450878e-05,
+      "loss": 0.3917,
+      "step": 880
+    },
+    {
+      "epoch": 1.0291928185666326,
+      "grad_norm": 0.09038885215386036,
+      "learning_rate": 1.1241187944276088e-05,
+      "loss": 0.3831,
+      "step": 881
+    },
+    {
+      "epoch": 1.0303605313092978,
+      "grad_norm": 0.09489881395801665,
+      "learning_rate": 1.12209431687416e-05,
+      "loss": 0.3694,
+      "step": 882
+    },
+    {
+      "epoch": 1.0315282440519633,
+      "grad_norm": 0.0980284700414046,
+      "learning_rate": 1.1200693312155545e-05,
+      "loss": 0.4387,
+      "step": 883
+    },
+    {
+      "epoch": 1.0326959567946286,
+      "grad_norm": 0.09519934271443758,
+      "learning_rate": 1.1180438458789305e-05,
+      "loss": 0.3912,
+      "step": 884
+    },
+    {
+      "epoch": 1.0338636695372938,
+      "grad_norm": 0.09097342487978029,
+      "learning_rate": 1.1160178692935054e-05,
+      "loss": 0.3852,
+      "step": 885
+    },
+    {
+      "epoch": 1.035031382279959,
+      "grad_norm": 0.09243485142336327,
+      "learning_rate": 1.1139914098905406e-05,
+      "loss": 0.3674,
+      "step": 886
+    },
+    {
+      "epoch": 1.0361990950226245,
+      "grad_norm": 0.09920918757448359,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 0.3904,
+      "step": 887
+    },
+    {
+      "epoch": 1.0373668077652898,
+      "grad_norm": 0.09805487140551557,
+      "learning_rate": 1.1099370763670523e-05,
+      "loss": 0.408,
+      "step": 888
+    },
+    {
+      "epoch": 1.038534520507955,
+      "grad_norm": 0.09783260971824685,
+      "learning_rate": 1.107909219118958e-05,
+      "loss": 0.3665,
+      "step": 889
+    },
+    {
+      "epoch": 1.0397022332506203,
+      "grad_norm": 0.0934808809531099,
+      "learning_rate": 1.1058809127981133e-05,
+      "loss": 0.3665,
+      "step": 890
+    },
+    {
+      "epoch": 1.0408699459932858,
+      "grad_norm": 0.09473383020780415,
+      "learning_rate": 1.1038521658454753e-05,
+      "loss": 0.3848,
+      "step": 891
+    },
+    {
+      "epoch": 1.042037658735951,
+      "grad_norm": 0.09615362954376974,
+      "learning_rate": 1.1018229867038358e-05,
+      "loss": 0.3807,
+      "step": 892
+    },
+    {
+      "epoch": 1.0432053714786163,
+      "grad_norm": 0.09726694205258642,
+      "learning_rate": 1.0997933838177828e-05,
+      "loss": 0.4016,
+      "step": 893
+    },
+    {
+      "epoch": 1.0443730842212815,
+      "grad_norm": 0.09557480612653117,
+      "learning_rate": 1.0977633656336706e-05,
+      "loss": 0.4068,
+      "step": 894
+    },
+    {
+      "epoch": 1.045540796963947,
+      "grad_norm": 0.09250651244953627,
+      "learning_rate": 1.0957329405995794e-05,
+      "loss": 0.3783,
+      "step": 895
+    },
+    {
+      "epoch": 1.0467085097066122,
+      "grad_norm": 0.09800475244938166,
+      "learning_rate": 1.0937021171652842e-05,
+      "loss": 0.3745,
+      "step": 896
+    },
+    {
+      "epoch": 1.0478762224492775,
+      "grad_norm": 0.09186411687230699,
+      "learning_rate": 1.0916709037822172e-05,
+      "loss": 0.3717,
+      "step": 897
+    },
+    {
+      "epoch": 1.0490439351919427,
+      "grad_norm": 0.08789177409833958,
+      "learning_rate": 1.0896393089034336e-05,
+      "loss": 0.3568,
+      "step": 898
+    },
+    {
+      "epoch": 1.0502116479346082,
+      "grad_norm": 0.0949544101136822,
+      "learning_rate": 1.0876073409835762e-05,
+      "loss": 0.3576,
+      "step": 899
+    },
+    {
+      "epoch": 1.0513793606772734,
+      "grad_norm": 0.09201189165269388,
+      "learning_rate": 1.08557500847884e-05,
+      "loss": 0.3881,
+      "step": 900
+    },
+    {
+      "epoch": 1.0525470734199387,
+      "grad_norm": 0.09036576507454733,
+      "learning_rate": 1.0835423198469373e-05,
+      "loss": 0.3565,
+      "step": 901
+    },
+    {
+      "epoch": 1.053714786162604,
+      "grad_norm": 0.09986780274167059,
+      "learning_rate": 1.0815092835470633e-05,
+      "loss": 0.4053,
+      "step": 902
+    },
+    {
+      "epoch": 1.0548824989052692,
+      "grad_norm": 0.09454541889103038,
+      "learning_rate": 1.0794759080398587e-05,
+      "loss": 0.3884,
+      "step": 903
+    },
+    {
+      "epoch": 1.0560502116479347,
+      "grad_norm": 0.09779455585663654,
+      "learning_rate": 1.077442201787377e-05,
+      "loss": 0.3652,
+      "step": 904
+    },
+    {
+      "epoch": 1.0572179243906,
+      "grad_norm": 0.09604086348057415,
+      "learning_rate": 1.0754081732530473e-05,
+      "loss": 0.3957,
+      "step": 905
+    },
+    {
+      "epoch": 1.0583856371332652,
+      "grad_norm": 0.09460777164791803,
+      "learning_rate": 1.0733738309016402e-05,
+      "loss": 0.4029,
+      "step": 906
+    },
+    {
+      "epoch": 1.0595533498759304,
+      "grad_norm": 0.09266789815148346,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.3874,
+      "step": 907
+    },
+    {
+      "epoch": 1.060721062618596,
+      "grad_norm": 0.0977478457011073,
+      "learning_rate": 1.0693042386131713e-05,
+      "loss": 0.396,
+      "step": 908
+    },
+    {
+      "epoch": 1.0618887753612611,
+      "grad_norm": 0.09279050253355389,
+      "learning_rate": 1.0672690056120398e-05,
+      "loss": 0.3711,
+      "step": 909
+    },
+    {
+      "epoch": 1.0630564881039264,
+      "grad_norm": 0.09777401216975849,
+      "learning_rate": 1.0652334926656209e-05,
+      "loss": 0.3912,
+      "step": 910
+    },
+    {
+      "epoch": 1.0642242008465916,
+      "grad_norm": 0.10161102094910146,
+      "learning_rate": 1.0631977082448626e-05,
+      "loss": 0.3932,
+      "step": 911
+    },
+    {
+      "epoch": 1.0653919135892571,
+      "grad_norm": 0.09723663789267635,
+      "learning_rate": 1.0611616608218429e-05,
+      "loss": 0.374,
+      "step": 912
+    },
+    {
+      "epoch": 1.0665596263319224,
+      "grad_norm": 0.09947729269803318,
+      "learning_rate": 1.059125358869734e-05,
+      "loss": 0.3935,
+      "step": 913
+    },
+    {
+      "epoch": 1.0677273390745876,
+      "grad_norm": 0.10063468961781528,
+      "learning_rate": 1.0570888108627682e-05,
+      "loss": 0.3861,
+      "step": 914
+    },
+    {
+      "epoch": 1.0688950518172529,
+      "grad_norm": 0.09774158643148059,
+      "learning_rate": 1.0550520252762002e-05,
+      "loss": 0.4096,
+      "step": 915
+    },
+    {
+      "epoch": 1.0700627645599183,
+      "grad_norm": 0.09745786109455548,
+      "learning_rate": 1.0530150105862748e-05,
+      "loss": 0.3949,
+      "step": 916
+    },
+    {
+      "epoch": 1.0712304773025836,
+      "grad_norm": 0.09779095018434988,
+      "learning_rate": 1.05097777527019e-05,
+      "loss": 0.4068,
+      "step": 917
+    },
+    {
+      "epoch": 1.0723981900452488,
+      "grad_norm": 0.09179041981291437,
+      "learning_rate": 1.0489403278060613e-05,
+      "loss": 0.369,
+      "step": 918
+    },
+    {
+      "epoch": 1.073565902787914,
+      "grad_norm": 0.09397863418838655,
+      "learning_rate": 1.0469026766728876e-05,
+      "loss": 0.3616,
+      "step": 919
+    },
+    {
+      "epoch": 1.0747336155305796,
+      "grad_norm": 0.09432971607971319,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.3752,
+      "step": 920
+    },
+    {
+      "epoch": 1.0759013282732448,
+      "grad_norm": 0.09488309688697125,
+      "learning_rate": 1.0428267973196027e-05,
+      "loss": 0.3875,
+      "step": 921
+    },
+    {
+      "epoch": 1.07706904101591,
+      "grad_norm": 0.0941479477360919,
+      "learning_rate": 1.0407885860615859e-05,
+      "loss": 0.3933,
+      "step": 922
+    },
+    {
+      "epoch": 1.0782367537585753,
+      "grad_norm": 0.09520477177170705,
+      "learning_rate": 1.0387502050586418e-05,
+      "loss": 0.3825,
+      "step": 923
+    },
+    {
+      "epoch": 1.0794044665012408,
+      "grad_norm": 0.09752896070696108,
+      "learning_rate": 1.0367116627936549e-05,
+      "loss": 0.372,
+      "step": 924
+    },
+    {
+      "epoch": 1.080572179243906,
+      "grad_norm": 0.09793729984867162,
+      "learning_rate": 1.0346729677501793e-05,
+      "loss": 0.3986,
+      "step": 925
+    },
+    {
+      "epoch": 1.0817398919865713,
+      "grad_norm": 1.3724626183544528,
+      "learning_rate": 1.0326341284124061e-05,
+      "loss": 0.5223,
+      "step": 926
+    },
+    {
+      "epoch": 1.0829076047292365,
+      "grad_norm": 0.101698923293553,
+      "learning_rate": 1.0305951532651262e-05,
+      "loss": 0.3834,
+      "step": 927
+    },
+    {
+      "epoch": 1.084075317471902,
+      "grad_norm": 0.10590282035249167,
+      "learning_rate": 1.0285560507936962e-05,
+      "loss": 0.3909,
+      "step": 928
+    },
+    {
+      "epoch": 1.0852430302145673,
+      "grad_norm": 0.13476487614838073,
+      "learning_rate": 1.0265168294840025e-05,
+      "loss": 0.3882,
+      "step": 929
+    },
+    {
+      "epoch": 1.0864107429572325,
+      "grad_norm": 0.4862061947198677,
+      "learning_rate": 1.0244774978224254e-05,
+      "loss": 0.4162,
+      "step": 930
+    },
+    {
+      "epoch": 1.0875784556998978,
+      "grad_norm": 0.11018726715695439,
+      "learning_rate": 1.0224380642958052e-05,
+      "loss": 0.4025,
+      "step": 931
+    },
+    {
+      "epoch": 1.0887461684425632,
+      "grad_norm": 0.11902669074473962,
+      "learning_rate": 1.0203985373914056e-05,
+      "loss": 0.3973,
+      "step": 932
+    },
+    {
+      "epoch": 1.0899138811852285,
+      "grad_norm": 0.12793532400439037,
+      "learning_rate": 1.018358925596879e-05,
+      "loss": 0.3742,
+      "step": 933
+    },
+    {
+      "epoch": 1.0910815939278937,
+      "grad_norm": 0.1017281569659894,
+      "learning_rate": 1.016319237400232e-05,
+      "loss": 0.3766,
+      "step": 934
+    },
+    {
+      "epoch": 1.092249306670559,
+      "grad_norm": 0.10421250165733273,
+      "learning_rate": 1.0142794812897874e-05,
+      "loss": 0.3942,
+      "step": 935
+    },
+    {
+      "epoch": 1.0934170194132244,
+      "grad_norm": 0.10731480818722398,
+      "learning_rate": 1.0122396657541522e-05,
+      "loss": 0.3724,
+      "step": 936
+    },
+    {
+      "epoch": 1.0945847321558897,
+      "grad_norm": 0.0998553275286067,
+      "learning_rate": 1.0101997992821798e-05,
+      "loss": 0.3982,
+      "step": 937
+    },
+    {
+      "epoch": 1.095752444898555,
+      "grad_norm": 0.09999804067555541,
+      "learning_rate": 1.0081598903629359e-05,
+      "loss": 0.3938,
+      "step": 938
+    },
+    {
+      "epoch": 1.0969201576412202,
+      "grad_norm": 0.11613002986140951,
+      "learning_rate": 1.0061199474856628e-05,
+      "loss": 0.3909,
+      "step": 939
+    },
+    {
+      "epoch": 1.0980878703838857,
+      "grad_norm": 0.1096065066146652,
+      "learning_rate": 1.0040799791397444e-05,
+      "loss": 0.3739,
+      "step": 940
+    },
+    {
+      "epoch": 1.099255583126551,
+      "grad_norm": 0.10347194245536699,
+      "learning_rate": 1.0020399938146701e-05,
+      "loss": 0.3958,
+      "step": 941
+    },
+    {
+      "epoch": 1.1004232958692162,
+      "grad_norm": 0.1061734773549622,
+      "learning_rate": 1e-05,
+      "loss": 0.4025,
+      "step": 942
+    },
+    {
+      "epoch": 1.1015910086118814,
+      "grad_norm": 0.09466270719368831,
+      "learning_rate": 9.979600061853302e-06,
+      "loss": 0.3754,
+      "step": 943
+    },
+    {
+      "epoch": 1.102758721354547,
+      "grad_norm": 0.10613337172593779,
+      "learning_rate": 9.95920020860256e-06,
+      "loss": 0.3981,
+      "step": 944
+    },
+    {
+      "epoch": 1.1039264340972121,
+      "grad_norm": 0.09783608243389717,
+      "learning_rate": 9.938800525143374e-06,
+      "loss": 0.3883,
+      "step": 945
+    },
+    {
+      "epoch": 1.1050941468398774,
+      "grad_norm": 0.09585197655082116,
+      "learning_rate": 9.918401096370644e-06,
+      "loss": 0.354,
+      "step": 946
+    },
+    {
+      "epoch": 1.1062618595825426,
+      "grad_norm": 0.10471428141893593,
+      "learning_rate": 9.898002007178204e-06,
+      "loss": 0.4031,
+      "step": 947
+    },
+    {
+      "epoch": 1.107429572325208,
+      "grad_norm": 0.10095083756343486,
+      "learning_rate": 9.877603342458483e-06,
+      "loss": 0.3718,
+      "step": 948
+    },
+    {
+      "epoch": 1.1085972850678734,
+      "grad_norm": 0.09500645141076981,
+      "learning_rate": 9.85720518710213e-06,
+      "loss": 0.3707,
+      "step": 949
+    },
+    {
+      "epoch": 1.1097649978105386,
+      "grad_norm": 0.09621322087212691,
+      "learning_rate": 9.836807625997683e-06,
+      "loss": 0.3763,
+      "step": 950
+    },
+    {
+      "epoch": 1.1109327105532039,
+      "grad_norm": 0.10372552877634751,
+      "learning_rate": 9.81641074403121e-06,
+      "loss": 0.3837,
+      "step": 951
+    },
+    {
+      "epoch": 1.1121004232958693,
+      "grad_norm": 0.10009878005581078,
+      "learning_rate": 9.79601462608595e-06,
+      "loss": 0.4361,
+      "step": 952
+    },
+    {
+      "epoch": 1.1132681360385346,
+      "grad_norm": 0.09594334046770711,
+      "learning_rate": 9.775619357041952e-06,
+      "loss": 0.39,
+      "step": 953
+    },
+    {
+      "epoch": 1.1144358487811998,
+      "grad_norm": 0.1072860547731465,
+      "learning_rate": 9.75522502177575e-06,
+      "loss": 0.4231,
+      "step": 954
+    },
+    {
+      "epoch": 1.115603561523865,
+      "grad_norm": 0.10240362214736846,
+      "learning_rate": 9.734831705159975e-06,
+      "loss": 0.3983,
+      "step": 955
+    },
+    {
+      "epoch": 1.1167712742665303,
+      "grad_norm": 0.094344257634451,
+      "learning_rate": 9.71443949206304e-06,
+      "loss": 0.3918,
+      "step": 956
+    },
+    {
+      "epoch": 1.1179389870091958,
+      "grad_norm": 0.0960460350544273,
+      "learning_rate": 9.69404846734874e-06,
+      "loss": 0.3872,
+      "step": 957
+    },
+    {
+      "epoch": 1.119106699751861,
+      "grad_norm": 0.09939792687028334,
+      "learning_rate": 9.67365871587594e-06,
+      "loss": 0.3934,
+      "step": 958
+    },
+    {
+      "epoch": 1.1202744124945263,
+      "grad_norm": 0.09676706752533512,
+      "learning_rate": 9.653270322498209e-06,
+      "loss": 0.3792,
+      "step": 959
+    },
+    {
+      "epoch": 1.1214421252371916,
+      "grad_norm": 0.09444160158396059,
+      "learning_rate": 9.632883372063458e-06,
+      "loss": 0.3808,
+      "step": 960
+    },
+    {
+      "epoch": 1.122609837979857,
+      "grad_norm": 0.09384811669477879,
+      "learning_rate": 9.612497949413586e-06,
+      "loss": 0.3818,
+      "step": 961
+    },
+    {
+      "epoch": 1.1237775507225223,
+      "grad_norm": 0.09941213515962642,
+      "learning_rate": 9.592114139384146e-06,
+      "loss": 0.3998,
+      "step": 962
+    },
+    {
+      "epoch": 1.1249452634651875,
+      "grad_norm": 0.1017720846845437,
+      "learning_rate": 9.571732026803978e-06,
+      "loss": 0.4102,
+      "step": 963
+    },
+    {
+      "epoch": 1.1261129762078528,
+      "grad_norm": 0.09804075564166953,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.387,
+      "step": 964
+    },
+    {
+      "epoch": 1.1272806889505183,
+      "grad_norm": 0.09346927923249171,
+      "learning_rate": 9.530973233271127e-06,
+      "loss": 0.3708,
+      "step": 965
+    },
+    {
+      "epoch": 1.1284484016931835,
+      "grad_norm": 0.10996408781363641,
+      "learning_rate": 9.51059672193939e-06,
+      "loss": 0.406,
+      "step": 966
+    },
+    {
+      "epoch": 1.1296161144358488,
+      "grad_norm": 0.10075821727956762,
+      "learning_rate": 9.4902222472981e-06,
+      "loss": 0.4033,
+      "step": 967
+    },
+    {
+      "epoch": 1.130783827178514,
+      "grad_norm": 0.09108500032163587,
+      "learning_rate": 9.469849894137254e-06,
+      "loss": 0.3677,
+      "step": 968
+    },
+    {
+      "epoch": 1.1319515399211795,
+      "grad_norm": 0.09730173505902022,
+      "learning_rate": 9.449479747238001e-06,
+      "loss": 0.4003,
+      "step": 969
+    },
+    {
+      "epoch": 1.1331192526638447,
+      "grad_norm": 0.09815164376789498,
+      "learning_rate": 9.42911189137232e-06,
+      "loss": 0.3824,
+      "step": 970
+    },
+    {
+      "epoch": 1.13428696540651,
+      "grad_norm": 0.09425815911972142,
+      "learning_rate": 9.408746411302661e-06,
+      "loss": 0.3643,
+      "step": 971
+    },
+    {
+      "epoch": 1.1354546781491752,
+      "grad_norm": 0.10099212428596535,
+      "learning_rate": 9.388383391781576e-06,
+      "loss": 0.4083,
+      "step": 972
+    },
+    {
+      "epoch": 1.1366223908918407,
+      "grad_norm": 0.0991699212958655,
+      "learning_rate": 9.368022917551377e-06,
+      "loss": 0.4087,
+      "step": 973
+    },
+    {
+      "epoch": 1.137790103634506,
+      "grad_norm": 0.0983527591656301,
+      "learning_rate": 9.347665073343795e-06,
+      "loss": 0.3955,
+      "step": 974
+    },
+    {
+      "epoch": 1.1389578163771712,
+      "grad_norm": 0.09413156408526462,
+      "learning_rate": 9.327309943879604e-06,
+      "loss": 0.3897,
+      "step": 975
+    },
+    {
+      "epoch": 1.1401255291198364,
+      "grad_norm": 0.09750689827059382,
+      "learning_rate": 9.306957613868292e-06,
+      "loss": 0.379,
+      "step": 976
+    },
+    {
+      "epoch": 1.141293241862502,
+      "grad_norm": 0.091998120845641,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.3868,
+      "step": 977
+    },
+    {
+      "epoch": 1.1424609546051672,
+      "grad_norm": 0.10107905947422102,
+      "learning_rate": 9.266261690983601e-06,
+      "loss": 0.3915,
+      "step": 978
+    },
+    {
+      "epoch": 1.1436286673478324,
+      "grad_norm": 0.09330197866406084,
+      "learning_rate": 9.245918267469528e-06,
+      "loss": 0.3715,
+      "step": 979
+    },
+    {
+      "epoch": 1.1447963800904977,
+      "grad_norm": 0.10301084241167073,
+      "learning_rate": 9.225577982126234e-06,
+      "loss": 0.4081,
+      "step": 980
+    },
+    {
+      "epoch": 1.1459640928331631,
+      "grad_norm": 0.10124913178446235,
+      "learning_rate": 9.205240919601417e-06,
+      "loss": 0.4093,
+      "step": 981
+    },
+    {
+      "epoch": 1.1471318055758284,
+      "grad_norm": 0.09348823082815333,
+      "learning_rate": 9.184907164529369e-06,
+      "loss": 0.3662,
+      "step": 982
+    },
+    {
+      "epoch": 1.1482995183184936,
+      "grad_norm": 0.09613718284411132,
+      "learning_rate": 9.164576801530628e-06,
+      "loss": 0.3719,
+      "step": 983
+    },
+    {
+      "epoch": 1.149467231061159,
+      "grad_norm": 0.09413172469704524,
+      "learning_rate": 9.144249915211605e-06,
+      "loss": 0.3837,
+      "step": 984
+    },
+    {
+      "epoch": 1.1506349438038241,
+      "grad_norm": 0.09686785739058323,
+      "learning_rate": 9.123926590164241e-06,
+      "loss": 0.3806,
+      "step": 985
+    },
+    {
+      "epoch": 1.1518026565464896,
+      "grad_norm": 0.0963513340016892,
+      "learning_rate": 9.103606910965666e-06,
+      "loss": 0.3728,
+      "step": 986
+    },
+    {
+      "epoch": 1.1529703692891549,
+      "grad_norm": 0.09561075737126681,
+      "learning_rate": 9.083290962177828e-06,
+      "loss": 0.372,
+      "step": 987
+    },
+    {
+      "epoch": 1.1541380820318201,
+      "grad_norm": 0.08855261789006977,
+      "learning_rate": 9.06297882834716e-06,
+      "loss": 0.3675,
+      "step": 988
+    },
+    {
+      "epoch": 1.1553057947744856,
+      "grad_norm": 0.09600831390579805,
+      "learning_rate": 9.04267059400421e-06,
+      "loss": 0.3832,
+      "step": 989
+    },
+    {
+      "epoch": 1.1564735075171508,
+      "grad_norm": 0.09644444836944494,
+      "learning_rate": 9.022366343663299e-06,
+      "loss": 0.3907,
+      "step": 990
+    },
+    {
+      "epoch": 1.157641220259816,
+      "grad_norm": 0.10036691547130368,
+      "learning_rate": 9.002066161822174e-06,
+      "loss": 0.4005,
+      "step": 991
+    },
+    {
+      "epoch": 1.1588089330024813,
+      "grad_norm": 0.10220868204875526,
+      "learning_rate": 8.981770132961649e-06,
+      "loss": 0.3912,
+      "step": 992
+    },
+    {
+      "epoch": 1.1599766457451466,
+      "grad_norm": 0.09774528914984215,
+      "learning_rate": 8.961478341545248e-06,
+      "loss": 0.3825,
+      "step": 993
+    },
+    {
+      "epoch": 1.161144358487812,
+      "grad_norm": 0.09533234448917559,
+      "learning_rate": 8.94119087201887e-06,
+      "loss": 0.3722,
+      "step": 994
+    },
+    {
+      "epoch": 1.1623120712304773,
+      "grad_norm": 0.10503286753394957,
+      "learning_rate": 8.920907808810423e-06,
+      "loss": 0.41,
+      "step": 995
+    },
+    {
+      "epoch": 1.1634797839731426,
+      "grad_norm": 0.09326235118739942,
+      "learning_rate": 8.900629236329482e-06,
+      "loss": 0.3658,
+      "step": 996
+    },
+    {
+      "epoch": 1.164647496715808,
+      "grad_norm": 0.09723423239030093,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.3871,
+      "step": 997
+    },
+    {
+      "epoch": 1.1658152094584733,
+      "grad_norm": 0.10193028128453424,
+      "learning_rate": 8.860085901094595e-06,
+      "loss": 0.3859,
+      "step": 998
+    },
+    {
+      "epoch": 1.1669829222011385,
+      "grad_norm": 0.10052720830807603,
+      "learning_rate": 8.839821307064948e-06,
+      "loss": 0.3854,
+      "step": 999
+    },
+    {
+      "epoch": 1.1681506349438038,
+      "grad_norm": 0.09674431881287524,
+      "learning_rate": 8.819561541210698e-06,
+      "loss": 0.371,
+      "step": 1000
+    },
+    {
+      "epoch": 1.169318347686469,
+      "grad_norm": 0.09597350463724497,
+      "learning_rate": 8.799306687844459e-06,
+      "loss": 0.3749,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1704860604291345,
+      "grad_norm": 0.09470816267049023,
+      "learning_rate": 8.779056831258402e-06,
+      "loss": 0.3868,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1716537731717998,
+      "grad_norm": 0.09623591852091334,
+      "learning_rate": 8.758812055723912e-06,
+      "loss": 0.3788,
+      "step": 1003
+    },
+    {
+      "epoch": 1.172821485914465,
+      "grad_norm": 0.09214039030227648,
+      "learning_rate": 8.738572445491225e-06,
+      "loss": 0.3703,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1739891986571303,
+      "grad_norm": 0.09893553021440614,
+      "learning_rate": 8.718338084789074e-06,
+      "loss": 0.4031,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1751569113997957,
+      "grad_norm": 0.09797097610872796,
+      "learning_rate": 8.698109057824354e-06,
+      "loss": 0.3964,
+      "step": 1006
+    },
+    {
+      "epoch": 1.176324624142461,
+      "grad_norm": 0.10879486956458007,
+      "learning_rate": 8.67788544878176e-06,
+      "loss": 0.4094,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1774923368851262,
+      "grad_norm": 0.10580464186326233,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.3748,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1786600496277915,
+      "grad_norm": 0.09305162797355758,
+      "learning_rate": 8.637454821088657e-06,
+      "loss": 0.3669,
+      "step": 1009
+    },
+    {
+      "epoch": 1.179827762370457,
+      "grad_norm": 0.09503316667464466,
+      "learning_rate": 8.6172479706934e-06,
+      "loss": 0.3833,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1809954751131222,
+      "grad_norm": 0.10370903961913584,
+      "learning_rate": 8.597046874730074e-06,
+      "loss": 0.4075,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1821631878557874,
+      "grad_norm": 0.10438290056122883,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3951,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1833309005984527,
+      "grad_norm": 0.10042365052380034,
+      "learning_rate": 8.556662282348782e-06,
+      "loss": 0.387,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1844986133411182,
+      "grad_norm": 0.09691900840771812,
+      "learning_rate": 8.53647895399448e-06,
+      "loss": 0.375,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1856663260837834,
+      "grad_norm": 0.0913316520200236,
+      "learning_rate": 8.516301716198769e-06,
+      "loss": 0.3889,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1868340388264487,
+      "grad_norm": 0.09551697762763234,
+      "learning_rate": 8.496130652930818e-06,
+      "loss": 0.3974,
+      "step": 1016
+    },
+    {
+      "epoch": 1.188001751569114,
+      "grad_norm": 0.09774110486845207,
+      "learning_rate": 8.4759658481341e-06,
+      "loss": 0.388,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1891694643117794,
+      "grad_norm": 0.09475102434178309,
+      "learning_rate": 8.455807385726047e-06,
+      "loss": 0.3792,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1903371770544446,
+      "grad_norm": 0.09341002971107742,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.3879,
+      "step": 1019
+    },
+    {
+      "epoch": 1.19150488979711,
+      "grad_norm": 0.0913441240503179,
+      "learning_rate": 8.415509823613332e-06,
+      "loss": 0.3593,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1926726025397751,
+      "grad_norm": 0.08944500740864503,
+      "learning_rate": 8.395370891610162e-06,
+      "loss": 0.3694,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1938403152824406,
+      "grad_norm": 0.09296240588794494,
+      "learning_rate": 8.375238637397942e-06,
+      "loss": 0.3659,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1950080280251059,
+      "grad_norm": 0.0900102014488785,
+      "learning_rate": 8.355113144758642e-06,
+      "loss": 0.3684,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1961757407677711,
+      "grad_norm": 0.09456721628222196,
+      "learning_rate": 8.334994497446091e-06,
+      "loss": 0.3794,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1973434535104364,
+      "grad_norm": 0.10154045039615404,
+      "learning_rate": 8.31488277918563e-06,
+      "loss": 0.3819,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1985111662531018,
+      "grad_norm": 0.10153717027815708,
+      "learning_rate": 8.294778073673762e-06,
+      "loss": 0.403,
+      "step": 1026
+    },
+    {
+      "epoch": 1.199678878995767,
+      "grad_norm": 0.09865604458148998,
+      "learning_rate": 8.274680464577816e-06,
+      "loss": 0.3932,
+      "step": 1027
+    },
+    {
+      "epoch": 1.2008465917384323,
+      "grad_norm": 0.09551702130680097,
+      "learning_rate": 8.25459003553558e-06,
+      "loss": 0.3799,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2020143044810976,
+      "grad_norm": 0.09314319585780066,
+      "learning_rate": 8.234506870154956e-06,
+      "loss": 0.3786,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2031820172237628,
+      "grad_norm": 0.09034937288020167,
+      "learning_rate": 8.214431052013636e-06,
+      "loss": 0.376,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2043497299664283,
+      "grad_norm": 0.0970946938893389,
+      "learning_rate": 8.194362664658715e-06,
+      "loss": 0.3773,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2055174427090936,
+      "grad_norm": 0.09602335261367298,
+      "learning_rate": 8.174301791606384e-06,
+      "loss": 0.3887,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2066851554517588,
+      "grad_norm": 0.09889357298046583,
+      "learning_rate": 8.154248516341547e-06,
+      "loss": 0.3878,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2078528681944243,
+      "grad_norm": 0.09360779584988053,
+      "learning_rate": 8.134202922317496e-06,
+      "loss": 0.392,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2090205809370895,
+      "grad_norm": 0.0958383279813103,
+      "learning_rate": 8.114165092955556e-06,
+      "loss": 0.396,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2101882936797548,
+      "grad_norm": 0.09476862543956316,
+      "learning_rate": 8.094135111644741e-06,
+      "loss": 0.4038,
+      "step": 1036
+    },
+    {
+      "epoch": 1.21135600642242,
+      "grad_norm": 0.10034862769397959,
+      "learning_rate": 8.074113061741397e-06,
+      "loss": 0.3807,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2125237191650853,
+      "grad_norm": 0.0987861424538886,
+      "learning_rate": 8.054099026568874e-06,
+      "loss": 0.3746,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2136914319077508,
+      "grad_norm": 0.10045422667773264,
+      "learning_rate": 8.03409308941716e-06,
+      "loss": 0.4115,
+      "step": 1039
+    },
+    {
+      "epoch": 1.214859144650416,
+      "grad_norm": 0.09646741079149977,
+      "learning_rate": 8.014095333542548e-06,
+      "loss": 0.3868,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2160268573930813,
+      "grad_norm": 0.09818486825455776,
+      "learning_rate": 7.994105842167274e-06,
+      "loss": 0.3961,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2171945701357467,
+      "grad_norm": 0.09829967458270988,
+      "learning_rate": 7.974124698479192e-06,
+      "loss": 0.3813,
+      "step": 1042
+    },
+    {
+      "epoch": 1.218362282878412,
+      "grad_norm": 0.1008783959126934,
+      "learning_rate": 7.954151985631409e-06,
+      "loss": 0.3844,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2195299956210772,
+      "grad_norm": 0.09686370884148608,
+      "learning_rate": 7.934187786741956e-06,
+      "loss": 0.3821,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2206977083637425,
+      "grad_norm": 0.1016772873852054,
+      "learning_rate": 7.914232184893418e-06,
+      "loss": 0.4061,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2218654211064077,
+      "grad_norm": 0.10056189371629586,
+      "learning_rate": 7.894285263132612e-06,
+      "loss": 0.3945,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2230331338490732,
+      "grad_norm": 0.09718578186681956,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.392,
+      "step": 1047
+    },
+    {
+      "epoch": 1.2242008465917384,
+      "grad_norm": 0.10160388615892617,
+      "learning_rate": 7.854417791880508e-06,
+      "loss": 0.3985,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2253685593344037,
+      "grad_norm": 0.0929429914191325,
+      "learning_rate": 7.834497408300839e-06,
+      "loss": 0.3702,
+      "step": 1049
+    },
+    {
+      "epoch": 1.226536272077069,
+      "grad_norm": 0.09230499455415049,
+      "learning_rate": 7.814586036631485e-06,
+      "loss": 0.3825,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2277039848197344,
+      "grad_norm": 0.09222149628394136,
+      "learning_rate": 7.79468375973519e-06,
+      "loss": 0.3636,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2288716975623997,
+      "grad_norm": 0.09090032991281599,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.3782,
+      "step": 1052
+    },
+    {
+      "epoch": 1.230039410305065,
+      "grad_norm": 0.10003120748493209,
+      "learning_rate": 7.75490682152319e-06,
+      "loss": 0.3874,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2312071230477302,
+      "grad_norm": 0.09832016358268297,
+      "learning_rate": 7.735032325742355e-06,
+      "loss": 0.378,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2323748357903956,
+      "grad_norm": 0.09678576461387102,
+      "learning_rate": 7.715167255803643e-06,
+      "loss": 0.3749,
+      "step": 1055
+    },
+    {
+      "epoch": 1.233542548533061,
+      "grad_norm": 0.09745879273422486,
+      "learning_rate": 7.695311694377116e-06,
+      "loss": 0.3912,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2347102612757261,
+      "grad_norm": 0.10355940472076414,
+      "learning_rate": 7.675465724093254e-06,
+      "loss": 0.4034,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2358779740183914,
+      "grad_norm": 0.0987593860736555,
+      "learning_rate": 7.655629427542636e-06,
+      "loss": 0.3805,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2370456867610569,
+      "grad_norm": 0.09764078851913165,
+      "learning_rate": 7.635802887275582e-06,
+      "loss": 0.3944,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2382133995037221,
+      "grad_norm": 0.09910480156295406,
+      "learning_rate": 7.615986185801807e-06,
+      "loss": 0.3818,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2393811122463874,
+      "grad_norm": 0.09355718872447116,
+      "learning_rate": 7.596179405590076e-06,
+      "loss": 0.3686,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2405488249890526,
+      "grad_norm": 0.09766024066026631,
+      "learning_rate": 7.576382629067877e-06,
+      "loss": 0.3837,
+      "step": 1062
+    },
+    {
+      "epoch": 1.241716537731718,
+      "grad_norm": 0.09701720404230303,
+      "learning_rate": 7.556595938621058e-06,
+      "loss": 0.388,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2428842504743833,
+      "grad_norm": 0.10004445085505433,
+      "learning_rate": 7.536819416593504e-06,
+      "loss": 0.3942,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2440519632170486,
+      "grad_norm": 0.0911522304718376,
+      "learning_rate": 7.517053145286766e-06,
+      "loss": 0.3606,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2452196759597138,
+      "grad_norm": 0.09292838455724968,
+      "learning_rate": 7.497297206959746e-06,
+      "loss": 0.3787,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2463873887023793,
+      "grad_norm": 0.08912154216844467,
+      "learning_rate": 7.47755168382835e-06,
+      "loss": 0.3499,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2475551014450446,
+      "grad_norm": 0.09726418996265268,
+      "learning_rate": 7.4578166580651335e-06,
+      "loss": 0.38,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2487228141877098,
+      "grad_norm": 0.09594528413716297,
+      "learning_rate": 7.438092211798959e-06,
+      "loss": 0.3787,
+      "step": 1069
+    },
+    {
+      "epoch": 1.249890526930375,
+      "grad_norm": 0.09661439049784143,
+      "learning_rate": 7.4183784271146796e-06,
+      "loss": 0.395,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2510582396730405,
+      "grad_norm": 0.09504074565984576,
+      "learning_rate": 7.398675386052767e-06,
+      "loss": 0.3711,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2522259524157058,
+      "grad_norm": 0.09343915745833874,
+      "learning_rate": 7.378983170608982e-06,
+      "loss": 0.3544,
+      "step": 1072
+    },
+    {
+      "epoch": 1.253393665158371,
+      "grad_norm": 0.10024926620491191,
+      "learning_rate": 7.359301862734034e-06,
+      "loss": 0.3804,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2545613779010363,
+      "grad_norm": 0.09304437073427886,
+      "learning_rate": 7.33963154433325e-06,
+      "loss": 0.366,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2557290906437015,
+      "grad_norm": 0.09766018922071627,
+      "learning_rate": 7.319972297266215e-06,
+      "loss": 0.3813,
+      "step": 1075
+    },
+    {
+      "epoch": 1.256896803386367,
+      "grad_norm": 0.099453983389873,
+      "learning_rate": 7.3003242033464314e-06,
+      "loss": 0.3772,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2580645161290323,
+      "grad_norm": 0.0956999685980025,
+      "learning_rate": 7.280687344341006e-06,
+      "loss": 0.3836,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2592322288716975,
+      "grad_norm": 0.10089413050056119,
+      "learning_rate": 7.2610618019702776e-06,
+      "loss": 0.3781,
+      "step": 1078
+    },
+    {
+      "epoch": 1.260399941614363,
+      "grad_norm": 0.09315998976482322,
+      "learning_rate": 7.241447657907494e-06,
+      "loss": 0.3669,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2615676543570282,
+      "grad_norm": 0.0908401499644227,
+      "learning_rate": 7.221844993778464e-06,
+      "loss": 0.381,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2627353670996935,
+      "grad_norm": 0.4219897409909474,
+      "learning_rate": 7.202253891161225e-06,
+      "loss": 0.3815,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2639030798423587,
+      "grad_norm": 0.09787806970129576,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3906,
+      "step": 1082
+    },
+    {
+      "epoch": 1.265070792585024,
+      "grad_norm": 0.10033770193363811,
+      "learning_rate": 7.1631066965333705e-06,
+      "loss": 0.3911,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2662385053276894,
+      "grad_norm": 0.09332913365109877,
+      "learning_rate": 7.143550767436894e-06,
+      "loss": 0.3618,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2674062180703547,
+      "grad_norm": 0.10126402190248236,
+      "learning_rate": 7.124006725679828e-06,
+      "loss": 0.3917,
+      "step": 1085
+    },
+    {
+      "epoch": 1.26857393081302,
+      "grad_norm": 0.10114792217471164,
+      "learning_rate": 7.1044746525962446e-06,
+      "loss": 0.3855,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2697416435556854,
+      "grad_norm": 0.09864248540069014,
+      "learning_rate": 7.084954629470417e-06,
+      "loss": 0.3637,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2709093562983507,
+      "grad_norm": 0.09915784665109799,
+      "learning_rate": 7.065446737536455e-06,
+      "loss": 0.3986,
+      "step": 1088
+    },
+    {
+      "epoch": 1.272077069041016,
+      "grad_norm": 0.09952941895022283,
+      "learning_rate": 7.045951057978001e-06,
+      "loss": 0.394,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2732447817836812,
+      "grad_norm": 0.101815592494143,
+      "learning_rate": 7.026467671927863e-06,
+      "loss": 0.3943,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2744124945263464,
+      "grad_norm": 0.11059831928821681,
+      "learning_rate": 7.0069966604676996e-06,
+      "loss": 0.3794,
+      "step": 1091
+    },
+    {
+      "epoch": 1.275580207269012,
+      "grad_norm": 0.09711709490485887,
+      "learning_rate": 6.9875381046276605e-06,
+      "loss": 0.347,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2767479200116771,
+      "grad_norm": 0.10189891906717742,
+      "learning_rate": 6.968092085386063e-06,
+      "loss": 0.3888,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2779156327543424,
+      "grad_norm": 0.10230828560123235,
+      "learning_rate": 6.948658683669056e-06,
+      "loss": 0.4057,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2790833454970079,
+      "grad_norm": 0.10480789418548657,
+      "learning_rate": 6.929237980350283e-06,
+      "loss": 0.3869,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2802510582396731,
+      "grad_norm": 0.09197549935811657,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3677,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2814187709823384,
+      "grad_norm": 0.09416693189093944,
+      "learning_rate": 6.8904349921374065e-06,
+      "loss": 0.3911,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2825864837250036,
+      "grad_norm": 0.09579083407380812,
+      "learning_rate": 6.871052868725012e-06,
+      "loss": 0.3825,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2837541964676689,
+      "grad_norm": 0.09563046613582157,
+      "learning_rate": 6.851683766673587e-06,
+      "loss": 0.3924,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2849219092103343,
+      "grad_norm": 0.09578547776308792,
+      "learning_rate": 6.832327766589177e-06,
+      "loss": 0.3763,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2860896219529996,
+      "grad_norm": 0.09396861118734533,
+      "learning_rate": 6.812984949023307e-06,
+      "loss": 0.3752,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2872573346956648,
+      "grad_norm": 0.09083075202793507,
+      "learning_rate": 6.793655394472645e-06,
+      "loss": 0.3573,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2884250474383303,
+      "grad_norm": 0.09496216984322674,
+      "learning_rate": 6.774339183378663e-06,
+      "loss": 0.3605,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2895927601809956,
+      "grad_norm": 0.09824777185541565,
+      "learning_rate": 6.755036396127297e-06,
+      "loss": 0.3936,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2907604729236608,
+      "grad_norm": 0.09575630068147027,
+      "learning_rate": 6.73574711304862e-06,
+      "loss": 0.3772,
+      "step": 1105
+    },
+    {
+      "epoch": 1.291928185666326,
+      "grad_norm": 0.10403727267645094,
+      "learning_rate": 6.716471414416519e-06,
+      "loss": 0.4161,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2930958984089913,
+      "grad_norm": 0.10581478795971565,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 0.4076,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2942636111516568,
+      "grad_norm": 0.1012665609242961,
+      "learning_rate": 6.6779610913045344e-06,
+      "loss": 0.3805,
+      "step": 1108
+    },
+    {
+      "epoch": 1.295431323894322,
+      "grad_norm": 0.0988010921669972,
+      "learning_rate": 6.6587266270884095e-06,
+      "loss": 0.3824,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2965990366369873,
+      "grad_norm": 0.09584050609525407,
+      "learning_rate": 6.639506067845698e-06,
+      "loss": 0.3796,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2977667493796525,
+      "grad_norm": 0.12122719273776812,
+      "learning_rate": 6.620299493564283e-06,
+      "loss": 0.3743,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2989344621223178,
+      "grad_norm": 0.09890111433780889,
+      "learning_rate": 6.601106984173835e-06,
+      "loss": 0.3823,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3001021748649833,
+      "grad_norm": 0.09185298461788174,
+      "learning_rate": 6.581928619545501e-06,
+      "loss": 0.3368,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3012698876076485,
+      "grad_norm": 0.09389716637345087,
+      "learning_rate": 6.562764479491566e-06,
+      "loss": 0.3761,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3024376003503138,
+      "grad_norm": 0.09367069524884113,
+      "learning_rate": 6.543614643765116e-06,
+      "loss": 0.385,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3036053130929792,
+      "grad_norm": 0.09663614819737174,
+      "learning_rate": 6.524479192059699e-06,
+      "loss": 0.3495,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3047730258356445,
+      "grad_norm": 0.0914227687708427,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.3532,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3059407385783097,
+      "grad_norm": 0.09926701796739341,
+      "learning_rate": 6.486251759186573e-06,
+      "loss": 0.3755,
+      "step": 1118
+    },
+    {
+      "epoch": 1.307108451320975,
+      "grad_norm": 0.0948341427532342,
+      "learning_rate": 6.467159937105352e-06,
+      "loss": 0.3908,
+      "step": 1119
+    },
+    {
+      "epoch": 1.3082761640636402,
+      "grad_norm": 0.10046243163083776,
+      "learning_rate": 6.4480828172174714e-06,
+      "loss": 0.395,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3094438768063057,
+      "grad_norm": 0.09753122021122666,
+      "learning_rate": 6.429020478913885e-06,
+      "loss": 0.375,
+      "step": 1121
+    },
+    {
+      "epoch": 1.310611589548971,
+      "grad_norm": 0.10244979465239348,
+      "learning_rate": 6.409973001524012e-06,
+      "loss": 0.3872,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3117793022916362,
+      "grad_norm": 0.09497406975538417,
+      "learning_rate": 6.390940464315448e-06,
+      "loss": 0.3659,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3129470150343017,
+      "grad_norm": 0.09510573215158989,
+      "learning_rate": 6.3719229464935915e-06,
+      "loss": 0.3819,
+      "step": 1124
+    },
+    {
+      "epoch": 1.314114727776967,
+      "grad_norm": 0.09586249970012513,
+      "learning_rate": 6.352920527201349e-06,
+      "loss": 0.3957,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3152824405196322,
+      "grad_norm": 0.09118474043242411,
+      "learning_rate": 6.333933285518797e-06,
+      "loss": 0.3703,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3164501532622974,
+      "grad_norm": 0.09443604283338651,
+      "learning_rate": 6.314961300462841e-06,
+      "loss": 0.3835,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3176178660049627,
+      "grad_norm": 0.16165317617542796,
+      "learning_rate": 6.29600465098689e-06,
+      "loss": 0.3752,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3187855787476281,
+      "grad_norm": 0.08786271798778474,
+      "learning_rate": 6.277063415980549e-06,
+      "loss": 0.3589,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3199532914902934,
+      "grad_norm": 0.09450896243758541,
+      "learning_rate": 6.258137674269262e-06,
+      "loss": 0.3839,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3211210042329586,
+      "grad_norm": 0.0996212193070755,
+      "learning_rate": 6.239227504614004e-06,
+      "loss": 0.3937,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3222887169756241,
+      "grad_norm": 0.09958265909374302,
+      "learning_rate": 6.220332985710936e-06,
+      "loss": 0.3775,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3234564297182894,
+      "grad_norm": 0.09766500036039029,
+      "learning_rate": 6.201454196191095e-06,
+      "loss": 0.3818,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3246241424609546,
+      "grad_norm": 0.09516149330923178,
+      "learning_rate": 6.182591214620057e-06,
+      "loss": 0.386,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3257918552036199,
+      "grad_norm": 0.09479693712304701,
+      "learning_rate": 6.1637441194976145e-06,
+      "loss": 0.3656,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3269595679462851,
+      "grad_norm": 0.10353354900343523,
+      "learning_rate": 6.144912989257441e-06,
+      "loss": 0.3902,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3281272806889506,
+      "grad_norm": 0.10891814819445178,
+      "learning_rate": 6.126097902266772e-06,
+      "loss": 0.3824,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3292949934316158,
+      "grad_norm": 0.10376967635096404,
+      "learning_rate": 6.107298936826086e-06,
+      "loss": 0.4085,
+      "step": 1138
+    },
+    {
+      "epoch": 1.330462706174281,
+      "grad_norm": 0.09840322610904288,
+      "learning_rate": 6.088516171168763e-06,
+      "loss": 0.3872,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3316304189169466,
+      "grad_norm": 0.0957982343231223,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.3606,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3327981316596118,
+      "grad_norm": 0.09153528552098161,
+      "learning_rate": 6.050999551800318e-06,
+      "loss": 0.3505,
+      "step": 1141
+    },
+    {
+      "epoch": 1.333965844402277,
+      "grad_norm": 0.09311171028186024,
+      "learning_rate": 6.032265854217574e-06,
+      "loss": 0.3744,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3351335571449423,
+      "grad_norm": 0.09537723781937188,
+      "learning_rate": 6.013548668674305e-06,
+      "loss": 0.3677,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3363012698876076,
+      "grad_norm": 0.0957494126833448,
+      "learning_rate": 5.994848073063552e-06,
+      "loss": 0.3901,
+      "step": 1144
+    },
+    {
+      "epoch": 1.337468982630273,
+      "grad_norm": 0.09183061439106642,
+      "learning_rate": 5.9761641452093225e-06,
+      "loss": 0.366,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3386366953729383,
+      "grad_norm": 0.09561001624829013,
+      "learning_rate": 5.957496962866263e-06,
+      "loss": 0.3862,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3398044081156035,
+      "grad_norm": 0.09080630935888652,
+      "learning_rate": 5.938846603719333e-06,
+      "loss": 0.3556,
+      "step": 1147
+    },
+    {
+      "epoch": 1.340972120858269,
+      "grad_norm": 0.09171856251234368,
+      "learning_rate": 5.9202131453834664e-06,
+      "loss": 0.3567,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3421398336009343,
+      "grad_norm": 0.0972213220298501,
+      "learning_rate": 5.901596665403284e-06,
+      "loss": 0.3915,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3433075463435995,
+      "grad_norm": 0.09377676022797841,
+      "learning_rate": 5.8829972412527325e-06,
+      "loss": 0.3815,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3444752590862648,
+      "grad_norm": 0.09106393348567257,
+      "learning_rate": 5.864414950334796e-06,
+      "loss": 0.3557,
+      "step": 1151
+    },
+    {
+      "epoch": 1.34564297182893,
+      "grad_norm": 0.09151377434725003,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3655,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3468106845715955,
+      "grad_norm": 0.09535808454573214,
+      "learning_rate": 5.827302077451813e-06,
+      "loss": 0.403,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3479783973142607,
+      "grad_norm": 0.09471568990027776,
+      "learning_rate": 5.808771649934924e-06,
+      "loss": 0.4031,
+      "step": 1154
+    },
+    {
+      "epoch": 1.349146110056926,
+      "grad_norm": 0.09528081092112367,
+      "learning_rate": 5.790258664546318e-06,
+      "loss": 0.3746,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3503138227995912,
+      "grad_norm": 0.09376067279567021,
+      "learning_rate": 5.7717631983292375e-06,
+      "loss": 0.3736,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3514815355422565,
+      "grad_norm": 0.09795478040347316,
+      "learning_rate": 5.753285328254039e-06,
+      "loss": 0.3821,
+      "step": 1157
+    },
+    {
+      "epoch": 1.352649248284922,
+      "grad_norm": 0.09458967451304481,
+      "learning_rate": 5.73482513121783e-06,
+      "loss": 0.3552,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3538169610275872,
+      "grad_norm": 0.09698429235908317,
+      "learning_rate": 5.716382684044191e-06,
+      "loss": 0.3879,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3549846737702524,
+      "grad_norm": 0.0949677072025779,
+      "learning_rate": 5.6979580634828125e-06,
+      "loss": 0.3705,
+      "step": 1160
+    },
+    {
+      "epoch": 1.356152386512918,
+      "grad_norm": 0.09517023449918667,
+      "learning_rate": 5.679551346209218e-06,
+      "loss": 0.3568,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3573200992555832,
+      "grad_norm": 0.09213503948003247,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.373,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3584878119982484,
+      "grad_norm": 0.09301045555951051,
+      "learning_rate": 5.642791927854599e-06,
+      "loss": 0.3759,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3596555247409137,
+      "grad_norm": 0.09868214230426851,
+      "learning_rate": 5.6244393797507944e-06,
+      "loss": 0.3845,
+      "step": 1164
+    },
+    {
+      "epoch": 1.360823237483579,
+      "grad_norm": 0.09957405307784416,
+      "learning_rate": 5.60610504088859e-06,
+      "loss": 0.3702,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3619909502262444,
+      "grad_norm": 0.09986366616491628,
+      "learning_rate": 5.587788987567785e-06,
+      "loss": 0.3977,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3631586629689096,
+      "grad_norm": 0.0960032378976153,
+      "learning_rate": 5.569491296012095e-06,
+      "loss": 0.3807,
+      "step": 1167
+    },
+    {
+      "epoch": 1.364326375711575,
+      "grad_norm": 0.09327882124647721,
+      "learning_rate": 5.551212042368792e-06,
+      "loss": 0.3811,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3654940884542404,
+      "grad_norm": 0.09459567615598503,
+      "learning_rate": 5.532951302708447e-06,
+      "loss": 0.3671,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3666618011969056,
+      "grad_norm": 0.09789489350455229,
+      "learning_rate": 5.514709153024571e-06,
+      "loss": 0.3773,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3678295139395709,
+      "grad_norm": 0.09576191296902924,
+      "learning_rate": 5.496485669233315e-06,
+      "loss": 0.3747,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3689972266822361,
+      "grad_norm": 0.10308442234697685,
+      "learning_rate": 5.478280927173145e-06,
+      "loss": 0.3776,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3701649394249014,
+      "grad_norm": 0.09762881865012467,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.3649,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3713326521675668,
+      "grad_norm": 0.1420055770949279,
+      "learning_rate": 5.441927971209644e-06,
+      "loss": 0.3777,
+      "step": 1174
+    },
+    {
+      "epoch": 1.372500364910232,
+      "grad_norm": 0.09596718812842064,
+      "learning_rate": 5.423779908592022e-06,
+      "loss": 0.3577,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3736680776528973,
+      "grad_norm": 0.09526473373276642,
+      "learning_rate": 5.405650890276255e-06,
+      "loss": 0.3873,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3748357903955628,
+      "grad_norm": 0.09490482662090931,
+      "learning_rate": 5.387540991707697e-06,
+      "loss": 0.353,
+      "step": 1177
+    },
+    {
+      "epoch": 1.376003503138228,
+      "grad_norm": 0.09222876695143915,
+      "learning_rate": 5.369450288252113e-06,
+      "loss": 0.3573,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3771712158808933,
+      "grad_norm": 0.0954912601046934,
+      "learning_rate": 5.351378855195405e-06,
+      "loss": 0.3679,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3783389286235586,
+      "grad_norm": 0.09767798237069118,
+      "learning_rate": 5.333326767743263e-06,
+      "loss": 0.3667,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3795066413662238,
+      "grad_norm": 0.09280058390441205,
+      "learning_rate": 5.315294101020882e-06,
+      "loss": 0.3576,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3806743541088893,
+      "grad_norm": 0.09725865889793178,
+      "learning_rate": 5.297280930072632e-06,
+      "loss": 0.3707,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3818420668515545,
+      "grad_norm": 0.09874361077699359,
+      "learning_rate": 5.279287329861749e-06,
+      "loss": 0.3925,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3830097795942198,
+      "grad_norm": 0.09117592761682212,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.3651,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3841774923368853,
+      "grad_norm": 0.09709388299004261,
+      "learning_rate": 5.243359141097469e-06,
+      "loss": 0.3594,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3853452050795505,
+      "grad_norm": 0.10603723174712727,
+      "learning_rate": 5.225424702062082e-06,
+      "loss": 0.4008,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3865129178222158,
+      "grad_norm": 0.09322742493284092,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.3839,
+      "step": 1187
+    },
+    {
+      "epoch": 1.387680630564881,
+      "grad_norm": 0.09602849273762974,
+      "learning_rate": 5.1896155078624225e-06,
+      "loss": 0.3828,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3888483433075463,
+      "grad_norm": 0.10148596432889277,
+      "learning_rate": 5.171740901720943e-06,
+      "loss": 0.3941,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3900160560502117,
+      "grad_norm": 0.09441461388181713,
+      "learning_rate": 5.153886388761586e-06,
+      "loss": 0.3445,
+      "step": 1190
+    },
+    {
+      "epoch": 1.391183768792877,
+      "grad_norm": 0.09205421490862604,
+      "learning_rate": 5.136052043287324e-06,
+      "loss": 0.3554,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3923514815355422,
+      "grad_norm": 0.0903575708854221,
+      "learning_rate": 5.11823793951719e-06,
+      "loss": 0.3484,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3935191942782077,
+      "grad_norm": 0.10267533022890149,
+      "learning_rate": 5.100444151585982e-06,
+      "loss": 0.3766,
+      "step": 1193
+    },
+    {
+      "epoch": 1.394686907020873,
+      "grad_norm": 0.09012781240426788,
+      "learning_rate": 5.082670753543961e-06,
+      "loss": 0.3654,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3958546197635382,
+      "grad_norm": 0.10100467788292229,
+      "learning_rate": 5.064917819356532e-06,
+      "loss": 0.379,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3970223325062034,
+      "grad_norm": 0.09846092212951825,
+      "learning_rate": 5.0471854229039286e-06,
+      "loss": 0.3864,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3981900452488687,
+      "grad_norm": 0.09446048755854573,
+      "learning_rate": 5.029473637980926e-06,
+      "loss": 0.3595,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3993577579915342,
+      "grad_norm": 0.10476142561232651,
+      "learning_rate": 5.011782538296512e-06,
+      "loss": 0.3765,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4005254707341994,
+      "grad_norm": 0.10942477816997212,
+      "learning_rate": 4.994112197473605e-06,
+      "loss": 0.4003,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4016931834768647,
+      "grad_norm": 0.09755326746045137,
+      "learning_rate": 4.976462689048718e-06,
+      "loss": 0.3785,
+      "step": 1200
+    },
+    {
+      "epoch": 1.40286089621953,
+      "grad_norm": 0.09159067111204743,
+      "learning_rate": 4.958834086471683e-06,
+      "loss": 0.372,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4040286089621952,
+      "grad_norm": 0.09910913368472578,
+      "learning_rate": 4.941226463105322e-06,
+      "loss": 0.3734,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4051963217048606,
+      "grad_norm": 0.09777166043553917,
+      "learning_rate": 4.923639892225147e-06,
+      "loss": 0.3972,
+      "step": 1203
+    },
+    {
+      "epoch": 1.406364034447526,
+      "grad_norm": 0.10014709433191654,
+      "learning_rate": 4.906074447019068e-06,
+      "loss": 0.3856,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4075317471901911,
+      "grad_norm": 0.08885606542742455,
+      "learning_rate": 4.888530200587075e-06,
+      "loss": 0.3598,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4086994599328566,
+      "grad_norm": 0.09057055759057761,
+      "learning_rate": 4.87100722594094e-06,
+      "loss": 0.3722,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4098671726755219,
+      "grad_norm": 0.09976678435339434,
+      "learning_rate": 4.8535055960039046e-06,
+      "loss": 0.3839,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4110348854181871,
+      "grad_norm": 0.09492857814234625,
+      "learning_rate": 4.836025383610382e-06,
+      "loss": 0.3487,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4122025981608524,
+      "grad_norm": 0.09925390484479202,
+      "learning_rate": 4.818566661505663e-06,
+      "loss": 0.3843,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4133703109035176,
+      "grad_norm": 0.09356124319644056,
+      "learning_rate": 4.801129502345605e-06,
+      "loss": 0.3708,
+      "step": 1210
+    },
+    {
+      "epoch": 1.414538023646183,
+      "grad_norm": 0.09681153638413802,
+      "learning_rate": 4.783713978696315e-06,
+      "loss": 0.3808,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4157057363888483,
+      "grad_norm": 0.0949264295507538,
+      "learning_rate": 4.766320163033882e-06,
+      "loss": 0.3551,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4168734491315136,
+      "grad_norm": 0.09333278996508942,
+      "learning_rate": 4.748948127744037e-06,
+      "loss": 0.3587,
+      "step": 1213
+    },
+    {
+      "epoch": 1.418041161874179,
+      "grad_norm": 0.09461943667352425,
+      "learning_rate": 4.7315979451218866e-06,
+      "loss": 0.3811,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4192088746168443,
+      "grad_norm": 0.10859464124355514,
+      "learning_rate": 4.714269687371581e-06,
+      "loss": 0.3976,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4203765873595096,
+      "grad_norm": 0.09640895207446186,
+      "learning_rate": 4.696963426606041e-06,
+      "loss": 0.3809,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4215443001021748,
+      "grad_norm": 0.09119264196368795,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.3457,
+      "step": 1217
+    },
+    {
+      "epoch": 1.42271201284484,
+      "grad_norm": 0.09511927414522674,
+      "learning_rate": 4.6624171840229005e-06,
+      "loss": 0.3672,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4238797255875055,
+      "grad_norm": 0.09668218065435247,
+      "learning_rate": 4.645177345972218e-06,
+      "loss": 0.3667,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4250474383301708,
+      "grad_norm": 0.08741579806507122,
+      "learning_rate": 4.6279597924395434e-06,
+      "loss": 0.3508,
+      "step": 1220
+    },
+    {
+      "epoch": 1.426215151072836,
+      "grad_norm": 0.09901027110973522,
+      "learning_rate": 4.61076459507709e-06,
+      "loss": 0.3944,
+      "step": 1221
+    },
+    {
+      "epoch": 1.4273828638155015,
+      "grad_norm": 0.09239257058807426,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.3656,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4285505765581668,
+      "grad_norm": 0.09617360699752654,
+      "learning_rate": 4.576441555006193e-06,
+      "loss": 0.372,
+      "step": 1223
+    },
+    {
+      "epoch": 1.429718289300832,
+      "grad_norm": 0.09851992639439981,
+      "learning_rate": 4.559313855135795e-06,
+      "loss": 0.3743,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4308860020434973,
+      "grad_norm": 0.10233929730183906,
+      "learning_rate": 4.542208797111113e-06,
+      "loss": 0.3642,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4320537147861625,
+      "grad_norm": 0.09499747570201827,
+      "learning_rate": 4.525126452116201e-06,
+      "loss": 0.3824,
+      "step": 1226
+    },
+    {
+      "epoch": 1.433221427528828,
+      "grad_norm": 0.09709059517096928,
+      "learning_rate": 4.508066891240586e-06,
+      "loss": 0.3691,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4343891402714932,
+      "grad_norm": 0.09430312267021133,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.3703,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4355568530141585,
+      "grad_norm": 0.09667982222131408,
+      "learning_rate": 4.474016405730973e-06,
+      "loss": 0.3767,
+      "step": 1229
+    },
+    {
+      "epoch": 1.436724565756824,
+      "grad_norm": 0.0885809584199242,
+      "learning_rate": 4.457025622800771e-06,
+      "loss": 0.34,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4378922784994892,
+      "grad_norm": 0.09307007167385968,
+      "learning_rate": 4.440057907396854e-06,
+      "loss": 0.3671,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4390599912421544,
+      "grad_norm": 0.0962468976604288,
+      "learning_rate": 4.423113330131708e-06,
+      "loss": 0.3798,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4402277039848197,
+      "grad_norm": 0.09737635264712234,
+      "learning_rate": 4.406191961521533e-06,
+      "loss": 0.3781,
+      "step": 1233
+    },
+    {
+      "epoch": 1.441395416727485,
+      "grad_norm": 0.09028824872394711,
+      "learning_rate": 4.389293871985949e-06,
+      "loss": 0.3414,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4425631294701504,
+      "grad_norm": 0.09860774718434533,
+      "learning_rate": 4.372419131847684e-06,
+      "loss": 0.3776,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4437308422128157,
+      "grad_norm": 0.0964425545621925,
+      "learning_rate": 4.355567811332311e-06,
+      "loss": 0.3795,
+      "step": 1236
+    },
+    {
+      "epoch": 1.444898554955481,
+      "grad_norm": 0.09131830565712312,
+      "learning_rate": 4.338739980567926e-06,
+      "loss": 0.3607,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4460662676981464,
+      "grad_norm": 0.0963231444265032,
+      "learning_rate": 4.321935709584884e-06,
+      "loss": 0.3543,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4472339804408116,
+      "grad_norm": 0.09956734962509264,
+      "learning_rate": 4.305155068315481e-06,
+      "loss": 0.3719,
+      "step": 1239
+    },
+    {
+      "epoch": 1.448401693183477,
+      "grad_norm": 0.09681588278053553,
+      "learning_rate": 4.2883981265936884e-06,
+      "loss": 0.3671,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4495694059261421,
+      "grad_norm": 0.09616622893573795,
+      "learning_rate": 4.271664954154845e-06,
+      "loss": 0.3823,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4507371186688074,
+      "grad_norm": 0.09127956210751918,
+      "learning_rate": 4.254955620635371e-06,
+      "loss": 0.3498,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4519048314114729,
+      "grad_norm": 0.09392618579515659,
+      "learning_rate": 4.2382701955724724e-06,
+      "loss": 0.368,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4530725441541381,
+      "grad_norm": 0.09869671328677897,
+      "learning_rate": 4.221608748403872e-06,
+      "loss": 0.3646,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4542402568968034,
+      "grad_norm": 0.0963526374923208,
+      "learning_rate": 4.204971348467498e-06,
+      "loss": 0.3498,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4554079696394686,
+      "grad_norm": 0.09777616722135353,
+      "learning_rate": 4.188358065001215e-06,
+      "loss": 0.3589,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4565756823821339,
+      "grad_norm": 0.10237926936733373,
+      "learning_rate": 4.1717689671425e-06,
+      "loss": 0.3672,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4577433951247993,
+      "grad_norm": 0.09078185547976898,
+      "learning_rate": 4.155204123928205e-06,
+      "loss": 0.361,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4589111078674646,
+      "grad_norm": 0.08851697243373527,
+      "learning_rate": 4.138663604294233e-06,
+      "loss": 0.3599,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4600788206101298,
+      "grad_norm": 0.09661916606123462,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.3752,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4612465333527953,
+      "grad_norm": 0.09577628083405451,
+      "learning_rate": 4.105655811004481e-06,
+      "loss": 0.3946,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4624142460954606,
+      "grad_norm": 0.09494178519461606,
+      "learning_rate": 4.0891886747132356e-06,
+      "loss": 0.3912,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4635819588381258,
+      "grad_norm": 0.09684977121629534,
+      "learning_rate": 4.072746136730828e-06,
+      "loss": 0.3666,
+      "step": 1253
+    },
+    {
+      "epoch": 1.464749671580791,
+      "grad_norm": 0.09393730513534997,
+      "learning_rate": 4.056328265484184e-06,
+      "loss": 0.3833,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4659173843234563,
+      "grad_norm": 0.10461261325211733,
+      "learning_rate": 4.039935129297565e-06,
+      "loss": 0.4008,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4670850970661218,
+      "grad_norm": 0.09352324387635584,
+      "learning_rate": 4.023566796392313e-06,
+      "loss": 0.3644,
+      "step": 1256
+    },
+    {
+      "epoch": 1.468252809808787,
+      "grad_norm": 0.0957166934516256,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.3788,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4694205225514523,
+      "grad_norm": 0.09778712429520477,
+      "learning_rate": 3.990904812794835e-06,
+      "loss": 0.3856,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 0.09544447771809779,
+      "learning_rate": 3.974611298028037e-06,
+      "loss": 0.3776,
+      "step": 1259
+    },
+    {
+      "epoch": 1.471755948036783,
+      "grad_norm": 0.09263582112910754,
+      "learning_rate": 3.958342858392893e-06,
+      "loss": 0.3838,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4729236607794483,
+      "grad_norm": 0.09342472046777786,
+      "learning_rate": 3.942099561591802e-06,
+      "loss": 0.3619,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4740913735221135,
+      "grad_norm": 0.09926577446071275,
+      "learning_rate": 3.925881475222528e-06,
+      "loss": 0.3586,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4752590862647788,
+      "grad_norm": 0.09108641967028386,
+      "learning_rate": 3.909688666777914e-06,
+      "loss": 0.3697,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4764267990074442,
+      "grad_norm": 0.09389805560460875,
+      "learning_rate": 3.893521203645618e-06,
+      "loss": 0.3765,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4775945117501095,
+      "grad_norm": 0.10238978071934751,
+      "learning_rate": 3.877379153107821e-06,
+      "loss": 0.3847,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4787622244927747,
+      "grad_norm": 0.09490876884354875,
+      "learning_rate": 3.8612625823409365e-06,
+      "loss": 0.3741,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4799299372354402,
+      "grad_norm": 0.09907799394707499,
+      "learning_rate": 3.845171558415348e-06,
+      "loss": 0.3821,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4810976499781054,
+      "grad_norm": 0.09804596224162061,
+      "learning_rate": 3.829106148295127e-06,
+      "loss": 0.3778,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4822653627207707,
+      "grad_norm": 0.09738387092582243,
+      "learning_rate": 3.8130664188377464e-06,
+      "loss": 0.3863,
+      "step": 1269
+    },
+    {
+      "epoch": 1.483433075463436,
+      "grad_norm": 0.09998777904349948,
+      "learning_rate": 3.7970524367938146e-06,
+      "loss": 0.3905,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4846007882061012,
+      "grad_norm": 0.08952752635466311,
+      "learning_rate": 3.78106426880678e-06,
+      "loss": 0.365,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4857685009487667,
+      "grad_norm": 0.09288530630907671,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.3667,
+      "step": 1272
+    },
+    {
+      "epoch": 1.486936213691432,
+      "grad_norm": 0.0986090562064486,
+      "learning_rate": 3.749165641039796e-06,
+      "loss": 0.3746,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4881039264340972,
+      "grad_norm": 0.08915303175381803,
+      "learning_rate": 3.733255314008516e-06,
+      "loss": 0.358,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4892716391767626,
+      "grad_norm": 0.09664597430944794,
+      "learning_rate": 3.7173710665309017e-06,
+      "loss": 0.3759,
+      "step": 1275
+    },
+    {
+      "epoch": 1.490439351919428,
+      "grad_norm": 0.093421123190077,
+      "learning_rate": 3.701512964710513e-06,
+      "loss": 0.3478,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4916070646620931,
+      "grad_norm": 0.09629219370383887,
+      "learning_rate": 3.685681074542089e-06,
+      "loss": 0.3774,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4927747774047584,
+      "grad_norm": 0.10051426133947577,
+      "learning_rate": 3.6698754619112974e-06,
+      "loss": 0.3831,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4939424901474236,
+      "grad_norm": 0.09816164932029396,
+      "learning_rate": 3.654096192594442e-06,
+      "loss": 0.3956,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4951102028900891,
+      "grad_norm": 0.09333023677589218,
+      "learning_rate": 3.6383433322582028e-06,
+      "loss": 0.3738,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4962779156327544,
+      "grad_norm": 0.09546289093646682,
+      "learning_rate": 3.62261694645935e-06,
+      "loss": 0.3647,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4974456283754196,
+      "grad_norm": 0.09397662165506086,
+      "learning_rate": 3.6069171006444882e-06,
+      "loss": 0.3663,
+      "step": 1282
+    },
+    {
+      "epoch": 1.498613341118085,
+      "grad_norm": 0.1010305622768237,
+      "learning_rate": 3.591243860149759e-06,
+      "loss": 0.3751,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4997810538607503,
+      "grad_norm": 0.10009361449789315,
+      "learning_rate": 3.575597290200599e-06,
+      "loss": 0.3859,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5009487666034156,
+      "grad_norm": 0.08928429112118803,
+      "learning_rate": 3.5599774559114475e-06,
+      "loss": 0.3506,
+      "step": 1285
+    },
+    {
+      "epoch": 1.5021164793460808,
+      "grad_norm": 0.08890279304017477,
+      "learning_rate": 3.544384422285477e-06,
+      "loss": 0.3467,
+      "step": 1286
+    },
+    {
+      "epoch": 1.503284192088746,
+      "grad_norm": 0.09148823076513532,
+      "learning_rate": 3.528818254214329e-06,
+      "loss": 0.3685,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5044519048314116,
+      "grad_norm": 0.1001530685479951,
+      "learning_rate": 3.513279016477844e-06,
+      "loss": 0.3942,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5056196175740768,
+      "grad_norm": 0.09805646926757054,
+      "learning_rate": 3.4977667737437905e-06,
+      "loss": 0.3675,
+      "step": 1289
+    },
+    {
+      "epoch": 1.506787330316742,
+      "grad_norm": 0.09310816969176963,
+      "learning_rate": 3.4822815905675955e-06,
+      "loss": 0.3749,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5079550430594075,
+      "grad_norm": 0.09222450312240804,
+      "learning_rate": 3.4668235313920707e-06,
+      "loss": 0.3522,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5091227558020726,
+      "grad_norm": 0.09554768488495534,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.3708,
+      "step": 1292
+    },
+    {
+      "epoch": 1.510290468544738,
+      "grad_norm": 0.09068820567736799,
+      "learning_rate": 3.4359890422496245e-06,
+      "loss": 0.3733,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5114581812874033,
+      "grad_norm": 0.09318731115370502,
+      "learning_rate": 3.4206127406028744e-06,
+      "loss": 0.3635,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5126258940300685,
+      "grad_norm": 0.09250426809999598,
+      "learning_rate": 3.4052638195965924e-06,
+      "loss": 0.3502,
+      "step": 1295
+    },
+    {
+      "epoch": 1.513793606772734,
+      "grad_norm": 0.09458586814588292,
+      "learning_rate": 3.3899423431065215e-06,
+      "loss": 0.3766,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5149613195153993,
+      "grad_norm": 0.10472853677459545,
+      "learning_rate": 3.374648374894204e-06,
+      "loss": 0.368,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5161290322580645,
+      "grad_norm": 0.09353612305578374,
+      "learning_rate": 3.3593819786067005e-06,
+      "loss": 0.3662,
+      "step": 1298
+    },
+    {
+      "epoch": 1.51729674500073,
+      "grad_norm": 0.08689410022884632,
+      "learning_rate": 3.344143217776319e-06,
+      "loss": 0.3528,
+      "step": 1299
+    },
+    {
+      "epoch": 1.518464457743395,
+      "grad_norm": 0.09060599989607804,
+      "learning_rate": 3.3289321558203767e-06,
+      "loss": 0.3458,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5196321704860605,
+      "grad_norm": 0.1004099780012222,
+      "learning_rate": 3.3137488560409027e-06,
+      "loss": 0.3745,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5207998832287257,
+      "grad_norm": 0.09824981205211233,
+      "learning_rate": 3.2985933816244065e-06,
+      "loss": 0.3662,
+      "step": 1302
+    },
+    {
+      "epoch": 1.521967595971391,
+      "grad_norm": 0.09705573359707163,
+      "learning_rate": 3.283465795641588e-06,
+      "loss": 0.3752,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5231353087140564,
+      "grad_norm": 0.09452562719468359,
+      "learning_rate": 3.268366161047096e-06,
+      "loss": 0.362,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5243030214567217,
+      "grad_norm": 0.09517746340849446,
+      "learning_rate": 3.2532945406792573e-06,
+      "loss": 0.3635,
+      "step": 1305
+    },
+    {
+      "epoch": 1.525470734199387,
+      "grad_norm": 0.10038110161406752,
+      "learning_rate": 3.2382509972598087e-06,
+      "loss": 0.3828,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5266384469420524,
+      "grad_norm": 0.10054227808612394,
+      "learning_rate": 3.223235593393642e-06,
+      "loss": 0.3783,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5278061596847174,
+      "grad_norm": 0.10066875117998414,
+      "learning_rate": 3.2082483915685526e-06,
+      "loss": 0.3724,
+      "step": 1308
+    },
+    {
+      "epoch": 1.528973872427383,
+      "grad_norm": 0.10012959775126513,
+      "learning_rate": 3.1932894541549643e-06,
+      "loss": 0.3904,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5301415851700482,
+      "grad_norm": 0.09804841466425596,
+      "learning_rate": 3.1783588434056845e-06,
+      "loss": 0.3843,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5313092979127134,
+      "grad_norm": 0.09671559804736235,
+      "learning_rate": 3.163456621455625e-06,
+      "loss": 0.3469,
+      "step": 1311
+    },
+    {
+      "epoch": 1.532477010655379,
+      "grad_norm": 0.08839353374066848,
+      "learning_rate": 3.1485828503215588e-06,
+      "loss": 0.3636,
+      "step": 1312
+    },
+    {
+      "epoch": 1.533644723398044,
+      "grad_norm": 0.09234948921323062,
+      "learning_rate": 3.133737591901864e-06,
+      "loss": 0.3638,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5348124361407094,
+      "grad_norm": 0.09657183068072787,
+      "learning_rate": 3.1189209079762606e-06,
+      "loss": 0.3728,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5359801488833746,
+      "grad_norm": 0.09819446930387776,
+      "learning_rate": 3.104132860205549e-06,
+      "loss": 0.3781,
+      "step": 1315
+    },
+    {
+      "epoch": 1.53714786162604,
+      "grad_norm": 0.0964164162445765,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.3669,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5383155743687054,
+      "grad_norm": 0.09735831397199647,
+      "learning_rate": 3.074642919175883e-06,
+      "loss": 0.3953,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5394832871113706,
+      "grad_norm": 0.09511927553375048,
+      "learning_rate": 3.0599411486416587e-06,
+      "loss": 0.3702,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5406509998540359,
+      "grad_norm": 0.0997221143954931,
+      "learning_rate": 3.045268259711256e-06,
+      "loss": 0.3858,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5418187125967013,
+      "grad_norm": 0.09269286193367425,
+      "learning_rate": 3.0306243134470668e-06,
+      "loss": 0.3616,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5429864253393664,
+      "grad_norm": 0.09405518756834742,
+      "learning_rate": 3.0160093707910275e-06,
+      "loss": 0.3639,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5441541380820318,
+      "grad_norm": 0.09539202364246008,
+      "learning_rate": 3.001423492564384e-06,
+      "loss": 0.3748,
+      "step": 1322
+    },
+    {
+      "epoch": 1.545321850824697,
+      "grad_norm": 0.09349464430793186,
+      "learning_rate": 2.9868667394674143e-06,
+      "loss": 0.3838,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5464895635673623,
+      "grad_norm": 0.09149700061770734,
+      "learning_rate": 2.972339172079204e-06,
+      "loss": 0.3644,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5476572763100278,
+      "grad_norm": 0.09041844006647969,
+      "learning_rate": 2.957840850857373e-06,
+      "loss": 0.3515,
+      "step": 1325
+    },
+    {
+      "epoch": 1.548824989052693,
+      "grad_norm": 0.1020257853941481,
+      "learning_rate": 2.9433718361378326e-06,
+      "loss": 0.3871,
+      "step": 1326
+    },
+    {
+      "epoch": 1.5499927017953583,
+      "grad_norm": 0.09272480401552746,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.3345,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5511604145380238,
+      "grad_norm": 0.09766630956158603,
+      "learning_rate": 2.9145219669391944e-06,
+      "loss": 0.3801,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5523281272806888,
+      "grad_norm": 0.0954329024212816,
+      "learning_rate": 2.9001412325211163e-06,
+      "loss": 0.3575,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5534958400233543,
+      "grad_norm": 0.09341947771791971,
+      "learning_rate": 2.885790044726853e-06,
+      "loss": 0.3734,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5546635527660195,
+      "grad_norm": 0.09524304633215917,
+      "learning_rate": 2.871468463280003e-06,
+      "loss": 0.3604,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5558312655086848,
+      "grad_norm": 0.09753242087268299,
+      "learning_rate": 2.8571765477809645e-06,
+      "loss": 0.3676,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5569989782513503,
+      "grad_norm": 0.09735215699511077,
+      "learning_rate": 2.8429143577066764e-06,
+      "loss": 0.3769,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5581666909940155,
+      "grad_norm": 0.09418432881808961,
+      "learning_rate": 2.8286819524103657e-06,
+      "loss": 0.3725,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5593344037366808,
+      "grad_norm": 0.09452504273879686,
+      "learning_rate": 2.8144793911213174e-06,
+      "loss": 0.3737,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5605021164793462,
+      "grad_norm": 0.09585997772859713,
+      "learning_rate": 2.80030673294461e-06,
+      "loss": 0.3822,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5616698292220113,
+      "grad_norm": 0.09654031127811696,
+      "learning_rate": 2.7861640368608846e-06,
+      "loss": 0.3683,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5628375419646767,
+      "grad_norm": 0.09567793443641003,
+      "learning_rate": 2.7720513617260857e-06,
+      "loss": 0.372,
+      "step": 1338
+    },
+    {
+      "epoch": 1.564005254707342,
+      "grad_norm": 0.09813172338886775,
+      "learning_rate": 2.7579687662712285e-06,
+      "loss": 0.3736,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5651729674500072,
+      "grad_norm": 0.08995815200021313,
+      "learning_rate": 2.7439163091021525e-06,
+      "loss": 0.3802,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5663406801926727,
+      "grad_norm": 0.09059540930653838,
+      "learning_rate": 2.7298940486992654e-06,
+      "loss": 0.3477,
+      "step": 1341
+    },
+    {
+      "epoch": 1.567508392935338,
+      "grad_norm": 0.09807881626735997,
+      "learning_rate": 2.71590204341731e-06,
+      "loss": 0.3782,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5686761056780032,
+      "grad_norm": 0.0971582185570604,
+      "learning_rate": 2.7019403514851274e-06,
+      "loss": 0.3681,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5698438184206687,
+      "grad_norm": 0.09318010015381444,
+      "learning_rate": 2.688009031005403e-06,
+      "loss": 0.3619,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5710115311633337,
+      "grad_norm": 0.09629050608890191,
+      "learning_rate": 2.674108139954429e-06,
+      "loss": 0.3839,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5721792439059992,
+      "grad_norm": 0.09093706253985726,
+      "learning_rate": 2.6602377361818576e-06,
+      "loss": 0.377,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5733469566486644,
+      "grad_norm": 0.09770834767045823,
+      "learning_rate": 2.646397877410476e-06,
+      "loss": 0.3552,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5745146693913297,
+      "grad_norm": 0.09589223795480563,
+      "learning_rate": 2.6325886212359496e-06,
+      "loss": 0.3698,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5756823821339951,
+      "grad_norm": 0.09388353084976483,
+      "learning_rate": 2.6188100251265947e-06,
+      "loss": 0.3536,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5768500948766604,
+      "grad_norm": 0.09417216803225362,
+      "learning_rate": 2.6050621464231243e-06,
+      "loss": 0.3525,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5780178076193256,
+      "grad_norm": 0.0977362053942121,
+      "learning_rate": 2.591345042338421e-06,
+      "loss": 0.3882,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5791855203619911,
+      "grad_norm": 0.09341582848319586,
+      "learning_rate": 2.5776587699573007e-06,
+      "loss": 0.3637,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5803532331046561,
+      "grad_norm": 0.09838040693389137,
+      "learning_rate": 2.564003386236271e-06,
+      "loss": 0.3741,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5815209458473216,
+      "grad_norm": 0.10200468301501524,
+      "learning_rate": 2.550378948003287e-06,
+      "loss": 0.3687,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5826886585899869,
+      "grad_norm": 0.0972299305377385,
+      "learning_rate": 2.5367855119575314e-06,
+      "loss": 0.3856,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5838563713326521,
+      "grad_norm": 0.09859777898435446,
+      "learning_rate": 2.523223134669157e-06,
+      "loss": 0.3741,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5850240840753176,
+      "grad_norm": 0.09590712049242321,
+      "learning_rate": 2.5096918725790754e-06,
+      "loss": 0.3609,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5861917968179826,
+      "grad_norm": 0.09250290442669705,
+      "learning_rate": 2.496191781998698e-06,
+      "loss": 0.3592,
+      "step": 1358
+    },
+    {
+      "epoch": 1.587359509560648,
+      "grad_norm": 0.09181435539204466,
+      "learning_rate": 2.4827229191097236e-06,
+      "loss": 0.3359,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5885272223033133,
+      "grad_norm": 0.0991608784057147,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.3745,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5896949350459786,
+      "grad_norm": 0.09576530825252226,
+      "learning_rate": 2.4558791004827487e-06,
+      "loss": 0.3587,
+      "step": 1361
+    },
+    {
+      "epoch": 1.590862647788644,
+      "grad_norm": 0.09566035552298277,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.3751,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5920303605313093,
+      "grad_norm": 0.0962845409004749,
+      "learning_rate": 2.4291608635483733e-06,
+      "loss": 0.3808,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5931980732739746,
+      "grad_norm": 0.09473584128044839,
+      "learning_rate": 2.4158489772852035e-06,
+      "loss": 0.3636,
+      "step": 1364
+    },
+    {
+      "epoch": 1.59436578601664,
+      "grad_norm": 0.0923952716241333,
+      "learning_rate": 2.4025686530663716e-06,
+      "loss": 0.3423,
+      "step": 1365
+    },
+    {
+      "epoch": 1.595533498759305,
+      "grad_norm": 0.09633916063512732,
+      "learning_rate": 2.3893199461589943e-06,
+      "loss": 0.3761,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5967012115019705,
+      "grad_norm": 0.09139098855698184,
+      "learning_rate": 2.376102911698618e-06,
+      "loss": 0.3536,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5978689242446358,
+      "grad_norm": 0.09140658949031343,
+      "learning_rate": 2.3629176046889755e-06,
+      "loss": 0.3285,
+      "step": 1368
+    },
+    {
+      "epoch": 1.599036636987301,
+      "grad_norm": 0.09103408172579729,
+      "learning_rate": 2.3497640800017687e-06,
+      "loss": 0.3477,
+      "step": 1369
+    },
+    {
+      "epoch": 1.6002043497299665,
+      "grad_norm": 0.0976374374739043,
+      "learning_rate": 2.3366423923764268e-06,
+      "loss": 0.3906,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6013720624726318,
+      "grad_norm": 0.09680722226746691,
+      "learning_rate": 2.323552596419889e-06,
+      "loss": 0.3799,
+      "step": 1371
+    },
+    {
+      "epoch": 1.602539775215297,
+      "grad_norm": 0.09226341639005319,
+      "learning_rate": 2.3104947466063785e-06,
+      "loss": 0.368,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6037074879579625,
+      "grad_norm": 0.09903428571788879,
+      "learning_rate": 2.2974688972771743e-06,
+      "loss": 0.3777,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6048752007006275,
+      "grad_norm": 0.09463053206810111,
+      "learning_rate": 2.284475102640371e-06,
+      "loss": 0.3722,
+      "step": 1374
+    },
+    {
+      "epoch": 1.606042913443293,
+      "grad_norm": 0.09538385570507717,
+      "learning_rate": 2.27151341677068e-06,
+      "loss": 0.3753,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6072106261859582,
+      "grad_norm": 0.09310044715718335,
+      "learning_rate": 2.2585838936091753e-06,
+      "loss": 0.3501,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6083783389286235,
+      "grad_norm": 0.09512854839578291,
+      "learning_rate": 2.2456865869630974e-06,
+      "loss": 0.3679,
+      "step": 1377
+    },
+    {
+      "epoch": 1.609546051671289,
+      "grad_norm": 0.09367195666781424,
+      "learning_rate": 2.2328215505056004e-06,
+      "loss": 0.3606,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6107137644139542,
+      "grad_norm": 0.09414307291251012,
+      "learning_rate": 2.2199888377755573e-06,
+      "loss": 0.3571,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6118814771566194,
+      "grad_norm": 0.10168473138219913,
+      "learning_rate": 2.207188502177313e-06,
+      "loss": 0.3826,
+      "step": 1380
+    },
+    {
+      "epoch": 1.613049189899285,
+      "grad_norm": 0.09365484090495793,
+      "learning_rate": 2.194420596980481e-06,
+      "loss": 0.3523,
+      "step": 1381
+    },
+    {
+      "epoch": 1.61421690264195,
+      "grad_norm": 0.09335878858198102,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.3592,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.0885808019163929,
+      "learning_rate": 2.1689822901944456e-06,
+      "loss": 0.3379,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6165523281272807,
+      "grad_norm": 0.09648311842162939,
+      "learning_rate": 2.156311994468774e-06,
+      "loss": 0.3543,
+      "step": 1384
+    },
+    {
+      "epoch": 1.617720040869946,
+      "grad_norm": 0.09425566139181153,
+      "learning_rate": 2.1436743408711234e-06,
+      "loss": 0.368,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6188877536126114,
+      "grad_norm": 0.10010978227954366,
+      "learning_rate": 2.1310693819940842e-06,
+      "loss": 0.3851,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6200554663552766,
+      "grad_norm": 0.10372881965324054,
+      "learning_rate": 2.1184971702941947e-06,
+      "loss": 0.3791,
+      "step": 1387
+    },
+    {
+      "epoch": 1.621223179097942,
+      "grad_norm": 0.09774118735927279,
+      "learning_rate": 2.1059577580917067e-06,
+      "loss": 0.356,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6223908918406074,
+      "grad_norm": 0.09955144056625372,
+      "learning_rate": 2.0934511975703795e-06,
+      "loss": 0.3626,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6235586045832724,
+      "grad_norm": 0.10215804324902913,
+      "learning_rate": 2.0809775407772505e-06,
+      "loss": 0.396,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6247263173259379,
+      "grad_norm": 0.9017191990220045,
+      "learning_rate": 2.068536839622427e-06,
+      "loss": 0.4181,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6258940300686031,
+      "grad_norm": 0.10225171762110219,
+      "learning_rate": 2.0561291458788736e-06,
+      "loss": 0.3982,
+      "step": 1392
+    },
+    {
+      "epoch": 1.6270617428112684,
+      "grad_norm": 0.10029479128283883,
+      "learning_rate": 2.043754511182191e-06,
+      "loss": 0.3621,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6282294555539338,
+      "grad_norm": 0.09287952095812832,
+      "learning_rate": 2.031412987030398e-06,
+      "loss": 0.3619,
+      "step": 1394
+    },
+    {
+      "epoch": 1.629397168296599,
+      "grad_norm": 0.09527830129636024,
+      "learning_rate": 2.01910462478372e-06,
+      "loss": 0.3734,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6305648810392643,
+      "grad_norm": 0.09723439113249231,
+      "learning_rate": 2.0068294756643846e-06,
+      "loss": 0.3636,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6317325937819298,
+      "grad_norm": 0.09852464670530955,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.3876,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6329003065245948,
+      "grad_norm": 0.09976308409305476,
+      "learning_rate": 1.982379021005325e-06,
+      "loss": 0.3816,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6340680192672603,
+      "grad_norm": 0.09310125560302326,
+      "learning_rate": 1.9702038172181016e-06,
+      "loss": 0.3632,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6352357320099256,
+      "grad_norm": 0.09354550462383626,
+      "learning_rate": 1.958062030062795e-06,
+      "loss": 0.3661,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6364034447525908,
+      "grad_norm": 0.09772678612840353,
+      "learning_rate": 1.94595371006842e-06,
+      "loss": 0.3766,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6375711574952563,
+      "grad_norm": 0.09510581505190932,
+      "learning_rate": 1.9338789076247e-06,
+      "loss": 0.3621,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6387388702379213,
+      "grad_norm": 0.09679953205896041,
+      "learning_rate": 1.9218376729818842e-06,
+      "loss": 0.366,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6399065829805868,
+      "grad_norm": 0.09818657616881495,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.3761,
+      "step": 1404
+    },
+    {
+      "epoch": 1.641074295723252,
+      "grad_norm": 0.1011603025367264,
+      "learning_rate": 1.8978561074012702e-06,
+      "loss": 0.3936,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6422420084659173,
+      "grad_norm": 0.1809039354011843,
+      "learning_rate": 1.8859158762646468e-06,
+      "loss": 0.3771,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6434097212085828,
+      "grad_norm": 0.10007683629032452,
+      "learning_rate": 1.8740094125308773e-06,
+      "loss": 0.3716,
+      "step": 1407
+    },
+    {
+      "epoch": 1.644577433951248,
+      "grad_norm": 0.10263598128871035,
+      "learning_rate": 1.8621367657496504e-06,
+      "loss": 0.4027,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6457451466939133,
+      "grad_norm": 0.09014451264110687,
+      "learning_rate": 1.850297985329932e-06,
+      "loss": 0.3543,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6469128594365787,
+      "grad_norm": 0.09697077990807151,
+      "learning_rate": 1.8384931205397305e-06,
+      "loss": 0.359,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6480805721792438,
+      "grad_norm": 0.09600159229648492,
+      "learning_rate": 1.826722220505931e-06,
+      "loss": 0.3532,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6492482849219092,
+      "grad_norm": 0.09668831266325659,
+      "learning_rate": 1.8149853342140644e-06,
+      "loss": 0.3806,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6504159976645745,
+      "grad_norm": 0.0915111038815517,
+      "learning_rate": 1.8032825105081141e-06,
+      "loss": 0.3604,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6515837104072397,
+      "grad_norm": 0.09502239525097328,
+      "learning_rate": 1.7916137980903047e-06,
+      "loss": 0.3681,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6527514231499052,
+      "grad_norm": 0.0981264012808681,
+      "learning_rate": 1.7799792455209019e-06,
+      "loss": 0.3817,
+      "step": 1415
+    },
+    {
+      "epoch": 1.6539191358925704,
+      "grad_norm": 0.09116347760173368,
+      "learning_rate": 1.7683789012180196e-06,
+      "loss": 0.358,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6550868486352357,
+      "grad_norm": 0.09653957762770053,
+      "learning_rate": 1.7568128134574114e-06,
+      "loss": 0.3675,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6562545613779012,
+      "grad_norm": 0.0935697944614011,
+      "learning_rate": 1.74528103037226e-06,
+      "loss": 0.3444,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6574222741205662,
+      "grad_norm": 0.09789426999023643,
+      "learning_rate": 1.7337835999529995e-06,
+      "loss": 0.3748,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6585899868632317,
+      "grad_norm": 0.10068200850392607,
+      "learning_rate": 1.722320570047089e-06,
+      "loss": 0.3788,
+      "step": 1420
+    },
+    {
+      "epoch": 1.659757699605897,
+      "grad_norm": 0.09739937299053956,
+      "learning_rate": 1.7108919883588404e-06,
+      "loss": 0.3785,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6609254123485622,
+      "grad_norm": 0.09737914616008829,
+      "learning_rate": 1.6994979024491943e-06,
+      "loss": 0.3802,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6620931250912276,
+      "grad_norm": 0.10092633917601493,
+      "learning_rate": 1.6881383597355438e-06,
+      "loss": 0.3594,
+      "step": 1423
+    },
+    {
+      "epoch": 1.663260837833893,
+      "grad_norm": 0.09540950525628525,
+      "learning_rate": 1.6768134074915277e-06,
+      "loss": 0.3747,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6644285505765581,
+      "grad_norm": 0.09547323049776985,
+      "learning_rate": 1.6655230928468257e-06,
+      "loss": 0.3721,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6655962633192236,
+      "grad_norm": 0.09608696051496808,
+      "learning_rate": 1.6542674627869738e-06,
+      "loss": 0.3524,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6667639760618886,
+      "grad_norm": 0.10038539398626577,
+      "learning_rate": 1.6430465641531701e-06,
+      "loss": 0.3636,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6679316888045541,
+      "grad_norm": 0.09330985092841829,
+      "learning_rate": 1.6318604436420738e-06,
+      "loss": 0.3465,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6690994015472194,
+      "grad_norm": 0.0943920491662281,
+      "learning_rate": 1.6207091478056059e-06,
+      "loss": 0.3557,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6702671142898846,
+      "grad_norm": 0.09519395567489186,
+      "learning_rate": 1.609592723050767e-06,
+      "loss": 0.3598,
+      "step": 1430
+    },
+    {
+      "epoch": 1.67143482703255,
+      "grad_norm": 0.09908701690078711,
+      "learning_rate": 1.5985112156394368e-06,
+      "loss": 0.3921,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6726025397752153,
+      "grad_norm": 0.10165701027640732,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.3749,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6737702525178806,
+      "grad_norm": 0.09135023882441025,
+      "learning_rate": 1.5764531371680859e-06,
+      "loss": 0.36,
+      "step": 1433
+    },
+    {
+      "epoch": 1.674937965260546,
+      "grad_norm": 0.0971829064274695,
+      "learning_rate": 1.5654766579045033e-06,
+      "loss": 0.3689,
+      "step": 1434
+    },
+    {
+      "epoch": 1.676105678003211,
+      "grad_norm": 0.10188915112172657,
+      "learning_rate": 1.5545352795769231e-06,
+      "loss": 0.3767,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6772733907458766,
+      "grad_norm": 0.09703800197067587,
+      "learning_rate": 1.5436290477187589e-06,
+      "loss": 0.3624,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6784411034885418,
+      "grad_norm": 0.09718790325775568,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.3796,
+      "step": 1437
+    },
+    {
+      "epoch": 1.679608816231207,
+      "grad_norm": 0.09675233398812697,
+      "learning_rate": 1.5219222048128125e-06,
+      "loss": 0.3692,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6807765289738725,
+      "grad_norm": 0.09134195127249442,
+      "learning_rate": 1.5111216840997745e-06,
+      "loss": 0.3498,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6819442417165378,
+      "grad_norm": 0.09912674632972783,
+      "learning_rate": 1.500356490525261e-06,
+      "loss": 0.3673,
+      "step": 1440
+    },
+    {
+      "epoch": 1.683111954459203,
+      "grad_norm": 0.08929513254925396,
+      "learning_rate": 1.4896266688894823e-06,
+      "loss": 0.3462,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6842796672018685,
+      "grad_norm": 0.09782382462905835,
+      "learning_rate": 1.478932263845435e-06,
+      "loss": 0.3804,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6854473799445335,
+      "grad_norm": 0.08947727258166059,
+      "learning_rate": 1.4682733198987353e-06,
+      "loss": 0.3507,
+      "step": 1443
+    },
+    {
+      "epoch": 1.686615092687199,
+      "grad_norm": 0.09465725647023682,
+      "learning_rate": 1.457649881407417e-06,
+      "loss": 0.3761,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6877828054298643,
+      "grad_norm": 0.09628091322979697,
+      "learning_rate": 1.4470619925817654e-06,
+      "loss": 0.381,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6889505181725295,
+      "grad_norm": 0.09135003910864857,
+      "learning_rate": 1.436509697484111e-06,
+      "loss": 0.354,
+      "step": 1446
+    },
+    {
+      "epoch": 1.690118230915195,
+      "grad_norm": 0.09288423286644555,
+      "learning_rate": 1.425993040028667e-06,
+      "loss": 0.3482,
+      "step": 1447
+    },
+    {
+      "epoch": 1.69128594365786,
+      "grad_norm": 0.09687339589361757,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.364,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6924536564005255,
+      "grad_norm": 0.0932105671391047,
+      "learning_rate": 1.4050668129595346e-06,
+      "loss": 0.3651,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6936213691431907,
+      "grad_norm": 0.0956365204504552,
+      "learning_rate": 1.39465733043199e-06,
+      "loss": 0.3681,
+      "step": 1450
+    },
+    {
+      "epoch": 1.694789081885856,
+      "grad_norm": 0.09388204128542417,
+      "learning_rate": 1.3842836597185916e-06,
+      "loss": 0.3662,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6959567946285214,
+      "grad_norm": 0.10581244602515943,
+      "learning_rate": 1.373945843990192e-06,
+      "loss": 0.3811,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6971245073711867,
+      "grad_norm": 0.09791780973990746,
+      "learning_rate": 1.3636439262684299e-06,
+      "loss": 0.377,
+      "step": 1453
+    },
+    {
+      "epoch": 1.698292220113852,
+      "grad_norm": 0.09513106996627814,
+      "learning_rate": 1.3533779494255483e-06,
+      "loss": 0.3531,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6994599328565174,
+      "grad_norm": 0.09600339016892241,
+      "learning_rate": 1.343147956184221e-06,
+      "loss": 0.3719,
+      "step": 1455
+    },
+    {
+      "epoch": 1.7006276455991824,
+      "grad_norm": 0.0958574556421162,
+      "learning_rate": 1.332953989117377e-06,
+      "loss": 0.3642,
+      "step": 1456
+    },
+    {
+      "epoch": 1.701795358341848,
+      "grad_norm": 0.09374023435122916,
+      "learning_rate": 1.322796090648013e-06,
+      "loss": 0.3559,
+      "step": 1457
+    },
+    {
+      "epoch": 1.7029630710845132,
+      "grad_norm": 0.09141408823770272,
+      "learning_rate": 1.3126743030490307e-06,
+      "loss": 0.3571,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7041307838271784,
+      "grad_norm": 0.09385813768007341,
+      "learning_rate": 1.3025886684430467e-06,
+      "loss": 0.3635,
+      "step": 1459
+    },
+    {
+      "epoch": 1.705298496569844,
+      "grad_norm": 0.08974197996117664,
+      "learning_rate": 1.2925392288022299e-06,
+      "loss": 0.3379,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7064662093125091,
+      "grad_norm": 0.09267830571889124,
+      "learning_rate": 1.2825260259481143e-06,
+      "loss": 0.3611,
+      "step": 1461
+    },
+    {
+      "epoch": 1.7076339220551744,
+      "grad_norm": 0.0956294578280517,
+      "learning_rate": 1.272549101551438e-06,
+      "loss": 0.3705,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7088016347978399,
+      "grad_norm": 0.09056395176528022,
+      "learning_rate": 1.2626084971319641e-06,
+      "loss": 0.3525,
+      "step": 1463
+    },
+    {
+      "epoch": 1.709969347540505,
+      "grad_norm": 0.0999486454241795,
+      "learning_rate": 1.2527042540583e-06,
+      "loss": 0.3736,
+      "step": 1464
+    },
+    {
+      "epoch": 1.7111370602831704,
+      "grad_norm": 0.09396141051760422,
+      "learning_rate": 1.2428364135477356e-06,
+      "loss": 0.3791,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7123047730258356,
+      "grad_norm": 0.08817608972829283,
+      "learning_rate": 1.2330050166660712e-06,
+      "loss": 0.3449,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7134724857685009,
+      "grad_norm": 0.09012635106070117,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.3417,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7146401985111663,
+      "grad_norm": 0.09258023598897828,
+      "learning_rate": 1.2134517172941563e-06,
+      "loss": 0.3649,
+      "step": 1468
+    },
+    {
+      "epoch": 1.7158079112538316,
+      "grad_norm": 0.09468016691957576,
+      "learning_rate": 1.2037298961765065e-06,
+      "loss": 0.3706,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7169756239964968,
+      "grad_norm": 0.09043325502684234,
+      "learning_rate": 1.19404468143262e-06,
+      "loss": 0.3661,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7181433367391623,
+      "grad_norm": 0.09410232788842793,
+      "learning_rate": 1.184396113368288e-06,
+      "loss": 0.3667,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7193110494818273,
+      "grad_norm": 0.0916223175575254,
+      "learning_rate": 1.1747842321367886e-06,
+      "loss": 0.3656,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7204787622244928,
+      "grad_norm": 0.09836633688550374,
+      "learning_rate": 1.1652090777387248e-06,
+      "loss": 0.374,
+      "step": 1473
+    },
+    {
+      "epoch": 1.721646474967158,
+      "grad_norm": 0.09432868127628295,
+      "learning_rate": 1.1556706900218572e-06,
+      "loss": 0.3468,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7228141877098233,
+      "grad_norm": 0.09448143675734637,
+      "learning_rate": 1.1461691086809434e-06,
+      "loss": 0.349,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7239819004524888,
+      "grad_norm": 0.09166445897271126,
+      "learning_rate": 1.1367043732575666e-06,
+      "loss": 0.3408,
+      "step": 1476
+    },
+    {
+      "epoch": 1.725149613195154,
+      "grad_norm": 0.10319452698339615,
+      "learning_rate": 1.1272765231399684e-06,
+      "loss": 0.3797,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7263173259378193,
+      "grad_norm": 0.09035204514237555,
+      "learning_rate": 1.1178855975628967e-06,
+      "loss": 0.3453,
+      "step": 1478
+    },
+    {
+      "epoch": 1.7274850386804848,
+      "grad_norm": 0.09158035307848489,
+      "learning_rate": 1.1085316356074272e-06,
+      "loss": 0.3416,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7286527514231498,
+      "grad_norm": 0.09691052151296135,
+      "learning_rate": 1.099214676200816e-06,
+      "loss": 0.3606,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7298204641658153,
+      "grad_norm": 0.09341114717440614,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.3623,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7309881769084805,
+      "grad_norm": 0.10168835283581132,
+      "learning_rate": 1.0806919199730614e-06,
+      "loss": 0.3709,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7321558896511458,
+      "grad_norm": 0.10278221272490709,
+      "learning_rate": 1.0714862002358385e-06,
+      "loss": 0.394,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7333236023938112,
+      "grad_norm": 0.09492701998787048,
+      "learning_rate": 1.0623176372149802e-06,
+      "loss": 0.3736,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7344913151364765,
+      "grad_norm": 0.09305131162976767,
+      "learning_rate": 1.0531862690661865e-06,
+      "loss": 0.3602,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7356590278791417,
+      "grad_norm": 0.08971349892848529,
+      "learning_rate": 1.0440921337903699e-06,
+      "loss": 0.3371,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7368267406218072,
+      "grad_norm": 0.09631349459322053,
+      "learning_rate": 1.035035269233493e-06,
+      "loss": 0.3651,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7379944533644722,
+      "grad_norm": 0.09511691172475335,
+      "learning_rate": 1.0260157130864178e-06,
+      "loss": 0.3436,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7391621661071377,
+      "grad_norm": 0.09727847544481745,
+      "learning_rate": 1.0170335028847345e-06,
+      "loss": 0.3666,
+      "step": 1489
+    },
+    {
+      "epoch": 1.740329878849803,
+      "grad_norm": 0.09933836698100705,
+      "learning_rate": 1.0080886760086227e-06,
+      "loss": 0.3845,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7414975915924682,
+      "grad_norm": 0.09211261856685542,
+      "learning_rate": 9.991812696826887e-07,
+      "loss": 0.3531,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7426653043351337,
+      "grad_norm": 0.0954016137929733,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.3486,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7438330170777987,
+      "grad_norm": 0.09636372635929485,
+      "learning_rate": 9.814788668009766e-07,
+      "loss": 0.3692,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7450007298204642,
+      "grad_norm": 0.09649808097255291,
+      "learning_rate": 9.726839439151447e-07,
+      "loss": 0.3805,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7461684425631294,
+      "grad_norm": 0.09568847131649606,
+      "learning_rate": 9.63926588919083e-07,
+      "loss": 0.3683,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7473361553057947,
+      "grad_norm": 0.6136966025219838,
+      "learning_rate": 9.55206838257219e-07,
+      "loss": 0.462,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7485038680484601,
+      "grad_norm": 0.09569580259734461,
+      "learning_rate": 9.465247282174806e-07,
+      "loss": 0.3691,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7496715807911254,
+      "grad_norm": 0.09092100319726626,
+      "learning_rate": 9.378802949311583e-07,
+      "loss": 0.3452,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7508392935337906,
+      "grad_norm": 0.09243708431727536,
+      "learning_rate": 9.292735743727432e-07,
+      "loss": 0.3595,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7520070062764561,
+      "grad_norm": 0.09561244332326518,
+      "learning_rate": 9.207046023597866e-07,
+      "loss": 0.3714,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7531747190191211,
+      "grad_norm": 0.0930592311437806,
+      "learning_rate": 9.121734145527394e-07,
+      "loss": 0.3669,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7543424317617866,
+      "grad_norm": 0.0981790533150985,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.3655,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7555101445044519,
+      "grad_norm": 0.097578341133992,
+      "learning_rate": 8.952245334118415e-07,
+      "loss": 0.3673,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7566778572471171,
+      "grad_norm": 0.09364479831877834,
+      "learning_rate": 8.868069106121002e-07,
+      "loss": 0.3696,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7578455699897826,
+      "grad_norm": 0.09728651376968718,
+      "learning_rate": 8.784272130861937e-07,
+      "loss": 0.3781,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7590132827324478,
+      "grad_norm": 0.08922434076656491,
+      "learning_rate": 8.700854757068988e-07,
+      "loss": 0.3702,
+      "step": 1506
+    },
+    {
+      "epoch": 1.760180995475113,
+      "grad_norm": 0.09610419561601122,
+      "learning_rate": 8.617817331890155e-07,
+      "loss": 0.3588,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7613487082177786,
+      "grad_norm": 0.09554338075755513,
+      "learning_rate": 8.535160200892234e-07,
+      "loss": 0.3804,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7625164209604436,
+      "grad_norm": 0.09095855067039994,
+      "learning_rate": 8.4528837080594e-07,
+      "loss": 0.3579,
+      "step": 1509
+    },
+    {
+      "epoch": 1.763684133703109,
+      "grad_norm": 0.09478040314719811,
+      "learning_rate": 8.370988195791807e-07,
+      "loss": 0.3691,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7648518464457743,
+      "grad_norm": 0.10005152724846832,
+      "learning_rate": 8.289474004904108e-07,
+      "loss": 0.3605,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7660195591884396,
+      "grad_norm": 0.09950834663566707,
+      "learning_rate": 8.208341474624071e-07,
+      "loss": 0.3719,
+      "step": 1512
+    },
+    {
+      "epoch": 1.767187271931105,
+      "grad_norm": 0.09161075302313332,
+      "learning_rate": 8.127590942591112e-07,
+      "loss": 0.3733,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7683549846737703,
+      "grad_norm": 0.09377857720603688,
+      "learning_rate": 8.047222744854943e-07,
+      "loss": 0.3542,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7695226974164355,
+      "grad_norm": 0.09857451825064223,
+      "learning_rate": 7.96723721587419e-07,
+      "loss": 0.3874,
+      "step": 1515
+    },
+    {
+      "epoch": 1.770690410159101,
+      "grad_norm": 0.09406591857362909,
+      "learning_rate": 7.887634688515e-07,
+      "loss": 0.3564,
+      "step": 1516
+    },
+    {
+      "epoch": 1.771858122901766,
+      "grad_norm": 0.0919815154469294,
+      "learning_rate": 7.808415494049515e-07,
+      "loss": 0.3561,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7730258356444315,
+      "grad_norm": 0.09044079959091988,
+      "learning_rate": 7.729579962154743e-07,
+      "loss": 0.342,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7741935483870968,
+      "grad_norm": 0.09175523739342407,
+      "learning_rate": 7.651128420910925e-07,
+      "loss": 0.3516,
+      "step": 1519
+    },
+    {
+      "epoch": 1.775361261129762,
+      "grad_norm": 0.10029630762624403,
+      "learning_rate": 7.573061196800414e-07,
+      "loss": 0.3792,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7765289738724275,
+      "grad_norm": 0.09524227287613211,
+      "learning_rate": 7.495378614706095e-07,
+      "loss": 0.3441,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7776966866150927,
+      "grad_norm": 0.09408123898746534,
+      "learning_rate": 7.418080997910204e-07,
+      "loss": 0.3526,
+      "step": 1522
+    },
+    {
+      "epoch": 1.778864399357758,
+      "grad_norm": 0.09306998941796975,
+      "learning_rate": 7.341168668092857e-07,
+      "loss": 0.3478,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7800321121004234,
+      "grad_norm": 0.09343022670924185,
+      "learning_rate": 7.26464194533083e-07,
+      "loss": 0.359,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7811998248430885,
+      "grad_norm": 0.09970501583044715,
+      "learning_rate": 7.188501148096117e-07,
+      "loss": 0.3825,
+      "step": 1525
+    },
+    {
+      "epoch": 1.782367537585754,
+      "grad_norm": 0.08893639857795682,
+      "learning_rate": 7.11274659325465e-07,
+      "loss": 0.358,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7835352503284192,
+      "grad_norm": 0.09707481629788065,
+      "learning_rate": 7.037378596065047e-07,
+      "loss": 0.3706,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7847029630710844,
+      "grad_norm": 0.09744441512796331,
+      "learning_rate": 6.962397470177162e-07,
+      "loss": 0.3492,
+      "step": 1528
+    },
+    {
+      "epoch": 1.78587067581375,
+      "grad_norm": 0.09131548515041593,
+      "learning_rate": 6.887803527630865e-07,
+      "loss": 0.3715,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7870383885564152,
+      "grad_norm": 0.09832310612364485,
+      "learning_rate": 6.813597078854772e-07,
+      "loss": 0.3593,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7882061012990804,
+      "grad_norm": 0.09139787935322442,
+      "learning_rate": 6.739778432664867e-07,
+      "loss": 0.3497,
+      "step": 1531
+    },
+    {
+      "epoch": 1.789373814041746,
+      "grad_norm": 0.08972409530333175,
+      "learning_rate": 6.666347896263326e-07,
+      "loss": 0.3492,
+      "step": 1532
+    },
+    {
+      "epoch": 1.790541526784411,
+      "grad_norm": 0.0939306053376951,
+      "learning_rate": 6.593305775237102e-07,
+      "loss": 0.368,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7917092395270764,
+      "grad_norm": 0.09569517281607347,
+      "learning_rate": 6.520652373556746e-07,
+      "loss": 0.3644,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7928769522697416,
+      "grad_norm": 0.0991725162188733,
+      "learning_rate": 6.448387993575155e-07,
+      "loss": 0.3523,
+      "step": 1535
+    },
+    {
+      "epoch": 1.794044665012407,
+      "grad_norm": 0.08771666942901221,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.337,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7952123777550724,
+      "grad_norm": 0.09316970564456332,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.375,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7963800904977374,
+      "grad_norm": 0.09362896611685759,
+      "learning_rate": 6.233931983060104e-07,
+      "loss": 0.3746,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7975478032404029,
+      "grad_norm": 0.09898459497443035,
+      "learning_rate": 6.16322668100472e-07,
+      "loss": 0.3777,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7987155159830681,
+      "grad_norm": 0.09669925482078409,
+      "learning_rate": 6.092911888103404e-07,
+      "loss": 0.3387,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7998832287257334,
+      "grad_norm": 0.0931018287408337,
+      "learning_rate": 6.022987896976706e-07,
+      "loss": 0.3576,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8010509414683988,
+      "grad_norm": 0.09675480900151948,
+      "learning_rate": 5.953454998618857e-07,
+      "loss": 0.3633,
+      "step": 1542
+    },
+    {
+      "epoch": 1.802218654211064,
+      "grad_norm": 0.09885623911089705,
+      "learning_rate": 5.884313482396497e-07,
+      "loss": 0.3711,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8033863669537293,
+      "grad_norm": 0.08769249736145526,
+      "learning_rate": 5.815563636047539e-07,
+      "loss": 0.3114,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8045540796963948,
+      "grad_norm": 0.08848327487014311,
+      "learning_rate": 5.747205745679885e-07,
+      "loss": 0.3388,
+      "step": 1545
+    },
+    {
+      "epoch": 1.8057217924390598,
+      "grad_norm": 0.09339099434570758,
+      "learning_rate": 5.6792400957703e-07,
+      "loss": 0.3639,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8068895051817253,
+      "grad_norm": 0.33282817701561246,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.3984,
+      "step": 1547
+    },
+    {
+      "epoch": 1.8080572179243906,
+      "grad_norm": 0.09840098450387896,
+      "learning_rate": 5.544486647069614e-07,
+      "loss": 0.3656,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8092249306670558,
+      "grad_norm": 0.0940376579022807,
+      "learning_rate": 5.477699409065607e-07,
+      "loss": 0.3596,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8103926434097213,
+      "grad_norm": 0.09408872672561333,
+      "learning_rate": 5.411305533091604e-07,
+      "loss": 0.3568,
+      "step": 1550
+    },
+    {
+      "epoch": 1.8115603561523865,
+      "grad_norm": 0.09352272510498026,
+      "learning_rate": 5.345305295450997e-07,
+      "loss": 0.3623,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8127280688950518,
+      "grad_norm": 0.0964340374766019,
+      "learning_rate": 5.279698970809011e-07,
+      "loss": 0.35,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8138957816377173,
+      "grad_norm": 0.10016460260532058,
+      "learning_rate": 5.21448683219149e-07,
+      "loss": 0.3669,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8150634943803823,
+      "grad_norm": 0.09310738846289589,
+      "learning_rate": 5.149669150983938e-07,
+      "loss": 0.3649,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8162312071230478,
+      "grad_norm": 0.0960195508699629,
+      "learning_rate": 5.085246196930271e-07,
+      "loss": 0.3443,
+      "step": 1555
+    },
+    {
+      "epoch": 1.817398919865713,
+      "grad_norm": 0.09084411190525205,
+      "learning_rate": 5.02121823813172e-07,
+      "loss": 0.3476,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8185666326083783,
+      "grad_norm": 0.09317639736155481,
+      "learning_rate": 4.957585541045684e-07,
+      "loss": 0.3662,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8197343453510437,
+      "grad_norm": 0.09307450188842942,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.3552,
+      "step": 1558
+    },
+    {
+      "epoch": 1.820902058093709,
+      "grad_norm": 0.09378087236464318,
+      "learning_rate": 4.831506989615109e-07,
+      "loss": 0.3532,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8220697708363742,
+      "grad_norm": 0.09853887168487113,
+      "learning_rate": 4.769061659956464e-07,
+      "loss": 0.3592,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8232374835790397,
+      "grad_norm": 0.0895703213807493,
+      "learning_rate": 4.707012641379871e-07,
+      "loss": 0.3464,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8244051963217047,
+      "grad_norm": 0.09513249722995187,
+      "learning_rate": 4.64536019210724e-07,
+      "loss": 0.3617,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8255729090643702,
+      "grad_norm": 0.0894746077666333,
+      "learning_rate": 4.5841045687101017e-07,
+      "loss": 0.3489,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8267406218070354,
+      "grad_norm": 0.09329331543418164,
+      "learning_rate": 4.5232460261085966e-07,
+      "loss": 0.3437,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8279083345497007,
+      "grad_norm": 0.09807307477798957,
+      "learning_rate": 4.4627848175703315e-07,
+      "loss": 0.3781,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8290760472923662,
+      "grad_norm": 0.0902415191415804,
+      "learning_rate": 4.402721194709436e-07,
+      "loss": 0.3638,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8302437600350314,
+      "grad_norm": 0.09781240862701146,
+      "learning_rate": 4.34305540748543e-07,
+      "loss": 0.3781,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8314114727776967,
+      "grad_norm": 0.09120184776252946,
+      "learning_rate": 4.2837877042021915e-07,
+      "loss": 0.3515,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8325791855203621,
+      "grad_norm": 0.08798523489321085,
+      "learning_rate": 4.224918331506955e-07,
+      "loss": 0.3446,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8337468982630272,
+      "grad_norm": 0.09850984686241776,
+      "learning_rate": 4.1664475343892727e-07,
+      "loss": 0.385,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8349146110056926,
+      "grad_norm": 0.09166331676398091,
+      "learning_rate": 4.1083755561799997e-07,
+      "loss": 0.3825,
+      "step": 1571
+    },
+    {
+      "epoch": 1.836082323748358,
+      "grad_norm": 0.09680824974050209,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.3732,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8372500364910231,
+      "grad_norm": 0.09341341834306582,
+      "learning_rate": 3.9934290215104754e-07,
+      "loss": 0.371,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8384177492336886,
+      "grad_norm": 0.09600516442078894,
+      "learning_rate": 3.9365549434092987e-07,
+      "loss": 0.3649,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8395854619763539,
+      "grad_norm": 0.09121087952297006,
+      "learning_rate": 3.8800806409327263e-07,
+      "loss": 0.3509,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8407531747190191,
+      "grad_norm": 0.09354314647562445,
+      "learning_rate": 3.82400634910306e-07,
+      "loss": 0.3655,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8419208874616846,
+      "grad_norm": 0.0969042090957331,
+      "learning_rate": 3.768332301277866e-07,
+      "loss": 0.3563,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8430886002043496,
+      "grad_norm": 0.09259419589886143,
+      "learning_rate": 3.7130587291490995e-07,
+      "loss": 0.359,
+      "step": 1578
+    },
+    {
+      "epoch": 1.844256312947015,
+      "grad_norm": 0.09340095246809603,
+      "learning_rate": 3.658185862742103e-07,
+      "loss": 0.368,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8454240256896803,
+      "grad_norm": 0.09578272195369947,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.3556,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8465917384323456,
+      "grad_norm": 0.1011309498209991,
+      "learning_rate": 3.549643158856042e-07,
+      "loss": 0.3885,
+      "step": 1581
+    },
+    {
+      "epoch": 1.847759451175011,
+      "grad_norm": 0.09252607237667726,
+      "learning_rate": 3.4959737730860144e-07,
+      "loss": 0.3641,
+      "step": 1582
+    },
+    {
+      "epoch": 1.848927163917676,
+      "grad_norm": 0.08674035568331366,
+      "learning_rate": 3.4427059964539654e-07,
+      "loss": 0.3396,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8500948766603416,
+      "grad_norm": 0.0959456599433292,
+      "learning_rate": 3.3898400506379937e-07,
+      "loss": 0.3699,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8512625894030068,
+      "grad_norm": 0.09025619505784578,
+      "learning_rate": 3.337376155643879e-07,
+      "loss": 0.3402,
+      "step": 1585
+    },
+    {
+      "epoch": 1.852430302145672,
+      "grad_norm": 0.09126040473726447,
+      "learning_rate": 3.2853145298042954e-07,
+      "loss": 0.3605,
+      "step": 1586
+    },
+    {
+      "epoch": 1.8535980148883375,
+      "grad_norm": 0.08748312942904526,
+      "learning_rate": 3.2336553897778014e-07,
+      "loss": 0.3391,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8547657276310028,
+      "grad_norm": 0.10068794759764646,
+      "learning_rate": 3.1823989505479934e-07,
+      "loss": 0.3722,
+      "step": 1588
+    },
+    {
+      "epoch": 1.855933440373668,
+      "grad_norm": 0.08730692802197484,
+      "learning_rate": 3.1315454254226105e-07,
+      "loss": 0.3473,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8571011531163335,
+      "grad_norm": 0.09261790469290701,
+      "learning_rate": 3.081095026032599e-07,
+      "loss": 0.3561,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8582688658589985,
+      "grad_norm": 0.09888538693313259,
+      "learning_rate": 3.0310479623313125e-07,
+      "loss": 0.3727,
+      "step": 1591
+    },
+    {
+      "epoch": 1.859436578601664,
+      "grad_norm": 0.09282662069502769,
+      "learning_rate": 2.9814044425935605e-07,
+      "loss": 0.3594,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8606042913443293,
+      "grad_norm": 0.10241781296073642,
+      "learning_rate": 2.93216467341475e-07,
+      "loss": 0.3846,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8617720040869945,
+      "grad_norm": 0.09101787046467757,
+      "learning_rate": 2.8833288597100996e-07,
+      "loss": 0.3581,
+      "step": 1594
+    },
+    {
+      "epoch": 1.86293971682966,
+      "grad_norm": 0.09636106718131139,
+      "learning_rate": 2.8348972047137155e-07,
+      "loss": 0.36,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8641074295723252,
+      "grad_norm": 0.09663797447690749,
+      "learning_rate": 2.7868699099777295e-07,
+      "loss": 0.3627,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8652751423149905,
+      "grad_norm": 0.09952312994529225,
+      "learning_rate": 2.7392471753715623e-07,
+      "loss": 0.381,
+      "step": 1597
+    },
+    {
+      "epoch": 1.866442855057656,
+      "grad_norm": 0.09374205546702222,
+      "learning_rate": 2.69202919908097e-07,
+      "loss": 0.3736,
+      "step": 1598
+    },
+    {
+      "epoch": 1.867610567800321,
+      "grad_norm": 0.09427555700583168,
+      "learning_rate": 2.645216177607313e-07,
+      "loss": 0.3637,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8687782805429864,
+      "grad_norm": 0.09754156160189956,
+      "learning_rate": 2.5988083057666534e-07,
+      "loss": 0.3767,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8699459932856517,
+      "grad_norm": 0.09967824814346742,
+      "learning_rate": 2.552805776689049e-07,
+      "loss": 0.376,
+      "step": 1601
+    },
+    {
+      "epoch": 1.871113706028317,
+      "grad_norm": 0.0989029687673235,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.3874,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8722814187709824,
+      "grad_norm": 0.09263682086165874,
+      "learning_rate": 2.462017510907955e-07,
+      "loss": 0.3479,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8734491315136477,
+      "grad_norm": 0.09358655485307513,
+      "learning_rate": 2.417232152027016e-07,
+      "loss": 0.3652,
+      "step": 1604
+    },
+    {
+      "epoch": 1.874616844256313,
+      "grad_norm": 0.09414639933549811,
+      "learning_rate": 2.3728528915526462e-07,
+      "loss": 0.3665,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8757845569989784,
+      "grad_norm": 0.09433327463537104,
+      "learning_rate": 2.3288799141726548e-07,
+      "loss": 0.3699,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8769522697416434,
+      "grad_norm": 0.09057496504084202,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.335,
+      "step": 1607
+    },
+    {
+      "epoch": 1.878119982484309,
+      "grad_norm": 0.09753032737418249,
+      "learning_rate": 2.242153538992331e-07,
+      "loss": 0.3637,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8792876952269741,
+      "grad_norm": 0.0999901526361326,
+      "learning_rate": 2.1994005021106713e-07,
+      "loss": 0.3751,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8804554079696394,
+      "grad_norm": 0.09449883977041862,
+      "learning_rate": 2.1570544701592254e-07,
+      "loss": 0.3659,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8816231207123049,
+      "grad_norm": 0.09193102627429028,
+      "learning_rate": 2.1151156193643496e-07,
+      "loss": 0.3396,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8827908334549701,
+      "grad_norm": 0.09534927040813856,
+      "learning_rate": 2.0735841242578992e-07,
+      "loss": 0.3654,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8839585461976354,
+      "grad_norm": 0.09564422662833709,
+      "learning_rate": 2.0324601576764525e-07,
+      "loss": 0.3595,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8851262589403008,
+      "grad_norm": 0.09351314128764587,
+      "learning_rate": 1.9917438907606556e-07,
+      "loss": 0.353,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8862939716829659,
+      "grad_norm": 0.09663353762310097,
+      "learning_rate": 1.951435492954501e-07,
+      "loss": 0.351,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8874616844256313,
+      "grad_norm": 0.09462182080123754,
+      "learning_rate": 1.9115351320045495e-07,
+      "loss": 0.373,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8886293971682966,
+      "grad_norm": 0.09599401596772898,
+      "learning_rate": 1.8720429739592983e-07,
+      "loss": 0.3747,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8897971099109618,
+      "grad_norm": 0.093571109749288,
+      "learning_rate": 1.8329591831685146e-07,
+      "loss": 0.3602,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8909648226536273,
+      "grad_norm": 0.09565814905840625,
+      "learning_rate": 1.7942839222824804e-07,
+      "loss": 0.3653,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8921325353962926,
+      "grad_norm": 0.09174113909532629,
+      "learning_rate": 1.7560173522513268e-07,
+      "loss": 0.3507,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8933002481389578,
+      "grad_norm": 0.09472079555729801,
+      "learning_rate": 1.7181596323244453e-07,
+      "loss": 0.3716,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8944679608816233,
+      "grad_norm": 0.09211301169375857,
+      "learning_rate": 1.6807109200496995e-07,
+      "loss": 0.3389,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8956356736242883,
+      "grad_norm": 0.09195892885117068,
+      "learning_rate": 1.6436713712729036e-07,
+      "loss": 0.3598,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8968033863669538,
+      "grad_norm": 0.08792730966506045,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.3434,
+      "step": 1624
+    },
+    {
+      "epoch": 1.897971099109619,
+      "grad_norm": 0.09479902191644857,
+      "learning_rate": 1.5708203790817166e-07,
+      "loss": 0.3661,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8991388118522843,
+      "grad_norm": 0.09171032301070867,
+      "learning_rate": 1.5350092388425108e-07,
+      "loss": 0.3449,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9003065245949498,
+      "grad_norm": 0.09833122366054534,
+      "learning_rate": 1.4996078684503145e-07,
+      "loss": 0.3931,
+      "step": 1627
+    },
+    {
+      "epoch": 1.9014742373376148,
+      "grad_norm": 0.09489808908713028,
+      "learning_rate": 1.464616415230702e-07,
+      "loss": 0.3616,
+      "step": 1628
+    },
+    {
+      "epoch": 1.9026419500802803,
+      "grad_norm": 0.08984334798381828,
+      "learning_rate": 1.4300350248034002e-07,
+      "loss": 0.339,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9038096628229455,
+      "grad_norm": 0.0872707552102835,
+      "learning_rate": 1.3958638410815905e-07,
+      "loss": 0.3377,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9049773755656108,
+      "grad_norm": 0.09602187557555454,
+      "learning_rate": 1.3621030062713759e-07,
+      "loss": 0.3602,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9061450883082762,
+      "grad_norm": 0.09198834459131468,
+      "learning_rate": 1.3287526608711132e-07,
+      "loss": 0.3561,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9073128010509415,
+      "grad_norm": 0.09670092755472738,
+      "learning_rate": 1.2958129436709154e-07,
+      "loss": 0.3562,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9084805137936067,
+      "grad_norm": 0.09381491595184102,
+      "learning_rate": 1.2632839917520179e-07,
+      "loss": 0.363,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9096482265362722,
+      "grad_norm": 0.09815367142841175,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.384,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9108159392789372,
+      "grad_norm": 0.1343134975295095,
+      "learning_rate": 1.1994589235353682e-07,
+      "loss": 0.3623,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9119836520216027,
+      "grad_norm": 0.10143568587591245,
+      "learning_rate": 1.1681630728506699e-07,
+      "loss": 0.3774,
+      "step": 1637
+    },
+    {
+      "epoch": 1.913151364764268,
+      "grad_norm": 0.09759112538734832,
+      "learning_rate": 1.1372785186723135e-07,
+      "loss": 0.3598,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9143190775069332,
+      "grad_norm": 0.09425566435833203,
+      "learning_rate": 1.1068053895288089e-07,
+      "loss": 0.3583,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9154867902495987,
+      "grad_norm": 0.10184025494514068,
+      "learning_rate": 1.0767438122364914e-07,
+      "loss": 0.3897,
+      "step": 1640
+    },
+    {
+      "epoch": 1.916654502992264,
+      "grad_norm": 0.09258427149306978,
+      "learning_rate": 1.0470939118989997e-07,
+      "loss": 0.3618,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9178222157349292,
+      "grad_norm": 0.0940175879465826,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.3582,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9189899284775946,
+      "grad_norm": 0.09710447314865601,
+      "learning_rate": 9.890296339363448e-08,
+      "loss": 0.3787,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9201576412202597,
+      "grad_norm": 0.09825933715531684,
+      "learning_rate": 9.606154979502791e-08,
+      "loss": 0.357,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9213253539629251,
+      "grad_norm": 0.09545438741216841,
+      "learning_rate": 9.326135221961907e-08,
+      "loss": 0.3767,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9224930667055904,
+      "grad_norm": 0.0907002923303512,
+      "learning_rate": 9.0502382320653e-08,
+      "loss": 0.3545,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9236607794482556,
+      "grad_norm": 0.09806378115101612,
+      "learning_rate": 8.778465157979976e-08,
+      "loss": 0.3724,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9248284921909211,
+      "grad_norm": 0.09557504495280691,
+      "learning_rate": 8.510817130711224e-08,
+      "loss": 0.3641,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9259962049335864,
+      "grad_norm": 0.08957841341920765,
+      "learning_rate": 8.247295264097288e-08,
+      "loss": 0.3339,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9271639176762516,
+      "grad_norm": 0.09772961944807744,
+      "learning_rate": 7.987900654805481e-08,
+      "loss": 0.361,
+      "step": 1650
+    },
+    {
+      "epoch": 1.928331630418917,
+      "grad_norm": 0.09907696211932171,
+      "learning_rate": 7.732634382326965e-08,
+      "loss": 0.3797,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9294993431615821,
+      "grad_norm": 0.09831747899101093,
+      "learning_rate": 7.481497508972313e-08,
+      "loss": 0.3798,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9306670559042476,
+      "grad_norm": 0.09325349892482146,
+      "learning_rate": 7.234491079867623e-08,
+      "loss": 0.3502,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9318347686469128,
+      "grad_norm": 0.09953661188406146,
+      "learning_rate": 6.991616122949629e-08,
+      "loss": 0.3787,
+      "step": 1654
+    },
+    {
+      "epoch": 1.933002481389578,
+      "grad_norm": 0.09306200569938558,
+      "learning_rate": 6.752873648961821e-08,
+      "loss": 0.3539,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9341701941322436,
+      "grad_norm": 0.10098655418191428,
+      "learning_rate": 6.51826465144978e-08,
+      "loss": 0.3784,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9353379068749088,
+      "grad_norm": 0.09650067978860849,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.3551,
+      "step": 1657
+    },
+    {
+      "epoch": 1.936505619617574,
+      "grad_norm": 0.09579745963745084,
+      "learning_rate": 6.061450974022775e-08,
+      "loss": 0.3782,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9376733323602395,
+      "grad_norm": 0.09971554861555072,
+      "learning_rate": 5.839248195174008e-08,
+      "loss": 0.3707,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9388410451029046,
+      "grad_norm": 0.09523019983267962,
+      "learning_rate": 5.621182694925731e-08,
+      "loss": 0.3701,
+      "step": 1660
+    },
+    {
+      "epoch": 1.94000875784557,
+      "grad_norm": 0.09549161175832406,
+      "learning_rate": 5.407255380774579e-08,
+      "loss": 0.3604,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 0.09582900816668524,
+      "learning_rate": 5.19746714299596e-08,
+      "loss": 0.3724,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9423441833309005,
+      "grad_norm": 0.09561261913308941,
+      "learning_rate": 4.991818854640396e-08,
+      "loss": 0.3684,
+      "step": 1663
+    },
+    {
+      "epoch": 1.943511896073566,
+      "grad_norm": 0.09395489458793399,
+      "learning_rate": 4.79031137152941e-08,
+      "loss": 0.3739,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9446796088162313,
+      "grad_norm": 0.09510569060104347,
+      "learning_rate": 4.592945532252424e-08,
+      "loss": 0.36,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9458473215588965,
+      "grad_norm": 0.09411585129596566,
+      "learning_rate": 4.399722158162867e-08,
+      "loss": 0.3408,
+      "step": 1666
+    },
+    {
+      "epoch": 1.947015034301562,
+      "grad_norm": 0.09089837124254001,
+      "learning_rate": 4.210642053375069e-08,
+      "loss": 0.3467,
+      "step": 1667
+    },
+    {
+      "epoch": 1.948182747044227,
+      "grad_norm": 0.10061043579570687,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.3631,
+      "step": 1668
+    },
+    {
+      "epoch": 1.9493504597868925,
+      "grad_norm": 0.09261954878288876,
+      "learning_rate": 3.844914781946485e-08,
+      "loss": 0.3461,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9505181725295577,
+      "grad_norm": 0.09338438280807018,
+      "learning_rate": 3.668269137308666e-08,
+      "loss": 0.3456,
+      "step": 1670
+    },
+    {
+      "epoch": 1.951685885272223,
+      "grad_norm": 0.08951516299039715,
+      "learning_rate": 3.495769805972215e-08,
+      "loss": 0.3597,
+      "step": 1671
+    },
+    {
+      "epoch": 1.9528535980148884,
+      "grad_norm": 0.09264387806113662,
+      "learning_rate": 3.327417505806785e-08,
+      "loss": 0.3429,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9540213107575535,
+      "grad_norm": 0.09646166576170609,
+      "learning_rate": 3.163212937423832e-08,
+      "loss": 0.3673,
+      "step": 1673
+    },
+    {
+      "epoch": 1.955189023500219,
+      "grad_norm": 0.09261328050363513,
+      "learning_rate": 3.003156784173511e-08,
+      "loss": 0.3656,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9563567362428842,
+      "grad_norm": 0.08773353879776341,
+      "learning_rate": 2.8472497121424524e-08,
+      "loss": 0.3464,
+      "step": 1675
+    },
+    {
+      "epoch": 1.9575244489855494,
+      "grad_norm": 0.09964025300701589,
+      "learning_rate": 2.695492370149988e-08,
+      "loss": 0.3709,
+      "step": 1676
+    },
+    {
+      "epoch": 1.958692161728215,
+      "grad_norm": 0.09115302025877647,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.3604,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9598598744708802,
+      "grad_norm": 0.0911975937867763,
+      "learning_rate": 2.4044293852099055e-08,
+      "loss": 0.3442,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9610275872135454,
+      "grad_norm": 0.10046926994562616,
+      "learning_rate": 2.265124953543918e-08,
+      "loss": 0.3788,
+      "step": 1679
+    },
+    {
+      "epoch": 1.962195299956211,
+      "grad_norm": 0.09560222325231486,
+      "learning_rate": 2.1299726744747896e-08,
+      "loss": 0.3572,
+      "step": 1680
+    },
+    {
+      "epoch": 1.963363012698876,
+      "grad_norm": 0.09330216841009756,
+      "learning_rate": 1.998973110449498e-08,
+      "loss": 0.3571,
+      "step": 1681
+    },
+    {
+      "epoch": 1.9645307254415414,
+      "grad_norm": 0.09520195660818714,
+      "learning_rate": 1.872126806633068e-08,
+      "loss": 0.382,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9656984381842066,
+      "grad_norm": 0.09820582068685688,
+      "learning_rate": 1.749434290906571e-08,
+      "loss": 0.3659,
+      "step": 1683
+    },
+    {
+      "epoch": 1.966866150926872,
+      "grad_norm": 0.0986222724326869,
+      "learning_rate": 1.630896073864352e-08,
+      "loss": 0.3713,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9680338636695374,
+      "grad_norm": 0.09145005245155467,
+      "learning_rate": 1.5165126488126957e-08,
+      "loss": 0.3514,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9692015764122026,
+      "grad_norm": 0.09751836179749522,
+      "learning_rate": 1.4062844917672735e-08,
+      "loss": 0.382,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9703692891548679,
+      "grad_norm": 0.09381312592404245,
+      "learning_rate": 1.300212061451367e-08,
+      "loss": 0.3362,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9715370018975333,
+      "grad_norm": 0.10565185499292257,
+      "learning_rate": 1.1982957992936472e-08,
+      "loss": 0.376,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9727047146401984,
+      "grad_norm": 0.08973751034644811,
+      "learning_rate": 1.1005361294268435e-08,
+      "loss": 0.3583,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9738724273828638,
+      "grad_norm": 0.0928639193897308,
+      "learning_rate": 1.0069334586854106e-08,
+      "loss": 0.3501,
+      "step": 1690
+    },
+    {
+      "epoch": 1.975040140125529,
+      "grad_norm": 0.09323240259974658,
+      "learning_rate": 9.174881766043086e-09,
+      "loss": 0.352,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9762078528681943,
+      "grad_norm": 0.09291631722667865,
+      "learning_rate": 8.322006554171147e-09,
+      "loss": 0.3567,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9773755656108598,
+      "grad_norm": 0.09420551743106956,
+      "learning_rate": 7.510712500546913e-09,
+      "loss": 0.3713,
+      "step": 1693
+    },
+    {
+      "epoch": 1.978543278353525,
+      "grad_norm": 0.09426153425001181,
+      "learning_rate": 6.741002981435207e-09,
+      "loss": 0.3637,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9797109910961903,
+      "grad_norm": 0.09118020556733614,
+      "learning_rate": 6.0128812000415096e-09,
+      "loss": 0.3502,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9808787038388558,
+      "grad_norm": 0.09596001311100454,
+      "learning_rate": 5.326350186503071e-09,
+      "loss": 0.3666,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9820464165815208,
+      "grad_norm": 0.09796803217366595,
+      "learning_rate": 4.681412797872265e-09,
+      "loss": 0.3952,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9832141293241863,
+      "grad_norm": 0.09237254603074536,
+      "learning_rate": 4.0780717181077015e-09,
+      "loss": 0.3515,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9843818420668515,
+      "grad_norm": 0.09135396874977618,
+      "learning_rate": 3.5163294580609076e-09,
+      "loss": 0.3614,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9855495548095168,
+      "grad_norm": 0.09563631854202738,
+      "learning_rate": 2.9961883554674443e-09,
+      "loss": 0.3663,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9867172675521823,
+      "grad_norm": 0.095178172673331,
+      "learning_rate": 2.5176505749346937e-09,
+      "loss": 0.3396,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9878849802948475,
+      "grad_norm": 0.0992619501989855,
+      "learning_rate": 2.080718107935198e-09,
+      "loss": 0.3771,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9890526930375128,
+      "grad_norm": 0.09764703208634054,
+      "learning_rate": 1.6853927727988883e-09,
+      "loss": 0.3732,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9902204057801782,
+      "grad_norm": 0.09584494815501875,
+      "learning_rate": 1.3316762147030925e-09,
+      "loss": 0.3626,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9913881185228433,
+      "grad_norm": 0.08868863609715774,
+      "learning_rate": 1.019569905666984e-09,
+      "loss": 0.3501,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9925558312655087,
+      "grad_norm": 0.09611959557096654,
+      "learning_rate": 7.490751445449196e-10,
+      "loss": 0.3669,
+      "step": 1706
+    },
+    {
+      "epoch": 1.993723544008174,
+      "grad_norm": 0.09411796617653201,
+      "learning_rate": 5.201930570242209e-10,
+      "loss": 0.3589,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9948912567508392,
+      "grad_norm": 0.09282937809856932,
+      "learning_rate": 3.3292459561518055e-10,
+      "loss": 0.3523,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9960589694935047,
+      "grad_norm": 0.08611806666987679,
+      "learning_rate": 1.872705396488428e-10,
+      "loss": 0.341,
+      "step": 1709
+    },
+    {
+      "epoch": 1.99722668223617,
+      "grad_norm": 0.09183621978198979,
+      "learning_rate": 8.323149527811325e-11,
+      "loss": 0.3703,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9983943949788352,
+      "grad_norm": 0.09610967946697352,
+      "learning_rate": 2.080789546776707e-11,
+      "loss": 0.3772,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9995621077215007,
+      "grad_norm": 0.09549495391412142,
+      "learning_rate": 0.0,
+      "loss": 0.3653,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9995621077215007,
+      "step": 1712,
+      "total_flos": 210186493870080.0,
+      "train_loss": 0.1864510835540072,
+      "train_runtime": 3004.1025,
+      "train_samples_per_second": 72.968,
+      "train_steps_per_second": 0.57
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1712,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 210186493870080.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}