diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.670758292249388,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.70758292249388e-05,
+      "grad_norm": 0.2088892071184168,
+      "learning_rate": 2e-05,
+      "loss": 5.5031,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001341516584498776,
+      "grad_norm": 0.17321255452342477,
+      "learning_rate": 2e-05,
+      "loss": 5.4741,
+      "step": 2
+    },
+    {
+      "epoch": 0.00020122748767481639,
+      "grad_norm": 0.17705717953835415,
+      "learning_rate": 2e-05,
+      "loss": 5.5811,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002683033168997552,
+      "grad_norm": 0.17260535045397474,
+      "learning_rate": 2e-05,
+      "loss": 5.4813,
+      "step": 4
+    },
+    {
+      "epoch": 0.00033537914612469396,
+      "grad_norm": 0.19166054132312318,
+      "learning_rate": 2e-05,
+      "loss": 5.7439,
+      "step": 5
+    },
+    {
+      "epoch": 0.00040245497534963277,
+      "grad_norm": 0.17062805312294974,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 6
+    },
+    {
+      "epoch": 0.00046953080457457153,
+      "grad_norm": 0.1441720310500711,
+      "learning_rate": 2e-05,
+      "loss": 5.6432,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005366066337995104,
+      "grad_norm": 0.18782153299505971,
+      "learning_rate": 2e-05,
+      "loss": 5.6218,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006036824630244491,
+      "grad_norm": 0.16853907715806088,
+      "learning_rate": 2e-05,
+      "loss": 5.6835,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006707582922493879,
+      "grad_norm": 0.1473332819625821,
+      "learning_rate": 2e-05,
+      "loss": 5.6875,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007378341214743267,
+      "grad_norm": 0.16391970465739825,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008049099506992655,
+      "grad_norm": 0.16220219477867096,
+      "learning_rate": 2e-05,
+      "loss": 5.5406,
+      "step": 12
+    },
+    {
+      "epoch": 0.0008719857799242044,
+      "grad_norm": 0.1453858067106413,
+      "learning_rate": 2e-05,
+      "loss": 5.5102,
+      "step": 13
+    },
+    {
+      "epoch": 0.0009390616091491431,
+      "grad_norm": 0.14084321490627466,
+      "learning_rate": 2e-05,
+      "loss": 5.5568,
+      "step": 14
+    },
+    {
+      "epoch": 0.0010061374383740819,
+      "grad_norm": 0.14000646727236543,
+      "learning_rate": 2e-05,
+      "loss": 5.5108,
+      "step": 15
+    },
+    {
+      "epoch": 0.0010732132675990208,
+      "grad_norm": 0.15977868406169066,
+      "learning_rate": 2e-05,
+      "loss": 5.5685,
+      "step": 16
+    },
+    {
+      "epoch": 0.0011402890968239595,
+      "grad_norm": 0.13950272788338686,
+      "learning_rate": 2e-05,
+      "loss": 5.4368,
+      "step": 17
+    },
+    {
+      "epoch": 0.0012073649260488982,
+      "grad_norm": 0.1423850995627629,
+      "learning_rate": 2e-05,
+      "loss": 5.5593,
+      "step": 18
+    },
+    {
+      "epoch": 0.0012744407552738371,
+      "grad_norm": 0.14591573013494366,
+      "learning_rate": 2e-05,
+      "loss": 5.5718,
+      "step": 19
+    },
+    {
+      "epoch": 0.0013415165844987758,
+      "grad_norm": 0.13782840770835594,
+      "learning_rate": 2e-05,
+      "loss": 5.5443,
+      "step": 20
+    },
+    {
+      "epoch": 0.0014085924137237148,
+      "grad_norm": 0.13978648384610254,
+      "learning_rate": 2e-05,
+      "loss": 5.408,
+      "step": 21
+    },
+    {
+      "epoch": 0.0014756682429486535,
+      "grad_norm": 0.13512627725813864,
+      "learning_rate": 2e-05,
+      "loss": 5.5443,
+      "step": 22
+    },
+    {
+      "epoch": 0.0015427440721735922,
+      "grad_norm": 0.1312106460504487,
+      "learning_rate": 2e-05,
+      "loss": 5.5331,
+      "step": 23
+    },
+    {
+      "epoch": 0.001609819901398531,
+      "grad_norm": 0.14056683946857726,
+      "learning_rate": 2e-05,
+      "loss": 5.5272,
+      "step": 24
+    },
+    {
+      "epoch": 0.0016768957306234698,
+      "grad_norm": 0.13231311902534845,
+      "learning_rate": 2e-05,
+      "loss": 5.5305,
+      "step": 25
+    },
+    {
+      "epoch": 0.0017439715598484087,
+      "grad_norm": 0.13784679736013092,
+      "learning_rate": 2e-05,
+      "loss": 5.5801,
+      "step": 26
+    },
+    {
+      "epoch": 0.0018110473890733474,
+      "grad_norm": 0.14221597170047004,
+      "learning_rate": 2e-05,
+      "loss": 5.4854,
+      "step": 27
+    },
+    {
+      "epoch": 0.0018781232182982861,
+      "grad_norm": 0.13640226046082501,
+      "learning_rate": 2e-05,
+      "loss": 5.6401,
+      "step": 28
+    },
+    {
+      "epoch": 0.001945199047523225,
+      "grad_norm": 0.13092587702477357,
+      "learning_rate": 2e-05,
+      "loss": 5.5403,
+      "step": 29
+    },
+    {
+      "epoch": 0.0020122748767481637,
+      "grad_norm": 0.13344689085780878,
+      "learning_rate": 2e-05,
+      "loss": 5.6236,
+      "step": 30
+    },
+    {
+      "epoch": 0.0020793507059731027,
+      "grad_norm": 0.1335214434649408,
+      "learning_rate": 2e-05,
+      "loss": 5.5047,
+      "step": 31
+    },
+    {
+      "epoch": 0.0021464265351980416,
+      "grad_norm": 0.1328396399147632,
+      "learning_rate": 2e-05,
+      "loss": 5.5275,
+      "step": 32
+    },
+    {
+      "epoch": 0.00221350236442298,
+      "grad_norm": 0.13535467521879146,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 33
+    },
+    {
+      "epoch": 0.002280578193647919,
+      "grad_norm": 0.13414149168866382,
+      "learning_rate": 2e-05,
+      "loss": 5.33,
+      "step": 34
+    },
+    {
+      "epoch": 0.002347654022872858,
+      "grad_norm": 0.13780529129927843,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 35
+    },
+    {
+      "epoch": 0.0024147298520977964,
+      "grad_norm": 0.13604219553292732,
+      "learning_rate": 2e-05,
+      "loss": 5.5357,
+      "step": 36
+    },
+    {
+      "epoch": 0.0024818056813227353,
+      "grad_norm": 0.1318214035103109,
+      "learning_rate": 2e-05,
+      "loss": 5.3572,
+      "step": 37
+    },
+    {
+      "epoch": 0.0025488815105476743,
+      "grad_norm": 0.1351630493431948,
+      "learning_rate": 2e-05,
+      "loss": 5.5322,
+      "step": 38
+    },
+    {
+      "epoch": 0.0026159573397726127,
+      "grad_norm": 0.1284795748342403,
+      "learning_rate": 2e-05,
+      "loss": 5.6209,
+      "step": 39
+    },
+    {
+      "epoch": 0.0026830331689975517,
+      "grad_norm": 0.13753129488003146,
+      "learning_rate": 2e-05,
+      "loss": 5.5629,
+      "step": 40
+    },
+    {
+      "epoch": 0.0027501089982224906,
+      "grad_norm": 0.1336923415931152,
+      "learning_rate": 2e-05,
+      "loss": 5.698,
+      "step": 41
+    },
+    {
+      "epoch": 0.0028171848274474295,
+      "grad_norm": 0.1353122557488069,
+      "learning_rate": 2e-05,
+      "loss": 5.5382,
+      "step": 42
+    },
+    {
+      "epoch": 0.002884260656672368,
+      "grad_norm": 0.13239176265106772,
+      "learning_rate": 2e-05,
+      "loss": 5.58,
+      "step": 43
+    },
+    {
+      "epoch": 0.002951336485897307,
+      "grad_norm": 0.12996921152033858,
+      "learning_rate": 2e-05,
+      "loss": 5.3952,
+      "step": 44
+    },
+    {
+      "epoch": 0.003018412315122246,
+      "grad_norm": 0.13451266328355627,
+      "learning_rate": 2e-05,
+      "loss": 5.5147,
+      "step": 45
+    },
+    {
+      "epoch": 0.0030854881443471843,
+      "grad_norm": 0.12953006544419737,
+      "learning_rate": 2e-05,
+      "loss": 5.5425,
+      "step": 46
+    },
+    {
+      "epoch": 0.0031525639735721232,
+      "grad_norm": 0.13190036666236674,
+      "learning_rate": 2e-05,
+      "loss": 5.4969,
+      "step": 47
+    },
+    {
+      "epoch": 0.003219639802797062,
+      "grad_norm": 0.12580813137406593,
+      "learning_rate": 2e-05,
+      "loss": 5.5555,
+      "step": 48
+    },
+    {
+      "epoch": 0.0032867156320220007,
+      "grad_norm": 0.134240834895453,
+      "learning_rate": 2e-05,
+      "loss": 5.5609,
+      "step": 49
+    },
+    {
+      "epoch": 0.0033537914612469396,
+      "grad_norm": 0.13509934480058702,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 50
+    },
+    {
+      "epoch": 0.0034208672904718785,
+      "grad_norm": 0.12911760934863037,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 51
+    },
+    {
+      "epoch": 0.0034879431196968174,
+      "grad_norm": 0.1289304855467342,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 52
+    },
+    {
+      "epoch": 0.003555018948921756,
+      "grad_norm": 0.13558938400867776,
+      "learning_rate": 2e-05,
+      "loss": 5.6586,
+      "step": 53
+    },
+    {
+      "epoch": 0.003622094778146695,
+      "grad_norm": 0.13951889305413906,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 54
+    },
+    {
+      "epoch": 0.0036891706073716338,
+      "grad_norm": 0.1310176371439046,
+      "learning_rate": 2e-05,
+      "loss": 5.4154,
+      "step": 55
+    },
+    {
+      "epoch": 0.0037562464365965722,
+      "grad_norm": 0.13468151741173084,
+      "learning_rate": 2e-05,
+      "loss": 5.4743,
+      "step": 56
+    },
+    {
+      "epoch": 0.003823322265821511,
+      "grad_norm": 0.13657775759268048,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 57
+    },
+    {
+      "epoch": 0.00389039809504645,
+      "grad_norm": 0.1377174064931782,
+      "learning_rate": 2e-05,
+      "loss": 5.5572,
+      "step": 58
+    },
+    {
+      "epoch": 0.003957473924271389,
+      "grad_norm": 0.13134450915923954,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 59
+    },
+    {
+      "epoch": 0.0040245497534963275,
+      "grad_norm": 0.13550158007127727,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 60
+    },
+    {
+      "epoch": 0.004091625582721266,
+      "grad_norm": 0.14353127480952524,
+      "learning_rate": 2e-05,
+      "loss": 5.5321,
+      "step": 61
+    },
+    {
+      "epoch": 0.004158701411946205,
+      "grad_norm": 0.13339324787974144,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 62
+    },
+    {
+      "epoch": 0.004225777241171144,
+      "grad_norm": 0.1325785575514609,
+      "learning_rate": 2e-05,
+      "loss": 5.5095,
+      "step": 63
+    },
+    {
+      "epoch": 0.004292853070396083,
+      "grad_norm": 0.13439259042034174,
+      "learning_rate": 2e-05,
+      "loss": 5.4382,
+      "step": 64
+    },
+    {
+      "epoch": 0.004359928899621022,
+      "grad_norm": 0.13099643230056837,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 65
+    },
+    {
+      "epoch": 0.00442700472884596,
+      "grad_norm": 0.1372781932082803,
+      "learning_rate": 2e-05,
+      "loss": 5.5851,
+      "step": 66
+    },
+    {
+      "epoch": 0.0044940805580708995,
+      "grad_norm": 0.13618270342101546,
+      "learning_rate": 2e-05,
+      "loss": 5.6073,
+      "step": 67
+    },
+    {
+      "epoch": 0.004561156387295838,
+      "grad_norm": 0.1281776153813227,
+      "learning_rate": 2e-05,
+      "loss": 5.5436,
+      "step": 68
+    },
+    {
+      "epoch": 0.0046282322165207765,
+      "grad_norm": 0.13014390480646368,
+      "learning_rate": 2e-05,
+      "loss": 5.5412,
+      "step": 69
+    },
+    {
+      "epoch": 0.004695308045745716,
+      "grad_norm": 0.13117541464138138,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 70
+    },
+    {
+      "epoch": 0.004762383874970654,
+      "grad_norm": 0.13532574317570706,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 71
+    },
+    {
+      "epoch": 0.004829459704195593,
+      "grad_norm": 0.13728600500531574,
+      "learning_rate": 2e-05,
+      "loss": 5.5811,
+      "step": 72
+    },
+    {
+      "epoch": 0.004896535533420532,
+      "grad_norm": 0.13883210697177115,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 73
+    },
+    {
+      "epoch": 0.004963611362645471,
+      "grad_norm": 0.13866279536440218,
+      "learning_rate": 2e-05,
+      "loss": 5.5829,
+      "step": 74
+    },
+    {
+      "epoch": 0.005030687191870409,
+      "grad_norm": 0.13919952789506831,
+      "learning_rate": 2e-05,
+      "loss": 5.4849,
+      "step": 75
+    },
+    {
+      "epoch": 0.0050977630210953485,
+      "grad_norm": 0.13470403306264467,
+      "learning_rate": 2e-05,
+      "loss": 5.5243,
+      "step": 76
+    },
+    {
+      "epoch": 0.005164838850320287,
+      "grad_norm": 0.13656360198853124,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 77
+    },
+    {
+      "epoch": 0.0052319146795452255,
+      "grad_norm": 0.13414689167296273,
+      "learning_rate": 2e-05,
+      "loss": 5.5352,
+      "step": 78
+    },
+    {
+      "epoch": 0.005298990508770165,
+      "grad_norm": 0.13548879642087264,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 79
+    },
+    {
+      "epoch": 0.005366066337995103,
+      "grad_norm": 0.13470819997700828,
+      "learning_rate": 2e-05,
+      "loss": 5.5881,
+      "step": 80
+    },
+    {
+      "epoch": 0.005433142167220043,
+      "grad_norm": 0.13946108633411297,
+      "learning_rate": 2e-05,
+      "loss": 5.5732,
+      "step": 81
+    },
+    {
+      "epoch": 0.005500217996444981,
+      "grad_norm": 0.14148673158288796,
+      "learning_rate": 2e-05,
+      "loss": 5.4718,
+      "step": 82
+    },
+    {
+      "epoch": 0.00556729382566992,
+      "grad_norm": 0.13123561751639057,
+      "learning_rate": 2e-05,
+      "loss": 5.4608,
+      "step": 83
+    },
+    {
+      "epoch": 0.005634369654894859,
+      "grad_norm": 0.14453362096375322,
+      "learning_rate": 2e-05,
+      "loss": 5.5689,
+      "step": 84
+    },
+    {
+      "epoch": 0.0057014454841197975,
+      "grad_norm": 0.1390134333564256,
+      "learning_rate": 2e-05,
+      "loss": 5.5596,
+      "step": 85
+    },
+    {
+      "epoch": 0.005768521313344736,
+      "grad_norm": 0.13086907602934444,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 86
+    },
+    {
+      "epoch": 0.005835597142569675,
+      "grad_norm": 0.1376140994825975,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 87
+    },
+    {
+      "epoch": 0.005902672971794614,
+      "grad_norm": 0.13940797532680824,
+      "learning_rate": 2e-05,
+      "loss": 5.5844,
+      "step": 88
+    },
+    {
+      "epoch": 0.005969748801019552,
+      "grad_norm": 0.1380753874103704,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 89
+    },
+    {
+      "epoch": 0.006036824630244492,
+      "grad_norm": 0.13230094784586252,
+      "learning_rate": 2e-05,
+      "loss": 5.4703,
+      "step": 90
+    },
+    {
+      "epoch": 0.00610390045946943,
+      "grad_norm": 0.14069642600745544,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 91
+    },
+    {
+      "epoch": 0.006170976288694369,
+      "grad_norm": 0.14169739074173623,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 92
+    },
+    {
+      "epoch": 0.006238052117919308,
+      "grad_norm": 0.1367517162249026,
+      "learning_rate": 2e-05,
+      "loss": 5.4307,
+      "step": 93
+    },
+    {
+      "epoch": 0.0063051279471442465,
+      "grad_norm": 0.13627396736285932,
+      "learning_rate": 2e-05,
+      "loss": 5.4903,
+      "step": 94
+    },
+    {
+      "epoch": 0.006372203776369185,
+      "grad_norm": 0.13699244564898574,
+      "learning_rate": 2e-05,
+      "loss": 5.6385,
+      "step": 95
+    },
+    {
+      "epoch": 0.006439279605594124,
+      "grad_norm": 0.13593832219752122,
+      "learning_rate": 2e-05,
+      "loss": 5.5313,
+      "step": 96
+    },
+    {
+      "epoch": 0.006506355434819063,
+      "grad_norm": 0.13857135414659413,
+      "learning_rate": 2e-05,
+      "loss": 5.6319,
+      "step": 97
+    },
+    {
+      "epoch": 0.006573431264044001,
+      "grad_norm": 0.13710855531989966,
+      "learning_rate": 2e-05,
+      "loss": 5.4892,
+      "step": 98
+    },
+    {
+      "epoch": 0.006640507093268941,
+      "grad_norm": 0.13379942072765424,
+      "learning_rate": 2e-05,
+      "loss": 5.5565,
+      "step": 99
+    },
+    {
+      "epoch": 0.006707582922493879,
+      "grad_norm": 0.13095575454186265,
+      "learning_rate": 2e-05,
+      "loss": 5.5431,
+      "step": 100
+    },
+    {
+      "epoch": 0.0067746587517188185,
+      "grad_norm": 0.12974376269591142,
+      "learning_rate": 2e-05,
+      "loss": 5.5068,
+      "step": 101
+    },
+    {
+      "epoch": 0.006841734580943757,
+      "grad_norm": 0.13412128703491133,
+      "learning_rate": 2e-05,
+      "loss": 5.5711,
+      "step": 102
+    },
+    {
+      "epoch": 0.0069088104101686955,
+      "grad_norm": 0.13271313772769053,
+      "learning_rate": 2e-05,
+      "loss": 5.634,
+      "step": 103
+    },
+    {
+      "epoch": 0.006975886239393635,
+      "grad_norm": 0.13641168267393963,
+      "learning_rate": 2e-05,
+      "loss": 5.5622,
+      "step": 104
+    },
+    {
+      "epoch": 0.007042962068618573,
+      "grad_norm": 0.1364249978035869,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 105
+    },
+    {
+      "epoch": 0.007110037897843512,
+      "grad_norm": 0.13342358315955516,
+      "learning_rate": 2e-05,
+      "loss": 5.5035,
+      "step": 106
+    },
+    {
+      "epoch": 0.007177113727068451,
+      "grad_norm": 0.14070835799735137,
+      "learning_rate": 2e-05,
+      "loss": 5.5564,
+      "step": 107
+    },
+    {
+      "epoch": 0.00724418955629339,
+      "grad_norm": 0.13328238411806956,
+      "learning_rate": 2e-05,
+      "loss": 5.4184,
+      "step": 108
+    },
+    {
+      "epoch": 0.007311265385518328,
+      "grad_norm": 0.13665788614083904,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 109
+    },
+    {
+      "epoch": 0.0073783412147432675,
+      "grad_norm": 0.13844396321247493,
+      "learning_rate": 2e-05,
+      "loss": 5.6078,
+      "step": 110
+    },
+    {
+      "epoch": 0.007445417043968206,
+      "grad_norm": 0.13675685111602764,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 111
+    },
+    {
+      "epoch": 0.0075124928731931445,
+      "grad_norm": 0.1403964012168049,
+      "learning_rate": 2e-05,
+      "loss": 5.3716,
+      "step": 112
+    },
+    {
+      "epoch": 0.007579568702418084,
+      "grad_norm": 0.13406503259528899,
+      "learning_rate": 2e-05,
+      "loss": 5.4193,
+      "step": 113
+    },
+    {
+      "epoch": 0.007646644531643022,
+      "grad_norm": 0.14028550665961523,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 114
+    },
+    {
+      "epoch": 0.007713720360867961,
+      "grad_norm": 0.13288525597029105,
+      "learning_rate": 2e-05,
+      "loss": 5.5415,
+      "step": 115
+    },
+    {
+      "epoch": 0.0077807961900929,
+      "grad_norm": 0.13356413724016333,
+      "learning_rate": 2e-05,
+      "loss": 5.5839,
+      "step": 116
+    },
+    {
+      "epoch": 0.00784787201931784,
+      "grad_norm": 0.13313072816239066,
+      "learning_rate": 2e-05,
+      "loss": 5.535,
+      "step": 117
+    },
+    {
+      "epoch": 0.007914947848542778,
+      "grad_norm": 0.14387497257630388,
+      "learning_rate": 2e-05,
+      "loss": 5.4184,
+      "step": 118
+    },
+    {
+      "epoch": 0.007982023677767716,
+      "grad_norm": 0.14451243030966593,
+      "learning_rate": 2e-05,
+      "loss": 5.501,
+      "step": 119
+    },
+    {
+      "epoch": 0.008049099506992655,
+      "grad_norm": 0.14232739108862444,
+      "learning_rate": 2e-05,
+      "loss": 5.5989,
+      "step": 120
+    },
+    {
+      "epoch": 0.008116175336217593,
+      "grad_norm": 0.13421891292311736,
+      "learning_rate": 2e-05,
+      "loss": 5.4069,
+      "step": 121
+    },
+    {
+      "epoch": 0.008183251165442532,
+      "grad_norm": 0.13500935263732758,
+      "learning_rate": 2e-05,
+      "loss": 5.5524,
+      "step": 122
+    },
+    {
+      "epoch": 0.008250326994667472,
+      "grad_norm": 0.14099518588784138,
+      "learning_rate": 2e-05,
+      "loss": 5.4164,
+      "step": 123
+    },
+    {
+      "epoch": 0.00831740282389241,
+      "grad_norm": 0.15075974471607823,
+      "learning_rate": 2e-05,
+      "loss": 5.486,
+      "step": 124
+    },
+    {
+      "epoch": 0.00838447865311735,
+      "grad_norm": 0.13783896093144649,
+      "learning_rate": 2e-05,
+      "loss": 5.5206,
+      "step": 125
+    },
+    {
+      "epoch": 0.008451554482342288,
+      "grad_norm": 0.1344609185715249,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 126
+    },
+    {
+      "epoch": 0.008518630311567226,
+      "grad_norm": 0.138208337990743,
+      "learning_rate": 2e-05,
+      "loss": 5.4815,
+      "step": 127
+    },
+    {
+      "epoch": 0.008585706140792166,
+      "grad_norm": 0.1393046417378179,
+      "learning_rate": 2e-05,
+      "loss": 5.4908,
+      "step": 128
+    },
+    {
+      "epoch": 0.008652781970017105,
+      "grad_norm": 0.1373988230937368,
+      "learning_rate": 2e-05,
+      "loss": 5.5015,
+      "step": 129
+    },
+    {
+      "epoch": 0.008719857799242043,
+      "grad_norm": 0.13285684151509292,
+      "learning_rate": 2e-05,
+      "loss": 5.5537,
+      "step": 130
+    },
+    {
+      "epoch": 0.008786933628466982,
+      "grad_norm": 0.13873899198440434,
+      "learning_rate": 2e-05,
+      "loss": 5.4701,
+      "step": 131
+    },
+    {
+      "epoch": 0.00885400945769192,
+      "grad_norm": 0.14461458297305468,
+      "learning_rate": 2e-05,
+      "loss": 5.5739,
+      "step": 132
+    },
+    {
+      "epoch": 0.008921085286916859,
+      "grad_norm": 0.1405940558904193,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 133
+    },
+    {
+      "epoch": 0.008988161116141799,
+      "grad_norm": 0.14286201262384682,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 134
+    },
+    {
+      "epoch": 0.009055236945366738,
+      "grad_norm": 0.13628491120224992,
+      "learning_rate": 2e-05,
+      "loss": 5.5413,
+      "step": 135
+    },
+    {
+      "epoch": 0.009122312774591676,
+      "grad_norm": 0.13920410242894715,
+      "learning_rate": 2e-05,
+      "loss": 5.5742,
+      "step": 136
+    },
+    {
+      "epoch": 0.009189388603816614,
+      "grad_norm": 0.15057892765641118,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 137
+    },
+    {
+      "epoch": 0.009256464433041553,
+      "grad_norm": 0.13250022431579417,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 138
+    },
+    {
+      "epoch": 0.009323540262266491,
+      "grad_norm": 0.13147301419695562,
+      "learning_rate": 2e-05,
+      "loss": 5.4712,
+      "step": 139
+    },
+    {
+      "epoch": 0.009390616091491432,
+      "grad_norm": 0.1412971415975928,
+      "learning_rate": 2e-05,
+      "loss": 5.5326,
+      "step": 140
+    },
+    {
+      "epoch": 0.00945769192071637,
+      "grad_norm": 0.13276962627372726,
+      "learning_rate": 2e-05,
+      "loss": 5.4725,
+      "step": 141
+    },
+    {
+      "epoch": 0.009524767749941309,
+      "grad_norm": 0.13525965020820085,
+      "learning_rate": 2e-05,
+      "loss": 5.5783,
+      "step": 142
+    },
+    {
+      "epoch": 0.009591843579166247,
+      "grad_norm": 0.13470426157166554,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 143
+    },
+    {
+      "epoch": 0.009658919408391186,
+      "grad_norm": 0.1387203005348155,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 144
+    },
+    {
+      "epoch": 0.009725995237616126,
+      "grad_norm": 0.13336201227369365,
+      "learning_rate": 2e-05,
+      "loss": 5.5642,
+      "step": 145
+    },
+    {
+      "epoch": 0.009793071066841064,
+      "grad_norm": 0.13909228255806302,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 146
+    },
+    {
+      "epoch": 0.009860146896066003,
+      "grad_norm": 0.13257100160349114,
+      "learning_rate": 2e-05,
+      "loss": 5.5663,
+      "step": 147
+    },
+    {
+      "epoch": 0.009927222725290941,
+      "grad_norm": 0.12991537279833248,
+      "learning_rate": 2e-05,
+      "loss": 5.4793,
+      "step": 148
+    },
+    {
+      "epoch": 0.00999429855451588,
+      "grad_norm": 0.13389375820706778,
+      "learning_rate": 2e-05,
+      "loss": 5.545,
+      "step": 149
+    },
+    {
+      "epoch": 0.010061374383740818,
+      "grad_norm": 0.13125567956322495,
+      "learning_rate": 2e-05,
+      "loss": 5.597,
+      "step": 150
+    },
+    {
+      "epoch": 0.010128450212965759,
+      "grad_norm": 0.13741817412973506,
+      "learning_rate": 2e-05,
+      "loss": 5.5423,
+      "step": 151
+    },
+    {
+      "epoch": 0.010195526042190697,
+      "grad_norm": 0.13251126955987788,
+      "learning_rate": 2e-05,
+      "loss": 5.6468,
+      "step": 152
+    },
+    {
+      "epoch": 0.010262601871415635,
+      "grad_norm": 0.13831460050034583,
+      "learning_rate": 2e-05,
+      "loss": 5.6094,
+      "step": 153
+    },
+    {
+      "epoch": 0.010329677700640574,
+      "grad_norm": 0.13343651134367654,
+      "learning_rate": 2e-05,
+      "loss": 5.5306,
+      "step": 154
+    },
+    {
+      "epoch": 0.010396753529865512,
+      "grad_norm": 0.13831155697627867,
+      "learning_rate": 2e-05,
+      "loss": 5.541,
+      "step": 155
+    },
+    {
+      "epoch": 0.010463829359090451,
+      "grad_norm": 0.14066572771824454,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 156
+    },
+    {
+      "epoch": 0.010530905188315391,
+      "grad_norm": 0.13792681398040846,
+      "learning_rate": 2e-05,
+      "loss": 5.6223,
+      "step": 157
+    },
+    {
+      "epoch": 0.01059798101754033,
+      "grad_norm": 0.13170243667857998,
+      "learning_rate": 2e-05,
+      "loss": 5.4,
+      "step": 158
+    },
+    {
+      "epoch": 0.010665056846765268,
+      "grad_norm": 0.1380771037852864,
+      "learning_rate": 2e-05,
+      "loss": 5.5203,
+      "step": 159
+    },
+    {
+      "epoch": 0.010732132675990207,
+      "grad_norm": 0.1433553427121612,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 160
+    },
+    {
+      "epoch": 0.010799208505215145,
+      "grad_norm": 0.13479448057617996,
+      "learning_rate": 2e-05,
+      "loss": 5.5257,
+      "step": 161
+    },
+    {
+      "epoch": 0.010866284334440085,
+      "grad_norm": 0.13159147735932109,
+      "learning_rate": 2e-05,
+      "loss": 5.522,
+      "step": 162
+    },
+    {
+      "epoch": 0.010933360163665024,
+      "grad_norm": 0.1369316227280448,
+      "learning_rate": 2e-05,
+      "loss": 5.5891,
+      "step": 163
+    },
+    {
+      "epoch": 0.011000435992889962,
+      "grad_norm": 0.14076615488304237,
+      "learning_rate": 2e-05,
+      "loss": 5.5571,
+      "step": 164
+    },
+    {
+      "epoch": 0.0110675118221149,
+      "grad_norm": 0.13283102030874372,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 165
+    },
+    {
+      "epoch": 0.01113458765133984,
+      "grad_norm": 0.13325588559324944,
+      "learning_rate": 2e-05,
+      "loss": 5.3261,
+      "step": 166
+    },
+    {
+      "epoch": 0.011201663480564778,
+      "grad_norm": 0.1299736453291738,
+      "learning_rate": 2e-05,
+      "loss": 5.6576,
+      "step": 167
+    },
+    {
+      "epoch": 0.011268739309789718,
+      "grad_norm": 0.13118665201796945,
+      "learning_rate": 2e-05,
+      "loss": 5.3818,
+      "step": 168
+    },
+    {
+      "epoch": 0.011335815139014657,
+      "grad_norm": 0.13653431212504638,
+      "learning_rate": 2e-05,
+      "loss": 5.5639,
+      "step": 169
+    },
+    {
+      "epoch": 0.011402890968239595,
+      "grad_norm": 0.13356055178008214,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 170
+    },
+    {
+      "epoch": 0.011469966797464533,
+      "grad_norm": 0.1354015940184796,
+      "learning_rate": 2e-05,
+      "loss": 5.6393,
+      "step": 171
+    },
+    {
+      "epoch": 0.011537042626689472,
+      "grad_norm": 0.13384596507250812,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 172
+    },
+    {
+      "epoch": 0.01160411845591441,
+      "grad_norm": 0.1335560443666149,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 173
+    },
+    {
+      "epoch": 0.01167119428513935,
+      "grad_norm": 0.1353461241003476,
+      "learning_rate": 2e-05,
+      "loss": 5.3689,
+      "step": 174
+    },
+    {
+      "epoch": 0.01173827011436429,
+      "grad_norm": 0.13310993121743173,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 175
+    },
+    {
+      "epoch": 0.011805345943589228,
+      "grad_norm": 0.14181164663352963,
+      "learning_rate": 2e-05,
+      "loss": 5.4778,
+      "step": 176
+    },
+    {
+      "epoch": 0.011872421772814166,
+      "grad_norm": 0.13123002552600915,
+      "learning_rate": 2e-05,
+      "loss": 5.5789,
+      "step": 177
+    },
+    {
+      "epoch": 0.011939497602039105,
+      "grad_norm": 0.1348766702462207,
+      "learning_rate": 2e-05,
+      "loss": 5.4131,
+      "step": 178
+    },
+    {
+      "epoch": 0.012006573431264045,
+      "grad_norm": 0.13218692198518966,
+      "learning_rate": 2e-05,
+      "loss": 5.5337,
+      "step": 179
+    },
+    {
+      "epoch": 0.012073649260488983,
+      "grad_norm": 0.13463891318118337,
+      "learning_rate": 2e-05,
+      "loss": 5.6098,
+      "step": 180
+    },
+    {
+      "epoch": 0.012140725089713922,
+      "grad_norm": 0.13349456709801627,
+      "learning_rate": 2e-05,
+      "loss": 5.4946,
+      "step": 181
+    },
+    {
+      "epoch": 0.01220780091893886,
+      "grad_norm": 0.13394579668557335,
+      "learning_rate": 2e-05,
+      "loss": 5.4568,
+      "step": 182
+    },
+    {
+      "epoch": 0.012274876748163799,
+      "grad_norm": 0.15453286629635815,
+      "learning_rate": 2e-05,
+      "loss": 5.5621,
+      "step": 183
+    },
+    {
+      "epoch": 0.012341952577388737,
+      "grad_norm": 0.13779559853420217,
+      "learning_rate": 2e-05,
+      "loss": 5.4652,
+      "step": 184
+    },
+    {
+      "epoch": 0.012409028406613678,
+      "grad_norm": 0.13886472144296014,
+      "learning_rate": 2e-05,
+      "loss": 5.5631,
+      "step": 185
+    },
+    {
+      "epoch": 0.012476104235838616,
+      "grad_norm": 0.1496746172659623,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 186
+    },
+    {
+      "epoch": 0.012543180065063554,
+      "grad_norm": 0.14208723005689966,
+      "learning_rate": 2e-05,
+      "loss": 5.6048,
+      "step": 187
+    },
+    {
+      "epoch": 0.012610255894288493,
+      "grad_norm": 0.13166535250834968,
+      "learning_rate": 2e-05,
+      "loss": 5.5068,
+      "step": 188
+    },
+    {
+      "epoch": 0.012677331723513431,
+      "grad_norm": 0.12913640357768716,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 189
+    },
+    {
+      "epoch": 0.01274440755273837,
+      "grad_norm": 0.13757135673412796,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 190
+    },
+    {
+      "epoch": 0.01281148338196331,
+      "grad_norm": 0.13655506597832912,
+      "learning_rate": 2e-05,
+      "loss": 5.4447,
+      "step": 191
+    },
+    {
+      "epoch": 0.012878559211188249,
+      "grad_norm": 0.1375066054180872,
+      "learning_rate": 2e-05,
+      "loss": 5.5012,
+      "step": 192
+    },
+    {
+      "epoch": 0.012945635040413187,
+      "grad_norm": 0.15610482037412207,
+      "learning_rate": 2e-05,
+      "loss": 5.521,
+      "step": 193
+    },
+    {
+      "epoch": 0.013012710869638126,
+      "grad_norm": 0.1367635037421754,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 194
+    },
+    {
+      "epoch": 0.013079786698863064,
+      "grad_norm": 0.1369003034390947,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 195
+    },
+    {
+      "epoch": 0.013146862528088003,
+      "grad_norm": 0.1380794970424235,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 196
+    },
+    {
+      "epoch": 0.013213938357312943,
+      "grad_norm": 0.13373877250050198,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 197
+    },
+    {
+      "epoch": 0.013281014186537881,
+      "grad_norm": 0.14906325760735176,
+      "learning_rate": 2e-05,
+      "loss": 5.5528,
+      "step": 198
+    },
+    {
+      "epoch": 0.01334809001576282,
+      "grad_norm": 0.15025471798102816,
+      "learning_rate": 2e-05,
+      "loss": 5.5543,
+      "step": 199
+    },
+    {
+      "epoch": 0.013415165844987758,
+      "grad_norm": 0.14253606046227107,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 200
+    },
+    {
+      "epoch": 0.013482241674212697,
+      "grad_norm": 0.13288747521549985,
+      "learning_rate": 2e-05,
+      "loss": 5.5899,
+      "step": 201
+    },
+    {
+      "epoch": 0.013549317503437637,
+      "grad_norm": 0.13694648081151692,
+      "learning_rate": 2e-05,
+      "loss": 5.5104,
+      "step": 202
+    },
+    {
+      "epoch": 0.013616393332662576,
+      "grad_norm": 0.13940060362621973,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 203
+    },
+    {
+      "epoch": 0.013683469161887514,
+      "grad_norm": 0.1551342233966156,
+      "learning_rate": 2e-05,
+      "loss": 5.4671,
+      "step": 204
+    },
+    {
+      "epoch": 0.013750544991112452,
+      "grad_norm": 0.13798361523391936,
+      "learning_rate": 2e-05,
+      "loss": 5.5551,
+      "step": 205
+    },
+    {
+      "epoch": 0.013817620820337391,
+      "grad_norm": 0.14112620854433694,
+      "learning_rate": 2e-05,
+      "loss": 5.5151,
+      "step": 206
+    },
+    {
+      "epoch": 0.01388469664956233,
+      "grad_norm": 0.14368054643929828,
+      "learning_rate": 2e-05,
+      "loss": 5.5385,
+      "step": 207
+    },
+    {
+      "epoch": 0.01395177247878727,
+      "grad_norm": 0.13492428125481046,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 208
+    },
+    {
+      "epoch": 0.014018848308012208,
+      "grad_norm": 0.1359437016514597,
+      "learning_rate": 2e-05,
+      "loss": 5.4821,
+      "step": 209
+    },
+    {
+      "epoch": 0.014085924137237147,
+      "grad_norm": 0.1422595482933864,
+      "learning_rate": 2e-05,
+      "loss": 5.6255,
+      "step": 210
+    },
+    {
+      "epoch": 0.014152999966462085,
+      "grad_norm": 0.13721004058447053,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 211
+    },
+    {
+      "epoch": 0.014220075795687024,
+      "grad_norm": 0.1395545772069129,
+      "learning_rate": 2e-05,
+      "loss": 5.5288,
+      "step": 212
+    },
+    {
+      "epoch": 0.014287151624911962,
+      "grad_norm": 0.12836343210391182,
+      "learning_rate": 2e-05,
+      "loss": 5.544,
+      "step": 213
+    },
+    {
+      "epoch": 0.014354227454136902,
+      "grad_norm": 0.15185543409605295,
+      "learning_rate": 2e-05,
+      "loss": 5.5624,
+      "step": 214
+    },
+    {
+      "epoch": 0.01442130328336184,
+      "grad_norm": 0.13534643795646575,
+      "learning_rate": 2e-05,
+      "loss": 5.5096,
+      "step": 215
+    },
+    {
+      "epoch": 0.01448837911258678,
+      "grad_norm": 0.13470870363624546,
+      "learning_rate": 2e-05,
+      "loss": 5.627,
+      "step": 216
+    },
+    {
+      "epoch": 0.014555454941811718,
+      "grad_norm": 0.1360087244048781,
+      "learning_rate": 2e-05,
+      "loss": 5.5024,
+      "step": 217
+    },
+    {
+      "epoch": 0.014622530771036656,
+      "grad_norm": 0.14024312273251197,
+      "learning_rate": 2e-05,
+      "loss": 5.5826,
+      "step": 218
+    },
+    {
+      "epoch": 0.014689606600261597,
+      "grad_norm": 0.1376420471114841,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 219
+    },
+    {
+      "epoch": 0.014756682429486535,
+      "grad_norm": 0.139656095792711,
+      "learning_rate": 2e-05,
+      "loss": 5.407,
+      "step": 220
+    },
+    {
+      "epoch": 0.014823758258711473,
+      "grad_norm": 0.1349880464633375,
+      "learning_rate": 2e-05,
+      "loss": 5.5401,
+      "step": 221
+    },
+    {
+      "epoch": 0.014890834087936412,
+      "grad_norm": 0.1435660108693425,
+      "learning_rate": 2e-05,
+      "loss": 5.5442,
+      "step": 222
+    },
+    {
+      "epoch": 0.01495790991716135,
+      "grad_norm": 0.13520150286281724,
+      "learning_rate": 2e-05,
+      "loss": 5.5113,
+      "step": 223
+    },
+    {
+      "epoch": 0.015024985746386289,
+      "grad_norm": 0.13382236365345912,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 224
+    },
+    {
+      "epoch": 0.01509206157561123,
+      "grad_norm": 0.1376846345964698,
+      "learning_rate": 2e-05,
+      "loss": 5.5181,
+      "step": 225
+    },
+    {
+      "epoch": 0.015159137404836168,
+      "grad_norm": 0.14407632997232372,
+      "learning_rate": 2e-05,
+      "loss": 5.5495,
+      "step": 226
+    },
+    {
+      "epoch": 0.015226213234061106,
+      "grad_norm": 0.14018701078938195,
+      "learning_rate": 2e-05,
+      "loss": 5.3041,
+      "step": 227
+    },
+    {
+      "epoch": 0.015293289063286045,
+      "grad_norm": 0.1322376057261667,
+      "learning_rate": 2e-05,
+      "loss": 5.5604,
+      "step": 228
+    },
+    {
+      "epoch": 0.015360364892510983,
+      "grad_norm": 0.1345121512748179,
+      "learning_rate": 2e-05,
+      "loss": 5.5028,
+      "step": 229
+    },
+    {
+      "epoch": 0.015427440721735922,
+      "grad_norm": 0.13198319874561015,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 230
+    },
+    {
+      "epoch": 0.015494516550960862,
+      "grad_norm": 0.13474593983822677,
+      "learning_rate": 2e-05,
+      "loss": 5.5025,
+      "step": 231
+    },
+    {
+      "epoch": 0.0155615923801858,
+      "grad_norm": 0.13388538276090686,
+      "learning_rate": 2e-05,
+      "loss": 5.5226,
+      "step": 232
+    },
+    {
+      "epoch": 0.01562866820941074,
+      "grad_norm": 0.1347084236920801,
+      "learning_rate": 2e-05,
+      "loss": 5.4754,
+      "step": 233
+    },
+    {
+      "epoch": 0.01569574403863568,
+      "grad_norm": 0.1379395338535944,
+      "learning_rate": 2e-05,
+      "loss": 5.549,
+      "step": 234
+    },
+    {
+      "epoch": 0.015762819867860618,
+      "grad_norm": 0.13428076935551114,
+      "learning_rate": 2e-05,
+      "loss": 5.5567,
+      "step": 235
+    },
+    {
+      "epoch": 0.015829895697085556,
+      "grad_norm": 0.13790232793247126,
+      "learning_rate": 2e-05,
+      "loss": 5.5109,
+      "step": 236
+    },
+    {
+      "epoch": 0.015896971526310495,
+      "grad_norm": 0.13042042688748032,
+      "learning_rate": 2e-05,
+      "loss": 5.5566,
+      "step": 237
+    },
+    {
+      "epoch": 0.015964047355535433,
+      "grad_norm": 0.1330663519900655,
+      "learning_rate": 2e-05,
+      "loss": 5.4465,
+      "step": 238
+    },
+    {
+      "epoch": 0.01603112318476037,
+      "grad_norm": 0.13385255377819233,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 239
+    },
+    {
+      "epoch": 0.01609819901398531,
+      "grad_norm": 0.12993697207416155,
+      "learning_rate": 2e-05,
+      "loss": 5.5471,
+      "step": 240
+    },
+    {
+      "epoch": 0.01616527484321025,
+      "grad_norm": 0.1396024743841848,
+      "learning_rate": 2e-05,
+      "loss": 5.6155,
+      "step": 241
+    },
+    {
+      "epoch": 0.016232350672435187,
+      "grad_norm": 0.13989956866798828,
+      "learning_rate": 2e-05,
+      "loss": 5.3373,
+      "step": 242
+    },
+    {
+      "epoch": 0.016299426501660125,
+      "grad_norm": 0.13976501449136522,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 243
+    },
+    {
+      "epoch": 0.016366502330885064,
+      "grad_norm": 0.13165882002358276,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 244
+    },
+    {
+      "epoch": 0.016433578160110006,
+      "grad_norm": 0.13488722967310984,
+      "learning_rate": 2e-05,
+      "loss": 5.6121,
+      "step": 245
+    },
+    {
+      "epoch": 0.016500653989334944,
+      "grad_norm": 0.1336981550894159,
+      "learning_rate": 2e-05,
+      "loss": 5.5589,
+      "step": 246
+    },
+    {
+      "epoch": 0.016567729818559883,
+      "grad_norm": 0.1440148401620789,
+      "learning_rate": 2e-05,
+      "loss": 5.3725,
+      "step": 247
+    },
+    {
+      "epoch": 0.01663480564778482,
+      "grad_norm": 0.1338616154763599,
+      "learning_rate": 2e-05,
+      "loss": 5.5469,
+      "step": 248
+    },
+    {
+      "epoch": 0.01670188147700976,
+      "grad_norm": 0.13554269305080055,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 249
+    },
+    {
+      "epoch": 0.0167689573062347,
+      "grad_norm": 0.14681796374747594,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 250
+    },
+    {
+      "epoch": 0.016836033135459637,
+      "grad_norm": 0.13703003196924554,
+      "learning_rate": 2e-05,
+      "loss": 5.5859,
+      "step": 251
+    },
+    {
+      "epoch": 0.016903108964684575,
+      "grad_norm": 0.14008819175775072,
+      "learning_rate": 2e-05,
+      "loss": 5.6353,
+      "step": 252
+    },
+    {
+      "epoch": 0.016970184793909514,
+      "grad_norm": 0.13891502890897356,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 253
+    },
+    {
+      "epoch": 0.017037260623134452,
+      "grad_norm": 0.13924088981573363,
+      "learning_rate": 2e-05,
+      "loss": 5.5485,
+      "step": 254
+    },
+    {
+      "epoch": 0.01710433645235939,
+      "grad_norm": 0.13673406685267592,
+      "learning_rate": 2e-05,
+      "loss": 5.3589,
+      "step": 255
+    },
+    {
+      "epoch": 0.017171412281584333,
+      "grad_norm": 0.13320380977224103,
+      "learning_rate": 2e-05,
+      "loss": 5.5116,
+      "step": 256
+    },
+    {
+      "epoch": 0.01723848811080927,
+      "grad_norm": 0.1410488335732908,
+      "learning_rate": 2e-05,
+      "loss": 5.4918,
+      "step": 257
+    },
+    {
+      "epoch": 0.01730556394003421,
+      "grad_norm": 0.13298435533886696,
+      "learning_rate": 2e-05,
+      "loss": 5.3208,
+      "step": 258
+    },
+    {
+      "epoch": 0.017372639769259148,
+      "grad_norm": 0.13988287243353253,
+      "learning_rate": 2e-05,
+      "loss": 5.4744,
+      "step": 259
+    },
+    {
+      "epoch": 0.017439715598484087,
+      "grad_norm": 0.14053031175611752,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 260
+    },
+    {
+      "epoch": 0.017506791427709025,
+      "grad_norm": 0.13757472499958526,
+      "learning_rate": 2e-05,
+      "loss": 5.5816,
+      "step": 261
+    },
+    {
+      "epoch": 0.017573867256933964,
+      "grad_norm": 0.13749854484353066,
+      "learning_rate": 2e-05,
+      "loss": 5.4426,
+      "step": 262
+    },
+    {
+      "epoch": 0.017640943086158902,
+      "grad_norm": 0.13923202465722923,
+      "learning_rate": 2e-05,
+      "loss": 5.5661,
+      "step": 263
+    },
+    {
+      "epoch": 0.01770801891538384,
+      "grad_norm": 0.1358334843327064,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 264
+    },
+    {
+      "epoch": 0.01777509474460878,
+      "grad_norm": 0.13869653732550644,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 265
+    },
+    {
+      "epoch": 0.017842170573833718,
+      "grad_norm": 0.14308039464116132,
+      "learning_rate": 2e-05,
+      "loss": 5.475,
+      "step": 266
+    },
+    {
+      "epoch": 0.017909246403058656,
+      "grad_norm": 0.14289446053175728,
+      "learning_rate": 2e-05,
+      "loss": 5.3703,
+      "step": 267
+    },
+    {
+      "epoch": 0.017976322232283598,
+      "grad_norm": 0.13954656437029866,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 268
+    },
+    {
+      "epoch": 0.018043398061508537,
+      "grad_norm": 0.13770164428473664,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 269
+    },
+    {
+      "epoch": 0.018110473890733475,
+      "grad_norm": 0.13593856485860595,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 270
+    },
+    {
+      "epoch": 0.018177549719958414,
+      "grad_norm": 0.1378056846456878,
+      "learning_rate": 2e-05,
+      "loss": 5.4144,
+      "step": 271
+    },
+    {
+      "epoch": 0.018244625549183352,
+      "grad_norm": 0.14677081976064846,
+      "learning_rate": 2e-05,
+      "loss": 5.5786,
+      "step": 272
+    },
+    {
+      "epoch": 0.01831170137840829,
+      "grad_norm": 0.1333997160128434,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 273
+    },
+    {
+      "epoch": 0.01837877720763323,
+      "grad_norm": 0.1334431474865435,
+      "learning_rate": 2e-05,
+      "loss": 5.5029,
+      "step": 274
+    },
+    {
+      "epoch": 0.018445853036858167,
+      "grad_norm": 0.14112418165203575,
+      "learning_rate": 2e-05,
+      "loss": 5.4597,
+      "step": 275
+    },
+    {
+      "epoch": 0.018512928866083106,
+      "grad_norm": 0.14547318226177025,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 276
+    },
+    {
+      "epoch": 0.018580004695308044,
+      "grad_norm": 0.14044227296642572,
+      "learning_rate": 2e-05,
+      "loss": 5.3426,
+      "step": 277
+    },
+    {
+      "epoch": 0.018647080524532983,
+      "grad_norm": 0.13315530974306822,
+      "learning_rate": 2e-05,
+      "loss": 5.5762,
+      "step": 278
+    },
+    {
+      "epoch": 0.018714156353757925,
+      "grad_norm": 0.13927811230900217,
+      "learning_rate": 2e-05,
+      "loss": 5.4824,
+      "step": 279
+    },
+    {
+      "epoch": 0.018781232182982863,
+      "grad_norm": 0.13636044567814218,
+      "learning_rate": 2e-05,
+      "loss": 5.3487,
+      "step": 280
+    },
+    {
+      "epoch": 0.018848308012207802,
+      "grad_norm": 0.1393901566987908,
+      "learning_rate": 2e-05,
+      "loss": 5.5384,
+      "step": 281
+    },
+    {
+      "epoch": 0.01891538384143274,
+      "grad_norm": 0.14015357810907544,
+      "learning_rate": 2e-05,
+      "loss": 5.5201,
+      "step": 282
+    },
+    {
+      "epoch": 0.01898245967065768,
+      "grad_norm": 0.13503157673832053,
+      "learning_rate": 2e-05,
+      "loss": 5.4651,
+      "step": 283
+    },
+    {
+      "epoch": 0.019049535499882617,
+      "grad_norm": 0.13649017002960714,
+      "learning_rate": 2e-05,
+      "loss": 5.6387,
+      "step": 284
+    },
+    {
+      "epoch": 0.019116611329107556,
+      "grad_norm": 0.14050803821643626,
+      "learning_rate": 2e-05,
+      "loss": 5.3334,
+      "step": 285
+    },
+    {
+      "epoch": 0.019183687158332494,
+      "grad_norm": 0.13275357391343542,
+      "learning_rate": 2e-05,
+      "loss": 5.6031,
+      "step": 286
+    },
+    {
+      "epoch": 0.019250762987557433,
+      "grad_norm": 0.13513545725223935,
+      "learning_rate": 2e-05,
+      "loss": 5.5803,
+      "step": 287
+    },
+    {
+      "epoch": 0.01931783881678237,
+      "grad_norm": 0.13942077888191706,
+      "learning_rate": 2e-05,
+      "loss": 5.5644,
+      "step": 288
+    },
+    {
+      "epoch": 0.01938491464600731,
+      "grad_norm": 0.1317177502262839,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 289
+    },
+    {
+      "epoch": 0.01945199047523225,
+      "grad_norm": 0.14018213104286506,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 290
+    },
+    {
+      "epoch": 0.01951906630445719,
+      "grad_norm": 0.1302789741300026,
+      "learning_rate": 2e-05,
+      "loss": 5.4,
+      "step": 291
+    },
+    {
+      "epoch": 0.01958614213368213,
+      "grad_norm": 0.13573053088025885,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 292
+    },
+    {
+      "epoch": 0.019653217962907067,
+      "grad_norm": 0.1298532622388913,
+      "learning_rate": 2e-05,
+      "loss": 5.5202,
+      "step": 293
+    },
+    {
+      "epoch": 0.019720293792132006,
+      "grad_norm": 0.1387368585183225,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 294
+    },
+    {
+      "epoch": 0.019787369621356944,
+      "grad_norm": 0.13286945880633255,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 295
+    },
+    {
+      "epoch": 0.019854445450581883,
+      "grad_norm": 0.13460463802703496,
+      "learning_rate": 2e-05,
+      "loss": 5.6098,
+      "step": 296
+    },
+    {
+      "epoch": 0.01992152127980682,
+      "grad_norm": 0.13266495300108314,
+      "learning_rate": 2e-05,
+      "loss": 5.5612,
+      "step": 297
+    },
+    {
+      "epoch": 0.01998859710903176,
+      "grad_norm": 0.1390013145487187,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 298
+    },
+    {
+      "epoch": 0.020055672938256698,
+      "grad_norm": 0.13036775755700042,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 299
+    },
+    {
+      "epoch": 0.020122748767481637,
+      "grad_norm": 0.1351561306033971,
+      "learning_rate": 2e-05,
+      "loss": 5.5295,
+      "step": 300
+    },
+    {
+      "epoch": 0.020189824596706575,
+      "grad_norm": 0.13068843395631943,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 301
+    },
+    {
+      "epoch": 0.020256900425931517,
+      "grad_norm": 0.13486391073328413,
+      "learning_rate": 2e-05,
+      "loss": 5.4127,
+      "step": 302
+    },
+    {
+      "epoch": 0.020323976255156456,
+      "grad_norm": 0.14064396051095973,
+      "learning_rate": 2e-05,
+      "loss": 5.4195,
+      "step": 303
+    },
+    {
+      "epoch": 0.020391052084381394,
+      "grad_norm": 0.14354945962603494,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 304
+    },
+    {
+      "epoch": 0.020458127913606333,
+      "grad_norm": 0.14099937008619579,
+      "learning_rate": 2e-05,
+      "loss": 5.3195,
+      "step": 305
+    },
+    {
+      "epoch": 0.02052520374283127,
+      "grad_norm": 0.13190265555710654,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 306
+    },
+    {
+      "epoch": 0.02059227957205621,
+      "grad_norm": 0.13464959866358975,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 307
+    },
+    {
+      "epoch": 0.020659355401281148,
+      "grad_norm": 0.1367295230234042,
+      "learning_rate": 2e-05,
+      "loss": 5.5423,
+      "step": 308
+    },
+    {
+      "epoch": 0.020726431230506086,
+      "grad_norm": 0.13914460779187957,
+      "learning_rate": 2e-05,
+      "loss": 5.6471,
+      "step": 309
+    },
+    {
+      "epoch": 0.020793507059731025,
+      "grad_norm": 0.1332943654584558,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 310
+    },
+    {
+      "epoch": 0.020860582888955963,
+      "grad_norm": 0.12766112720947653,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 311
+    },
+    {
+      "epoch": 0.020927658718180902,
+      "grad_norm": 0.13588223042602018,
+      "learning_rate": 2e-05,
+      "loss": 5.4093,
+      "step": 312
+    },
+    {
+      "epoch": 0.020994734547405844,
+      "grad_norm": 0.14136836470907532,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 313
+    },
+    {
+      "epoch": 0.021061810376630782,
+      "grad_norm": 0.1393880208383422,
+      "learning_rate": 2e-05,
+      "loss": 5.581,
+      "step": 314
+    },
+    {
+      "epoch": 0.02112888620585572,
+      "grad_norm": 0.13164162031839544,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 315
+    },
+    {
+      "epoch": 0.02119596203508066,
+      "grad_norm": 0.14010905667045967,
+      "learning_rate": 2e-05,
+      "loss": 5.6313,
+      "step": 316
+    },
+    {
+      "epoch": 0.021263037864305598,
+      "grad_norm": 0.14526862125687154,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 317
+    },
+    {
+      "epoch": 0.021330113693530536,
+      "grad_norm": 0.14093553081245067,
+      "learning_rate": 2e-05,
+      "loss": 5.528,
+      "step": 318
+    },
+    {
+      "epoch": 0.021397189522755475,
+      "grad_norm": 0.1358165356915093,
+      "learning_rate": 2e-05,
+      "loss": 5.4931,
+      "step": 319
+    },
+    {
+      "epoch": 0.021464265351980413,
+      "grad_norm": 0.14073772926918005,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 320
+    },
+    {
+      "epoch": 0.021531341181205352,
+      "grad_norm": 0.13652977553817378,
+      "learning_rate": 2e-05,
+      "loss": 5.4774,
+      "step": 321
+    },
+    {
+      "epoch": 0.02159841701043029,
+      "grad_norm": 0.14055402703216985,
+      "learning_rate": 2e-05,
+      "loss": 5.4304,
+      "step": 322
+    },
+    {
+      "epoch": 0.02166549283965523,
+      "grad_norm": 0.13776074331038693,
+      "learning_rate": 2e-05,
+      "loss": 5.4917,
+      "step": 323
+    },
+    {
+      "epoch": 0.02173256866888017,
+      "grad_norm": 0.14047674601848228,
+      "learning_rate": 2e-05,
+      "loss": 5.491,
+      "step": 324
+    },
+    {
+      "epoch": 0.02179964449810511,
+      "grad_norm": 0.13892914026490785,
+      "learning_rate": 2e-05,
+      "loss": 5.4963,
+      "step": 325
+    },
+    {
+      "epoch": 0.021866720327330048,
+      "grad_norm": 0.1382115436600762,
+      "learning_rate": 2e-05,
+      "loss": 5.5455,
+      "step": 326
+    },
+    {
+      "epoch": 0.021933796156554986,
+      "grad_norm": 0.13932573949417545,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 327
+    },
+    {
+      "epoch": 0.022000871985779925,
+      "grad_norm": 0.14107431052446137,
+      "learning_rate": 2e-05,
+      "loss": 5.4786,
+      "step": 328
+    },
+    {
+      "epoch": 0.022067947815004863,
+      "grad_norm": 0.13943872404016952,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 329
+    },
+    {
+      "epoch": 0.0221350236442298,
+      "grad_norm": 0.1417150365329029,
+      "learning_rate": 2e-05,
+      "loss": 5.5967,
+      "step": 330
+    },
+    {
+      "epoch": 0.02220209947345474,
+      "grad_norm": 0.1422942300595576,
+      "learning_rate": 2e-05,
+      "loss": 5.4733,
+      "step": 331
+    },
+    {
+      "epoch": 0.02226917530267968,
+      "grad_norm": 0.13801404901597958,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 332
+    },
+    {
+      "epoch": 0.022336251131904617,
+      "grad_norm": 0.14334325976263998,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 333
+    },
+    {
+      "epoch": 0.022403326961129556,
+      "grad_norm": 0.13734915017630128,
+      "learning_rate": 2e-05,
+      "loss": 5.669,
+      "step": 334
+    },
+    {
+      "epoch": 0.022470402790354494,
+      "grad_norm": 0.13302604328883835,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 335
+    },
+    {
+      "epoch": 0.022537478619579436,
+      "grad_norm": 0.13585129129115542,
+      "learning_rate": 2e-05,
+      "loss": 5.4531,
+      "step": 336
+    },
+    {
+      "epoch": 0.022604554448804375,
+      "grad_norm": 0.1478783628958156,
+      "learning_rate": 2e-05,
+      "loss": 5.3855,
+      "step": 337
+    },
+    {
+      "epoch": 0.022671630278029313,
+      "grad_norm": 0.14580947668953922,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 338
+    },
+    {
+      "epoch": 0.02273870610725425,
+      "grad_norm": 0.13237653135928842,
+      "learning_rate": 2e-05,
+      "loss": 5.493,
+      "step": 339
+    },
+    {
+      "epoch": 0.02280578193647919,
+      "grad_norm": 0.13720980295398222,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 340
+    },
+    {
+      "epoch": 0.02287285776570413,
+      "grad_norm": 0.13860990487488198,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 341
+    },
+    {
+      "epoch": 0.022939933594929067,
+      "grad_norm": 0.13744223044929282,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 342
+    },
+    {
+      "epoch": 0.023007009424154005,
+      "grad_norm": 0.13584639927592604,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 343
+    },
+    {
+      "epoch": 0.023074085253378944,
+      "grad_norm": 0.13395627502911653,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 344
+    },
+    {
+      "epoch": 0.023141161082603882,
+      "grad_norm": 0.13944307274388457,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 345
+    },
+    {
+      "epoch": 0.02320823691182882,
+      "grad_norm": 0.1398135732658755,
+      "learning_rate": 2e-05,
+      "loss": 5.47,
+      "step": 346
+    },
+    {
+      "epoch": 0.023275312741053763,
+      "grad_norm": 0.1404305164744192,
+      "learning_rate": 2e-05,
+      "loss": 5.4382,
+      "step": 347
+    },
+    {
+      "epoch": 0.0233423885702787,
+      "grad_norm": 0.13787092697983222,
+      "learning_rate": 2e-05,
+      "loss": 5.4405,
+      "step": 348
+    },
+    {
+      "epoch": 0.02340946439950364,
+      "grad_norm": 0.14044190484129454,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 349
+    },
+    {
+      "epoch": 0.02347654022872858,
+      "grad_norm": 0.1366848205371932,
+      "learning_rate": 2e-05,
+      "loss": 5.4977,
+      "step": 350
+    },
+    {
+      "epoch": 0.023543616057953517,
+      "grad_norm": 0.13577791584019544,
+      "learning_rate": 2e-05,
+      "loss": 5.3942,
+      "step": 351
+    },
+    {
+      "epoch": 0.023610691887178455,
+      "grad_norm": 0.1406626418408333,
+      "learning_rate": 2e-05,
+      "loss": 5.3176,
+      "step": 352
+    },
+    {
+      "epoch": 0.023677767716403394,
+      "grad_norm": 0.1354885761881078,
+      "learning_rate": 2e-05,
+      "loss": 5.6091,
+      "step": 353
+    },
+    {
+      "epoch": 0.023744843545628332,
+      "grad_norm": 0.1329458810127426,
+      "learning_rate": 2e-05,
+      "loss": 5.4475,
+      "step": 354
+    },
+    {
+      "epoch": 0.02381191937485327,
+      "grad_norm": 0.1396865316754293,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 355
+    },
+    {
+      "epoch": 0.02387899520407821,
+      "grad_norm": 0.13921113920897119,
+      "learning_rate": 2e-05,
+      "loss": 5.3942,
+      "step": 356
+    },
+    {
+      "epoch": 0.023946071033303148,
+      "grad_norm": 0.13167856019188462,
+      "learning_rate": 2e-05,
+      "loss": 5.4746,
+      "step": 357
+    },
+    {
+      "epoch": 0.02401314686252809,
+      "grad_norm": 0.1364284648900273,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 358
+    },
+    {
+      "epoch": 0.024080222691753028,
+      "grad_norm": 0.13555591249091348,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 359
+    },
+    {
+      "epoch": 0.024147298520977967,
+      "grad_norm": 0.14056587633398115,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 360
+    },
+    {
+      "epoch": 0.024214374350202905,
+      "grad_norm": 0.13581930293444056,
+      "learning_rate": 2e-05,
+      "loss": 5.3824,
+      "step": 361
+    },
+    {
+      "epoch": 0.024281450179427844,
+      "grad_norm": 0.1373287534406947,
+      "learning_rate": 2e-05,
+      "loss": 5.545,
+      "step": 362
+    },
+    {
+      "epoch": 0.024348526008652782,
+      "grad_norm": 0.1446479521870619,
+      "learning_rate": 2e-05,
+      "loss": 5.6046,
+      "step": 363
+    },
+    {
+      "epoch": 0.02441560183787772,
+      "grad_norm": 0.14134614280417196,
+      "learning_rate": 2e-05,
+      "loss": 5.4769,
+      "step": 364
+    },
+    {
+      "epoch": 0.02448267766710266,
+      "grad_norm": 0.13703265868085493,
+      "learning_rate": 2e-05,
+      "loss": 5.5608,
+      "step": 365
+    },
+    {
+      "epoch": 0.024549753496327598,
+      "grad_norm": 0.14173530188852576,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 366
+    },
+    {
+      "epoch": 0.024616829325552536,
+      "grad_norm": 0.13537935824715508,
+      "learning_rate": 2e-05,
+      "loss": 5.5082,
+      "step": 367
+    },
+    {
+      "epoch": 0.024683905154777475,
+      "grad_norm": 0.13640541237042714,
+      "learning_rate": 2e-05,
+      "loss": 5.5298,
+      "step": 368
+    },
+    {
+      "epoch": 0.024750980984002413,
+      "grad_norm": 0.1424194501022318,
+      "learning_rate": 2e-05,
+      "loss": 5.5066,
+      "step": 369
+    },
+    {
+      "epoch": 0.024818056813227355,
+      "grad_norm": 0.13785459031620195,
+      "learning_rate": 2e-05,
+      "loss": 5.4622,
+      "step": 370
+    },
+    {
+      "epoch": 0.024885132642452294,
+      "grad_norm": 0.1496021987416581,
+      "learning_rate": 2e-05,
+      "loss": 5.5762,
+      "step": 371
+    },
+    {
+      "epoch": 0.024952208471677232,
+      "grad_norm": 0.14758782327148273,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 372
+    },
+    {
+      "epoch": 0.02501928430090217,
+      "grad_norm": 0.14388897365700223,
+      "learning_rate": 2e-05,
+      "loss": 5.5665,
+      "step": 373
+    },
+    {
+      "epoch": 0.02508636013012711,
+      "grad_norm": 0.13748350279103597,
+      "learning_rate": 2e-05,
+      "loss": 5.3806,
+      "step": 374
+    },
+    {
+      "epoch": 0.025153435959352047,
+      "grad_norm": 0.13728178957568166,
+      "learning_rate": 2e-05,
+      "loss": 5.4678,
+      "step": 375
+    },
+    {
+      "epoch": 0.025220511788576986,
+      "grad_norm": 0.15219734643771088,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 376
+    },
+    {
+      "epoch": 0.025287587617801924,
+      "grad_norm": 0.13776093616107832,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 377
+    },
+    {
+      "epoch": 0.025354663447026863,
+      "grad_norm": 0.1323313019917841,
+      "learning_rate": 2e-05,
+      "loss": 5.4284,
+      "step": 378
+    },
+    {
+      "epoch": 0.0254217392762518,
+      "grad_norm": 0.1497693646313545,
+      "learning_rate": 2e-05,
+      "loss": 5.5278,
+      "step": 379
+    },
+    {
+      "epoch": 0.02548881510547674,
+      "grad_norm": 0.13581608804947878,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 380
+    },
+    {
+      "epoch": 0.025555890934701682,
+      "grad_norm": 0.13702156264001938,
+      "learning_rate": 2e-05,
+      "loss": 5.5198,
+      "step": 381
+    },
+    {
+      "epoch": 0.02562296676392662,
+      "grad_norm": 0.1479831250398658,
+      "learning_rate": 2e-05,
+      "loss": 5.4233,
+      "step": 382
+    },
+    {
+      "epoch": 0.02569004259315156,
+      "grad_norm": 0.14023110295322808,
+      "learning_rate": 2e-05,
+      "loss": 5.4498,
+      "step": 383
+    },
+    {
+      "epoch": 0.025757118422376497,
+      "grad_norm": 0.13907866409465663,
+      "learning_rate": 2e-05,
+      "loss": 5.4888,
+      "step": 384
+    },
+    {
+      "epoch": 0.025824194251601436,
+      "grad_norm": 0.13929828284213425,
+      "learning_rate": 2e-05,
+      "loss": 5.4757,
+      "step": 385
+    },
+    {
+      "epoch": 0.025891270080826374,
+      "grad_norm": 0.13525869424923675,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 386
+    },
+    {
+      "epoch": 0.025958345910051313,
+      "grad_norm": 0.1427802337933376,
+      "learning_rate": 2e-05,
+      "loss": 5.6668,
+      "step": 387
+    },
+    {
+      "epoch": 0.02602542173927625,
+      "grad_norm": 0.13126390847250383,
+      "learning_rate": 2e-05,
+      "loss": 5.535,
+      "step": 388
+    },
+    {
+      "epoch": 0.02609249756850119,
+      "grad_norm": 0.14213752720124426,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 389
+    },
+    {
+      "epoch": 0.02615957339772613,
+      "grad_norm": 0.13989910666508423,
+      "learning_rate": 2e-05,
+      "loss": 5.4791,
+      "step": 390
+    },
+    {
+      "epoch": 0.026226649226951067,
+      "grad_norm": 0.13430225912784785,
+      "learning_rate": 2e-05,
+      "loss": 5.3793,
+      "step": 391
+    },
+    {
+      "epoch": 0.026293725056176005,
+      "grad_norm": 0.14461781740490645,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 392
+    },
+    {
+      "epoch": 0.026360800885400947,
+      "grad_norm": 0.14446861354900634,
+      "learning_rate": 2e-05,
+      "loss": 5.3391,
+      "step": 393
+    },
+    {
+      "epoch": 0.026427876714625886,
+      "grad_norm": 0.14356768600109318,
+      "learning_rate": 2e-05,
+      "loss": 5.5977,
+      "step": 394
+    },
+    {
+      "epoch": 0.026494952543850824,
+      "grad_norm": 0.1362516471769211,
+      "learning_rate": 2e-05,
+      "loss": 5.5295,
+      "step": 395
+    },
+    {
+      "epoch": 0.026562028373075763,
+      "grad_norm": 0.15600174473100523,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 396
+    },
+    {
+      "epoch": 0.0266291042023007,
+      "grad_norm": 0.13853467608782233,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 397
+    },
+    {
+      "epoch": 0.02669618003152564,
+      "grad_norm": 0.13542811149476883,
+      "learning_rate": 2e-05,
+      "loss": 5.5862,
+      "step": 398
+    },
+    {
+      "epoch": 0.026763255860750578,
+      "grad_norm": 0.14308228641773624,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 399
+    },
+    {
+      "epoch": 0.026830331689975517,
+      "grad_norm": 0.14388696098957304,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 400
+    },
+    {
+      "epoch": 0.026897407519200455,
+      "grad_norm": 0.13311492221933083,
+      "learning_rate": 2e-05,
+      "loss": 5.6049,
+      "step": 401
+    },
+    {
+      "epoch": 0.026964483348425394,
+      "grad_norm": 0.15068428191963382,
+      "learning_rate": 2e-05,
+      "loss": 5.6092,
+      "step": 402
+    },
+    {
+      "epoch": 0.027031559177650332,
+      "grad_norm": 0.14149604609631475,
+      "learning_rate": 2e-05,
+      "loss": 5.5962,
+      "step": 403
+    },
+    {
+      "epoch": 0.027098635006875274,
+      "grad_norm": 0.13321908144926464,
+      "learning_rate": 2e-05,
+      "loss": 5.2231,
+      "step": 404
+    },
+    {
+      "epoch": 0.027165710836100213,
+      "grad_norm": 0.13859466222445496,
+      "learning_rate": 2e-05,
+      "loss": 5.5049,
+      "step": 405
+    },
+    {
+      "epoch": 0.02723278666532515,
+      "grad_norm": 0.15478622697011363,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 406
+    },
+    {
+      "epoch": 0.02729986249455009,
+      "grad_norm": 0.13575694755907047,
+      "learning_rate": 2e-05,
+      "loss": 5.5904,
+      "step": 407
+    },
+    {
+      "epoch": 0.027366938323775028,
+      "grad_norm": 0.14148167111382318,
+      "learning_rate": 2e-05,
+      "loss": 5.7243,
+      "step": 408
+    },
+    {
+      "epoch": 0.027434014152999966,
+      "grad_norm": 0.14287791773438815,
+      "learning_rate": 2e-05,
+      "loss": 5.4441,
+      "step": 409
+    },
+    {
+      "epoch": 0.027501089982224905,
+      "grad_norm": 0.13619881969751138,
+      "learning_rate": 2e-05,
+      "loss": 5.4259,
+      "step": 410
+    },
+    {
+      "epoch": 0.027568165811449843,
+      "grad_norm": 0.1446271165834697,
+      "learning_rate": 2e-05,
+      "loss": 5.28,
+      "step": 411
+    },
+    {
+      "epoch": 0.027635241640674782,
+      "grad_norm": 0.14155968313445627,
+      "learning_rate": 2e-05,
+      "loss": 5.4784,
+      "step": 412
+    },
+    {
+      "epoch": 0.02770231746989972,
+      "grad_norm": 0.13282673248736107,
+      "learning_rate": 2e-05,
+      "loss": 5.4283,
+      "step": 413
+    },
+    {
+      "epoch": 0.02776939329912466,
+      "grad_norm": 0.13853298832320834,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 414
+    },
+    {
+      "epoch": 0.0278364691283496,
+      "grad_norm": 0.1365273323961682,
+      "learning_rate": 2e-05,
+      "loss": 5.5315,
+      "step": 415
+    },
+    {
+      "epoch": 0.02790354495757454,
+      "grad_norm": 0.13571835930531323,
+      "learning_rate": 2e-05,
+      "loss": 5.3449,
+      "step": 416
+    },
+    {
+      "epoch": 0.027970620786799478,
+      "grad_norm": 0.13469511106014387,
+      "learning_rate": 2e-05,
+      "loss": 5.4744,
+      "step": 417
+    },
+    {
+      "epoch": 0.028037696616024416,
+      "grad_norm": 0.13662464602102403,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 418
+    },
+    {
+      "epoch": 0.028104772445249355,
+      "grad_norm": 0.13581649762198095,
+      "learning_rate": 2e-05,
+      "loss": 5.4875,
+      "step": 419
+    },
+    {
+      "epoch": 0.028171848274474293,
+      "grad_norm": 0.1330851772069556,
+      "learning_rate": 2e-05,
+      "loss": 5.4274,
+      "step": 420
+    },
+    {
+      "epoch": 0.028238924103699232,
+      "grad_norm": 0.14282359284803453,
+      "learning_rate": 2e-05,
+      "loss": 5.5418,
+      "step": 421
+    },
+    {
+      "epoch": 0.02830599993292417,
+      "grad_norm": 0.1342083795099212,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 422
+    },
+    {
+      "epoch": 0.02837307576214911,
+      "grad_norm": 0.13482010494998253,
+      "learning_rate": 2e-05,
+      "loss": 5.6325,
+      "step": 423
+    },
+    {
+      "epoch": 0.028440151591374047,
+      "grad_norm": 0.1394152554487557,
+      "learning_rate": 2e-05,
+      "loss": 5.5611,
+      "step": 424
+    },
+    {
+      "epoch": 0.028507227420598986,
+      "grad_norm": 0.1399139304195542,
+      "learning_rate": 2e-05,
+      "loss": 5.6259,
+      "step": 425
+    },
+    {
+      "epoch": 0.028574303249823924,
+      "grad_norm": 0.1392994407976783,
+      "learning_rate": 2e-05,
+      "loss": 5.3537,
+      "step": 426
+    },
+    {
+      "epoch": 0.028641379079048866,
+      "grad_norm": 0.13396664094567637,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 427
+    },
+    {
+      "epoch": 0.028708454908273805,
+      "grad_norm": 0.13541701659867753,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 428
+    },
+    {
+      "epoch": 0.028775530737498743,
+      "grad_norm": 0.13497212300887781,
+      "learning_rate": 2e-05,
+      "loss": 5.3372,
+      "step": 429
+    },
+    {
+      "epoch": 0.02884260656672368,
+      "grad_norm": 0.13949643353385802,
+      "learning_rate": 2e-05,
+      "loss": 5.5618,
+      "step": 430
+    },
+    {
+      "epoch": 0.02890968239594862,
+      "grad_norm": 0.13940262215484756,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 431
+    },
+    {
+      "epoch": 0.02897675822517356,
+      "grad_norm": 0.14071923837378633,
+      "learning_rate": 2e-05,
+      "loss": 5.4975,
+      "step": 432
+    },
+    {
+      "epoch": 0.029043834054398497,
+      "grad_norm": 0.14205565868695072,
+      "learning_rate": 2e-05,
+      "loss": 5.4597,
+      "step": 433
+    },
+    {
+      "epoch": 0.029110909883623436,
+      "grad_norm": 0.135123026600456,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 434
+    },
+    {
+      "epoch": 0.029177985712848374,
+      "grad_norm": 0.13233426812366825,
+      "learning_rate": 2e-05,
+      "loss": 5.4244,
+      "step": 435
+    },
+    {
+      "epoch": 0.029245061542073313,
+      "grad_norm": 0.1347367216940183,
+      "learning_rate": 2e-05,
+      "loss": 5.5574,
+      "step": 436
+    },
+    {
+      "epoch": 0.02931213737129825,
+      "grad_norm": 0.13488120836842218,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 437
+    },
+    {
+      "epoch": 0.029379213200523193,
+      "grad_norm": 0.13741477915485753,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 438
+    },
+    {
+      "epoch": 0.02944628902974813,
+      "grad_norm": 0.13611619541380157,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 439
+    },
+    {
+      "epoch": 0.02951336485897307,
+      "grad_norm": 0.13931356000837053,
+      "learning_rate": 2e-05,
+      "loss": 5.4267,
+      "step": 440
+    },
+    {
+      "epoch": 0.02958044068819801,
+      "grad_norm": 0.14033129427387336,
+      "learning_rate": 2e-05,
+      "loss": 5.5426,
+      "step": 441
+    },
+    {
+      "epoch": 0.029647516517422947,
+      "grad_norm": 0.14237050493804299,
+      "learning_rate": 2e-05,
+      "loss": 5.5924,
+      "step": 442
+    },
+    {
+      "epoch": 0.029714592346647885,
+      "grad_norm": 0.13698646809414747,
+      "learning_rate": 2e-05,
+      "loss": 5.5143,
+      "step": 443
+    },
+    {
+      "epoch": 0.029781668175872824,
+      "grad_norm": 0.1441271803700368,
+      "learning_rate": 2e-05,
+      "loss": 5.6055,
+      "step": 444
+    },
+    {
+      "epoch": 0.029848744005097762,
+      "grad_norm": 0.13818238967287785,
+      "learning_rate": 2e-05,
+      "loss": 5.6279,
+      "step": 445
+    },
+    {
+      "epoch": 0.0299158198343227,
+      "grad_norm": 0.1396905579089585,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 446
+    },
+    {
+      "epoch": 0.02998289566354764,
+      "grad_norm": 0.1346278932208077,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 447
+    },
+    {
+      "epoch": 0.030049971492772578,
+      "grad_norm": 0.1472701382188289,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 448
+    },
+    {
+      "epoch": 0.03011704732199752,
+      "grad_norm": 0.13620646142772422,
+      "learning_rate": 2e-05,
+      "loss": 5.4278,
+      "step": 449
+    },
+    {
+      "epoch": 0.03018412315122246,
+      "grad_norm": 0.139769840144812,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 450
+    },
+    {
+      "epoch": 0.030251198980447397,
+      "grad_norm": 0.13431662015387102,
+      "learning_rate": 2e-05,
+      "loss": 5.3964,
+      "step": 451
+    },
+    {
+      "epoch": 0.030318274809672335,
+      "grad_norm": 0.13614222957040845,
+      "learning_rate": 2e-05,
+      "loss": 5.5446,
+      "step": 452
+    },
+    {
+      "epoch": 0.030385350638897274,
+      "grad_norm": 0.13434201916745456,
+      "learning_rate": 2e-05,
+      "loss": 5.5751,
+      "step": 453
+    },
+    {
+      "epoch": 0.030452426468122212,
+      "grad_norm": 0.13559922939738983,
+      "learning_rate": 2e-05,
+      "loss": 5.4403,
+      "step": 454
+    },
+    {
+      "epoch": 0.03051950229734715,
+      "grad_norm": 0.1384197423953293,
+      "learning_rate": 2e-05,
+      "loss": 5.5235,
+      "step": 455
+    },
+    {
+      "epoch": 0.03058657812657209,
+      "grad_norm": 0.13907468567870393,
+      "learning_rate": 2e-05,
+      "loss": 5.4258,
+      "step": 456
+    },
+    {
+      "epoch": 0.030653653955797028,
+      "grad_norm": 0.13083304836010035,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 457
+    },
+    {
+      "epoch": 0.030720729785021966,
+      "grad_norm": 0.13721686400553276,
+      "learning_rate": 2e-05,
+      "loss": 5.5321,
+      "step": 458
+    },
+    {
+      "epoch": 0.030787805614246905,
+      "grad_norm": 0.14075765171865992,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 459
+    },
+    {
+      "epoch": 0.030854881443471843,
+      "grad_norm": 0.13036528177136278,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 460
+    },
+    {
+      "epoch": 0.030921957272696785,
+      "grad_norm": 0.13855385060783146,
+      "learning_rate": 2e-05,
+      "loss": 5.5771,
+      "step": 461
+    },
+    {
+      "epoch": 0.030989033101921724,
+      "grad_norm": 0.1371557251618741,
+      "learning_rate": 2e-05,
+      "loss": 5.5021,
+      "step": 462
+    },
+    {
+      "epoch": 0.031056108931146662,
+      "grad_norm": 0.14539671600075732,
+      "learning_rate": 2e-05,
+      "loss": 5.4447,
+      "step": 463
+    },
+    {
+      "epoch": 0.0311231847603716,
+      "grad_norm": 0.14279042916220416,
+      "learning_rate": 2e-05,
+      "loss": 5.434,
+      "step": 464
+    },
+    {
+      "epoch": 0.03119026058959654,
+      "grad_norm": 0.1319844368777511,
+      "learning_rate": 2e-05,
+      "loss": 5.5692,
+      "step": 465
+    },
+    {
+      "epoch": 0.03125733641882148,
+      "grad_norm": 0.1383976266576286,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 466
+    },
+    {
+      "epoch": 0.03132441224804642,
+      "grad_norm": 0.13656057789942108,
+      "learning_rate": 2e-05,
+      "loss": 5.5564,
+      "step": 467
+    },
+    {
+      "epoch": 0.03139148807727136,
+      "grad_norm": 0.13778414743801168,
+      "learning_rate": 2e-05,
+      "loss": 5.4971,
+      "step": 468
+    },
+    {
+      "epoch": 0.0314585639064963,
+      "grad_norm": 0.13625650077260326,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 469
+    },
+    {
+      "epoch": 0.031525639735721235,
+      "grad_norm": 0.13728643391748266,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 470
+    },
+    {
+      "epoch": 0.031592715564946174,
+      "grad_norm": 0.13580328335771577,
+      "learning_rate": 2e-05,
+      "loss": 5.4821,
+      "step": 471
+    },
+    {
+      "epoch": 0.03165979139417111,
+      "grad_norm": 0.14140290882949022,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 472
+    },
+    {
+      "epoch": 0.03172686722339605,
+      "grad_norm": 0.14291201743267656,
+      "learning_rate": 2e-05,
+      "loss": 5.5432,
+      "step": 473
+    },
+    {
+      "epoch": 0.03179394305262099,
+      "grad_norm": 0.1436189244558032,
+      "learning_rate": 2e-05,
+      "loss": 5.4005,
+      "step": 474
+    },
+    {
+      "epoch": 0.03186101888184593,
+      "grad_norm": 0.13630264220904464,
+      "learning_rate": 2e-05,
+      "loss": 5.4826,
+      "step": 475
+    },
+    {
+      "epoch": 0.031928094711070866,
+      "grad_norm": 0.13495990310862327,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 476
+    },
+    {
+      "epoch": 0.031995170540295804,
+      "grad_norm": 0.1478330990886607,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 477
+    },
+    {
+      "epoch": 0.03206224636952074,
+      "grad_norm": 0.14829139583159276,
+      "learning_rate": 2e-05,
+      "loss": 5.5358,
+      "step": 478
+    },
+    {
+      "epoch": 0.03212932219874568,
+      "grad_norm": 0.13624472672881807,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 479
+    },
+    {
+      "epoch": 0.03219639802797062,
+      "grad_norm": 0.1477171651804662,
+      "learning_rate": 2e-05,
+      "loss": 5.4897,
+      "step": 480
+    },
+    {
+      "epoch": 0.03226347385719556,
+      "grad_norm": 0.13663957269957958,
+      "learning_rate": 2e-05,
+      "loss": 5.3581,
+      "step": 481
+    },
+    {
+      "epoch": 0.0323305496864205,
+      "grad_norm": 0.14135385647848828,
+      "learning_rate": 2e-05,
+      "loss": 5.6269,
+      "step": 482
+    },
+    {
+      "epoch": 0.032397625515645435,
+      "grad_norm": 0.13618749920416148,
+      "learning_rate": 2e-05,
+      "loss": 5.5629,
+      "step": 483
+    },
+    {
+      "epoch": 0.032464701344870374,
+      "grad_norm": 0.14598840683373648,
+      "learning_rate": 2e-05,
+      "loss": 5.5184,
+      "step": 484
+    },
+    {
+      "epoch": 0.03253177717409531,
+      "grad_norm": 0.14044988424340765,
+      "learning_rate": 2e-05,
+      "loss": 5.5307,
+      "step": 485
+    },
+    {
+      "epoch": 0.03259885300332025,
+      "grad_norm": 0.14063398831330848,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 486
+    },
+    {
+      "epoch": 0.03266592883254519,
+      "grad_norm": 0.14329169807758987,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 487
+    },
+    {
+      "epoch": 0.03273300466177013,
+      "grad_norm": 0.13955101041545528,
+      "learning_rate": 2e-05,
+      "loss": 5.4437,
+      "step": 488
+    },
+    {
+      "epoch": 0.03280008049099507,
+      "grad_norm": 0.14266305942757573,
+      "learning_rate": 2e-05,
+      "loss": 5.4165,
+      "step": 489
+    },
+    {
+      "epoch": 0.03286715632022001,
+      "grad_norm": 0.14106744970060633,
+      "learning_rate": 2e-05,
+      "loss": 5.5754,
+      "step": 490
+    },
+    {
+      "epoch": 0.03293423214944495,
+      "grad_norm": 0.14370074349163833,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 491
+    },
+    {
+      "epoch": 0.03300130797866989,
+      "grad_norm": 0.13963738976655854,
+      "learning_rate": 2e-05,
+      "loss": 5.54,
+      "step": 492
+    },
+    {
+      "epoch": 0.03306838380789483,
+      "grad_norm": 0.14146186981111028,
+      "learning_rate": 2e-05,
+      "loss": 5.4907,
+      "step": 493
+    },
+    {
+      "epoch": 0.033135459637119766,
+      "grad_norm": 0.1447318132736114,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 494
+    },
+    {
+      "epoch": 0.033202535466344704,
+      "grad_norm": 0.13850270496055583,
+      "learning_rate": 2e-05,
+      "loss": 5.4863,
+      "step": 495
+    },
+    {
+      "epoch": 0.03326961129556964,
+      "grad_norm": 0.14216069408613088,
+      "learning_rate": 2e-05,
+      "loss": 5.5329,
+      "step": 496
+    },
+    {
+      "epoch": 0.03333668712479458,
+      "grad_norm": 0.14427267437471356,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 497
+    },
+    {
+      "epoch": 0.03340376295401952,
+      "grad_norm": 0.13932140836676593,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 498
+    },
+    {
+      "epoch": 0.03347083878324446,
+      "grad_norm": 0.14367394276047166,
+      "learning_rate": 2e-05,
+      "loss": 5.4106,
+      "step": 499
+    },
+    {
+      "epoch": 0.0335379146124694,
+      "grad_norm": 0.1393874677318847,
+      "learning_rate": 2e-05,
+      "loss": 5.309,
+      "step": 500
+    },
+    {
+      "epoch": 0.033604990441694335,
+      "grad_norm": 0.14488851081875279,
+      "learning_rate": 2e-05,
+      "loss": 5.5565,
+      "step": 501
+    },
+    {
+      "epoch": 0.033672066270919274,
+      "grad_norm": 0.14793721090457707,
+      "learning_rate": 2e-05,
+      "loss": 5.363,
+      "step": 502
+    },
+    {
+      "epoch": 0.03373914210014421,
+      "grad_norm": 0.14146585464550626,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 503
+    },
+    {
+      "epoch": 0.03380621792936915,
+      "grad_norm": 0.13726726554930982,
+      "learning_rate": 2e-05,
+      "loss": 5.467,
+      "step": 504
+    },
+    {
+      "epoch": 0.03387329375859409,
+      "grad_norm": 0.1418633669922156,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 505
+    },
+    {
+      "epoch": 0.03394036958781903,
+      "grad_norm": 0.14688269695131473,
+      "learning_rate": 2e-05,
+      "loss": 5.5817,
+      "step": 506
+    },
+    {
+      "epoch": 0.034007445417043966,
+      "grad_norm": 0.139009095401521,
+      "learning_rate": 2e-05,
+      "loss": 5.3049,
+      "step": 507
+    },
+    {
+      "epoch": 0.034074521246268905,
+      "grad_norm": 0.14080696309834667,
+      "learning_rate": 2e-05,
+      "loss": 5.3923,
+      "step": 508
+    },
+    {
+      "epoch": 0.03414159707549384,
+      "grad_norm": 0.14123243394184842,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 509
+    },
+    {
+      "epoch": 0.03420867290471878,
+      "grad_norm": 0.14580127846579277,
+      "learning_rate": 2e-05,
+      "loss": 5.4411,
+      "step": 510
+    },
+    {
+      "epoch": 0.03427574873394372,
+      "grad_norm": 0.13350034824740667,
+      "learning_rate": 2e-05,
+      "loss": 5.2895,
+      "step": 511
+    },
+    {
+      "epoch": 0.034342824563168665,
+      "grad_norm": 0.13647000906202034,
+      "learning_rate": 2e-05,
+      "loss": 5.558,
+      "step": 512
+    },
+    {
+      "epoch": 0.034409900392393604,
+      "grad_norm": 0.13860675813577458,
+      "learning_rate": 2e-05,
+      "loss": 5.6502,
+      "step": 513
+    },
+    {
+      "epoch": 0.03447697622161854,
+      "grad_norm": 0.14524770125419068,
+      "learning_rate": 2e-05,
+      "loss": 5.5042,
+      "step": 514
+    },
+    {
+      "epoch": 0.03454405205084348,
+      "grad_norm": 0.14320345407262589,
+      "learning_rate": 2e-05,
+      "loss": 5.598,
+      "step": 515
+    },
+    {
+      "epoch": 0.03461112788006842,
+      "grad_norm": 0.13415672032083029,
+      "learning_rate": 2e-05,
+      "loss": 5.3863,
+      "step": 516
+    },
+    {
+      "epoch": 0.03467820370929336,
+      "grad_norm": 0.13179397485156813,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 517
+    },
+    {
+      "epoch": 0.034745279538518296,
+      "grad_norm": 0.14782105129313314,
+      "learning_rate": 2e-05,
+      "loss": 5.5203,
+      "step": 518
+    },
+    {
+      "epoch": 0.034812355367743235,
+      "grad_norm": 0.1422382968685032,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 519
+    },
+    {
+      "epoch": 0.03487943119696817,
+      "grad_norm": 0.13652408341433767,
+      "learning_rate": 2e-05,
+      "loss": 5.5191,
+      "step": 520
+    },
+    {
+      "epoch": 0.03494650702619311,
+      "grad_norm": 0.14191914614395676,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 521
+    },
+    {
+      "epoch": 0.03501358285541805,
+      "grad_norm": 0.14747583217972665,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 522
+    },
+    {
+      "epoch": 0.03508065868464299,
+      "grad_norm": 0.14136956577302243,
+      "learning_rate": 2e-05,
+      "loss": 5.3528,
+      "step": 523
+    },
+    {
+      "epoch": 0.03514773451386793,
+      "grad_norm": 0.1395719243790696,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 524
+    },
+    {
+      "epoch": 0.035214810343092866,
+      "grad_norm": 0.14389336811451958,
+      "learning_rate": 2e-05,
+      "loss": 5.4952,
+      "step": 525
+    },
+    {
+      "epoch": 0.035281886172317804,
+      "grad_norm": 0.14260740116233941,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 526
+    },
+    {
+      "epoch": 0.03534896200154274,
+      "grad_norm": 0.1348277216130678,
+      "learning_rate": 2e-05,
+      "loss": 5.3601,
+      "step": 527
+    },
+    {
+      "epoch": 0.03541603783076768,
+      "grad_norm": 0.14549668705189578,
+      "learning_rate": 2e-05,
+      "loss": 5.435,
+      "step": 528
+    },
+    {
+      "epoch": 0.03548311365999262,
+      "grad_norm": 0.1456737965017188,
+      "learning_rate": 2e-05,
+      "loss": 5.3944,
+      "step": 529
+    },
+    {
+      "epoch": 0.03555018948921756,
+      "grad_norm": 0.1356501642805981,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 530
+    },
+    {
+      "epoch": 0.0356172653184425,
+      "grad_norm": 0.14044511571468438,
+      "learning_rate": 2e-05,
+      "loss": 5.4358,
+      "step": 531
+    },
+    {
+      "epoch": 0.035684341147667435,
+      "grad_norm": 0.1481717153953622,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 532
+    },
+    {
+      "epoch": 0.035751416976892374,
+      "grad_norm": 0.13991426802378085,
+      "learning_rate": 2e-05,
+      "loss": 5.5592,
+      "step": 533
+    },
+    {
+      "epoch": 0.03581849280611731,
+      "grad_norm": 0.14385081832078775,
+      "learning_rate": 2e-05,
+      "loss": 5.5402,
+      "step": 534
+    },
+    {
+      "epoch": 0.03588556863534226,
+      "grad_norm": 0.13806055097884395,
+      "learning_rate": 2e-05,
+      "loss": 5.3843,
+      "step": 535
+    },
+    {
+      "epoch": 0.035952644464567196,
+      "grad_norm": 0.1427953199256475,
+      "learning_rate": 2e-05,
+      "loss": 5.3584,
+      "step": 536
+    },
+    {
+      "epoch": 0.036019720293792135,
+      "grad_norm": 0.13306325262425062,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 537
+    },
+    {
+      "epoch": 0.03608679612301707,
+      "grad_norm": 0.1408053056341624,
+      "learning_rate": 2e-05,
+      "loss": 5.6113,
+      "step": 538
+    },
+    {
+      "epoch": 0.03615387195224201,
+      "grad_norm": 0.1387742849455146,
+      "learning_rate": 2e-05,
+      "loss": 5.4181,
+      "step": 539
+    },
+    {
+      "epoch": 0.03622094778146695,
+      "grad_norm": 0.13643333235157482,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 540
+    },
+    {
+      "epoch": 0.03628802361069189,
+      "grad_norm": 0.13658421853188527,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 541
+    },
+    {
+      "epoch": 0.03635509943991683,
+      "grad_norm": 0.1435077876125611,
+      "learning_rate": 2e-05,
+      "loss": 5.3278,
+      "step": 542
+    },
+    {
+      "epoch": 0.036422175269141766,
+      "grad_norm": 0.13239409817519127,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 543
+    },
+    {
+      "epoch": 0.036489251098366704,
+      "grad_norm": 0.1451562967232421,
+      "learning_rate": 2e-05,
+      "loss": 5.414,
+      "step": 544
+    },
+    {
+      "epoch": 0.03655632692759164,
+      "grad_norm": 0.14682351561486612,
+      "learning_rate": 2e-05,
+      "loss": 5.3347,
+      "step": 545
+    },
+    {
+      "epoch": 0.03662340275681658,
+      "grad_norm": 0.13232980864620708,
+      "learning_rate": 2e-05,
+      "loss": 5.5624,
+      "step": 546
+    },
+    {
+      "epoch": 0.03669047858604152,
+      "grad_norm": 0.1434220840238176,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 547
+    },
+    {
+      "epoch": 0.03675755441526646,
+      "grad_norm": 0.14219675784544605,
+      "learning_rate": 2e-05,
+      "loss": 5.482,
+      "step": 548
+    },
+    {
+      "epoch": 0.036824630244491396,
+      "grad_norm": 0.14336874812697367,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 549
+    },
+    {
+      "epoch": 0.036891706073716335,
+      "grad_norm": 0.13536108402329605,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 550
+    },
+    {
+      "epoch": 0.03695878190294127,
+      "grad_norm": 0.14007259563567018,
+      "learning_rate": 2e-05,
+      "loss": 5.5944,
+      "step": 551
+    },
+    {
+      "epoch": 0.03702585773216621,
+      "grad_norm": 0.13942402023153103,
+      "learning_rate": 2e-05,
+      "loss": 5.5577,
+      "step": 552
+    },
+    {
+      "epoch": 0.03709293356139115,
+      "grad_norm": 0.1337767835563611,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 553
+    },
+    {
+      "epoch": 0.03716000939061609,
+      "grad_norm": 0.13960829021769286,
+      "learning_rate": 2e-05,
+      "loss": 5.5737,
+      "step": 554
+    },
+    {
+      "epoch": 0.03722708521984103,
+      "grad_norm": 0.1411448232472383,
+      "learning_rate": 2e-05,
+      "loss": 5.5698,
+      "step": 555
+    },
+    {
+      "epoch": 0.037294161049065966,
+      "grad_norm": 0.14256740146277125,
+      "learning_rate": 2e-05,
+      "loss": 5.4772,
+      "step": 556
+    },
+    {
+      "epoch": 0.03736123687829091,
+      "grad_norm": 0.13822539715550178,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 557
+    },
+    {
+      "epoch": 0.03742831270751585,
+      "grad_norm": 0.14097323248362797,
+      "learning_rate": 2e-05,
+      "loss": 5.5028,
+      "step": 558
+    },
+    {
+      "epoch": 0.03749538853674079,
+      "grad_norm": 0.1386736355812705,
+      "learning_rate": 2e-05,
+      "loss": 5.555,
+      "step": 559
+    },
+    {
+      "epoch": 0.03756246436596573,
+      "grad_norm": 0.13891426035135243,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 560
+    },
+    {
+      "epoch": 0.037629540195190665,
+      "grad_norm": 0.14606355783798464,
+      "learning_rate": 2e-05,
+      "loss": 5.4714,
+      "step": 561
+    },
+    {
+      "epoch": 0.037696616024415604,
+      "grad_norm": 0.1321918596749185,
+      "learning_rate": 2e-05,
+      "loss": 5.5481,
+      "step": 562
+    },
+    {
+      "epoch": 0.03776369185364054,
+      "grad_norm": 0.1351852769680835,
+      "learning_rate": 2e-05,
+      "loss": 5.4913,
+      "step": 563
+    },
+    {
+      "epoch": 0.03783076768286548,
+      "grad_norm": 0.1405655204708842,
+      "learning_rate": 2e-05,
+      "loss": 5.3741,
+      "step": 564
+    },
+    {
+      "epoch": 0.03789784351209042,
+      "grad_norm": 0.13636462962121554,
+      "learning_rate": 2e-05,
+      "loss": 5.4713,
+      "step": 565
+    },
+    {
+      "epoch": 0.03796491934131536,
+      "grad_norm": 0.13661207553159538,
+      "learning_rate": 2e-05,
+      "loss": 5.5205,
+      "step": 566
+    },
+    {
+      "epoch": 0.038031995170540296,
+      "grad_norm": 0.13980690103939206,
+      "learning_rate": 2e-05,
+      "loss": 5.2878,
+      "step": 567
+    },
+    {
+      "epoch": 0.038099070999765235,
+      "grad_norm": 0.13959467913809961,
+      "learning_rate": 2e-05,
+      "loss": 5.5552,
+      "step": 568
+    },
+    {
+      "epoch": 0.03816614682899017,
+      "grad_norm": 0.1396055052539937,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 569
+    },
+    {
+      "epoch": 0.03823322265821511,
+      "grad_norm": 0.13193367937539396,
+      "learning_rate": 2e-05,
+      "loss": 5.5218,
+      "step": 570
+    },
+    {
+      "epoch": 0.03830029848744005,
+      "grad_norm": 0.13664550463911304,
+      "learning_rate": 2e-05,
+      "loss": 5.395,
+      "step": 571
+    },
+    {
+      "epoch": 0.03836737431666499,
+      "grad_norm": 0.13473126239557717,
+      "learning_rate": 2e-05,
+      "loss": 5.4586,
+      "step": 572
+    },
+    {
+      "epoch": 0.03843445014588993,
+      "grad_norm": 0.13164093228191315,
+      "learning_rate": 2e-05,
+      "loss": 5.4066,
+      "step": 573
+    },
+    {
+      "epoch": 0.038501525975114866,
+      "grad_norm": 0.13871402916426678,
+      "learning_rate": 2e-05,
+      "loss": 5.5186,
+      "step": 574
+    },
+    {
+      "epoch": 0.038568601804339804,
+      "grad_norm": 0.137519720494307,
+      "learning_rate": 2e-05,
+      "loss": 5.4445,
+      "step": 575
+    },
+    {
+      "epoch": 0.03863567763356474,
+      "grad_norm": 0.1355004671118137,
+      "learning_rate": 2e-05,
+      "loss": 5.5667,
+      "step": 576
+    },
+    {
+      "epoch": 0.03870275346278968,
+      "grad_norm": 0.14232664894195857,
+      "learning_rate": 2e-05,
+      "loss": 5.6236,
+      "step": 577
+    },
+    {
+      "epoch": 0.03876982929201462,
+      "grad_norm": 0.1493222915824619,
+      "learning_rate": 2e-05,
+      "loss": 5.3458,
+      "step": 578
+    },
+    {
+      "epoch": 0.03883690512123956,
+      "grad_norm": 0.13948887892224626,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 579
+    },
+    {
+      "epoch": 0.0389039809504645,
+      "grad_norm": 0.14402416350878355,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 580
+    },
+    {
+      "epoch": 0.03897105677968944,
+      "grad_norm": 0.14358669852650835,
+      "learning_rate": 2e-05,
+      "loss": 5.454,
+      "step": 581
+    },
+    {
+      "epoch": 0.03903813260891438,
+      "grad_norm": 0.13756924562596545,
+      "learning_rate": 2e-05,
+      "loss": 5.485,
+      "step": 582
+    },
+    {
+      "epoch": 0.03910520843813932,
+      "grad_norm": 0.1513360583916843,
+      "learning_rate": 2e-05,
+      "loss": 5.3952,
+      "step": 583
+    },
+    {
+      "epoch": 0.03917228426736426,
+      "grad_norm": 0.14091786816383903,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 584
+    },
+    {
+      "epoch": 0.039239360096589196,
+      "grad_norm": 0.1379135124585367,
+      "learning_rate": 2e-05,
+      "loss": 5.2596,
+      "step": 585
+    },
+    {
+      "epoch": 0.039306435925814134,
+      "grad_norm": 0.14466392612894816,
+      "learning_rate": 2e-05,
+      "loss": 5.6167,
+      "step": 586
+    },
+    {
+      "epoch": 0.03937351175503907,
+      "grad_norm": 0.14579166466124044,
+      "learning_rate": 2e-05,
+      "loss": 5.3335,
+      "step": 587
+    },
+    {
+      "epoch": 0.03944058758426401,
+      "grad_norm": 0.13616253236203202,
+      "learning_rate": 2e-05,
+      "loss": 5.3037,
+      "step": 588
+    },
+    {
+      "epoch": 0.03950766341348895,
+      "grad_norm": 0.13847137904328194,
+      "learning_rate": 2e-05,
+      "loss": 5.565,
+      "step": 589
+    },
+    {
+      "epoch": 0.03957473924271389,
+      "grad_norm": 0.1575811213534679,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 590
+    },
+    {
+      "epoch": 0.03964181507193883,
+      "grad_norm": 0.1410527173501351,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 591
+    },
+    {
+      "epoch": 0.039708890901163765,
+      "grad_norm": 0.13665623475849864,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 592
+    },
+    {
+      "epoch": 0.039775966730388704,
+      "grad_norm": 0.1338411862483947,
+      "learning_rate": 2e-05,
+      "loss": 5.5681,
+      "step": 593
+    },
+    {
+      "epoch": 0.03984304255961364,
+      "grad_norm": 0.14850151022804023,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 594
+    },
+    {
+      "epoch": 0.03991011838883858,
+      "grad_norm": 0.1449621747735679,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 595
+    },
+    {
+      "epoch": 0.03997719421806352,
+      "grad_norm": 0.13747463700667578,
+      "learning_rate": 2e-05,
+      "loss": 5.5933,
+      "step": 596
+    },
+    {
+      "epoch": 0.04004427004728846,
+      "grad_norm": 0.14247869152400375,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 597
+    },
+    {
+      "epoch": 0.040111345876513396,
+      "grad_norm": 0.14545732042052903,
+      "learning_rate": 2e-05,
+      "loss": 5.586,
+      "step": 598
+    },
+    {
+      "epoch": 0.040178421705738335,
+      "grad_norm": 0.14037656881537025,
+      "learning_rate": 2e-05,
+      "loss": 5.4353,
+      "step": 599
+    },
+    {
+      "epoch": 0.04024549753496327,
+      "grad_norm": 0.13890638141933984,
+      "learning_rate": 2e-05,
+      "loss": 5.5707,
+      "step": 600
+    },
+    {
+      "epoch": 0.04031257336418821,
+      "grad_norm": 0.14043548923391092,
+      "learning_rate": 2e-05,
+      "loss": 5.5406,
+      "step": 601
+    },
+    {
+      "epoch": 0.04037964919341315,
+      "grad_norm": 0.1410720869213143,
+      "learning_rate": 2e-05,
+      "loss": 5.4818,
+      "step": 602
+    },
+    {
+      "epoch": 0.040446725022638096,
+      "grad_norm": 0.14205292654560056,
+      "learning_rate": 2e-05,
+      "loss": 5.4011,
+      "step": 603
+    },
+    {
+      "epoch": 0.040513800851863034,
+      "grad_norm": 0.1405367716706814,
+      "learning_rate": 2e-05,
+      "loss": 5.4069,
+      "step": 604
+    },
+    {
+      "epoch": 0.04058087668108797,
+      "grad_norm": 0.1305995345486736,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 605
+    },
+    {
+      "epoch": 0.04064795251031291,
+      "grad_norm": 0.1438338887327888,
+      "learning_rate": 2e-05,
+      "loss": 5.5348,
+      "step": 606
+    },
+    {
+      "epoch": 0.04071502833953785,
+      "grad_norm": 0.1447802627138278,
+      "learning_rate": 2e-05,
+      "loss": 5.4614,
+      "step": 607
+    },
+    {
+      "epoch": 0.04078210416876279,
+      "grad_norm": 0.1386422869968883,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 608
+    },
+    {
+      "epoch": 0.04084917999798773,
+      "grad_norm": 0.14330857590760096,
+      "learning_rate": 2e-05,
+      "loss": 5.6255,
+      "step": 609
+    },
+    {
+      "epoch": 0.040916255827212665,
+      "grad_norm": 0.1482156192113495,
+      "learning_rate": 2e-05,
+      "loss": 5.5517,
+      "step": 610
+    },
+    {
+      "epoch": 0.040983331656437604,
+      "grad_norm": 0.1369358971036775,
+      "learning_rate": 2e-05,
+      "loss": 5.5285,
+      "step": 611
+    },
+    {
+      "epoch": 0.04105040748566254,
+      "grad_norm": 0.13821679904084444,
+      "learning_rate": 2e-05,
+      "loss": 5.5294,
+      "step": 612
+    },
+    {
+      "epoch": 0.04111748331488748,
+      "grad_norm": 0.14144677981017342,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 613
+    },
+    {
+      "epoch": 0.04118455914411242,
+      "grad_norm": 0.13824273464898712,
+      "learning_rate": 2e-05,
+      "loss": 5.5659,
+      "step": 614
+    },
+    {
+      "epoch": 0.04125163497333736,
+      "grad_norm": 0.14061187164486597,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 615
+    },
+    {
+      "epoch": 0.041318710802562296,
+      "grad_norm": 0.14095205119779503,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 616
+    },
+    {
+      "epoch": 0.041385786631787234,
+      "grad_norm": 0.14319504628953475,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 617
+    },
+    {
+      "epoch": 0.04145286246101217,
+      "grad_norm": 0.13836943767229448,
+      "learning_rate": 2e-05,
+      "loss": 5.4719,
+      "step": 618
+    },
+    {
+      "epoch": 0.04151993829023711,
+      "grad_norm": 0.13512784000399458,
+      "learning_rate": 2e-05,
+      "loss": 5.5451,
+      "step": 619
+    },
+    {
+      "epoch": 0.04158701411946205,
+      "grad_norm": 0.14189427112027952,
+      "learning_rate": 2e-05,
+      "loss": 5.6001,
+      "step": 620
+    },
+    {
+      "epoch": 0.04165408994868699,
+      "grad_norm": 0.1403513312690403,
+      "learning_rate": 2e-05,
+      "loss": 5.5655,
+      "step": 621
+    },
+    {
+      "epoch": 0.04172116577791193,
+      "grad_norm": 0.13945650528846082,
+      "learning_rate": 2e-05,
+      "loss": 5.4681,
+      "step": 622
+    },
+    {
+      "epoch": 0.041788241607136865,
+      "grad_norm": 0.13987003530335432,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 623
+    },
+    {
+      "epoch": 0.041855317436361804,
+      "grad_norm": 0.14274062855701655,
+      "learning_rate": 2e-05,
+      "loss": 5.4842,
+      "step": 624
+    },
+    {
+      "epoch": 0.04192239326558675,
+      "grad_norm": 0.1379136834023827,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 625
+    },
+    {
+      "epoch": 0.04198946909481169,
+      "grad_norm": 0.1331157866465741,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 626
+    },
+    {
+      "epoch": 0.042056544924036626,
+      "grad_norm": 0.1444552460966105,
+      "learning_rate": 2e-05,
+      "loss": 5.589,
+      "step": 627
+    },
+    {
+      "epoch": 0.042123620753261565,
+      "grad_norm": 0.1494344639472678,
+      "learning_rate": 2e-05,
+      "loss": 5.5014,
+      "step": 628
+    },
+    {
+      "epoch": 0.0421906965824865,
+      "grad_norm": 0.13779121503233616,
+      "learning_rate": 2e-05,
+      "loss": 5.4231,
+      "step": 629
+    },
+    {
+      "epoch": 0.04225777241171144,
+      "grad_norm": 0.1353682260026374,
+      "learning_rate": 2e-05,
+      "loss": 5.5386,
+      "step": 630
+    },
+    {
+      "epoch": 0.04232484824093638,
+      "grad_norm": 0.13505601004317694,
+      "learning_rate": 2e-05,
+      "loss": 5.6441,
+      "step": 631
+    },
+    {
+      "epoch": 0.04239192407016132,
+      "grad_norm": 0.13673407252093472,
+      "learning_rate": 2e-05,
+      "loss": 5.6584,
+      "step": 632
+    },
+    {
+      "epoch": 0.04245899989938626,
+      "grad_norm": 0.13952895671065724,
+      "learning_rate": 2e-05,
+      "loss": 5.5915,
+      "step": 633
+    },
+    {
+      "epoch": 0.042526075728611196,
+      "grad_norm": 0.13502645535908375,
+      "learning_rate": 2e-05,
+      "loss": 5.3987,
+      "step": 634
+    },
+    {
+      "epoch": 0.042593151557836134,
+      "grad_norm": 0.13722203420484583,
+      "learning_rate": 2e-05,
+      "loss": 5.4855,
+      "step": 635
+    },
+    {
+      "epoch": 0.04266022738706107,
+      "grad_norm": 0.13547254257428512,
+      "learning_rate": 2e-05,
+      "loss": 5.4538,
+      "step": 636
+    },
+    {
+      "epoch": 0.04272730321628601,
+      "grad_norm": 0.13346091985202946,
+      "learning_rate": 2e-05,
+      "loss": 5.5436,
+      "step": 637
+    },
+    {
+      "epoch": 0.04279437904551095,
+      "grad_norm": 0.1322312472443467,
+      "learning_rate": 2e-05,
+      "loss": 5.509,
+      "step": 638
+    },
+    {
+      "epoch": 0.04286145487473589,
+      "grad_norm": 0.13976357835377334,
+      "learning_rate": 2e-05,
+      "loss": 5.5889,
+      "step": 639
+    },
+    {
+      "epoch": 0.04292853070396083,
+      "grad_norm": 0.13533720945602512,
+      "learning_rate": 2e-05,
+      "loss": 5.4222,
+      "step": 640
+    },
+    {
+      "epoch": 0.042995606533185765,
+      "grad_norm": 0.13529379166141756,
+      "learning_rate": 2e-05,
+      "loss": 5.5203,
+      "step": 641
+    },
+    {
+      "epoch": 0.043062682362410704,
+      "grad_norm": 0.13405851885463874,
+      "learning_rate": 2e-05,
+      "loss": 5.472,
+      "step": 642
+    },
+    {
+      "epoch": 0.04312975819163564,
+      "grad_norm": 0.14299724362105026,
+      "learning_rate": 2e-05,
+      "loss": 5.5263,
+      "step": 643
+    },
+    {
+      "epoch": 0.04319683402086058,
+      "grad_norm": 0.13773847424131402,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 644
+    },
+    {
+      "epoch": 0.04326390985008552,
+      "grad_norm": 0.13941592754755283,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 645
+    },
+    {
+      "epoch": 0.04333098567931046,
+      "grad_norm": 0.14287175613530076,
+      "learning_rate": 2e-05,
+      "loss": 5.3087,
+      "step": 646
+    },
+    {
+      "epoch": 0.043398061508535396,
+      "grad_norm": 0.14332531094216394,
+      "learning_rate": 2e-05,
+      "loss": 5.4335,
+      "step": 647
+    },
+    {
+      "epoch": 0.04346513733776034,
+      "grad_norm": 0.13567264776547439,
+      "learning_rate": 2e-05,
+      "loss": 5.3195,
+      "step": 648
+    },
+    {
+      "epoch": 0.04353221316698528,
+      "grad_norm": 0.1358870056566697,
+      "learning_rate": 2e-05,
+      "loss": 5.5239,
+      "step": 649
+    },
+    {
+      "epoch": 0.04359928899621022,
+      "grad_norm": 0.14383192065048644,
+      "learning_rate": 2e-05,
+      "loss": 5.5391,
+      "step": 650
+    },
+    {
+      "epoch": 0.04366636482543516,
+      "grad_norm": 0.14205681708970144,
+      "learning_rate": 2e-05,
+      "loss": 5.3914,
+      "step": 651
+    },
+    {
+      "epoch": 0.043733440654660095,
+      "grad_norm": 0.14036834151940608,
+      "learning_rate": 2e-05,
+      "loss": 5.3794,
+      "step": 652
+    },
+    {
+      "epoch": 0.043800516483885034,
+      "grad_norm": 0.14128155321834157,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 653
+    },
+    {
+      "epoch": 0.04386759231310997,
+      "grad_norm": 0.1372517685943907,
+      "learning_rate": 2e-05,
+      "loss": 5.6727,
+      "step": 654
+    },
+    {
+      "epoch": 0.04393466814233491,
+      "grad_norm": 0.14416137682283492,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 655
+    },
+    {
+      "epoch": 0.04400174397155985,
+      "grad_norm": 0.14086294923237053,
+      "learning_rate": 2e-05,
+      "loss": 5.6039,
+      "step": 656
+    },
+    {
+      "epoch": 0.04406881980078479,
+      "grad_norm": 0.14131574559933985,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 657
+    },
+    {
+      "epoch": 0.044135895630009726,
+      "grad_norm": 0.13814460678025367,
+      "learning_rate": 2e-05,
+      "loss": 5.5273,
+      "step": 658
+    },
+    {
+      "epoch": 0.044202971459234665,
+      "grad_norm": 0.13444402320247367,
+      "learning_rate": 2e-05,
+      "loss": 5.4203,
+      "step": 659
+    },
+    {
+      "epoch": 0.0442700472884596,
+      "grad_norm": 0.13756833298978008,
+      "learning_rate": 2e-05,
+      "loss": 5.4767,
+      "step": 660
+    },
+    {
+      "epoch": 0.04433712311768454,
+      "grad_norm": 0.14084657154059518,
+      "learning_rate": 2e-05,
+      "loss": 5.6169,
+      "step": 661
+    },
+    {
+      "epoch": 0.04440419894690948,
+      "grad_norm": 0.13725206676233664,
+      "learning_rate": 2e-05,
+      "loss": 5.4911,
+      "step": 662
+    },
+    {
+      "epoch": 0.04447127477613442,
+      "grad_norm": 0.14671819805687436,
+      "learning_rate": 2e-05,
+      "loss": 5.4505,
+      "step": 663
+    },
+    {
+      "epoch": 0.04453835060535936,
+      "grad_norm": 0.13842636181004334,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 664
+    },
+    {
+      "epoch": 0.044605426434584296,
+      "grad_norm": 0.1356989081518579,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 665
+    },
+    {
+      "epoch": 0.044672502263809234,
+      "grad_norm": 0.143624327052737,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 666
+    },
+    {
+      "epoch": 0.04473957809303417,
+      "grad_norm": 0.14421885778439977,
+      "learning_rate": 2e-05,
+      "loss": 5.6708,
+      "step": 667
+    },
+    {
+      "epoch": 0.04480665392225911,
+      "grad_norm": 0.1427561940630554,
+      "learning_rate": 2e-05,
+      "loss": 5.5548,
+      "step": 668
+    },
+    {
+      "epoch": 0.04487372975148405,
+      "grad_norm": 0.13864440268106895,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 669
+    },
+    {
+      "epoch": 0.04494080558070899,
+      "grad_norm": 0.14245768057433805,
+      "learning_rate": 2e-05,
+      "loss": 5.4881,
+      "step": 670
+    },
+    {
+      "epoch": 0.045007881409933934,
+      "grad_norm": 0.14154681851686246,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 671
+    },
+    {
+      "epoch": 0.04507495723915887,
+      "grad_norm": 0.14527272461887336,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 672
+    },
+    {
+      "epoch": 0.04514203306838381,
+      "grad_norm": 0.13734141016394774,
+      "learning_rate": 2e-05,
+      "loss": 5.5022,
+      "step": 673
+    },
+    {
+      "epoch": 0.04520910889760875,
+      "grad_norm": 0.1465421696864576,
+      "learning_rate": 2e-05,
+      "loss": 5.5184,
+      "step": 674
+    },
+    {
+      "epoch": 0.04527618472683369,
+      "grad_norm": 0.13779435303068854,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 675
+    },
+    {
+      "epoch": 0.045343260556058626,
+      "grad_norm": 0.14178039715463103,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 676
+    },
+    {
+      "epoch": 0.045410336385283565,
+      "grad_norm": 0.14616190490145753,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 677
+    },
+    {
+      "epoch": 0.0454774122145085,
+      "grad_norm": 0.13763877213692075,
+      "learning_rate": 2e-05,
+      "loss": 5.4653,
+      "step": 678
+    },
+    {
+      "epoch": 0.04554448804373344,
+      "grad_norm": 0.13980016136474885,
+      "learning_rate": 2e-05,
+      "loss": 5.5444,
+      "step": 679
+    },
+    {
+      "epoch": 0.04561156387295838,
+      "grad_norm": 0.137715397390648,
+      "learning_rate": 2e-05,
+      "loss": 5.5387,
+      "step": 680
+    },
+    {
+      "epoch": 0.04567863970218332,
+      "grad_norm": 0.1372706413266507,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 681
+    },
+    {
+      "epoch": 0.04574571553140826,
+      "grad_norm": 0.1354507087608965,
+      "learning_rate": 2e-05,
+      "loss": 5.5454,
+      "step": 682
+    },
+    {
+      "epoch": 0.045812791360633195,
+      "grad_norm": 0.138702741054376,
+      "learning_rate": 2e-05,
+      "loss": 5.5524,
+      "step": 683
+    },
+    {
+      "epoch": 0.045879867189858134,
+      "grad_norm": 0.1439775365053947,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 684
+    },
+    {
+      "epoch": 0.04594694301908307,
+      "grad_norm": 0.13955574881868843,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 685
+    },
+    {
+      "epoch": 0.04601401884830801,
+      "grad_norm": 0.1427532255763558,
+      "learning_rate": 2e-05,
+      "loss": 5.5877,
+      "step": 686
+    },
+    {
+      "epoch": 0.04608109467753295,
+      "grad_norm": 0.14307774558633565,
+      "learning_rate": 2e-05,
+      "loss": 5.5149,
+      "step": 687
+    },
+    {
+      "epoch": 0.04614817050675789,
+      "grad_norm": 0.13561945813456946,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 688
+    },
+    {
+      "epoch": 0.046215246335982826,
+      "grad_norm": 0.13474997095301944,
+      "learning_rate": 2e-05,
+      "loss": 5.4804,
+      "step": 689
+    },
+    {
+      "epoch": 0.046282322165207765,
+      "grad_norm": 0.14295841961581612,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 690
+    },
+    {
+      "epoch": 0.0463493979944327,
+      "grad_norm": 0.13825600420584314,
+      "learning_rate": 2e-05,
+      "loss": 5.4524,
+      "step": 691
+    },
+    {
+      "epoch": 0.04641647382365764,
+      "grad_norm": 0.14042608783431912,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 692
+    },
+    {
+      "epoch": 0.04648354965288258,
+      "grad_norm": 0.13951528651867623,
+      "learning_rate": 2e-05,
+      "loss": 5.5752,
+      "step": 693
+    },
+    {
+      "epoch": 0.046550625482107526,
+      "grad_norm": 0.13793742248091767,
+      "learning_rate": 2e-05,
+      "loss": 5.4995,
+      "step": 694
+    },
+    {
+      "epoch": 0.046617701311332464,
+      "grad_norm": 0.1409577923361574,
+      "learning_rate": 2e-05,
+      "loss": 5.5288,
+      "step": 695
+    },
+    {
+      "epoch": 0.0466847771405574,
+      "grad_norm": 0.13977771961849436,
+      "learning_rate": 2e-05,
+      "loss": 5.3562,
+      "step": 696
+    },
+    {
+      "epoch": 0.04675185296978234,
+      "grad_norm": 0.14459618666743781,
+      "learning_rate": 2e-05,
+      "loss": 5.5676,
+      "step": 697
+    },
+    {
+      "epoch": 0.04681892879900728,
+      "grad_norm": 0.1354144056086241,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 698
+    },
+    {
+      "epoch": 0.04688600462823222,
+      "grad_norm": 0.14339030646872963,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 699
+    },
+    {
+      "epoch": 0.04695308045745716,
+      "grad_norm": 0.13708036185843223,
+      "learning_rate": 2e-05,
+      "loss": 5.5848,
+      "step": 700
+    },
+    {
+      "epoch": 0.047020156286682095,
+      "grad_norm": 0.13220596805939644,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 701
+    },
+    {
+      "epoch": 0.047087232115907034,
+      "grad_norm": 0.14054448501164102,
+      "learning_rate": 2e-05,
+      "loss": 5.5742,
+      "step": 702
+    },
+    {
+      "epoch": 0.04715430794513197,
+      "grad_norm": 0.1379018155485352,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 703
+    },
+    {
+      "epoch": 0.04722138377435691,
+      "grad_norm": 0.13604312786485726,
+      "learning_rate": 2e-05,
+      "loss": 5.6483,
+      "step": 704
+    },
+    {
+      "epoch": 0.04728845960358185,
+      "grad_norm": 0.1417843650138657,
+      "learning_rate": 2e-05,
+      "loss": 5.507,
+      "step": 705
+    },
+    {
+      "epoch": 0.04735553543280679,
+      "grad_norm": 0.14688767455070992,
+      "learning_rate": 2e-05,
+      "loss": 5.4784,
+      "step": 706
+    },
+    {
+      "epoch": 0.047422611262031726,
+      "grad_norm": 0.1424435900164657,
+      "learning_rate": 2e-05,
+      "loss": 5.3828,
+      "step": 707
+    },
+    {
+      "epoch": 0.047489687091256665,
+      "grad_norm": 0.14738016852072106,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 708
+    },
+    {
+      "epoch": 0.0475567629204816,
+      "grad_norm": 0.15536535033988336,
+      "learning_rate": 2e-05,
+      "loss": 5.4301,
+      "step": 709
+    },
+    {
+      "epoch": 0.04762383874970654,
+      "grad_norm": 0.14129097197304216,
+      "learning_rate": 2e-05,
+      "loss": 5.3782,
+      "step": 710
+    },
+    {
+      "epoch": 0.04769091457893148,
+      "grad_norm": 0.14731117823849502,
+      "learning_rate": 2e-05,
+      "loss": 5.4473,
+      "step": 711
+    },
+    {
+      "epoch": 0.04775799040815642,
+      "grad_norm": 0.14435206870439454,
+      "learning_rate": 2e-05,
+      "loss": 5.5475,
+      "step": 712
+    },
+    {
+      "epoch": 0.04782506623738136,
+      "grad_norm": 0.14288452454163153,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 713
+    },
+    {
+      "epoch": 0.047892142066606296,
+      "grad_norm": 0.14161191308170515,
+      "learning_rate": 2e-05,
+      "loss": 5.3787,
+      "step": 714
+    },
+    {
+      "epoch": 0.047959217895831234,
+      "grad_norm": 0.1418903743742546,
+      "learning_rate": 2e-05,
+      "loss": 5.4836,
+      "step": 715
+    },
+    {
+      "epoch": 0.04802629372505618,
+      "grad_norm": 0.14489603623091238,
+      "learning_rate": 2e-05,
+      "loss": 5.5585,
+      "step": 716
+    },
+    {
+      "epoch": 0.04809336955428112,
+      "grad_norm": 0.13924266255096693,
+      "learning_rate": 2e-05,
+      "loss": 5.479,
+      "step": 717
+    },
+    {
+      "epoch": 0.048160445383506056,
+      "grad_norm": 0.13940991905924555,
+      "learning_rate": 2e-05,
+      "loss": 5.5412,
+      "step": 718
+    },
+    {
+      "epoch": 0.048227521212730995,
+      "grad_norm": 0.14059649452273185,
+      "learning_rate": 2e-05,
+      "loss": 5.577,
+      "step": 719
+    },
+    {
+      "epoch": 0.04829459704195593,
+      "grad_norm": 0.156013079699878,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 720
+    },
+    {
+      "epoch": 0.04836167287118087,
+      "grad_norm": 0.13860426311076016,
+      "learning_rate": 2e-05,
+      "loss": 5.4887,
+      "step": 721
+    },
+    {
+      "epoch": 0.04842874870040581,
+      "grad_norm": 0.15058213786113767,
+      "learning_rate": 2e-05,
+      "loss": 5.4957,
+      "step": 722
+    },
+    {
+      "epoch": 0.04849582452963075,
+      "grad_norm": 0.1471991804658962,
+      "learning_rate": 2e-05,
+      "loss": 5.385,
+      "step": 723
+    },
+    {
+      "epoch": 0.04856290035885569,
+      "grad_norm": 0.13898185980430522,
+      "learning_rate": 2e-05,
+      "loss": 5.4351,
+      "step": 724
+    },
+    {
+      "epoch": 0.048629976188080626,
+      "grad_norm": 0.14474775848734892,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 725
+    },
+    {
+      "epoch": 0.048697052017305564,
+      "grad_norm": 0.14547265157225628,
+      "learning_rate": 2e-05,
+      "loss": 5.45,
+      "step": 726
+    },
+    {
+      "epoch": 0.0487641278465305,
+      "grad_norm": 0.15110964241263422,
+      "learning_rate": 2e-05,
+      "loss": 5.3932,
+      "step": 727
+    },
+    {
+      "epoch": 0.04883120367575544,
+      "grad_norm": 0.14568161529226067,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 728
+    },
+    {
+      "epoch": 0.04889827950498038,
+      "grad_norm": 0.14317519146667804,
+      "learning_rate": 2e-05,
+      "loss": 5.3993,
+      "step": 729
+    },
+    {
+      "epoch": 0.04896535533420532,
+      "grad_norm": 0.14368900396520803,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 730
+    },
+    {
+      "epoch": 0.04903243116343026,
+      "grad_norm": 0.14799592146466703,
+      "learning_rate": 2e-05,
+      "loss": 5.4929,
+      "step": 731
+    },
+    {
+      "epoch": 0.049099506992655195,
+      "grad_norm": 0.1374586908551005,
+      "learning_rate": 2e-05,
+      "loss": 5.4652,
+      "step": 732
+    },
+    {
+      "epoch": 0.049166582821880134,
+      "grad_norm": 0.14090424067704502,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 733
+    },
+    {
+      "epoch": 0.04923365865110507,
+      "grad_norm": 0.14512103945201657,
+      "learning_rate": 2e-05,
+      "loss": 5.5188,
+      "step": 734
+    },
+    {
+      "epoch": 0.04930073448033001,
+      "grad_norm": 0.13827428444801806,
+      "learning_rate": 2e-05,
+      "loss": 5.5551,
+      "step": 735
+    },
+    {
+      "epoch": 0.04936781030955495,
+      "grad_norm": 0.14675688280788232,
+      "learning_rate": 2e-05,
+      "loss": 5.343,
+      "step": 736
+    },
+    {
+      "epoch": 0.04943488613877989,
+      "grad_norm": 0.14638123039207257,
+      "learning_rate": 2e-05,
+      "loss": 5.5466,
+      "step": 737
+    },
+    {
+      "epoch": 0.049501961968004826,
+      "grad_norm": 0.1560724955556814,
+      "learning_rate": 2e-05,
+      "loss": 5.5679,
+      "step": 738
+    },
+    {
+      "epoch": 0.04956903779722977,
+      "grad_norm": 0.14466166873369335,
+      "learning_rate": 2e-05,
+      "loss": 5.3942,
+      "step": 739
+    },
+    {
+      "epoch": 0.04963611362645471,
+      "grad_norm": 0.13981762977797096,
+      "learning_rate": 2e-05,
+      "loss": 5.4849,
+      "step": 740
+    },
+    {
+      "epoch": 0.04970318945567965,
+      "grad_norm": 0.1459812432169665,
+      "learning_rate": 2e-05,
+      "loss": 5.5568,
+      "step": 741
+    },
+    {
+      "epoch": 0.04977026528490459,
+      "grad_norm": 0.14413644529517733,
+      "learning_rate": 2e-05,
+      "loss": 5.47,
+      "step": 742
+    },
+    {
+      "epoch": 0.049837341114129526,
+      "grad_norm": 0.13914012564303369,
+      "learning_rate": 2e-05,
+      "loss": 5.5178,
+      "step": 743
+    },
+    {
+      "epoch": 0.049904416943354464,
+      "grad_norm": 0.13910785363640651,
+      "learning_rate": 2e-05,
+      "loss": 5.3553,
+      "step": 744
+    },
+    {
+      "epoch": 0.0499714927725794,
+      "grad_norm": 0.15294409282184693,
+      "learning_rate": 2e-05,
+      "loss": 5.5345,
+      "step": 745
+    },
+    {
+      "epoch": 0.05003856860180434,
+      "grad_norm": 0.14262523649231573,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 746
+    },
+    {
+      "epoch": 0.05010564443102928,
+      "grad_norm": 0.13740834231732385,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 747
+    },
+    {
+      "epoch": 0.05017272026025422,
+      "grad_norm": 0.14381353132298083,
+      "learning_rate": 2e-05,
+      "loss": 5.5713,
+      "step": 748
+    },
+    {
+      "epoch": 0.050239796089479156,
+      "grad_norm": 0.14407248975999912,
+      "learning_rate": 2e-05,
+      "loss": 5.4998,
+      "step": 749
+    },
+    {
+      "epoch": 0.050306871918704095,
+      "grad_norm": 0.14299021645183782,
+      "learning_rate": 2e-05,
+      "loss": 5.6161,
+      "step": 750
+    },
+    {
+      "epoch": 0.05037394774792903,
+      "grad_norm": 0.1436514686554492,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 751
+    },
+    {
+      "epoch": 0.05044102357715397,
+      "grad_norm": 0.14349355455732365,
+      "learning_rate": 2e-05,
+      "loss": 5.5511,
+      "step": 752
+    },
+    {
+      "epoch": 0.05050809940637891,
+      "grad_norm": 0.15364827332811654,
+      "learning_rate": 2e-05,
+      "loss": 5.5485,
+      "step": 753
+    },
+    {
+      "epoch": 0.05057517523560385,
+      "grad_norm": 0.1395874557499621,
+      "learning_rate": 2e-05,
+      "loss": 5.5329,
+      "step": 754
+    },
+    {
+      "epoch": 0.05064225106482879,
+      "grad_norm": 0.13851607290094617,
+      "learning_rate": 2e-05,
+      "loss": 5.5744,
+      "step": 755
+    },
+    {
+      "epoch": 0.050709326894053726,
+      "grad_norm": 0.1482448145385073,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 756
+    },
+    {
+      "epoch": 0.050776402723278664,
+      "grad_norm": 0.14357580439103637,
+      "learning_rate": 2e-05,
+      "loss": 5.4076,
+      "step": 757
+    },
+    {
+      "epoch": 0.0508434785525036,
+      "grad_norm": 0.13615352609853693,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 758
+    },
+    {
+      "epoch": 0.05091055438172854,
+      "grad_norm": 0.14104478054752959,
+      "learning_rate": 2e-05,
+      "loss": 5.579,
+      "step": 759
+    },
+    {
+      "epoch": 0.05097763021095348,
+      "grad_norm": 0.14382269214670343,
+      "learning_rate": 2e-05,
+      "loss": 5.4899,
+      "step": 760
+    },
+    {
+      "epoch": 0.05104470604017842,
+      "grad_norm": 0.1453424860173766,
+      "learning_rate": 2e-05,
+      "loss": 5.5042,
+      "step": 761
+    },
+    {
+      "epoch": 0.051111781869403364,
+      "grad_norm": 0.1378142275682304,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 762
+    },
+    {
+      "epoch": 0.0511788576986283,
+      "grad_norm": 0.1407211767551057,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 763
+    },
+    {
+      "epoch": 0.05124593352785324,
+      "grad_norm": 0.14507754599168074,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 764
+    },
+    {
+      "epoch": 0.05131300935707818,
+      "grad_norm": 0.1371333172533457,
+      "learning_rate": 2e-05,
+      "loss": 5.4573,
+      "step": 765
+    },
+    {
+      "epoch": 0.05138008518630312,
+      "grad_norm": 0.14019084845144397,
+      "learning_rate": 2e-05,
+      "loss": 5.45,
+      "step": 766
+    },
+    {
+      "epoch": 0.051447161015528056,
+      "grad_norm": 0.14601344719671633,
+      "learning_rate": 2e-05,
+      "loss": 5.5222,
+      "step": 767
+    },
+    {
+      "epoch": 0.051514236844752995,
+      "grad_norm": 0.13709215476394632,
+      "learning_rate": 2e-05,
+      "loss": 5.5543,
+      "step": 768
+    },
+    {
+      "epoch": 0.05158131267397793,
+      "grad_norm": 0.13709275755688305,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 769
+    },
+    {
+      "epoch": 0.05164838850320287,
+      "grad_norm": 0.14582096377803794,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 770
+    },
+    {
+      "epoch": 0.05171546433242781,
+      "grad_norm": 0.14413621052224881,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 771
+    },
+    {
+      "epoch": 0.05178254016165275,
+      "grad_norm": 0.13946679017960892,
+      "learning_rate": 2e-05,
+      "loss": 5.4633,
+      "step": 772
+    },
+    {
+      "epoch": 0.05184961599087769,
+      "grad_norm": 0.13784149293242642,
+      "learning_rate": 2e-05,
+      "loss": 5.5561,
+      "step": 773
+    },
+    {
+      "epoch": 0.051916691820102626,
+      "grad_norm": 0.139259944074652,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 774
+    },
+    {
+      "epoch": 0.051983767649327564,
+      "grad_norm": 0.14244617490022182,
+      "learning_rate": 2e-05,
+      "loss": 5.5005,
+      "step": 775
+    },
+    {
+      "epoch": 0.0520508434785525,
+      "grad_norm": 0.1355845305077489,
+      "learning_rate": 2e-05,
+      "loss": 5.5592,
+      "step": 776
+    },
+    {
+      "epoch": 0.05211791930777744,
+      "grad_norm": 0.1422706649195537,
+      "learning_rate": 2e-05,
+      "loss": 5.4064,
+      "step": 777
+    },
+    {
+      "epoch": 0.05218499513700238,
+      "grad_norm": 0.14646043192603106,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 778
+    },
+    {
+      "epoch": 0.05225207096622732,
+      "grad_norm": 0.147456315844512,
+      "learning_rate": 2e-05,
+      "loss": 5.3064,
+      "step": 779
+    },
+    {
+      "epoch": 0.05231914679545226,
+      "grad_norm": 0.1470154653015442,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 780
+    },
+    {
+      "epoch": 0.052386222624677195,
+      "grad_norm": 0.14069580701895698,
+      "learning_rate": 2e-05,
+      "loss": 5.6507,
+      "step": 781
+    },
+    {
+      "epoch": 0.052453298453902134,
+      "grad_norm": 0.14032940746735406,
+      "learning_rate": 2e-05,
+      "loss": 5.4467,
+      "step": 782
+    },
+    {
+      "epoch": 0.05252037428312707,
+      "grad_norm": 0.13175202171608927,
+      "learning_rate": 2e-05,
+      "loss": 5.2898,
+      "step": 783
+    },
+    {
+      "epoch": 0.05258745011235201,
+      "grad_norm": 0.14175649744340305,
+      "learning_rate": 2e-05,
+      "loss": 5.4159,
+      "step": 784
+    },
+    {
+      "epoch": 0.052654525941576956,
+      "grad_norm": 0.14354654706079994,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 785
+    },
+    {
+      "epoch": 0.052721601770801894,
+      "grad_norm": 0.14651242905165499,
+      "learning_rate": 2e-05,
+      "loss": 5.3574,
+      "step": 786
+    },
+    {
+      "epoch": 0.05278867760002683,
+      "grad_norm": 0.13971193924898306,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 787
+    },
+    {
+      "epoch": 0.05285575342925177,
+      "grad_norm": 0.145747693978211,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 788
+    },
+    {
+      "epoch": 0.05292282925847671,
+      "grad_norm": 0.1415688699713498,
+      "learning_rate": 2e-05,
+      "loss": 5.519,
+      "step": 789
+    },
+    {
+      "epoch": 0.05298990508770165,
+      "grad_norm": 0.1336041732092004,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 790
+    },
+    {
+      "epoch": 0.05305698091692659,
+      "grad_norm": 0.1365943676391641,
+      "learning_rate": 2e-05,
+      "loss": 5.5071,
+      "step": 791
+    },
+    {
+      "epoch": 0.053124056746151525,
+      "grad_norm": 0.13602140990713338,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 792
+    },
+    {
+      "epoch": 0.053191132575376464,
+      "grad_norm": 0.1377099261404001,
+      "learning_rate": 2e-05,
+      "loss": 5.1798,
+      "step": 793
+    },
+    {
+      "epoch": 0.0532582084046014,
+      "grad_norm": 0.14286396454688055,
+      "learning_rate": 2e-05,
+      "loss": 5.6053,
+      "step": 794
+    },
+    {
+      "epoch": 0.05332528423382634,
+      "grad_norm": 0.13769427743235285,
+      "learning_rate": 2e-05,
+      "loss": 5.6173,
+      "step": 795
+    },
+    {
+      "epoch": 0.05339236006305128,
+      "grad_norm": 0.14037523303507846,
+      "learning_rate": 2e-05,
+      "loss": 5.4617,
+      "step": 796
+    },
+    {
+      "epoch": 0.05345943589227622,
+      "grad_norm": 0.13784999632976636,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 797
+    },
+    {
+      "epoch": 0.053526511721501156,
+      "grad_norm": 0.14077061853857362,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 798
+    },
+    {
+      "epoch": 0.053593587550726095,
+      "grad_norm": 0.13607626227235245,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 799
+    },
+    {
+      "epoch": 0.05366066337995103,
+      "grad_norm": 0.13974003766746196,
+      "learning_rate": 2e-05,
+      "loss": 5.5318,
+      "step": 800
+    },
+    {
+      "epoch": 0.05372773920917597,
+      "grad_norm": 0.13639664080985492,
+      "learning_rate": 2e-05,
+      "loss": 5.3341,
+      "step": 801
+    },
+    {
+      "epoch": 0.05379481503840091,
+      "grad_norm": 0.14167395023305554,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 802
+    },
+    {
+      "epoch": 0.05386189086762585,
+      "grad_norm": 0.1394741194712935,
+      "learning_rate": 2e-05,
+      "loss": 5.5296,
+      "step": 803
+    },
+    {
+      "epoch": 0.05392896669685079,
+      "grad_norm": 0.13712594140783416,
+      "learning_rate": 2e-05,
+      "loss": 5.3746,
+      "step": 804
+    },
+    {
+      "epoch": 0.053996042526075726,
+      "grad_norm": 0.14158707733158002,
+      "learning_rate": 2e-05,
+      "loss": 5.3914,
+      "step": 805
+    },
+    {
+      "epoch": 0.054063118355300664,
+      "grad_norm": 0.13486094025995213,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 806
+    },
+    {
+      "epoch": 0.05413019418452561,
+      "grad_norm": 0.14201804441816912,
+      "learning_rate": 2e-05,
+      "loss": 5.3864,
+      "step": 807
+    },
+    {
+      "epoch": 0.05419727001375055,
+      "grad_norm": 0.14257058862578786,
+      "learning_rate": 2e-05,
+      "loss": 5.5227,
+      "step": 808
+    },
+    {
+      "epoch": 0.05426434584297549,
+      "grad_norm": 0.140244101248035,
+      "learning_rate": 2e-05,
+      "loss": 5.4433,
+      "step": 809
+    },
+    {
+      "epoch": 0.054331421672200425,
+      "grad_norm": 0.14432063434507378,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 810
+    },
+    {
+      "epoch": 0.054398497501425364,
+      "grad_norm": 0.14269364021934897,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 811
+    },
+    {
+      "epoch": 0.0544655733306503,
+      "grad_norm": 0.1350589464942071,
+      "learning_rate": 2e-05,
+      "loss": 5.5264,
+      "step": 812
+    },
+    {
+      "epoch": 0.05453264915987524,
+      "grad_norm": 0.13664763358249052,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 813
+    },
+    {
+      "epoch": 0.05459972498910018,
+      "grad_norm": 0.14147306932220935,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 814
+    },
+    {
+      "epoch": 0.05466680081832512,
+      "grad_norm": 0.13878967516307975,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 815
+    },
+    {
+      "epoch": 0.054733876647550056,
+      "grad_norm": 0.13787894692276778,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 816
+    },
+    {
+      "epoch": 0.054800952476774994,
+      "grad_norm": 0.13774696079253215,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 817
+    },
+    {
+      "epoch": 0.05486802830599993,
+      "grad_norm": 0.13650675459088185,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 818
+    },
+    {
+      "epoch": 0.05493510413522487,
+      "grad_norm": 0.14158812692741873,
+      "learning_rate": 2e-05,
+      "loss": 5.5465,
+      "step": 819
+    },
+    {
+      "epoch": 0.05500217996444981,
+      "grad_norm": 0.1459840984810963,
+      "learning_rate": 2e-05,
+      "loss": 5.427,
+      "step": 820
+    },
+    {
+      "epoch": 0.05506925579367475,
+      "grad_norm": 0.1333773447411539,
+      "learning_rate": 2e-05,
+      "loss": 5.481,
+      "step": 821
+    },
+    {
+      "epoch": 0.05513633162289969,
+      "grad_norm": 0.1384659909074751,
+      "learning_rate": 2e-05,
+      "loss": 5.2546,
+      "step": 822
+    },
+    {
+      "epoch": 0.055203407452124625,
+      "grad_norm": 0.14398823490378715,
+      "learning_rate": 2e-05,
+      "loss": 5.4862,
+      "step": 823
+    },
+    {
+      "epoch": 0.055270483281349564,
+      "grad_norm": 0.13875110841844615,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 824
+    },
+    {
+      "epoch": 0.0553375591105745,
+      "grad_norm": 0.13948950139399105,
+      "learning_rate": 2e-05,
+      "loss": 5.5384,
+      "step": 825
+    },
+    {
+      "epoch": 0.05540463493979944,
+      "grad_norm": 0.15116020802467697,
+      "learning_rate": 2e-05,
+      "loss": 5.4948,
+      "step": 826
+    },
+    {
+      "epoch": 0.05547171076902438,
+      "grad_norm": 0.13819426660576553,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 827
+    },
+    {
+      "epoch": 0.05553878659824932,
+      "grad_norm": 0.14568352193832856,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 828
+    },
+    {
+      "epoch": 0.055605862427474256,
+      "grad_norm": 0.15529099787024003,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 829
+    },
+    {
+      "epoch": 0.0556729382566992,
+      "grad_norm": 0.13876326922565835,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 830
+    },
+    {
+      "epoch": 0.05574001408592414,
+      "grad_norm": 0.13577318238666272,
+      "learning_rate": 2e-05,
+      "loss": 5.4858,
+      "step": 831
+    },
+    {
+      "epoch": 0.05580708991514908,
+      "grad_norm": 0.13595378860625845,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 832
+    },
+    {
+      "epoch": 0.05587416574437402,
+      "grad_norm": 0.15779272648314802,
+      "learning_rate": 2e-05,
+      "loss": 5.5412,
+      "step": 833
+    },
+    {
+      "epoch": 0.055941241573598956,
+      "grad_norm": 0.1387678022712011,
+      "learning_rate": 2e-05,
+      "loss": 5.3525,
+      "step": 834
+    },
+    {
+      "epoch": 0.056008317402823894,
+      "grad_norm": 0.13807992148054876,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 835
+    },
+    {
+      "epoch": 0.05607539323204883,
+      "grad_norm": 0.14276025564396136,
+      "learning_rate": 2e-05,
+      "loss": 5.5332,
+      "step": 836
+    },
+    {
+      "epoch": 0.05614246906127377,
+      "grad_norm": 0.1409028275800159,
+      "learning_rate": 2e-05,
+      "loss": 5.5933,
+      "step": 837
+    },
+    {
+      "epoch": 0.05620954489049871,
+      "grad_norm": 0.14006034522749472,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 838
+    },
+    {
+      "epoch": 0.05627662071972365,
+      "grad_norm": 0.14767453023280888,
+      "learning_rate": 2e-05,
+      "loss": 5.5192,
+      "step": 839
+    },
+    {
+      "epoch": 0.05634369654894859,
+      "grad_norm": 0.14190206730998428,
+      "learning_rate": 2e-05,
+      "loss": 5.6028,
+      "step": 840
+    },
+    {
+      "epoch": 0.056410772378173525,
+      "grad_norm": 0.14087398642948445,
+      "learning_rate": 2e-05,
+      "loss": 5.3911,
+      "step": 841
+    },
+    {
+      "epoch": 0.056477848207398464,
+      "grad_norm": 0.14057267469250576,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 842
+    },
+    {
+      "epoch": 0.0565449240366234,
+      "grad_norm": 0.14076872251712083,
+      "learning_rate": 2e-05,
+      "loss": 5.3893,
+      "step": 843
+    },
+    {
+      "epoch": 0.05661199986584834,
+      "grad_norm": 0.1350938626348686,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 844
+    },
+    {
+      "epoch": 0.05667907569507328,
+      "grad_norm": 0.13915139457632292,
+      "learning_rate": 2e-05,
+      "loss": 5.5325,
+      "step": 845
+    },
+    {
+      "epoch": 0.05674615152429822,
+      "grad_norm": 0.13812758742751205,
+      "learning_rate": 2e-05,
+      "loss": 5.3681,
+      "step": 846
+    },
+    {
+      "epoch": 0.056813227353523156,
+      "grad_norm": 0.1353484945833019,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 847
+    },
+    {
+      "epoch": 0.056880303182748095,
+      "grad_norm": 0.1332408583701437,
+      "learning_rate": 2e-05,
+      "loss": 5.3802,
+      "step": 848
+    },
+    {
+      "epoch": 0.05694737901197303,
+      "grad_norm": 0.14174163270345783,
+      "learning_rate": 2e-05,
+      "loss": 5.3823,
+      "step": 849
+    },
+    {
+      "epoch": 0.05701445484119797,
+      "grad_norm": 0.13892567553248103,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 850
+    },
+    {
+      "epoch": 0.05708153067042291,
+      "grad_norm": 0.13734709444032592,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 851
+    },
+    {
+      "epoch": 0.05714860649964785,
+      "grad_norm": 0.14277060734479116,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 852
+    },
+    {
+      "epoch": 0.057215682328872794,
+      "grad_norm": 0.14130189619703876,
+      "learning_rate": 2e-05,
+      "loss": 5.5571,
+      "step": 853
+    },
+    {
+      "epoch": 0.05728275815809773,
+      "grad_norm": 0.13530427873091264,
+      "learning_rate": 2e-05,
+      "loss": 5.5739,
+      "step": 854
+    },
+    {
+      "epoch": 0.05734983398732267,
+      "grad_norm": 0.13945347479306888,
+      "learning_rate": 2e-05,
+      "loss": 5.653,
+      "step": 855
+    },
+    {
+      "epoch": 0.05741690981654761,
+      "grad_norm": 0.1374314318007774,
+      "learning_rate": 2e-05,
+      "loss": 5.3369,
+      "step": 856
+    },
+    {
+      "epoch": 0.05748398564577255,
+      "grad_norm": 0.1375300801579823,
+      "learning_rate": 2e-05,
+      "loss": 5.4263,
+      "step": 857
+    },
+    {
+      "epoch": 0.057551061474997486,
+      "grad_norm": 0.143143572627421,
+      "learning_rate": 2e-05,
+      "loss": 5.391,
+      "step": 858
+    },
+    {
+      "epoch": 0.057618137304222425,
+      "grad_norm": 0.14253496001737856,
+      "learning_rate": 2e-05,
+      "loss": 5.6284,
+      "step": 859
+    },
+    {
+      "epoch": 0.05768521313344736,
+      "grad_norm": 0.14241012072714998,
+      "learning_rate": 2e-05,
+      "loss": 5.5044,
+      "step": 860
+    },
+    {
+      "epoch": 0.0577522889626723,
+      "grad_norm": 0.1446386983727709,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 861
+    },
+    {
+      "epoch": 0.05781936479189724,
+      "grad_norm": 0.13753673113124346,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 862
+    },
+    {
+      "epoch": 0.05788644062112218,
+      "grad_norm": 0.13388397561799564,
+      "learning_rate": 2e-05,
+      "loss": 5.5171,
+      "step": 863
+    },
+    {
+      "epoch": 0.05795351645034712,
+      "grad_norm": 0.14387247898191605,
+      "learning_rate": 2e-05,
+      "loss": 5.351,
+      "step": 864
+    },
+    {
+      "epoch": 0.058020592279572056,
+      "grad_norm": 0.13846387917257807,
+      "learning_rate": 2e-05,
+      "loss": 5.3717,
+      "step": 865
+    },
+    {
+      "epoch": 0.058087668108796994,
+      "grad_norm": 0.1387772996752303,
+      "learning_rate": 2e-05,
+      "loss": 5.4604,
+      "step": 866
+    },
+    {
+      "epoch": 0.05815474393802193,
+      "grad_norm": 0.13740457747626075,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 867
+    },
+    {
+      "epoch": 0.05822181976724687,
+      "grad_norm": 0.1420323840381132,
+      "learning_rate": 2e-05,
+      "loss": 5.5668,
+      "step": 868
+    },
+    {
+      "epoch": 0.05828889559647181,
+      "grad_norm": 0.13722131968167312,
+      "learning_rate": 2e-05,
+      "loss": 5.4712,
+      "step": 869
+    },
+    {
+      "epoch": 0.05835597142569675,
+      "grad_norm": 0.13695411515133893,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 870
+    },
+    {
+      "epoch": 0.05842304725492169,
+      "grad_norm": 0.136261333389352,
+      "learning_rate": 2e-05,
+      "loss": 5.3434,
+      "step": 871
+    },
+    {
+      "epoch": 0.058490123084146625,
+      "grad_norm": 0.14596221144034294,
+      "learning_rate": 2e-05,
+      "loss": 5.6215,
+      "step": 872
+    },
+    {
+      "epoch": 0.058557198913371564,
+      "grad_norm": 0.13686208974725156,
+      "learning_rate": 2e-05,
+      "loss": 5.487,
+      "step": 873
+    },
+    {
+      "epoch": 0.0586242747425965,
+      "grad_norm": 0.14019161433698765,
+      "learning_rate": 2e-05,
+      "loss": 5.4383,
+      "step": 874
+    },
+    {
+      "epoch": 0.05869135057182145,
+      "grad_norm": 0.1426100777175223,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 875
+    },
+    {
+      "epoch": 0.058758426401046386,
+      "grad_norm": 0.13965488796368594,
+      "learning_rate": 2e-05,
+      "loss": 5.5507,
+      "step": 876
+    },
+    {
+      "epoch": 0.058825502230271325,
+      "grad_norm": 0.13639290650168254,
+      "learning_rate": 2e-05,
+      "loss": 5.4881,
+      "step": 877
+    },
+    {
+      "epoch": 0.05889257805949626,
+      "grad_norm": 0.13666268909296495,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 878
+    },
+    {
+      "epoch": 0.0589596538887212,
+      "grad_norm": 0.14413906456030398,
+      "learning_rate": 2e-05,
+      "loss": 5.5592,
+      "step": 879
+    },
+    {
+      "epoch": 0.05902672971794614,
+      "grad_norm": 0.1429403323240712,
+      "learning_rate": 2e-05,
+      "loss": 5.4641,
+      "step": 880
+    },
+    {
+      "epoch": 0.05909380554717108,
+      "grad_norm": 0.13512452658212443,
+      "learning_rate": 2e-05,
+      "loss": 5.6031,
+      "step": 881
+    },
+    {
+      "epoch": 0.05916088137639602,
+      "grad_norm": 0.1475140817477487,
+      "learning_rate": 2e-05,
+      "loss": 5.5272,
+      "step": 882
+    },
+    {
+      "epoch": 0.059227957205620956,
+      "grad_norm": 0.14940223612041237,
+      "learning_rate": 2e-05,
+      "loss": 5.6019,
+      "step": 883
+    },
+    {
+      "epoch": 0.059295033034845894,
+      "grad_norm": 0.13674108862846301,
+      "learning_rate": 2e-05,
+      "loss": 5.5681,
+      "step": 884
+    },
+    {
+      "epoch": 0.05936210886407083,
+      "grad_norm": 0.13940093999562198,
+      "learning_rate": 2e-05,
+      "loss": 5.5676,
+      "step": 885
+    },
+    {
+      "epoch": 0.05942918469329577,
+      "grad_norm": 0.14837945848325476,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 886
+    },
+    {
+      "epoch": 0.05949626052252071,
+      "grad_norm": 0.15011354648255704,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 887
+    },
+    {
+      "epoch": 0.05956333635174565,
+      "grad_norm": 0.1362209694144478,
+      "learning_rate": 2e-05,
+      "loss": 5.4284,
+      "step": 888
+    },
+    {
+      "epoch": 0.059630412180970586,
+      "grad_norm": 0.14684444411160688,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 889
+    },
+    {
+      "epoch": 0.059697488010195525,
+      "grad_norm": 0.1395625625161395,
+      "learning_rate": 2e-05,
+      "loss": 5.5442,
+      "step": 890
+    },
+    {
+      "epoch": 0.05976456383942046,
+      "grad_norm": 0.14254708197502527,
+      "learning_rate": 2e-05,
+      "loss": 5.5444,
+      "step": 891
+    },
+    {
+      "epoch": 0.0598316396686454,
+      "grad_norm": 0.1475803682858939,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 892
+    },
+    {
+      "epoch": 0.05989871549787034,
+      "grad_norm": 0.1406127834801801,
+      "learning_rate": 2e-05,
+      "loss": 5.2831,
+      "step": 893
+    },
+    {
+      "epoch": 0.05996579132709528,
+      "grad_norm": 0.14310900234828802,
+      "learning_rate": 2e-05,
+      "loss": 5.5871,
+      "step": 894
+    },
+    {
+      "epoch": 0.06003286715632022,
+      "grad_norm": 0.14093769764642358,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 895
+    },
+    {
+      "epoch": 0.060099942985545156,
+      "grad_norm": 0.14191864192436038,
+      "learning_rate": 2e-05,
+      "loss": 5.3113,
+      "step": 896
+    },
+    {
+      "epoch": 0.060167018814770094,
+      "grad_norm": 0.14235473367066245,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 897
+    },
+    {
+      "epoch": 0.06023409464399504,
+      "grad_norm": 0.15059353819172402,
+      "learning_rate": 2e-05,
+      "loss": 5.4607,
+      "step": 898
+    },
+    {
+      "epoch": 0.06030117047321998,
+      "grad_norm": 0.14899940698176392,
+      "learning_rate": 2e-05,
+      "loss": 5.4367,
+      "step": 899
+    },
+    {
+      "epoch": 0.06036824630244492,
+      "grad_norm": 0.1430397072684481,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 900
+    },
+    {
+      "epoch": 0.060435322131669855,
+      "grad_norm": 0.14136365852867017,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 901
+    },
+    {
+      "epoch": 0.060502397960894794,
+      "grad_norm": 0.1480207035572126,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 902
+    },
+    {
+      "epoch": 0.06056947379011973,
+      "grad_norm": 0.14587433880174241,
+      "learning_rate": 2e-05,
+      "loss": 5.4813,
+      "step": 903
+    },
+    {
+      "epoch": 0.06063654961934467,
+      "grad_norm": 0.13745118216058058,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 904
+    },
+    {
+      "epoch": 0.06070362544856961,
+      "grad_norm": 0.14815188248219527,
+      "learning_rate": 2e-05,
+      "loss": 5.5552,
+      "step": 905
+    },
+    {
+      "epoch": 0.06077070127779455,
+      "grad_norm": 0.15415144822204996,
+      "learning_rate": 2e-05,
+      "loss": 5.5095,
+      "step": 906
+    },
+    {
+      "epoch": 0.060837777107019486,
+      "grad_norm": 0.15181524962880177,
+      "learning_rate": 2e-05,
+      "loss": 5.3029,
+      "step": 907
+    },
+    {
+      "epoch": 0.060904852936244425,
+      "grad_norm": 0.1416912136458125,
+      "learning_rate": 2e-05,
+      "loss": 5.5007,
+      "step": 908
+    },
+    {
+      "epoch": 0.06097192876546936,
+      "grad_norm": 0.1438777775082393,
+      "learning_rate": 2e-05,
+      "loss": 5.4892,
+      "step": 909
+    },
+    {
+      "epoch": 0.0610390045946943,
+      "grad_norm": 0.14232780598974237,
+      "learning_rate": 2e-05,
+      "loss": 5.5196,
+      "step": 910
+    },
+    {
+      "epoch": 0.06110608042391924,
+      "grad_norm": 0.1358565494728242,
+      "learning_rate": 2e-05,
+      "loss": 5.5891,
+      "step": 911
+    },
+    {
+      "epoch": 0.06117315625314418,
+      "grad_norm": 0.1401310661087953,
+      "learning_rate": 2e-05,
+      "loss": 5.2692,
+      "step": 912
+    },
+    {
+      "epoch": 0.06124023208236912,
+      "grad_norm": 0.14689278941422543,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 913
+    },
+    {
+      "epoch": 0.061307307911594056,
+      "grad_norm": 0.13829368223283206,
+      "learning_rate": 2e-05,
+      "loss": 5.4825,
+      "step": 914
+    },
+    {
+      "epoch": 0.061374383740818994,
+      "grad_norm": 0.135609036348283,
+      "learning_rate": 2e-05,
+      "loss": 5.4353,
+      "step": 915
+    },
+    {
+      "epoch": 0.06144145957004393,
+      "grad_norm": 0.14145041358914442,
+      "learning_rate": 2e-05,
+      "loss": 5.4622,
+      "step": 916
+    },
+    {
+      "epoch": 0.06150853539926887,
+      "grad_norm": 0.14970412784364212,
+      "learning_rate": 2e-05,
+      "loss": 5.3994,
+      "step": 917
+    },
+    {
+      "epoch": 0.06157561122849381,
+      "grad_norm": 0.13605722210160578,
+      "learning_rate": 2e-05,
+      "loss": 5.5852,
+      "step": 918
+    },
+    {
+      "epoch": 0.06164268705771875,
+      "grad_norm": 0.1402338188085907,
+      "learning_rate": 2e-05,
+      "loss": 5.5012,
+      "step": 919
+    },
+    {
+      "epoch": 0.061709762886943686,
+      "grad_norm": 0.1386535419548301,
+      "learning_rate": 2e-05,
+      "loss": 5.5691,
+      "step": 920
+    },
+    {
+      "epoch": 0.06177683871616863,
+      "grad_norm": 0.14120651661213565,
+      "learning_rate": 2e-05,
+      "loss": 5.4144,
+      "step": 921
+    },
+    {
+      "epoch": 0.06184391454539357,
+      "grad_norm": 0.1418152694970351,
+      "learning_rate": 2e-05,
+      "loss": 5.475,
+      "step": 922
+    },
+    {
+      "epoch": 0.06191099037461851,
+      "grad_norm": 0.14081268526884533,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 923
+    },
+    {
+      "epoch": 0.06197806620384345,
+      "grad_norm": 0.1429539519572204,
+      "learning_rate": 2e-05,
+      "loss": 5.5652,
+      "step": 924
+    },
+    {
+      "epoch": 0.062045142033068386,
+      "grad_norm": 0.14014628197100631,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 925
+    },
+    {
+      "epoch": 0.062112217862293324,
+      "grad_norm": 0.1403602122795351,
+      "learning_rate": 2e-05,
+      "loss": 5.5362,
+      "step": 926
+    },
+    {
+      "epoch": 0.06217929369151826,
+      "grad_norm": 0.13766903314220935,
+      "learning_rate": 2e-05,
+      "loss": 5.5593,
+      "step": 927
+    },
+    {
+      "epoch": 0.0622463695207432,
+      "grad_norm": 0.14442357345855064,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 928
+    },
+    {
+      "epoch": 0.06231344534996814,
+      "grad_norm": 0.14637890184169824,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 929
+    },
+    {
+      "epoch": 0.06238052117919308,
+      "grad_norm": 0.14086066691010285,
+      "learning_rate": 2e-05,
+      "loss": 5.4162,
+      "step": 930
+    },
+    {
+      "epoch": 0.06244759700841802,
+      "grad_norm": 0.14469989861884466,
+      "learning_rate": 2e-05,
+      "loss": 5.5903,
+      "step": 931
+    },
+    {
+      "epoch": 0.06251467283764296,
+      "grad_norm": 0.15443996737079702,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 932
+    },
+    {
+      "epoch": 0.0625817486668679,
+      "grad_norm": 0.1447133409366118,
+      "learning_rate": 2e-05,
+      "loss": 5.502,
+      "step": 933
+    },
+    {
+      "epoch": 0.06264882449609284,
+      "grad_norm": 0.13714509914637055,
+      "learning_rate": 2e-05,
+      "loss": 5.5174,
+      "step": 934
+    },
+    {
+      "epoch": 0.06271590032531778,
+      "grad_norm": 0.14660932287603656,
+      "learning_rate": 2e-05,
+      "loss": 5.5223,
+      "step": 935
+    },
+    {
+      "epoch": 0.06278297615454272,
+      "grad_norm": 0.14713844137885843,
+      "learning_rate": 2e-05,
+      "loss": 5.4564,
+      "step": 936
+    },
+    {
+      "epoch": 0.06285005198376765,
+      "grad_norm": 0.1345358901580513,
+      "learning_rate": 2e-05,
+      "loss": 5.5006,
+      "step": 937
+    },
+    {
+      "epoch": 0.0629171278129926,
+      "grad_norm": 0.15090421113098856,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 938
+    },
+    {
+      "epoch": 0.06298420364221753,
+      "grad_norm": 0.13614640484954693,
+      "learning_rate": 2e-05,
+      "loss": 5.4222,
+      "step": 939
+    },
+    {
+      "epoch": 0.06305127947144247,
+      "grad_norm": 0.1406727021827449,
+      "learning_rate": 2e-05,
+      "loss": 5.4434,
+      "step": 940
+    },
+    {
+      "epoch": 0.06311835530066741,
+      "grad_norm": 0.14014131545298447,
+      "learning_rate": 2e-05,
+      "loss": 5.6228,
+      "step": 941
+    },
+    {
+      "epoch": 0.06318543112989235,
+      "grad_norm": 0.1413707632397799,
+      "learning_rate": 2e-05,
+      "loss": 5.5162,
+      "step": 942
+    },
+    {
+      "epoch": 0.06325250695911729,
+      "grad_norm": 0.13638836970443632,
+      "learning_rate": 2e-05,
+      "loss": 5.2534,
+      "step": 943
+    },
+    {
+      "epoch": 0.06331958278834222,
+      "grad_norm": 0.1416451593858626,
+      "learning_rate": 2e-05,
+      "loss": 5.5657,
+      "step": 944
+    },
+    {
+      "epoch": 0.06338665861756716,
+      "grad_norm": 0.1470180217638178,
+      "learning_rate": 2e-05,
+      "loss": 5.4488,
+      "step": 945
+    },
+    {
+      "epoch": 0.0634537344467921,
+      "grad_norm": 0.1402528261630909,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 946
+    },
+    {
+      "epoch": 0.06352081027601704,
+      "grad_norm": 0.14168013819452752,
+      "learning_rate": 2e-05,
+      "loss": 5.6216,
+      "step": 947
+    },
+    {
+      "epoch": 0.06358788610524198,
+      "grad_norm": 0.13599946813522276,
+      "learning_rate": 2e-05,
+      "loss": 5.3424,
+      "step": 948
+    },
+    {
+      "epoch": 0.06365496193446692,
+      "grad_norm": 0.1462738772591234,
+      "learning_rate": 2e-05,
+      "loss": 5.3495,
+      "step": 949
+    },
+    {
+      "epoch": 0.06372203776369186,
+      "grad_norm": 0.14257591062972663,
+      "learning_rate": 2e-05,
+      "loss": 5.5289,
+      "step": 950
+    },
+    {
+      "epoch": 0.0637891135929168,
+      "grad_norm": 0.13730199516767894,
+      "learning_rate": 2e-05,
+      "loss": 5.3791,
+      "step": 951
+    },
+    {
+      "epoch": 0.06385618942214173,
+      "grad_norm": 0.14071710088510572,
+      "learning_rate": 2e-05,
+      "loss": 5.6989,
+      "step": 952
+    },
+    {
+      "epoch": 0.06392326525136667,
+      "grad_norm": 0.13788605699179837,
+      "learning_rate": 2e-05,
+      "loss": 5.3702,
+      "step": 953
+    },
+    {
+      "epoch": 0.06399034108059161,
+      "grad_norm": 0.14253818887026737,
+      "learning_rate": 2e-05,
+      "loss": 5.5086,
+      "step": 954
+    },
+    {
+      "epoch": 0.06405741690981655,
+      "grad_norm": 0.1419512183340297,
+      "learning_rate": 2e-05,
+      "loss": 5.4139,
+      "step": 955
+    },
+    {
+      "epoch": 0.06412449273904149,
+      "grad_norm": 0.13641133373263273,
+      "learning_rate": 2e-05,
+      "loss": 5.6436,
+      "step": 956
+    },
+    {
+      "epoch": 0.06419156856826642,
+      "grad_norm": 0.1423878286069241,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 957
+    },
+    {
+      "epoch": 0.06425864439749136,
+      "grad_norm": 0.1343091580272714,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 958
+    },
+    {
+      "epoch": 0.0643257202267163,
+      "grad_norm": 0.1416301662054869,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 959
+    },
+    {
+      "epoch": 0.06439279605594124,
+      "grad_norm": 0.1406843573770194,
+      "learning_rate": 2e-05,
+      "loss": 5.6132,
+      "step": 960
+    },
+    {
+      "epoch": 0.06445987188516618,
+      "grad_norm": 0.1424090051868738,
+      "learning_rate": 2e-05,
+      "loss": 5.54,
+      "step": 961
+    },
+    {
+      "epoch": 0.06452694771439112,
+      "grad_norm": 0.1517475569009069,
+      "learning_rate": 2e-05,
+      "loss": 5.4407,
+      "step": 962
+    },
+    {
+      "epoch": 0.06459402354361606,
+      "grad_norm": 0.14808580189993836,
+      "learning_rate": 2e-05,
+      "loss": 5.5214,
+      "step": 963
+    },
+    {
+      "epoch": 0.064661099372841,
+      "grad_norm": 0.1415675119324065,
+      "learning_rate": 2e-05,
+      "loss": 5.5676,
+      "step": 964
+    },
+    {
+      "epoch": 0.06472817520206593,
+      "grad_norm": 0.1494174332805045,
+      "learning_rate": 2e-05,
+      "loss": 5.506,
+      "step": 965
+    },
+    {
+      "epoch": 0.06479525103129087,
+      "grad_norm": 0.1396771353250545,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 966
+    },
+    {
+      "epoch": 0.06486232686051581,
+      "grad_norm": 0.1431365155044209,
+      "learning_rate": 2e-05,
+      "loss": 5.4358,
+      "step": 967
+    },
+    {
+      "epoch": 0.06492940268974075,
+      "grad_norm": 0.15043753801926193,
+      "learning_rate": 2e-05,
+      "loss": 5.2945,
+      "step": 968
+    },
+    {
+      "epoch": 0.06499647851896569,
+      "grad_norm": 0.1457021416191956,
+      "learning_rate": 2e-05,
+      "loss": 5.4948,
+      "step": 969
+    },
+    {
+      "epoch": 0.06506355434819062,
+      "grad_norm": 0.14317023515833505,
+      "learning_rate": 2e-05,
+      "loss": 5.5581,
+      "step": 970
+    },
+    {
+      "epoch": 0.06513063017741556,
+      "grad_norm": 0.14392523134142407,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 971
+    },
+    {
+      "epoch": 0.0651977060066405,
+      "grad_norm": 0.13961506605718788,
+      "learning_rate": 2e-05,
+      "loss": 5.4761,
+      "step": 972
+    },
+    {
+      "epoch": 0.06526478183586544,
+      "grad_norm": 0.14661756815235763,
+      "learning_rate": 2e-05,
+      "loss": 5.4214,
+      "step": 973
+    },
+    {
+      "epoch": 0.06533185766509038,
+      "grad_norm": 0.1502094874218225,
+      "learning_rate": 2e-05,
+      "loss": 5.4245,
+      "step": 974
+    },
+    {
+      "epoch": 0.06539893349431532,
+      "grad_norm": 0.13807945735837462,
+      "learning_rate": 2e-05,
+      "loss": 5.5346,
+      "step": 975
+    },
+    {
+      "epoch": 0.06546600932354026,
+      "grad_norm": 0.14097145075217724,
+      "learning_rate": 2e-05,
+      "loss": 5.5025,
+      "step": 976
+    },
+    {
+      "epoch": 0.0655330851527652,
+      "grad_norm": 0.14699011978998494,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 977
+    },
+    {
+      "epoch": 0.06560016098199015,
+      "grad_norm": 0.13954138721060855,
+      "learning_rate": 2e-05,
+      "loss": 5.4361,
+      "step": 978
+    },
+    {
+      "epoch": 0.06566723681121509,
+      "grad_norm": 0.13462150953089716,
+      "learning_rate": 2e-05,
+      "loss": 5.5481,
+      "step": 979
+    },
+    {
+      "epoch": 0.06573431264044002,
+      "grad_norm": 0.14131618901622997,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 980
+    },
+    {
+      "epoch": 0.06580138846966496,
+      "grad_norm": 0.1410610627720264,
+      "learning_rate": 2e-05,
+      "loss": 5.3419,
+      "step": 981
+    },
+    {
+      "epoch": 0.0658684642988899,
+      "grad_norm": 0.14746307906368222,
+      "learning_rate": 2e-05,
+      "loss": 5.5164,
+      "step": 982
+    },
+    {
+      "epoch": 0.06593554012811484,
+      "grad_norm": 0.14345163280006756,
+      "learning_rate": 2e-05,
+      "loss": 5.5695,
+      "step": 983
+    },
+    {
+      "epoch": 0.06600261595733978,
+      "grad_norm": 0.13873894896339284,
+      "learning_rate": 2e-05,
+      "loss": 5.4458,
+      "step": 984
+    },
+    {
+      "epoch": 0.06606969178656472,
+      "grad_norm": 0.13815448538380473,
+      "learning_rate": 2e-05,
+      "loss": 5.4279,
+      "step": 985
+    },
+    {
+      "epoch": 0.06613676761578965,
+      "grad_norm": 0.14243536706924334,
+      "learning_rate": 2e-05,
+      "loss": 5.6864,
+      "step": 986
+    },
+    {
+      "epoch": 0.06620384344501459,
+      "grad_norm": 0.14804898457927465,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 987
+    },
+    {
+      "epoch": 0.06627091927423953,
+      "grad_norm": 0.14163182862408727,
+      "learning_rate": 2e-05,
+      "loss": 5.355,
+      "step": 988
+    },
+    {
+      "epoch": 0.06633799510346447,
+      "grad_norm": 0.14106974980590847,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 989
+    },
+    {
+      "epoch": 0.06640507093268941,
+      "grad_norm": 0.1463973211213733,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 990
+    },
+    {
+      "epoch": 0.06647214676191435,
+      "grad_norm": 0.13863246088588727,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 991
+    },
+    {
+      "epoch": 0.06653922259113929,
+      "grad_norm": 0.13820198703500147,
+      "learning_rate": 2e-05,
+      "loss": 5.5041,
+      "step": 992
+    },
+    {
+      "epoch": 0.06660629842036422,
+      "grad_norm": 0.14590687219591825,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 993
+    },
+    {
+      "epoch": 0.06667337424958916,
+      "grad_norm": 0.1441502716480331,
+      "learning_rate": 2e-05,
+      "loss": 5.5926,
+      "step": 994
+    },
+    {
+      "epoch": 0.0667404500788141,
+      "grad_norm": 0.1383324685051119,
+      "learning_rate": 2e-05,
+      "loss": 5.2819,
+      "step": 995
+    },
+    {
+      "epoch": 0.06680752590803904,
+      "grad_norm": 0.13736178852125686,
+      "learning_rate": 2e-05,
+      "loss": 5.5035,
+      "step": 996
+    },
+    {
+      "epoch": 0.06687460173726398,
+      "grad_norm": 0.13858731313062048,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 997
+    },
+    {
+      "epoch": 0.06694167756648892,
+      "grad_norm": 0.14505483746317052,
+      "learning_rate": 2e-05,
+      "loss": 5.5129,
+      "step": 998
+    },
+    {
+      "epoch": 0.06700875339571385,
+      "grad_norm": 0.1430188441396592,
+      "learning_rate": 2e-05,
+      "loss": 5.4757,
+      "step": 999
+    },
+    {
+      "epoch": 0.0670758292249388,
+      "grad_norm": 0.14405572597639804,
+      "learning_rate": 2e-05,
+      "loss": 5.5309,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06714290505416373,
+      "grad_norm": 0.14420543000214317,
+      "learning_rate": 2e-05,
+      "loss": 5.4797,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06720998088338867,
+      "grad_norm": 0.14851517857430163,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06727705671261361,
+      "grad_norm": 0.14583599683339143,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06734413254183855,
+      "grad_norm": 0.1555651851359692,
+      "learning_rate": 2e-05,
+      "loss": 5.4172,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06741120837106349,
+      "grad_norm": 0.13812575310833325,
+      "learning_rate": 2e-05,
+      "loss": 5.5737,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06747828420028842,
+      "grad_norm": 0.15711470856947715,
+      "learning_rate": 2e-05,
+      "loss": 5.5873,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06754536002951336,
+      "grad_norm": 0.13901211556871806,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0676124358587383,
+      "grad_norm": 0.13927864389636174,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06767951168796324,
+      "grad_norm": 0.14158542386555123,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06774658751718818,
+      "grad_norm": 0.14353046296885066,
+      "learning_rate": 2e-05,
+      "loss": 5.4718,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06781366334641312,
+      "grad_norm": 0.14204280486231566,
+      "learning_rate": 2e-05,
+      "loss": 5.513,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06788073917563806,
+      "grad_norm": 0.13444323188584192,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 1012
+    },
+    {
+      "epoch": 0.067947815004863,
+      "grad_norm": 0.13741491296283515,
+      "learning_rate": 2e-05,
+      "loss": 5.285,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06801489083408793,
+      "grad_norm": 0.13954649460686996,
+      "learning_rate": 2e-05,
+      "loss": 5.543,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06808196666331287,
+      "grad_norm": 0.14578867505470353,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06814904249253781,
+      "grad_norm": 0.14111504177081444,
+      "learning_rate": 2e-05,
+      "loss": 5.4701,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06821611832176275,
+      "grad_norm": 0.13955645828244653,
+      "learning_rate": 2e-05,
+      "loss": 5.5489,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06828319415098769,
+      "grad_norm": 0.14299837576813884,
+      "learning_rate": 2e-05,
+      "loss": 5.3295,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06835026998021262,
+      "grad_norm": 0.1418240516080218,
+      "learning_rate": 2e-05,
+      "loss": 5.5749,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06841734580943756,
+      "grad_norm": 0.14038716591908607,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0684844216386625,
+      "grad_norm": 0.1461168971838683,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06855149746788744,
+      "grad_norm": 0.14185332005520468,
+      "learning_rate": 2e-05,
+      "loss": 5.5131,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06861857329711239,
+      "grad_norm": 0.14319015260729892,
+      "learning_rate": 2e-05,
+      "loss": 5.5279,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06868564912633733,
+      "grad_norm": 0.13798329033655377,
+      "learning_rate": 2e-05,
+      "loss": 5.5017,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06875272495556227,
+      "grad_norm": 0.14199635737956007,
+      "learning_rate": 2e-05,
+      "loss": 5.4655,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06881980078478721,
+      "grad_norm": 0.13943549742355377,
+      "learning_rate": 2e-05,
+      "loss": 5.4443,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06888687661401215,
+      "grad_norm": 0.14198877913372665,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06895395244323708,
+      "grad_norm": 0.14212895181676885,
+      "learning_rate": 2e-05,
+      "loss": 5.5207,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06902102827246202,
+      "grad_norm": 0.14181549512601482,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06908810410168696,
+      "grad_norm": 0.13968785896108063,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0691551799309119,
+      "grad_norm": 0.1458894091883494,
+      "learning_rate": 2e-05,
+      "loss": 5.4799,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06922225576013684,
+      "grad_norm": 0.14662279348565962,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06928933158936178,
+      "grad_norm": 0.13688140857834163,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06935640741858672,
+      "grad_norm": 0.14181437484839982,
+      "learning_rate": 2e-05,
+      "loss": 5.4739,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06942348324781165,
+      "grad_norm": 0.1482169290577267,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06949055907703659,
+      "grad_norm": 0.14035351702567372,
+      "learning_rate": 2e-05,
+      "loss": 5.533,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06955763490626153,
+      "grad_norm": 0.14020746200098522,
+      "learning_rate": 2e-05,
+      "loss": 5.4794,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06962471073548647,
+      "grad_norm": 0.14544451998741595,
+      "learning_rate": 2e-05,
+      "loss": 5.4267,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06969178656471141,
+      "grad_norm": 0.14381063273595002,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06975886239393635,
+      "grad_norm": 0.14765003466443793,
+      "learning_rate": 2e-05,
+      "loss": 5.3775,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06982593822316129,
+      "grad_norm": 0.14156940712819174,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06989301405238622,
+      "grad_norm": 0.1430620947732063,
+      "learning_rate": 2e-05,
+      "loss": 5.3188,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06996008988161116,
+      "grad_norm": 0.15319921901894945,
+      "learning_rate": 2e-05,
+      "loss": 5.6305,
+      "step": 1043
+    },
+    {
+      "epoch": 0.0700271657108361,
+      "grad_norm": 0.14468005581288698,
+      "learning_rate": 2e-05,
+      "loss": 5.5308,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07009424154006104,
+      "grad_norm": 0.14272432692090467,
+      "learning_rate": 2e-05,
+      "loss": 5.4387,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07016131736928598,
+      "grad_norm": 0.1463843713355547,
+      "learning_rate": 2e-05,
+      "loss": 5.4086,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07022839319851092,
+      "grad_norm": 0.14723893957380496,
+      "learning_rate": 2e-05,
+      "loss": 5.5626,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07029546902773585,
+      "grad_norm": 0.13950860044380903,
+      "learning_rate": 2e-05,
+      "loss": 5.5123,
+      "step": 1048
+    },
+    {
+      "epoch": 0.0703625448569608,
+      "grad_norm": 0.1442010900037428,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07042962068618573,
+      "grad_norm": 0.15409751471477767,
+      "learning_rate": 2e-05,
+      "loss": 5.4824,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07049669651541067,
+      "grad_norm": 0.14509094467912714,
+      "learning_rate": 2e-05,
+      "loss": 5.3989,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07056377234463561,
+      "grad_norm": 0.14526710838882345,
+      "learning_rate": 2e-05,
+      "loss": 5.3232,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07063084817386055,
+      "grad_norm": 0.1478504773026071,
+      "learning_rate": 2e-05,
+      "loss": 5.3862,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07069792400308549,
+      "grad_norm": 0.14638817503052967,
+      "learning_rate": 2e-05,
+      "loss": 5.3418,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07076499983231042,
+      "grad_norm": 0.13586301346644258,
+      "learning_rate": 2e-05,
+      "loss": 5.49,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07083207566153536,
+      "grad_norm": 0.14654483443128458,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0708991514907603,
+      "grad_norm": 0.14078132247927755,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07096622731998524,
+      "grad_norm": 0.1407883127223292,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07103330314921018,
+      "grad_norm": 0.1404300390243042,
+      "learning_rate": 2e-05,
+      "loss": 5.629,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07110037897843512,
+      "grad_norm": 0.14367245900913395,
+      "learning_rate": 2e-05,
+      "loss": 5.6084,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07116745480766005,
+      "grad_norm": 0.14015981020280674,
+      "learning_rate": 2e-05,
+      "loss": 5.3851,
+      "step": 1061
+    },
+    {
+      "epoch": 0.071234530636885,
+      "grad_norm": 0.13598385520107845,
+      "learning_rate": 2e-05,
+      "loss": 5.5021,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07130160646610993,
+      "grad_norm": 0.1398234397174872,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07136868229533487,
+      "grad_norm": 0.14761029735908618,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07143575812455981,
+      "grad_norm": 0.13777721500850715,
+      "learning_rate": 2e-05,
+      "loss": 5.5455,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07150283395378475,
+      "grad_norm": 0.1510481568356787,
+      "learning_rate": 2e-05,
+      "loss": 5.5922,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07156990978300969,
+      "grad_norm": 0.13725662016717374,
+      "learning_rate": 2e-05,
+      "loss": 5.5206,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07163698561223462,
+      "grad_norm": 0.14034392248368274,
+      "learning_rate": 2e-05,
+      "loss": 5.5104,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07170406144145958,
+      "grad_norm": 0.13452173211377477,
+      "learning_rate": 2e-05,
+      "loss": 5.3784,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07177113727068452,
+      "grad_norm": 0.14377743296526738,
+      "learning_rate": 2e-05,
+      "loss": 5.412,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07183821309990945,
+      "grad_norm": 0.14796813518763766,
+      "learning_rate": 2e-05,
+      "loss": 5.4716,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07190528892913439,
+      "grad_norm": 0.13877900951054004,
+      "learning_rate": 2e-05,
+      "loss": 5.5335,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07197236475835933,
+      "grad_norm": 0.1418218083548077,
+      "learning_rate": 2e-05,
+      "loss": 5.4379,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07203944058758427,
+      "grad_norm": 0.13808535465538047,
+      "learning_rate": 2e-05,
+      "loss": 5.3946,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07210651641680921,
+      "grad_norm": 0.13787068354150553,
+      "learning_rate": 2e-05,
+      "loss": 5.4985,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07217359224603415,
+      "grad_norm": 0.14553252974202666,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07224066807525908,
+      "grad_norm": 0.14126519970976545,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07230774390448402,
+      "grad_norm": 0.14386066438497633,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07237481973370896,
+      "grad_norm": 0.14319137965131715,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0724418955629339,
+      "grad_norm": 0.145466636132673,
+      "learning_rate": 2e-05,
+      "loss": 5.4764,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07250897139215884,
+      "grad_norm": 0.1383770266311131,
+      "learning_rate": 2e-05,
+      "loss": 5.4225,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07257604722138378,
+      "grad_norm": 0.1388380094793619,
+      "learning_rate": 2e-05,
+      "loss": 5.5216,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07264312305060872,
+      "grad_norm": 0.143174754656856,
+      "learning_rate": 2e-05,
+      "loss": 5.6383,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07271019887983365,
+      "grad_norm": 0.14270063662045063,
+      "learning_rate": 2e-05,
+      "loss": 5.5725,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07277727470905859,
+      "grad_norm": 0.142513594273265,
+      "learning_rate": 2e-05,
+      "loss": 5.3141,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07284435053828353,
+      "grad_norm": 0.13813919071852465,
+      "learning_rate": 2e-05,
+      "loss": 5.5079,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07291142636750847,
+      "grad_norm": 0.14001492106441765,
+      "learning_rate": 2e-05,
+      "loss": 5.37,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07297850219673341,
+      "grad_norm": 0.14833141785339185,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07304557802595835,
+      "grad_norm": 0.14319809727780125,
+      "learning_rate": 2e-05,
+      "loss": 5.6504,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07311265385518328,
+      "grad_norm": 0.15023110031578885,
+      "learning_rate": 2e-05,
+      "loss": 5.4674,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07317972968440822,
+      "grad_norm": 0.14391542059478438,
+      "learning_rate": 2e-05,
+      "loss": 5.345,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07324680551363316,
+      "grad_norm": 0.13875135309894027,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0733138813428581,
+      "grad_norm": 0.14617946371935692,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07338095717208304,
+      "grad_norm": 0.14352086030089384,
+      "learning_rate": 2e-05,
+      "loss": 5.4158,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07344803300130798,
+      "grad_norm": 0.1407758648836727,
+      "learning_rate": 2e-05,
+      "loss": 5.5167,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07351510883053292,
+      "grad_norm": 0.13690388037278067,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07358218465975785,
+      "grad_norm": 0.14326613200988836,
+      "learning_rate": 2e-05,
+      "loss": 5.5469,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07364926048898279,
+      "grad_norm": 0.14025574217353903,
+      "learning_rate": 2e-05,
+      "loss": 5.4878,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07371633631820773,
+      "grad_norm": 0.13826395568501904,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07378341214743267,
+      "grad_norm": 0.1412223290548939,
+      "learning_rate": 2e-05,
+      "loss": 5.548,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07385048797665761,
+      "grad_norm": 0.13571291285795697,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07391756380588255,
+      "grad_norm": 0.13831213480833468,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07398463963510749,
+      "grad_norm": 0.14504910093075676,
+      "learning_rate": 2e-05,
+      "loss": 5.5017,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07405171546433242,
+      "grad_norm": 0.14168429083685968,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07411879129355736,
+      "grad_norm": 0.14469188290322638,
+      "learning_rate": 2e-05,
+      "loss": 5.5204,
+      "step": 1105
+    },
+    {
+      "epoch": 0.0741858671227823,
+      "grad_norm": 0.14185924515940823,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07425294295200724,
+      "grad_norm": 0.14612197526895698,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07432001878123218,
+      "grad_norm": 0.14121288580058683,
+      "learning_rate": 2e-05,
+      "loss": 5.3779,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07438709461045712,
+      "grad_norm": 0.14943595548179592,
+      "learning_rate": 2e-05,
+      "loss": 5.3419,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07445417043968205,
+      "grad_norm": 0.1396906667646618,
+      "learning_rate": 2e-05,
+      "loss": 5.2597,
+      "step": 1110
+    },
+    {
+      "epoch": 0.074521246268907,
+      "grad_norm": 0.14010939088213667,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07458832209813193,
+      "grad_norm": 0.14310651381279835,
+      "learning_rate": 2e-05,
+      "loss": 5.5552,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07465539792735687,
+      "grad_norm": 0.14154181175386166,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07472247375658182,
+      "grad_norm": 0.1388526904015761,
+      "learning_rate": 2e-05,
+      "loss": 5.4133,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07478954958580676,
+      "grad_norm": 0.13943142320190335,
+      "learning_rate": 2e-05,
+      "loss": 5.6327,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0748566254150317,
+      "grad_norm": 0.14196690009070131,
+      "learning_rate": 2e-05,
+      "loss": 5.5694,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07492370124425664,
+      "grad_norm": 0.14172971013563768,
+      "learning_rate": 2e-05,
+      "loss": 5.4827,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07499077707348158,
+      "grad_norm": 0.14921824667168682,
+      "learning_rate": 2e-05,
+      "loss": 5.5972,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07505785290270651,
+      "grad_norm": 0.14266500204850666,
+      "learning_rate": 2e-05,
+      "loss": 5.697,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07512492873193145,
+      "grad_norm": 0.1399058048528367,
+      "learning_rate": 2e-05,
+      "loss": 5.53,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07519200456115639,
+      "grad_norm": 0.1379672146718741,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07525908039038133,
+      "grad_norm": 0.13669976798965122,
+      "learning_rate": 2e-05,
+      "loss": 5.5489,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07532615621960627,
+      "grad_norm": 0.1498855010450185,
+      "learning_rate": 2e-05,
+      "loss": 5.4513,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07539323204883121,
+      "grad_norm": 0.141315877085081,
+      "learning_rate": 2e-05,
+      "loss": 5.4295,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07546030787805615,
+      "grad_norm": 0.14970094927335933,
+      "learning_rate": 2e-05,
+      "loss": 5.5392,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07552738370728108,
+      "grad_norm": 0.13937016528933227,
+      "learning_rate": 2e-05,
+      "loss": 5.4836,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07559445953650602,
+      "grad_norm": 0.14454906140057708,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07566153536573096,
+      "grad_norm": 0.14364119520212693,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 1128
+    },
+    {
+      "epoch": 0.0757286111949559,
+      "grad_norm": 0.14686814266625314,
+      "learning_rate": 2e-05,
+      "loss": 5.5364,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07579568702418084,
+      "grad_norm": 0.14205250678237097,
+      "learning_rate": 2e-05,
+      "loss": 5.3681,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07586276285340578,
+      "grad_norm": 0.1417902049569719,
+      "learning_rate": 2e-05,
+      "loss": 5.4931,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07592983868263072,
+      "grad_norm": 0.14015524254332487,
+      "learning_rate": 2e-05,
+      "loss": 5.454,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07599691451185565,
+      "grad_norm": 0.14006062815305553,
+      "learning_rate": 2e-05,
+      "loss": 5.4379,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07606399034108059,
+      "grad_norm": 0.1412175960450175,
+      "learning_rate": 2e-05,
+      "loss": 5.5073,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07613106617030553,
+      "grad_norm": 0.14836329420713437,
+      "learning_rate": 2e-05,
+      "loss": 5.4641,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07619814199953047,
+      "grad_norm": 0.13711186688293733,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07626521782875541,
+      "grad_norm": 0.15034168932998612,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07633229365798035,
+      "grad_norm": 0.14757455320995166,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07639936948720528,
+      "grad_norm": 0.14295354758394763,
+      "learning_rate": 2e-05,
+      "loss": 5.3849,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07646644531643022,
+      "grad_norm": 0.14051213201591073,
+      "learning_rate": 2e-05,
+      "loss": 5.5789,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07653352114565516,
+      "grad_norm": 0.1451991728939992,
+      "learning_rate": 2e-05,
+      "loss": 5.5705,
+      "step": 1141
+    },
+    {
+      "epoch": 0.0766005969748801,
+      "grad_norm": 0.14744993143529694,
+      "learning_rate": 2e-05,
+      "loss": 5.4072,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07666767280410504,
+      "grad_norm": 0.1359864047610627,
+      "learning_rate": 2e-05,
+      "loss": 5.5598,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07673474863332998,
+      "grad_norm": 0.14329474704277556,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07680182446255492,
+      "grad_norm": 0.1437953329730129,
+      "learning_rate": 2e-05,
+      "loss": 5.5751,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07686890029177985,
+      "grad_norm": 0.13566370020749074,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07693597612100479,
+      "grad_norm": 0.13865861667817359,
+      "learning_rate": 2e-05,
+      "loss": 5.5791,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07700305195022973,
+      "grad_norm": 0.14137836957841587,
+      "learning_rate": 2e-05,
+      "loss": 5.4182,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07707012777945467,
+      "grad_norm": 0.13438963155897532,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07713720360867961,
+      "grad_norm": 0.13971576616661888,
+      "learning_rate": 2e-05,
+      "loss": 5.5189,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07720427943790455,
+      "grad_norm": 0.15071537608823352,
+      "learning_rate": 2e-05,
+      "loss": 5.6116,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07727135526712949,
+      "grad_norm": 0.1497273942093903,
+      "learning_rate": 2e-05,
+      "loss": 5.4276,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07733843109635442,
+      "grad_norm": 0.14675851494809283,
+      "learning_rate": 2e-05,
+      "loss": 5.3975,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07740550692557936,
+      "grad_norm": 0.15775326150127558,
+      "learning_rate": 2e-05,
+      "loss": 5.3885,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0774725827548043,
+      "grad_norm": 0.15763459469359942,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07753965858402924,
+      "grad_norm": 0.14972605150511542,
+      "learning_rate": 2e-05,
+      "loss": 5.7013,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07760673441325418,
+      "grad_norm": 0.14941510795187188,
+      "learning_rate": 2e-05,
+      "loss": 5.6841,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07767381024247912,
+      "grad_norm": 0.14388876023524758,
+      "learning_rate": 2e-05,
+      "loss": 5.4145,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07774088607170405,
+      "grad_norm": 0.1414274638635389,
+      "learning_rate": 2e-05,
+      "loss": 5.5352,
+      "step": 1159
+    },
+    {
+      "epoch": 0.077807961900929,
+      "grad_norm": 0.1448866724512301,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07787503773015395,
+      "grad_norm": 0.14445148029316138,
+      "learning_rate": 2e-05,
+      "loss": 5.4582,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07794211355937888,
+      "grad_norm": 0.13867493141039325,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07800918938860382,
+      "grad_norm": 0.14241103154775353,
+      "learning_rate": 2e-05,
+      "loss": 5.5174,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07807626521782876,
+      "grad_norm": 0.137269020714671,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0781433410470537,
+      "grad_norm": 0.1416581383252918,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07821041687627864,
+      "grad_norm": 0.14282486212942994,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07827749270550358,
+      "grad_norm": 0.14576104632973472,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07834456853472851,
+      "grad_norm": 0.1439491106233289,
+      "learning_rate": 2e-05,
+      "loss": 5.3542,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07841164436395345,
+      "grad_norm": 0.14029559017199444,
+      "learning_rate": 2e-05,
+      "loss": 5.5309,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07847872019317839,
+      "grad_norm": 0.13861767576638004,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07854579602240333,
+      "grad_norm": 0.13605406254857547,
+      "learning_rate": 2e-05,
+      "loss": 5.402,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07861287185162827,
+      "grad_norm": 0.1460571393535904,
+      "learning_rate": 2e-05,
+      "loss": 5.5562,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07867994768085321,
+      "grad_norm": 0.1427100012264957,
+      "learning_rate": 2e-05,
+      "loss": 5.3891,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07874702351007815,
+      "grad_norm": 0.13862309871492792,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07881409933930308,
+      "grad_norm": 0.14165401441640516,
+      "learning_rate": 2e-05,
+      "loss": 5.6267,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07888117516852802,
+      "grad_norm": 0.14428629101664345,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07894825099775296,
+      "grad_norm": 0.13708697694439304,
+      "learning_rate": 2e-05,
+      "loss": 5.4609,
+      "step": 1177
+    },
+    {
+      "epoch": 0.0790153268269779,
+      "grad_norm": 0.14219321351245726,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07908240265620284,
+      "grad_norm": 0.14475176213082103,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07914947848542778,
+      "grad_norm": 0.14495323607053287,
+      "learning_rate": 2e-05,
+      "loss": 5.6,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07921655431465272,
+      "grad_norm": 0.1469052936398813,
+      "learning_rate": 2e-05,
+      "loss": 5.3686,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07928363014387765,
+      "grad_norm": 0.14016774976625285,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07935070597310259,
+      "grad_norm": 0.14581741620028651,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07941778180232753,
+      "grad_norm": 0.14561274071106803,
+      "learning_rate": 2e-05,
+      "loss": 5.4177,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07948485763155247,
+      "grad_norm": 0.14248876696428192,
+      "learning_rate": 2e-05,
+      "loss": 5.5295,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07955193346077741,
+      "grad_norm": 0.14258629549593332,
+      "learning_rate": 2e-05,
+      "loss": 5.5004,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07961900929000235,
+      "grad_norm": 0.15171497412873902,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07968608511922728,
+      "grad_norm": 0.13765355491737286,
+      "learning_rate": 2e-05,
+      "loss": 5.5511,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07975316094845222,
+      "grad_norm": 0.13331337694325643,
+      "learning_rate": 2e-05,
+      "loss": 5.45,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07982023677767716,
+      "grad_norm": 0.14369864811228147,
+      "learning_rate": 2e-05,
+      "loss": 5.4154,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0798873126069021,
+      "grad_norm": 0.13934538108135266,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07995438843612704,
+      "grad_norm": 0.1387065288897377,
+      "learning_rate": 2e-05,
+      "loss": 5.4245,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08002146426535198,
+      "grad_norm": 0.13725346030769553,
+      "learning_rate": 2e-05,
+      "loss": 5.4823,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08008854009457692,
+      "grad_norm": 0.14181522105240052,
+      "learning_rate": 2e-05,
+      "loss": 5.5834,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08015561592380185,
+      "grad_norm": 0.13933488499479074,
+      "learning_rate": 2e-05,
+      "loss": 5.5666,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08022269175302679,
+      "grad_norm": 0.1450222428363668,
+      "learning_rate": 2e-05,
+      "loss": 5.3183,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08028976758225173,
+      "grad_norm": 0.13456240498817187,
+      "learning_rate": 2e-05,
+      "loss": 5.5776,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08035684341147667,
+      "grad_norm": 0.13995869787569398,
+      "learning_rate": 2e-05,
+      "loss": 5.3833,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08042391924070161,
+      "grad_norm": 0.1356441202877071,
+      "learning_rate": 2e-05,
+      "loss": 5.418,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08049099506992655,
+      "grad_norm": 0.14863068702431556,
+      "learning_rate": 2e-05,
+      "loss": 5.3933,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08055807089915148,
+      "grad_norm": 0.15243129602460476,
+      "learning_rate": 2e-05,
+      "loss": 5.3849,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08062514672837642,
+      "grad_norm": 0.14559793057631934,
+      "learning_rate": 2e-05,
+      "loss": 5.4994,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08069222255760136,
+      "grad_norm": 0.14444955043448668,
+      "learning_rate": 2e-05,
+      "loss": 5.5834,
+      "step": 1203
+    },
+    {
+      "epoch": 0.0807592983868263,
+      "grad_norm": 0.13988687565641014,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08082637421605125,
+      "grad_norm": 0.14420819305763274,
+      "learning_rate": 2e-05,
+      "loss": 5.4773,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08089345004527619,
+      "grad_norm": 0.14742108493999984,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08096052587450113,
+      "grad_norm": 0.13974213138278402,
+      "learning_rate": 2e-05,
+      "loss": 5.4927,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08102760170372607,
+      "grad_norm": 0.14361355206249993,
+      "learning_rate": 2e-05,
+      "loss": 5.5512,
+      "step": 1208
+    },
+    {
+      "epoch": 0.081094677532951,
+      "grad_norm": 0.1439746585907705,
+      "learning_rate": 2e-05,
+      "loss": 5.4776,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08116175336217595,
+      "grad_norm": 0.14224476792250978,
+      "learning_rate": 2e-05,
+      "loss": 5.5649,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08122882919140088,
+      "grad_norm": 0.14180373490556625,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08129590502062582,
+      "grad_norm": 0.1421707939155885,
+      "learning_rate": 2e-05,
+      "loss": 5.4037,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08136298084985076,
+      "grad_norm": 0.15568110549071132,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0814300566790757,
+      "grad_norm": 0.14097309466557634,
+      "learning_rate": 2e-05,
+      "loss": 5.3046,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08149713250830064,
+      "grad_norm": 0.13690270879247768,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08156420833752558,
+      "grad_norm": 0.1383294024200251,
+      "learning_rate": 2e-05,
+      "loss": 5.374,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08163128416675051,
+      "grad_norm": 0.13731438562875722,
+      "learning_rate": 2e-05,
+      "loss": 5.5753,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08169835999597545,
+      "grad_norm": 0.1344968347752966,
+      "learning_rate": 2e-05,
+      "loss": 5.401,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08176543582520039,
+      "grad_norm": 0.14580935846297446,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08183251165442533,
+      "grad_norm": 0.13862608763428322,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08189958748365027,
+      "grad_norm": 0.13775352933912643,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08196666331287521,
+      "grad_norm": 0.13795105633831883,
+      "learning_rate": 2e-05,
+      "loss": 5.6114,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08203373914210015,
+      "grad_norm": 0.13882046307373774,
+      "learning_rate": 2e-05,
+      "loss": 5.516,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08210081497132508,
+      "grad_norm": 0.13726080734303064,
+      "learning_rate": 2e-05,
+      "loss": 5.3876,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08216789080055002,
+      "grad_norm": 0.1357022412430794,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08223496662977496,
+      "grad_norm": 0.1397268904644595,
+      "learning_rate": 2e-05,
+      "loss": 5.618,
+      "step": 1226
+    },
+    {
+      "epoch": 0.0823020424589999,
+      "grad_norm": 0.13905483205575644,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08236911828822484,
+      "grad_norm": 0.13957779800863448,
+      "learning_rate": 2e-05,
+      "loss": 5.3884,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08243619411744978,
+      "grad_norm": 0.14016284661818446,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08250326994667471,
+      "grad_norm": 0.1419329863054458,
+      "learning_rate": 2e-05,
+      "loss": 5.464,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08257034577589965,
+      "grad_norm": 0.1381730352621069,
+      "learning_rate": 2e-05,
+      "loss": 5.3382,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08263742160512459,
+      "grad_norm": 0.13902620071228722,
+      "learning_rate": 2e-05,
+      "loss": 5.5059,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08270449743434953,
+      "grad_norm": 0.14448608486257666,
+      "learning_rate": 2e-05,
+      "loss": 5.2856,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08277157326357447,
+      "grad_norm": 0.13700705800394947,
+      "learning_rate": 2e-05,
+      "loss": 5.4649,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08283864909279941,
+      "grad_norm": 0.13757637252375599,
+      "learning_rate": 2e-05,
+      "loss": 5.5843,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08290572492202435,
+      "grad_norm": 0.13850377329026067,
+      "learning_rate": 2e-05,
+      "loss": 5.3955,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08297280075124928,
+      "grad_norm": 0.14160081985817075,
+      "learning_rate": 2e-05,
+      "loss": 5.5125,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08303987658047422,
+      "grad_norm": 0.1322479742677396,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08310695240969916,
+      "grad_norm": 0.1454671179330801,
+      "learning_rate": 2e-05,
+      "loss": 5.4719,
+      "step": 1239
+    },
+    {
+      "epoch": 0.0831740282389241,
+      "grad_norm": 0.14266170253037294,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08324110406814904,
+      "grad_norm": 0.13585619447341524,
+      "learning_rate": 2e-05,
+      "loss": 5.5123,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08330817989737398,
+      "grad_norm": 0.14849820823452817,
+      "learning_rate": 2e-05,
+      "loss": 5.5173,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08337525572659892,
+      "grad_norm": 0.14543715953344263,
+      "learning_rate": 2e-05,
+      "loss": 5.5148,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08344233155582385,
+      "grad_norm": 0.1459546125475233,
+      "learning_rate": 2e-05,
+      "loss": 5.552,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08350940738504879,
+      "grad_norm": 0.1386335120351884,
+      "learning_rate": 2e-05,
+      "loss": 5.6138,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08357648321427373,
+      "grad_norm": 0.13960381761972612,
+      "learning_rate": 2e-05,
+      "loss": 5.6059,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08364355904349867,
+      "grad_norm": 0.141316925471431,
+      "learning_rate": 2e-05,
+      "loss": 5.2984,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08371063487272361,
+      "grad_norm": 0.1474974535827872,
+      "learning_rate": 2e-05,
+      "loss": 5.5959,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08377771070194855,
+      "grad_norm": 0.13451415781274684,
+      "learning_rate": 2e-05,
+      "loss": 5.5253,
+      "step": 1249
+    },
+    {
+      "epoch": 0.0838447865311735,
+      "grad_norm": 0.14660248870583265,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08391186236039844,
+      "grad_norm": 0.1366119473228817,
+      "learning_rate": 2e-05,
+      "loss": 5.5314,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08397893818962338,
+      "grad_norm": 0.1376404335669714,
+      "learning_rate": 2e-05,
+      "loss": 5.5085,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08404601401884831,
+      "grad_norm": 0.14697870564318216,
+      "learning_rate": 2e-05,
+      "loss": 5.5257,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08411308984807325,
+      "grad_norm": 0.1385388928126056,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08418016567729819,
+      "grad_norm": 0.13864248997072806,
+      "learning_rate": 2e-05,
+      "loss": 5.4451,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08424724150652313,
+      "grad_norm": 0.13928840466967118,
+      "learning_rate": 2e-05,
+      "loss": 5.5414,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08431431733574807,
+      "grad_norm": 0.1438527501394235,
+      "learning_rate": 2e-05,
+      "loss": 5.5946,
+      "step": 1257
+    },
+    {
+      "epoch": 0.084381393164973,
+      "grad_norm": 0.1483817494510357,
+      "learning_rate": 2e-05,
+      "loss": 5.5329,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08444846899419794,
+      "grad_norm": 0.1436334211930949,
+      "learning_rate": 2e-05,
+      "loss": 5.4315,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08451554482342288,
+      "grad_norm": 0.13794680458943412,
+      "learning_rate": 2e-05,
+      "loss": 5.3955,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08458262065264782,
+      "grad_norm": 0.14038165944998016,
+      "learning_rate": 2e-05,
+      "loss": 5.5834,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08464969648187276,
+      "grad_norm": 0.14930209685151472,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0847167723110977,
+      "grad_norm": 0.13962483816095564,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08478384814032264,
+      "grad_norm": 0.13859763317349108,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08485092396954758,
+      "grad_norm": 0.1464107327685371,
+      "learning_rate": 2e-05,
+      "loss": 5.3452,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08491799979877251,
+      "grad_norm": 0.14126231353237328,
+      "learning_rate": 2e-05,
+      "loss": 5.5179,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08498507562799745,
+      "grad_norm": 0.14505936282697152,
+      "learning_rate": 2e-05,
+      "loss": 5.5057,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08505215145722239,
+      "grad_norm": 0.14875912251143386,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08511922728644733,
+      "grad_norm": 0.1389754046925835,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08518630311567227,
+      "grad_norm": 0.13967068183000111,
+      "learning_rate": 2e-05,
+      "loss": 5.5094,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0852533789448972,
+      "grad_norm": 0.1488209200460791,
+      "learning_rate": 2e-05,
+      "loss": 5.3736,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08532045477412215,
+      "grad_norm": 0.14794142692671336,
+      "learning_rate": 2e-05,
+      "loss": 5.5669,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08538753060334708,
+      "grad_norm": 0.1422316034288501,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08545460643257202,
+      "grad_norm": 0.14379568630833592,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08552168226179696,
+      "grad_norm": 0.14849849902957413,
+      "learning_rate": 2e-05,
+      "loss": 5.3962,
+      "step": 1275
+    },
+    {
+      "epoch": 0.0855887580910219,
+      "grad_norm": 0.14141089016964428,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08565583392024684,
+      "grad_norm": 0.1471683328338236,
+      "learning_rate": 2e-05,
+      "loss": 5.4604,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08572290974947178,
+      "grad_norm": 0.1434382386596933,
+      "learning_rate": 2e-05,
+      "loss": 5.494,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08578998557869671,
+      "grad_norm": 0.15575258717989754,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08585706140792165,
+      "grad_norm": 0.1472395239208941,
+      "learning_rate": 2e-05,
+      "loss": 5.4326,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08592413723714659,
+      "grad_norm": 0.1536841213581419,
+      "learning_rate": 2e-05,
+      "loss": 5.4758,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08599121306637153,
+      "grad_norm": 0.15743232549146766,
+      "learning_rate": 2e-05,
+      "loss": 5.5082,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08605828889559647,
+      "grad_norm": 0.14011920242559417,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08612536472482141,
+      "grad_norm": 0.14984687688555404,
+      "learning_rate": 2e-05,
+      "loss": 5.4604,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08619244055404635,
+      "grad_norm": 0.1522986003276844,
+      "learning_rate": 2e-05,
+      "loss": 5.3815,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08625951638327128,
+      "grad_norm": 0.14845541629021203,
+      "learning_rate": 2e-05,
+      "loss": 5.535,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08632659221249622,
+      "grad_norm": 0.14804710367644944,
+      "learning_rate": 2e-05,
+      "loss": 5.2777,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08639366804172116,
+      "grad_norm": 0.14857635704508282,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 1288
+    },
+    {
+      "epoch": 0.0864607438709461,
+      "grad_norm": 0.1440843472293447,
+      "learning_rate": 2e-05,
+      "loss": 5.4168,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08652781970017104,
+      "grad_norm": 0.14929841899364843,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08659489552939598,
+      "grad_norm": 0.14950326813066775,
+      "learning_rate": 2e-05,
+      "loss": 5.3887,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08666197135862092,
+      "grad_norm": 0.14338849328799566,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08672904718784585,
+      "grad_norm": 0.14272238210600874,
+      "learning_rate": 2e-05,
+      "loss": 5.5214,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08679612301707079,
+      "grad_norm": 0.14483528809254628,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08686319884629573,
+      "grad_norm": 0.14826009932239873,
+      "learning_rate": 2e-05,
+      "loss": 5.4484,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08693027467552068,
+      "grad_norm": 0.1417323097171973,
+      "learning_rate": 2e-05,
+      "loss": 5.5627,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08699735050474562,
+      "grad_norm": 0.14416745275351678,
+      "learning_rate": 2e-05,
+      "loss": 5.4807,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08706442633397056,
+      "grad_norm": 0.1454610647293804,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 1298
+    },
+    {
+      "epoch": 0.0871315021631955,
+      "grad_norm": 0.1499431951937672,
+      "learning_rate": 2e-05,
+      "loss": 5.5077,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08719857799242044,
+      "grad_norm": 0.14067372191873329,
+      "learning_rate": 2e-05,
+      "loss": 5.5985,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08726565382164538,
+      "grad_norm": 0.14346860534227457,
+      "learning_rate": 2e-05,
+      "loss": 5.4961,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08733272965087031,
+      "grad_norm": 0.13948958944753448,
+      "learning_rate": 2e-05,
+      "loss": 5.4222,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08739980548009525,
+      "grad_norm": 0.14791815130034747,
+      "learning_rate": 2e-05,
+      "loss": 5.5149,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08746688130932019,
+      "grad_norm": 0.14026761410621832,
+      "learning_rate": 2e-05,
+      "loss": 5.3621,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08753395713854513,
+      "grad_norm": 0.14170758880151219,
+      "learning_rate": 2e-05,
+      "loss": 5.5562,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08760103296777007,
+      "grad_norm": 0.14270336080594997,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 1306
+    },
+    {
+      "epoch": 0.087668108796995,
+      "grad_norm": 0.14757315957068892,
+      "learning_rate": 2e-05,
+      "loss": 5.515,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08773518462621994,
+      "grad_norm": 0.1393286351501242,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08780226045544488,
+      "grad_norm": 0.13538465016769555,
+      "learning_rate": 2e-05,
+      "loss": 5.5524,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08786933628466982,
+      "grad_norm": 0.14315300209426393,
+      "learning_rate": 2e-05,
+      "loss": 5.4182,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08793641211389476,
+      "grad_norm": 0.14651034617221165,
+      "learning_rate": 2e-05,
+      "loss": 5.5472,
+      "step": 1311
+    },
+    {
+      "epoch": 0.0880034879431197,
+      "grad_norm": 0.14091980011421534,
+      "learning_rate": 2e-05,
+      "loss": 5.4431,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08807056377234464,
+      "grad_norm": 0.14962838203591108,
+      "learning_rate": 2e-05,
+      "loss": 5.4233,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08813763960156958,
+      "grad_norm": 0.15037937406690022,
+      "learning_rate": 2e-05,
+      "loss": 5.5357,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08820471543079451,
+      "grad_norm": 0.1385031835115251,
+      "learning_rate": 2e-05,
+      "loss": 5.4967,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08827179126001945,
+      "grad_norm": 0.14415758347376742,
+      "learning_rate": 2e-05,
+      "loss": 5.3607,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08833886708924439,
+      "grad_norm": 0.146394607543345,
+      "learning_rate": 2e-05,
+      "loss": 5.5424,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08840594291846933,
+      "grad_norm": 0.1464684196111146,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08847301874769427,
+      "grad_norm": 0.1446010863015858,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 1319
+    },
+    {
+      "epoch": 0.0885400945769192,
+      "grad_norm": 0.1409837732684613,
+      "learning_rate": 2e-05,
+      "loss": 5.5623,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08860717040614415,
+      "grad_norm": 0.14800656384437572,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08867424623536908,
+      "grad_norm": 0.13927893940429178,
+      "learning_rate": 2e-05,
+      "loss": 5.4167,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08874132206459402,
+      "grad_norm": 0.14336809332156725,
+      "learning_rate": 2e-05,
+      "loss": 5.4193,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08880839789381896,
+      "grad_norm": 0.146472791808987,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 1324
+    },
+    {
+      "epoch": 0.0888754737230439,
+      "grad_norm": 0.13980344274198794,
+      "learning_rate": 2e-05,
+      "loss": 5.3657,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08894254955226884,
+      "grad_norm": 0.13946413683056713,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08900962538149378,
+      "grad_norm": 0.13785693630517237,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08907670121071871,
+      "grad_norm": 0.14487621962101752,
+      "learning_rate": 2e-05,
+      "loss": 5.5019,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08914377703994365,
+      "grad_norm": 0.14655467997402982,
+      "learning_rate": 2e-05,
+      "loss": 5.3784,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08921085286916859,
+      "grad_norm": 0.14282406432797431,
+      "learning_rate": 2e-05,
+      "loss": 5.4847,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08927792869839353,
+      "grad_norm": 0.14070380212180883,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08934500452761847,
+      "grad_norm": 0.13683969247205371,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 1332
+    },
+    {
+      "epoch": 0.0894120803568434,
+      "grad_norm": 0.1415213448382773,
+      "learning_rate": 2e-05,
+      "loss": 5.3499,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08947915618606835,
+      "grad_norm": 0.1492196024626836,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08954623201529328,
+      "grad_norm": 0.14010746746139968,
+      "learning_rate": 2e-05,
+      "loss": 5.698,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08961330784451822,
+      "grad_norm": 0.1485465918122867,
+      "learning_rate": 2e-05,
+      "loss": 5.5733,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08968038367374316,
+      "grad_norm": 0.14883575388857279,
+      "learning_rate": 2e-05,
+      "loss": 5.4234,
+      "step": 1337
+    },
+    {
+      "epoch": 0.0897474595029681,
+      "grad_norm": 0.13837125847988602,
+      "learning_rate": 2e-05,
+      "loss": 5.3907,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08981453533219304,
+      "grad_norm": 0.14908854717470627,
+      "learning_rate": 2e-05,
+      "loss": 5.6591,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08988161116141798,
+      "grad_norm": 0.15029909036932895,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08994868699064293,
+      "grad_norm": 0.13659116517487482,
+      "learning_rate": 2e-05,
+      "loss": 5.6395,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09001576281986787,
+      "grad_norm": 0.1470585989345011,
+      "learning_rate": 2e-05,
+      "loss": 5.5943,
+      "step": 1342
+    },
+    {
+      "epoch": 0.0900828386490928,
+      "grad_norm": 0.1440970320916433,
+      "learning_rate": 2e-05,
+      "loss": 5.4318,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09014991447831774,
+      "grad_norm": 0.14544803253010719,
+      "learning_rate": 2e-05,
+      "loss": 5.4892,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09021699030754268,
+      "grad_norm": 0.14935013879638187,
+      "learning_rate": 2e-05,
+      "loss": 5.5691,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09028406613676762,
+      "grad_norm": 0.14070388530930056,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09035114196599256,
+      "grad_norm": 0.14402562059908094,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0904182177952175,
+      "grad_norm": 0.14552149564165368,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09048529362444244,
+      "grad_norm": 0.14310681695339325,
+      "learning_rate": 2e-05,
+      "loss": 5.5059,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09055236945366738,
+      "grad_norm": 0.13950865992279246,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09061944528289231,
+      "grad_norm": 0.14208513306434092,
+      "learning_rate": 2e-05,
+      "loss": 5.4837,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09068652111211725,
+      "grad_norm": 0.13982778232989176,
+      "learning_rate": 2e-05,
+      "loss": 5.4343,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09075359694134219,
+      "grad_norm": 0.14031506079600467,
+      "learning_rate": 2e-05,
+      "loss": 5.4884,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09082067277056713,
+      "grad_norm": 0.15088647973548525,
+      "learning_rate": 2e-05,
+      "loss": 5.3649,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09088774859979207,
+      "grad_norm": 0.13463564055952962,
+      "learning_rate": 2e-05,
+      "loss": 5.4304,
+      "step": 1355
+    },
+    {
+      "epoch": 0.090954824429017,
+      "grad_norm": 0.14149913910140366,
+      "learning_rate": 2e-05,
+      "loss": 5.4084,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09102190025824194,
+      "grad_norm": 0.1433544612229547,
+      "learning_rate": 2e-05,
+      "loss": 5.3883,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09108897608746688,
+      "grad_norm": 0.14346381541416334,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09115605191669182,
+      "grad_norm": 0.1365856383913997,
+      "learning_rate": 2e-05,
+      "loss": 5.5309,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09122312774591676,
+      "grad_norm": 0.1507498943639077,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 1360
+    },
+    {
+      "epoch": 0.0912902035751417,
+      "grad_norm": 0.1422688355936516,
+      "learning_rate": 2e-05,
+      "loss": 5.5128,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09135727940436664,
+      "grad_norm": 0.14632624161612934,
+      "learning_rate": 2e-05,
+      "loss": 5.483,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09142435523359158,
+      "grad_norm": 0.14134062101034955,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09149143106281651,
+      "grad_norm": 0.14993726814192915,
+      "learning_rate": 2e-05,
+      "loss": 5.4476,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09155850689204145,
+      "grad_norm": 0.14520501059016772,
+      "learning_rate": 2e-05,
+      "loss": 5.3734,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09162558272126639,
+      "grad_norm": 0.14209286145426178,
+      "learning_rate": 2e-05,
+      "loss": 5.4849,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09169265855049133,
+      "grad_norm": 0.14197068080504954,
+      "learning_rate": 2e-05,
+      "loss": 5.3586,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09175973437971627,
+      "grad_norm": 0.14567417513884753,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 1368
+    },
+    {
+      "epoch": 0.0918268102089412,
+      "grad_norm": 0.13846187603755725,
+      "learning_rate": 2e-05,
+      "loss": 5.4962,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09189388603816614,
+      "grad_norm": 0.13684859803714142,
+      "learning_rate": 2e-05,
+      "loss": 5.4262,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09196096186739108,
+      "grad_norm": 0.13547004176919145,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09202803769661602,
+      "grad_norm": 0.13704379942294992,
+      "learning_rate": 2e-05,
+      "loss": 5.4251,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09209511352584096,
+      "grad_norm": 0.14658912197712248,
+      "learning_rate": 2e-05,
+      "loss": 5.4321,
+      "step": 1373
+    },
+    {
+      "epoch": 0.0921621893550659,
+      "grad_norm": 0.14078754752370842,
+      "learning_rate": 2e-05,
+      "loss": 5.5168,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09222926518429084,
+      "grad_norm": 0.1373571887020972,
+      "learning_rate": 2e-05,
+      "loss": 5.3446,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09229634101351578,
+      "grad_norm": 0.13806212900216036,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09236341684274071,
+      "grad_norm": 0.1441362791309543,
+      "learning_rate": 2e-05,
+      "loss": 5.5375,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09243049267196565,
+      "grad_norm": 0.14221368434281853,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09249756850119059,
+      "grad_norm": 0.14251898868609914,
+      "learning_rate": 2e-05,
+      "loss": 5.5217,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09256464433041553,
+      "grad_norm": 0.13821790731105055,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09263172015964047,
+      "grad_norm": 0.1420218417801707,
+      "learning_rate": 2e-05,
+      "loss": 5.4498,
+      "step": 1381
+    },
+    {
+      "epoch": 0.0926987959888654,
+      "grad_norm": 0.1439368145498818,
+      "learning_rate": 2e-05,
+      "loss": 5.3649,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09276587181809035,
+      "grad_norm": 0.1478590258683936,
+      "learning_rate": 2e-05,
+      "loss": 5.4757,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09283294764731528,
+      "grad_norm": 0.14823345196361726,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09290002347654022,
+      "grad_norm": 0.1438452571782001,
+      "learning_rate": 2e-05,
+      "loss": 5.3932,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09296709930576516,
+      "grad_norm": 0.1351761783974815,
+      "learning_rate": 2e-05,
+      "loss": 5.3545,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09303417513499011,
+      "grad_norm": 0.1440153626235563,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09310125096421505,
+      "grad_norm": 0.15230903630949522,
+      "learning_rate": 2e-05,
+      "loss": 5.5669,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09316832679343999,
+      "grad_norm": 0.14587576696540575,
+      "learning_rate": 2e-05,
+      "loss": 5.3719,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09323540262266493,
+      "grad_norm": 0.1467190359080022,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09330247845188987,
+      "grad_norm": 0.15094003316282553,
+      "learning_rate": 2e-05,
+      "loss": 5.5228,
+      "step": 1391
+    },
+    {
+      "epoch": 0.0933695542811148,
+      "grad_norm": 0.1436135415786532,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09343663011033974,
+      "grad_norm": 0.1385222587434453,
+      "learning_rate": 2e-05,
+      "loss": 5.5118,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09350370593956468,
+      "grad_norm": 0.1470136271444273,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09357078176878962,
+      "grad_norm": 0.1536992646908051,
+      "learning_rate": 2e-05,
+      "loss": 5.4609,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09363785759801456,
+      "grad_norm": 0.14172583532555863,
+      "learning_rate": 2e-05,
+      "loss": 5.5036,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0937049334272395,
+      "grad_norm": 0.13956270672432364,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09377200925646444,
+      "grad_norm": 0.15048477968946694,
+      "learning_rate": 2e-05,
+      "loss": 5.5912,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09383908508568937,
+      "grad_norm": 0.14036206973798984,
+      "learning_rate": 2e-05,
+      "loss": 5.6266,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09390616091491431,
+      "grad_norm": 0.14789654039172342,
+      "learning_rate": 2e-05,
+      "loss": 5.4464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09397323674413925,
+      "grad_norm": 0.14304027797456179,
+      "learning_rate": 2e-05,
+      "loss": 5.3957,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09404031257336419,
+      "grad_norm": 0.13693673616388563,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09410738840258913,
+      "grad_norm": 0.1408985494160115,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09417446423181407,
+      "grad_norm": 0.14469354787296979,
+      "learning_rate": 2e-05,
+      "loss": 5.5589,
+      "step": 1404
+    },
+    {
+      "epoch": 0.094241540061039,
+      "grad_norm": 0.14906176205114333,
+      "learning_rate": 2e-05,
+      "loss": 5.5683,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09430861589026394,
+      "grad_norm": 0.13952881665570002,
+      "learning_rate": 2e-05,
+      "loss": 5.5546,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09437569171948888,
+      "grad_norm": 0.14160559105183865,
+      "learning_rate": 2e-05,
+      "loss": 5.4777,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09444276754871382,
+      "grad_norm": 0.1456131541734397,
+      "learning_rate": 2e-05,
+      "loss": 5.3941,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09450984337793876,
+      "grad_norm": 0.1436878539134071,
+      "learning_rate": 2e-05,
+      "loss": 5.4318,
+      "step": 1409
+    },
+    {
+      "epoch": 0.0945769192071637,
+      "grad_norm": 0.14379645516385584,
+      "learning_rate": 2e-05,
+      "loss": 5.3143,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09464399503638864,
+      "grad_norm": 0.14616398315369994,
+      "learning_rate": 2e-05,
+      "loss": 5.4936,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09471107086561358,
+      "grad_norm": 0.14309080850810996,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09477814669483851,
+      "grad_norm": 0.14277810194761908,
+      "learning_rate": 2e-05,
+      "loss": 5.5601,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09484522252406345,
+      "grad_norm": 0.14026133428598478,
+      "learning_rate": 2e-05,
+      "loss": 5.6291,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09491229835328839,
+      "grad_norm": 0.14347709978629164,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09497937418251333,
+      "grad_norm": 0.1372592112755479,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09504645001173827,
+      "grad_norm": 0.1452696323322186,
+      "learning_rate": 2e-05,
+      "loss": 5.5036,
+      "step": 1417
+    },
+    {
+      "epoch": 0.0951135258409632,
+      "grad_norm": 0.14146311632595904,
+      "learning_rate": 2e-05,
+      "loss": 5.4214,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09518060167018814,
+      "grad_norm": 0.13792383024976448,
+      "learning_rate": 2e-05,
+      "loss": 5.4614,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09524767749941308,
+      "grad_norm": 0.14556430157345474,
+      "learning_rate": 2e-05,
+      "loss": 5.4545,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09531475332863802,
+      "grad_norm": 0.1453958325506217,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09538182915786296,
+      "grad_norm": 0.15123047878587004,
+      "learning_rate": 2e-05,
+      "loss": 5.424,
+      "step": 1422
+    },
+    {
+      "epoch": 0.0954489049870879,
+      "grad_norm": 0.13693484174831866,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09551598081631284,
+      "grad_norm": 0.15194491067390015,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09558305664553778,
+      "grad_norm": 0.15595662181242484,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09565013247476271,
+      "grad_norm": 0.14376313069922783,
+      "learning_rate": 2e-05,
+      "loss": 5.4155,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09571720830398765,
+      "grad_norm": 0.14570914540211619,
+      "learning_rate": 2e-05,
+      "loss": 5.6691,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09578428413321259,
+      "grad_norm": 0.1501387387113468,
+      "learning_rate": 2e-05,
+      "loss": 5.5234,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09585135996243753,
+      "grad_norm": 0.1417732097367562,
+      "learning_rate": 2e-05,
+      "loss": 5.4644,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09591843579166247,
+      "grad_norm": 0.1461061811047711,
+      "learning_rate": 2e-05,
+      "loss": 5.4655,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0959855116208874,
+      "grad_norm": 0.14521329841012698,
+      "learning_rate": 2e-05,
+      "loss": 5.5397,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09605258745011236,
+      "grad_norm": 0.14250003719317686,
+      "learning_rate": 2e-05,
+      "loss": 5.4671,
+      "step": 1432
+    },
+    {
+      "epoch": 0.0961196632793373,
+      "grad_norm": 0.1444623618466229,
+      "learning_rate": 2e-05,
+      "loss": 5.5584,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09618673910856224,
+      "grad_norm": 0.1558963611723596,
+      "learning_rate": 2e-05,
+      "loss": 5.4584,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09625381493778717,
+      "grad_norm": 0.1425850999994635,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09632089076701211,
+      "grad_norm": 0.1443912791393191,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09638796659623705,
+      "grad_norm": 0.14682650254472662,
+      "learning_rate": 2e-05,
+      "loss": 5.5409,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09645504242546199,
+      "grad_norm": 0.15195608423093737,
+      "learning_rate": 2e-05,
+      "loss": 5.5503,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09652211825468693,
+      "grad_norm": 0.14121706913474383,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09658919408391187,
+      "grad_norm": 0.14697444500133303,
+      "learning_rate": 2e-05,
+      "loss": 5.5413,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0966562699131368,
+      "grad_norm": 0.13942281678168458,
+      "learning_rate": 2e-05,
+      "loss": 5.4538,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09672334574236174,
+      "grad_norm": 0.14104285155413987,
+      "learning_rate": 2e-05,
+      "loss": 5.5466,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09679042157158668,
+      "grad_norm": 0.1427125486775428,
+      "learning_rate": 2e-05,
+      "loss": 5.5444,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09685749740081162,
+      "grad_norm": 0.15264755065161484,
+      "learning_rate": 2e-05,
+      "loss": 5.3982,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09692457323003656,
+      "grad_norm": 0.1448888528103812,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 1445
+    },
+    {
+      "epoch": 0.0969916490592615,
+      "grad_norm": 0.14267306588088355,
+      "learning_rate": 2e-05,
+      "loss": 5.4476,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09705872488848644,
+      "grad_norm": 0.15051597184981824,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09712580071771137,
+      "grad_norm": 0.14202500258740233,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09719287654693631,
+      "grad_norm": 0.1463733107045511,
+      "learning_rate": 2e-05,
+      "loss": 5.5573,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09725995237616125,
+      "grad_norm": 0.13877765131952716,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09732702820538619,
+      "grad_norm": 0.14287613691378953,
+      "learning_rate": 2e-05,
+      "loss": 5.2484,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09739410403461113,
+      "grad_norm": 0.1402116935936757,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09746117986383607,
+      "grad_norm": 0.13685702564068633,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 1453
+    },
+    {
+      "epoch": 0.097528255693061,
+      "grad_norm": 0.14914382011286004,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09759533152228594,
+      "grad_norm": 0.14328685861448667,
+      "learning_rate": 2e-05,
+      "loss": 5.5092,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09766240735151088,
+      "grad_norm": 0.13996318480142278,
+      "learning_rate": 2e-05,
+      "loss": 5.41,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09772948318073582,
+      "grad_norm": 0.14510732309808208,
+      "learning_rate": 2e-05,
+      "loss": 5.5072,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09779655900996076,
+      "grad_norm": 0.14638870773696838,
+      "learning_rate": 2e-05,
+      "loss": 5.6076,
+      "step": 1458
+    },
+    {
+      "epoch": 0.0978636348391857,
+      "grad_norm": 0.1405136863450372,
+      "learning_rate": 2e-05,
+      "loss": 5.3722,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09793071066841064,
+      "grad_norm": 0.14090502202056135,
+      "learning_rate": 2e-05,
+      "loss": 5.5572,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09799778649763558,
+      "grad_norm": 0.14052988343236372,
+      "learning_rate": 2e-05,
+      "loss": 5.587,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09806486232686051,
+      "grad_norm": 0.1453986509363523,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09813193815608545,
+      "grad_norm": 0.13637432305133412,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09819901398531039,
+      "grad_norm": 0.14250382738529394,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09826608981453533,
+      "grad_norm": 0.1378262893888623,
+      "learning_rate": 2e-05,
+      "loss": 5.3691,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09833316564376027,
+      "grad_norm": 0.13439725142698983,
+      "learning_rate": 2e-05,
+      "loss": 5.3776,
+      "step": 1466
+    },
+    {
+      "epoch": 0.0984002414729852,
+      "grad_norm": 0.1471630990442746,
+      "learning_rate": 2e-05,
+      "loss": 5.5504,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09846731730221014,
+      "grad_norm": 0.14399568495256856,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09853439313143508,
+      "grad_norm": 0.1477748217555435,
+      "learning_rate": 2e-05,
+      "loss": 5.4054,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09860146896066002,
+      "grad_norm": 0.1532820043239232,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09866854478988496,
+      "grad_norm": 0.14055123999015937,
+      "learning_rate": 2e-05,
+      "loss": 5.5743,
+      "step": 1471
+    },
+    {
+      "epoch": 0.0987356206191099,
+      "grad_norm": 0.14586513009013108,
+      "learning_rate": 2e-05,
+      "loss": 5.5328,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09880269644833484,
+      "grad_norm": 0.1481414525560901,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09886977227755978,
+      "grad_norm": 0.14133102583847396,
+      "learning_rate": 2e-05,
+      "loss": 5.5989,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09893684810678471,
+      "grad_norm": 0.14808957932939534,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09900392393600965,
+      "grad_norm": 0.1499221993714861,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09907099976523459,
+      "grad_norm": 0.14754184921942953,
+      "learning_rate": 2e-05,
+      "loss": 5.3247,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09913807559445954,
+      "grad_norm": 0.14500279027375446,
+      "learning_rate": 2e-05,
+      "loss": 5.5994,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09920515142368448,
+      "grad_norm": 0.14176833655694496,
+      "learning_rate": 2e-05,
+      "loss": 5.6478,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09927222725290942,
+      "grad_norm": 0.1394116809594869,
+      "learning_rate": 2e-05,
+      "loss": 5.6334,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09933930308213436,
+      "grad_norm": 0.13881772708575743,
+      "learning_rate": 2e-05,
+      "loss": 5.4758,
+      "step": 1481
+    },
+    {
+      "epoch": 0.0994063789113593,
+      "grad_norm": 0.1437889938262072,
+      "learning_rate": 2e-05,
+      "loss": 5.4788,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09947345474058424,
+      "grad_norm": 0.1471051561515782,
+      "learning_rate": 2e-05,
+      "loss": 5.449,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09954053056980917,
+      "grad_norm": 0.13730453970681833,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09960760639903411,
+      "grad_norm": 0.14443814260612675,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09967468222825905,
+      "grad_norm": 0.15344638471921557,
+      "learning_rate": 2e-05,
+      "loss": 5.4765,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09974175805748399,
+      "grad_norm": 0.13818076111041894,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09980883388670893,
+      "grad_norm": 0.14890323978630507,
+      "learning_rate": 2e-05,
+      "loss": 5.482,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09987590971593387,
+      "grad_norm": 0.14416808613099583,
+      "learning_rate": 2e-05,
+      "loss": 5.6296,
+      "step": 1489
+    },
+    {
+      "epoch": 0.0999429855451588,
+      "grad_norm": 0.1353993174088175,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10001006137438374,
+      "grad_norm": 0.1424246198847154,
+      "learning_rate": 2e-05,
+      "loss": 5.4142,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10007713720360868,
+      "grad_norm": 0.14695274609391934,
+      "learning_rate": 2e-05,
+      "loss": 5.5465,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10014421303283362,
+      "grad_norm": 0.1405966075130509,
+      "learning_rate": 2e-05,
+      "loss": 5.5469,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10021128886205856,
+      "grad_norm": 0.13802340853919748,
+      "learning_rate": 2e-05,
+      "loss": 5.4378,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1002783646912835,
+      "grad_norm": 0.1391301305923459,
+      "learning_rate": 2e-05,
+      "loss": 5.5521,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10034544052050844,
+      "grad_norm": 0.14376622390767146,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 1496
+    },
+    {
+      "epoch": 0.10041251634973337,
+      "grad_norm": 0.13956750056836595,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 1497
+    },
+    {
+      "epoch": 0.10047959217895831,
+      "grad_norm": 0.14870804120934086,
+      "learning_rate": 2e-05,
+      "loss": 5.4101,
+      "step": 1498
+    },
+    {
+      "epoch": 0.10054666800818325,
+      "grad_norm": 0.1436567174772107,
+      "learning_rate": 2e-05,
+      "loss": 5.3582,
+      "step": 1499
+    },
+    {
+      "epoch": 0.10061374383740819,
+      "grad_norm": 0.13488798243158245,
+      "learning_rate": 2e-05,
+      "loss": 5.418,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10068081966663313,
+      "grad_norm": 0.13321629375527355,
+      "learning_rate": 2e-05,
+      "loss": 5.505,
+      "step": 1501
+    },
+    {
+      "epoch": 0.10074789549585807,
+      "grad_norm": 0.14931682206030206,
+      "learning_rate": 2e-05,
+      "loss": 5.3826,
+      "step": 1502
+    },
+    {
+      "epoch": 0.100814971325083,
+      "grad_norm": 0.14365824907490743,
+      "learning_rate": 2e-05,
+      "loss": 5.6168,
+      "step": 1503
+    },
+    {
+      "epoch": 0.10088204715430794,
+      "grad_norm": 0.14740139694631255,
+      "learning_rate": 2e-05,
+      "loss": 5.418,
+      "step": 1504
+    },
+    {
+      "epoch": 0.10094912298353288,
+      "grad_norm": 0.14514302943048846,
+      "learning_rate": 2e-05,
+      "loss": 5.5008,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10101619881275782,
+      "grad_norm": 0.1431642248898501,
+      "learning_rate": 2e-05,
+      "loss": 5.4949,
+      "step": 1506
+    },
+    {
+      "epoch": 0.10108327464198276,
+      "grad_norm": 0.14368899929172005,
+      "learning_rate": 2e-05,
+      "loss": 5.4035,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1011503504712077,
+      "grad_norm": 0.14077032725242883,
+      "learning_rate": 2e-05,
+      "loss": 5.4832,
+      "step": 1508
+    },
+    {
+      "epoch": 0.10121742630043264,
+      "grad_norm": 0.1331346339592145,
+      "learning_rate": 2e-05,
+      "loss": 5.4048,
+      "step": 1509
+    },
+    {
+      "epoch": 0.10128450212965757,
+      "grad_norm": 0.14817593911720636,
+      "learning_rate": 2e-05,
+      "loss": 5.4464,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10135157795888251,
+      "grad_norm": 0.13620501882606223,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 1511
+    },
+    {
+      "epoch": 0.10141865378810745,
+      "grad_norm": 0.1385327664836683,
+      "learning_rate": 2e-05,
+      "loss": 5.4458,
+      "step": 1512
+    },
+    {
+      "epoch": 0.10148572961733239,
+      "grad_norm": 0.13838402522671625,
+      "learning_rate": 2e-05,
+      "loss": 5.4552,
+      "step": 1513
+    },
+    {
+      "epoch": 0.10155280544655733,
+      "grad_norm": 0.1496067977321398,
+      "learning_rate": 2e-05,
+      "loss": 5.5641,
+      "step": 1514
+    },
+    {
+      "epoch": 0.10161988127578227,
+      "grad_norm": 0.1460906635242363,
+      "learning_rate": 2e-05,
+      "loss": 5.3561,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1016869571050072,
+      "grad_norm": 0.14706094594030525,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 1516
+    },
+    {
+      "epoch": 0.10175403293423214,
+      "grad_norm": 0.14243887042602202,
+      "learning_rate": 2e-05,
+      "loss": 5.5024,
+      "step": 1517
+    },
+    {
+      "epoch": 0.10182110876345708,
+      "grad_norm": 0.14099101347442228,
+      "learning_rate": 2e-05,
+      "loss": 5.4459,
+      "step": 1518
+    },
+    {
+      "epoch": 0.10188818459268202,
+      "grad_norm": 0.13656408160939715,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 1519
+    },
+    {
+      "epoch": 0.10195526042190696,
+      "grad_norm": 0.15422594105675147,
+      "learning_rate": 2e-05,
+      "loss": 5.561,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1020223362511319,
+      "grad_norm": 0.145138258002595,
+      "learning_rate": 2e-05,
+      "loss": 5.405,
+      "step": 1521
+    },
+    {
+      "epoch": 0.10208941208035684,
+      "grad_norm": 0.1354271071035837,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 1522
+    },
+    {
+      "epoch": 0.10215648790958179,
+      "grad_norm": 0.14356568361781288,
+      "learning_rate": 2e-05,
+      "loss": 5.4135,
+      "step": 1523
+    },
+    {
+      "epoch": 0.10222356373880673,
+      "grad_norm": 0.14228170418508898,
+      "learning_rate": 2e-05,
+      "loss": 5.6452,
+      "step": 1524
+    },
+    {
+      "epoch": 0.10229063956803167,
+      "grad_norm": 0.1407475056817347,
+      "learning_rate": 2e-05,
+      "loss": 5.5819,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1023577153972566,
+      "grad_norm": 0.1367415422805007,
+      "learning_rate": 2e-05,
+      "loss": 5.4516,
+      "step": 1526
+    },
+    {
+      "epoch": 0.10242479122648154,
+      "grad_norm": 0.14452418013102258,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 1527
+    },
+    {
+      "epoch": 0.10249186705570648,
+      "grad_norm": 0.14923818561048818,
+      "learning_rate": 2e-05,
+      "loss": 5.541,
+      "step": 1528
+    },
+    {
+      "epoch": 0.10255894288493142,
+      "grad_norm": 0.13814565566242995,
+      "learning_rate": 2e-05,
+      "loss": 5.5311,
+      "step": 1529
+    },
+    {
+      "epoch": 0.10262601871415636,
+      "grad_norm": 0.14219689525430027,
+      "learning_rate": 2e-05,
+      "loss": 5.5101,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1026930945433813,
+      "grad_norm": 0.14010004173628177,
+      "learning_rate": 2e-05,
+      "loss": 5.5177,
+      "step": 1531
+    },
+    {
+      "epoch": 0.10276017037260624,
+      "grad_norm": 0.148023726161561,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 1532
+    },
+    {
+      "epoch": 0.10282724620183117,
+      "grad_norm": 0.14288422052927535,
+      "learning_rate": 2e-05,
+      "loss": 5.5802,
+      "step": 1533
+    },
+    {
+      "epoch": 0.10289432203105611,
+      "grad_norm": 0.13586985331286183,
+      "learning_rate": 2e-05,
+      "loss": 5.2623,
+      "step": 1534
+    },
+    {
+      "epoch": 0.10296139786028105,
+      "grad_norm": 0.14666474655625442,
+      "learning_rate": 2e-05,
+      "loss": 5.5293,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10302847368950599,
+      "grad_norm": 0.1420314853632253,
+      "learning_rate": 2e-05,
+      "loss": 5.3172,
+      "step": 1536
+    },
+    {
+      "epoch": 0.10309554951873093,
+      "grad_norm": 0.14226017170205124,
+      "learning_rate": 2e-05,
+      "loss": 5.2921,
+      "step": 1537
+    },
+    {
+      "epoch": 0.10316262534795587,
+      "grad_norm": 0.13756992554008657,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1032297011771808,
+      "grad_norm": 0.14012281621931746,
+      "learning_rate": 2e-05,
+      "loss": 5.5935,
+      "step": 1539
+    },
+    {
+      "epoch": 0.10329677700640574,
+      "grad_norm": 0.14614080045313507,
+      "learning_rate": 2e-05,
+      "loss": 5.5484,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10336385283563068,
+      "grad_norm": 0.15049913958678665,
+      "learning_rate": 2e-05,
+      "loss": 5.39,
+      "step": 1541
+    },
+    {
+      "epoch": 0.10343092866485562,
+      "grad_norm": 0.14398375990120477,
+      "learning_rate": 2e-05,
+      "loss": 5.5998,
+      "step": 1542
+    },
+    {
+      "epoch": 0.10349800449408056,
+      "grad_norm": 0.14232048242289982,
+      "learning_rate": 2e-05,
+      "loss": 5.4925,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1035650803233055,
+      "grad_norm": 0.13749243783503928,
+      "learning_rate": 2e-05,
+      "loss": 5.5421,
+      "step": 1544
+    },
+    {
+      "epoch": 0.10363215615253044,
+      "grad_norm": 0.1449515872877285,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 1545
+    },
+    {
+      "epoch": 0.10369923198175537,
+      "grad_norm": 0.15096689226647395,
+      "learning_rate": 2e-05,
+      "loss": 5.3836,
+      "step": 1546
+    },
+    {
+      "epoch": 0.10376630781098031,
+      "grad_norm": 0.14168366064977186,
+      "learning_rate": 2e-05,
+      "loss": 5.459,
+      "step": 1547
+    },
+    {
+      "epoch": 0.10383338364020525,
+      "grad_norm": 0.15649980635578664,
+      "learning_rate": 2e-05,
+      "loss": 5.3599,
+      "step": 1548
+    },
+    {
+      "epoch": 0.10390045946943019,
+      "grad_norm": 0.13776447492028498,
+      "learning_rate": 2e-05,
+      "loss": 5.41,
+      "step": 1549
+    },
+    {
+      "epoch": 0.10396753529865513,
+      "grad_norm": 0.1364754513749412,
+      "learning_rate": 2e-05,
+      "loss": 5.3024,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10403461112788007,
+      "grad_norm": 0.13708255968508387,
+      "learning_rate": 2e-05,
+      "loss": 5.511,
+      "step": 1551
+    },
+    {
+      "epoch": 0.104101686957105,
+      "grad_norm": 0.1446875922306525,
+      "learning_rate": 2e-05,
+      "loss": 5.4001,
+      "step": 1552
+    },
+    {
+      "epoch": 0.10416876278632994,
+      "grad_norm": 0.1424665685699538,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 1553
+    },
+    {
+      "epoch": 0.10423583861555488,
+      "grad_norm": 0.14503345132812664,
+      "learning_rate": 2e-05,
+      "loss": 5.5307,
+      "step": 1554
+    },
+    {
+      "epoch": 0.10430291444477982,
+      "grad_norm": 0.1497884504962539,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 1555
+    },
+    {
+      "epoch": 0.10436999027400476,
+      "grad_norm": 0.15605988268366836,
+      "learning_rate": 2e-05,
+      "loss": 5.6675,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1044370661032297,
+      "grad_norm": 0.14137419045183583,
+      "learning_rate": 2e-05,
+      "loss": 5.6229,
+      "step": 1557
+    },
+    {
+      "epoch": 0.10450414193245464,
+      "grad_norm": 0.1436336763573979,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 1558
+    },
+    {
+      "epoch": 0.10457121776167957,
+      "grad_norm": 0.15059536305628135,
+      "learning_rate": 2e-05,
+      "loss": 5.5414,
+      "step": 1559
+    },
+    {
+      "epoch": 0.10463829359090451,
+      "grad_norm": 0.13972773358580998,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10470536942012945,
+      "grad_norm": 0.1416903880782502,
+      "learning_rate": 2e-05,
+      "loss": 5.4733,
+      "step": 1561
+    },
+    {
+      "epoch": 0.10477244524935439,
+      "grad_norm": 0.14669716622132847,
+      "learning_rate": 2e-05,
+      "loss": 5.3905,
+      "step": 1562
+    },
+    {
+      "epoch": 0.10483952107857933,
+      "grad_norm": 0.14481199075925796,
+      "learning_rate": 2e-05,
+      "loss": 5.4331,
+      "step": 1563
+    },
+    {
+      "epoch": 0.10490659690780427,
+      "grad_norm": 0.1396169563659074,
+      "learning_rate": 2e-05,
+      "loss": 5.4084,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1049736727370292,
+      "grad_norm": 0.15186513923788264,
+      "learning_rate": 2e-05,
+      "loss": 5.3706,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10504074856625414,
+      "grad_norm": 0.14350223124207856,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 1566
+    },
+    {
+      "epoch": 0.10510782439547908,
+      "grad_norm": 0.14641523299354378,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 1567
+    },
+    {
+      "epoch": 0.10517490022470402,
+      "grad_norm": 0.15058467741464682,
+      "learning_rate": 2e-05,
+      "loss": 5.5003,
+      "step": 1568
+    },
+    {
+      "epoch": 0.10524197605392897,
+      "grad_norm": 0.14952971975024862,
+      "learning_rate": 2e-05,
+      "loss": 5.3552,
+      "step": 1569
+    },
+    {
+      "epoch": 0.10530905188315391,
+      "grad_norm": 0.1447270362406306,
+      "learning_rate": 2e-05,
+      "loss": 5.3409,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10537612771237885,
+      "grad_norm": 0.14401496436437722,
+      "learning_rate": 2e-05,
+      "loss": 5.5279,
+      "step": 1571
+    },
+    {
+      "epoch": 0.10544320354160379,
+      "grad_norm": 0.15791995478912157,
+      "learning_rate": 2e-05,
+      "loss": 5.4671,
+      "step": 1572
+    },
+    {
+      "epoch": 0.10551027937082873,
+      "grad_norm": 0.14677332053639972,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 1573
+    },
+    {
+      "epoch": 0.10557735520005367,
+      "grad_norm": 0.14809925522160325,
+      "learning_rate": 2e-05,
+      "loss": 5.5621,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1056444310292786,
+      "grad_norm": 0.14443655329215635,
+      "learning_rate": 2e-05,
+      "loss": 5.5806,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10571150685850354,
+      "grad_norm": 0.15187316508938864,
+      "learning_rate": 2e-05,
+      "loss": 5.5074,
+      "step": 1576
+    },
+    {
+      "epoch": 0.10577858268772848,
+      "grad_norm": 0.14727212147372276,
+      "learning_rate": 2e-05,
+      "loss": 5.4041,
+      "step": 1577
+    },
+    {
+      "epoch": 0.10584565851695342,
+      "grad_norm": 0.1389546512875751,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 1578
+    },
+    {
+      "epoch": 0.10591273434617836,
+      "grad_norm": 0.14439034799852285,
+      "learning_rate": 2e-05,
+      "loss": 5.5473,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1059798101754033,
+      "grad_norm": 0.14507559965320502,
+      "learning_rate": 2e-05,
+      "loss": 5.5735,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10604688600462824,
+      "grad_norm": 0.14267571223864173,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 1581
+    },
+    {
+      "epoch": 0.10611396183385317,
+      "grad_norm": 0.14210478643676563,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 1582
+    },
+    {
+      "epoch": 0.10618103766307811,
+      "grad_norm": 0.13712446562584085,
+      "learning_rate": 2e-05,
+      "loss": 5.5484,
+      "step": 1583
+    },
+    {
+      "epoch": 0.10624811349230305,
+      "grad_norm": 0.15361039319791728,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 1584
+    },
+    {
+      "epoch": 0.10631518932152799,
+      "grad_norm": 0.14336898327100167,
+      "learning_rate": 2e-05,
+      "loss": 5.5685,
+      "step": 1585
+    },
+    {
+      "epoch": 0.10638226515075293,
+      "grad_norm": 0.1377696405133565,
+      "learning_rate": 2e-05,
+      "loss": 5.2445,
+      "step": 1586
+    },
+    {
+      "epoch": 0.10644934097997787,
+      "grad_norm": 0.14969591078211128,
+      "learning_rate": 2e-05,
+      "loss": 5.5519,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1065164168092028,
+      "grad_norm": 0.1445595093201612,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 1588
+    },
+    {
+      "epoch": 0.10658349263842774,
+      "grad_norm": 0.1432980178585363,
+      "learning_rate": 2e-05,
+      "loss": 5.4389,
+      "step": 1589
+    },
+    {
+      "epoch": 0.10665056846765268,
+      "grad_norm": 0.1415484202135085,
+      "learning_rate": 2e-05,
+      "loss": 5.3719,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10671764429687762,
+      "grad_norm": 0.14765808557992754,
+      "learning_rate": 2e-05,
+      "loss": 5.4103,
+      "step": 1591
+    },
+    {
+      "epoch": 0.10678472012610256,
+      "grad_norm": 0.1573044394493838,
+      "learning_rate": 2e-05,
+      "loss": 5.3392,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1068517959553275,
+      "grad_norm": 0.13156425781974831,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 1593
+    },
+    {
+      "epoch": 0.10691887178455244,
+      "grad_norm": 0.1399307448335487,
+      "learning_rate": 2e-05,
+      "loss": 5.6144,
+      "step": 1594
+    },
+    {
+      "epoch": 0.10698594761377737,
+      "grad_norm": 0.16623993319067248,
+      "learning_rate": 2e-05,
+      "loss": 5.484,
+      "step": 1595
+    },
+    {
+      "epoch": 0.10705302344300231,
+      "grad_norm": 0.1486480344029779,
+      "learning_rate": 2e-05,
+      "loss": 5.3755,
+      "step": 1596
+    },
+    {
+      "epoch": 0.10712009927222725,
+      "grad_norm": 0.1470066652907782,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 1597
+    },
+    {
+      "epoch": 0.10718717510145219,
+      "grad_norm": 0.14508658043720463,
+      "learning_rate": 2e-05,
+      "loss": 5.2721,
+      "step": 1598
+    },
+    {
+      "epoch": 0.10725425093067713,
+      "grad_norm": 0.15204217421002456,
+      "learning_rate": 2e-05,
+      "loss": 5.6199,
+      "step": 1599
+    },
+    {
+      "epoch": 0.10732132675990207,
+      "grad_norm": 0.1380977637575904,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 1600
+    },
+    {
+      "epoch": 0.107388402589127,
+      "grad_norm": 0.14082072021061,
+      "learning_rate": 2e-05,
+      "loss": 5.2922,
+      "step": 1601
+    },
+    {
+      "epoch": 0.10745547841835194,
+      "grad_norm": 0.14366300546598018,
+      "learning_rate": 2e-05,
+      "loss": 5.4276,
+      "step": 1602
+    },
+    {
+      "epoch": 0.10752255424757688,
+      "grad_norm": 0.14642812533400576,
+      "learning_rate": 2e-05,
+      "loss": 5.4281,
+      "step": 1603
+    },
+    {
+      "epoch": 0.10758963007680182,
+      "grad_norm": 0.14608200514108957,
+      "learning_rate": 2e-05,
+      "loss": 5.2989,
+      "step": 1604
+    },
+    {
+      "epoch": 0.10765670590602676,
+      "grad_norm": 0.14004022503958988,
+      "learning_rate": 2e-05,
+      "loss": 5.4821,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1077237817352517,
+      "grad_norm": 0.14733918204446922,
+      "learning_rate": 2e-05,
+      "loss": 5.4471,
+      "step": 1606
+    },
+    {
+      "epoch": 0.10779085756447664,
+      "grad_norm": 0.1464914525330342,
+      "learning_rate": 2e-05,
+      "loss": 5.4009,
+      "step": 1607
+    },
+    {
+      "epoch": 0.10785793339370157,
+      "grad_norm": 0.14675367350917357,
+      "learning_rate": 2e-05,
+      "loss": 5.3843,
+      "step": 1608
+    },
+    {
+      "epoch": 0.10792500922292651,
+      "grad_norm": 0.15322463407308098,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 1609
+    },
+    {
+      "epoch": 0.10799208505215145,
+      "grad_norm": 0.14673559051695673,
+      "learning_rate": 2e-05,
+      "loss": 5.5637,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10805916088137639,
+      "grad_norm": 0.1494705331046625,
+      "learning_rate": 2e-05,
+      "loss": 5.5287,
+      "step": 1611
+    },
+    {
+      "epoch": 0.10812623671060133,
+      "grad_norm": 0.14411773039872902,
+      "learning_rate": 2e-05,
+      "loss": 5.4967,
+      "step": 1612
+    },
+    {
+      "epoch": 0.10819331253982627,
+      "grad_norm": 0.14195931389543873,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 1613
+    },
+    {
+      "epoch": 0.10826038836905122,
+      "grad_norm": 0.15472984288834787,
+      "learning_rate": 2e-05,
+      "loss": 5.4513,
+      "step": 1614
+    },
+    {
+      "epoch": 0.10832746419827616,
+      "grad_norm": 0.1562775797511193,
+      "learning_rate": 2e-05,
+      "loss": 5.3802,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1083945400275011,
+      "grad_norm": 0.14585106005807363,
+      "learning_rate": 2e-05,
+      "loss": 5.6416,
+      "step": 1616
+    },
+    {
+      "epoch": 0.10846161585672603,
+      "grad_norm": 0.15043701513070074,
+      "learning_rate": 2e-05,
+      "loss": 5.5746,
+      "step": 1617
+    },
+    {
+      "epoch": 0.10852869168595097,
+      "grad_norm": 0.14931235041865532,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 1618
+    },
+    {
+      "epoch": 0.10859576751517591,
+      "grad_norm": 0.1444564169913332,
+      "learning_rate": 2e-05,
+      "loss": 5.5137,
+      "step": 1619
+    },
+    {
+      "epoch": 0.10866284334440085,
+      "grad_norm": 0.14341798887990245,
+      "learning_rate": 2e-05,
+      "loss": 5.3757,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10872991917362579,
+      "grad_norm": 0.13593163083512766,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 1621
+    },
+    {
+      "epoch": 0.10879699500285073,
+      "grad_norm": 0.1462960234488021,
+      "learning_rate": 2e-05,
+      "loss": 5.5694,
+      "step": 1622
+    },
+    {
+      "epoch": 0.10886407083207567,
+      "grad_norm": 0.1489646357403895,
+      "learning_rate": 2e-05,
+      "loss": 5.339,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1089311466613006,
+      "grad_norm": 0.13877426185852562,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 1624
+    },
+    {
+      "epoch": 0.10899822249052554,
+      "grad_norm": 0.1393355934197879,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 1625
+    },
+    {
+      "epoch": 0.10906529831975048,
+      "grad_norm": 0.13819675166658885,
+      "learning_rate": 2e-05,
+      "loss": 5.3163,
+      "step": 1626
+    },
+    {
+      "epoch": 0.10913237414897542,
+      "grad_norm": 0.1432720616667487,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 1627
+    },
+    {
+      "epoch": 0.10919944997820036,
+      "grad_norm": 0.15732635256014468,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1092665258074253,
+      "grad_norm": 0.14271893851940626,
+      "learning_rate": 2e-05,
+      "loss": 5.4247,
+      "step": 1629
+    },
+    {
+      "epoch": 0.10933360163665024,
+      "grad_norm": 0.13904309106942067,
+      "learning_rate": 2e-05,
+      "loss": 5.6436,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10940067746587517,
+      "grad_norm": 0.1458382128306477,
+      "learning_rate": 2e-05,
+      "loss": 5.3394,
+      "step": 1631
+    },
+    {
+      "epoch": 0.10946775329510011,
+      "grad_norm": 0.14542384974558076,
+      "learning_rate": 2e-05,
+      "loss": 5.571,
+      "step": 1632
+    },
+    {
+      "epoch": 0.10953482912432505,
+      "grad_norm": 0.1473579993103474,
+      "learning_rate": 2e-05,
+      "loss": 5.6455,
+      "step": 1633
+    },
+    {
+      "epoch": 0.10960190495354999,
+      "grad_norm": 0.14803626971590494,
+      "learning_rate": 2e-05,
+      "loss": 5.4905,
+      "step": 1634
+    },
+    {
+      "epoch": 0.10966898078277493,
+      "grad_norm": 0.1443703665206401,
+      "learning_rate": 2e-05,
+      "loss": 5.347,
+      "step": 1635
+    },
+    {
+      "epoch": 0.10973605661199987,
+      "grad_norm": 0.14615683191088877,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1098031324412248,
+      "grad_norm": 0.14542917194443583,
+      "learning_rate": 2e-05,
+      "loss": 5.4712,
+      "step": 1637
+    },
+    {
+      "epoch": 0.10987020827044974,
+      "grad_norm": 0.14241383508357144,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 1638
+    },
+    {
+      "epoch": 0.10993728409967468,
+      "grad_norm": 0.14099400607943796,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 1639
+    },
+    {
+      "epoch": 0.11000435992889962,
+      "grad_norm": 0.14372499176572798,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11007143575812456,
+      "grad_norm": 0.14484381290392995,
+      "learning_rate": 2e-05,
+      "loss": 5.5169,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1101385115873495,
+      "grad_norm": 0.13949303692361023,
+      "learning_rate": 2e-05,
+      "loss": 5.4244,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11020558741657444,
+      "grad_norm": 0.14020444405814161,
+      "learning_rate": 2e-05,
+      "loss": 5.4815,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11027266324579937,
+      "grad_norm": 0.14375005513463238,
+      "learning_rate": 2e-05,
+      "loss": 5.4014,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11033973907502431,
+      "grad_norm": 0.1434973797511504,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11040681490424925,
+      "grad_norm": 0.14174880273486512,
+      "learning_rate": 2e-05,
+      "loss": 5.3112,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11047389073347419,
+      "grad_norm": 0.14138289550912594,
+      "learning_rate": 2e-05,
+      "loss": 5.5642,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11054096656269913,
+      "grad_norm": 0.14479067614998764,
+      "learning_rate": 2e-05,
+      "loss": 5.2692,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11060804239192407,
+      "grad_norm": 0.1478340134098868,
+      "learning_rate": 2e-05,
+      "loss": 5.459,
+      "step": 1649
+    },
+    {
+      "epoch": 0.110675118221149,
+      "grad_norm": 0.13909230019320293,
+      "learning_rate": 2e-05,
+      "loss": 5.4042,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11074219405037394,
+      "grad_norm": 0.1474088816508054,
+      "learning_rate": 2e-05,
+      "loss": 5.3549,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11080926987959888,
+      "grad_norm": 0.14885096916718182,
+      "learning_rate": 2e-05,
+      "loss": 5.2909,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11087634570882382,
+      "grad_norm": 0.13788542139191265,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11094342153804876,
+      "grad_norm": 0.14121594324857864,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1110104973672737,
+      "grad_norm": 0.15019807667517873,
+      "learning_rate": 2e-05,
+      "loss": 5.5792,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11107757319649864,
+      "grad_norm": 0.14842626129979514,
+      "learning_rate": 2e-05,
+      "loss": 5.4296,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11114464902572357,
+      "grad_norm": 0.14511358120898157,
+      "learning_rate": 2e-05,
+      "loss": 5.3396,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11121172485494851,
+      "grad_norm": 0.14389158901967103,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11127880068417345,
+      "grad_norm": 0.14621390599378412,
+      "learning_rate": 2e-05,
+      "loss": 5.5877,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1113458765133984,
+      "grad_norm": 0.1367833181639195,
+      "learning_rate": 2e-05,
+      "loss": 5.5572,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11141295234262334,
+      "grad_norm": 0.1543871516887754,
+      "learning_rate": 2e-05,
+      "loss": 5.4674,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11148002817184828,
+      "grad_norm": 0.1532344016609872,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11154710400107322,
+      "grad_norm": 0.15281589254277947,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11161417983029816,
+      "grad_norm": 0.15642266862766188,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1116812556595231,
+      "grad_norm": 0.15260758744362116,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11174833148874803,
+      "grad_norm": 0.13960684956017316,
+      "learning_rate": 2e-05,
+      "loss": 5.5088,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11181540731797297,
+      "grad_norm": 0.14736728747289005,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11188248314719791,
+      "grad_norm": 0.1586010165621013,
+      "learning_rate": 2e-05,
+      "loss": 5.4917,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11194955897642285,
+      "grad_norm": 0.1389775995023264,
+      "learning_rate": 2e-05,
+      "loss": 5.6105,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11201663480564779,
+      "grad_norm": 0.14776175175617412,
+      "learning_rate": 2e-05,
+      "loss": 5.4752,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11208371063487273,
+      "grad_norm": 0.1590314186160879,
+      "learning_rate": 2e-05,
+      "loss": 5.33,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11215078646409767,
+      "grad_norm": 0.13827655770008004,
+      "learning_rate": 2e-05,
+      "loss": 5.4445,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1122178622933226,
+      "grad_norm": 0.13986179021167408,
+      "learning_rate": 2e-05,
+      "loss": 5.5222,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11228493812254754,
+      "grad_norm": 0.15401166275088443,
+      "learning_rate": 2e-05,
+      "loss": 5.5212,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11235201395177248,
+      "grad_norm": 0.14920027591484544,
+      "learning_rate": 2e-05,
+      "loss": 5.5593,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11241908978099742,
+      "grad_norm": 0.13860587985135756,
+      "learning_rate": 2e-05,
+      "loss": 5.5041,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11248616561022236,
+      "grad_norm": 0.14374442973273263,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1125532414394473,
+      "grad_norm": 0.15242830165851026,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11262031726867223,
+      "grad_norm": 0.13789350788196741,
+      "learning_rate": 2e-05,
+      "loss": 5.4437,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11268739309789717,
+      "grad_norm": 0.14727969116791859,
+      "learning_rate": 2e-05,
+      "loss": 5.4111,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11275446892712211,
+      "grad_norm": 0.14214987824960282,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11282154475634705,
+      "grad_norm": 0.14595970499774005,
+      "learning_rate": 2e-05,
+      "loss": 5.6435,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11288862058557199,
+      "grad_norm": 0.1434022753861315,
+      "learning_rate": 2e-05,
+      "loss": 5.5328,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11295569641479693,
+      "grad_norm": 0.1457297480536738,
+      "learning_rate": 2e-05,
+      "loss": 5.4568,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11302277224402187,
+      "grad_norm": 0.14987248755891258,
+      "learning_rate": 2e-05,
+      "loss": 5.4784,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1130898480732468,
+      "grad_norm": 0.13987902377617775,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11315692390247174,
+      "grad_norm": 0.1475009358990396,
+      "learning_rate": 2e-05,
+      "loss": 5.5332,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11322399973169668,
+      "grad_norm": 0.13862240008295632,
+      "learning_rate": 2e-05,
+      "loss": 5.3133,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11329107556092162,
+      "grad_norm": 0.14104596315377577,
+      "learning_rate": 2e-05,
+      "loss": 5.5168,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11335815139014656,
+      "grad_norm": 0.13613384468398085,
+      "learning_rate": 2e-05,
+      "loss": 5.4714,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1134252272193715,
+      "grad_norm": 0.15111878256758637,
+      "learning_rate": 2e-05,
+      "loss": 5.4363,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11349230304859644,
+      "grad_norm": 0.15563268135133437,
+      "learning_rate": 2e-05,
+      "loss": 5.2485,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11355937887782137,
+      "grad_norm": 0.14284752130755368,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11362645470704631,
+      "grad_norm": 0.14522601727915224,
+      "learning_rate": 2e-05,
+      "loss": 5.3625,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11369353053627125,
+      "grad_norm": 0.14841178320487514,
+      "learning_rate": 2e-05,
+      "loss": 5.5814,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11376060636549619,
+      "grad_norm": 0.15210437457114429,
+      "learning_rate": 2e-05,
+      "loss": 5.3907,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11382768219472113,
+      "grad_norm": 0.14710045648465214,
+      "learning_rate": 2e-05,
+      "loss": 5.5314,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11389475802394607,
+      "grad_norm": 0.14721650360640648,
+      "learning_rate": 2e-05,
+      "loss": 5.4857,
+      "step": 1698
+    },
+    {
+      "epoch": 0.113961833853171,
+      "grad_norm": 0.14009601987820763,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11402890968239594,
+      "grad_norm": 0.1426760604779711,
+      "learning_rate": 2e-05,
+      "loss": 5.4738,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11409598551162088,
+      "grad_norm": 0.14925255464763285,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11416306134084582,
+      "grad_norm": 0.1413761049364667,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11423013717007076,
+      "grad_norm": 0.14321421076611238,
+      "learning_rate": 2e-05,
+      "loss": 5.5736,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1142972129992957,
+      "grad_norm": 0.16049628905342794,
+      "learning_rate": 2e-05,
+      "loss": 5.3788,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11436428882852065,
+      "grad_norm": 0.148256803816772,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11443136465774559,
+      "grad_norm": 0.14364586016983996,
+      "learning_rate": 2e-05,
+      "loss": 5.4314,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11449844048697053,
+      "grad_norm": 0.1467029419685999,
+      "learning_rate": 2e-05,
+      "loss": 5.3013,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11456551631619546,
+      "grad_norm": 0.15522267515990815,
+      "learning_rate": 2e-05,
+      "loss": 5.3878,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1146325921454204,
+      "grad_norm": 0.14234215779710443,
+      "learning_rate": 2e-05,
+      "loss": 5.468,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11469966797464534,
+      "grad_norm": 0.14477630312220202,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11476674380387028,
+      "grad_norm": 0.1509622446590184,
+      "learning_rate": 2e-05,
+      "loss": 5.5572,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11483381963309522,
+      "grad_norm": 0.14139622715702943,
+      "learning_rate": 2e-05,
+      "loss": 5.4228,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11490089546232016,
+      "grad_norm": 0.13728283539232577,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1149679712915451,
+      "grad_norm": 0.14385065722903867,
+      "learning_rate": 2e-05,
+      "loss": 5.5163,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11503504712077003,
+      "grad_norm": 0.1469554980532785,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11510212294999497,
+      "grad_norm": 0.14154326326620378,
+      "learning_rate": 2e-05,
+      "loss": 5.5492,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11516919877921991,
+      "grad_norm": 0.14635400548053296,
+      "learning_rate": 2e-05,
+      "loss": 5.5457,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11523627460844485,
+      "grad_norm": 0.14685027467780085,
+      "learning_rate": 2e-05,
+      "loss": 5.6477,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11530335043766979,
+      "grad_norm": 0.1406667127728588,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11537042626689473,
+      "grad_norm": 0.141575020626003,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11543750209611967,
+      "grad_norm": 0.14438793775109762,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1155045779253446,
+      "grad_norm": 0.13945321539802924,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11557165375456954,
+      "grad_norm": 0.14060607284160484,
+      "learning_rate": 2e-05,
+      "loss": 5.4901,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11563872958379448,
+      "grad_norm": 0.1362904880946167,
+      "learning_rate": 2e-05,
+      "loss": 5.31,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11570580541301942,
+      "grad_norm": 0.14362229052016126,
+      "learning_rate": 2e-05,
+      "loss": 5.5493,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11577288124224436,
+      "grad_norm": 0.1417223455591491,
+      "learning_rate": 2e-05,
+      "loss": 5.4912,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1158399570714693,
+      "grad_norm": 0.13874346449573324,
+      "learning_rate": 2e-05,
+      "loss": 5.4304,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11590703290069423,
+      "grad_norm": 0.1402481347128242,
+      "learning_rate": 2e-05,
+      "loss": 5.4172,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11597410872991917,
+      "grad_norm": 0.14179166456469974,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11604118455914411,
+      "grad_norm": 0.14653321047772336,
+      "learning_rate": 2e-05,
+      "loss": 5.3887,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11610826038836905,
+      "grad_norm": 0.13824177741203886,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11617533621759399,
+      "grad_norm": 0.1342768519973022,
+      "learning_rate": 2e-05,
+      "loss": 5.4803,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11624241204681893,
+      "grad_norm": 0.1456812486907552,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11630948787604387,
+      "grad_norm": 0.1513258729106233,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1163765637052688,
+      "grad_norm": 0.1431748355942768,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11644363953449374,
+      "grad_norm": 0.14893919113240348,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11651071536371868,
+      "grad_norm": 0.14415972683589934,
+      "learning_rate": 2e-05,
+      "loss": 5.4222,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11657779119294362,
+      "grad_norm": 0.14288291873988201,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11664486702216856,
+      "grad_norm": 0.1430572554290256,
+      "learning_rate": 2e-05,
+      "loss": 5.3307,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1167119428513935,
+      "grad_norm": 0.1408172983349332,
+      "learning_rate": 2e-05,
+      "loss": 5.5201,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11677901868061843,
+      "grad_norm": 0.15343006715211654,
+      "learning_rate": 2e-05,
+      "loss": 5.3665,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11684609450984337,
+      "grad_norm": 0.14408194940862395,
+      "learning_rate": 2e-05,
+      "loss": 5.4308,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11691317033906831,
+      "grad_norm": 0.14272070304441373,
+      "learning_rate": 2e-05,
+      "loss": 5.3602,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11698024616829325,
+      "grad_norm": 0.14756400979045037,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11704732199751819,
+      "grad_norm": 0.15268672806362318,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11711439782674313,
+      "grad_norm": 0.1444870186418168,
+      "learning_rate": 2e-05,
+      "loss": 5.485,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11718147365596807,
+      "grad_norm": 0.15958947824192685,
+      "learning_rate": 2e-05,
+      "loss": 5.3768,
+      "step": 1747
+    },
+    {
+      "epoch": 0.117248549485193,
+      "grad_norm": 0.14276410168388082,
+      "learning_rate": 2e-05,
+      "loss": 5.4266,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11731562531441794,
+      "grad_norm": 0.14966506905791563,
+      "learning_rate": 2e-05,
+      "loss": 5.4444,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1173827011436429,
+      "grad_norm": 0.1570601019559566,
+      "learning_rate": 2e-05,
+      "loss": 5.6154,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11744977697286783,
+      "grad_norm": 0.14403242458640272,
+      "learning_rate": 2e-05,
+      "loss": 5.5484,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11751685280209277,
+      "grad_norm": 0.15491311043837644,
+      "learning_rate": 2e-05,
+      "loss": 5.5698,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11758392863131771,
+      "grad_norm": 0.15027091376162985,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11765100446054265,
+      "grad_norm": 0.14855468085812534,
+      "learning_rate": 2e-05,
+      "loss": 5.4722,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11771808028976759,
+      "grad_norm": 0.14665593020355294,
+      "learning_rate": 2e-05,
+      "loss": 5.3527,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11778515611899253,
+      "grad_norm": 0.14758459949186548,
+      "learning_rate": 2e-05,
+      "loss": 5.4116,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11785223194821746,
+      "grad_norm": 0.14748622661312627,
+      "learning_rate": 2e-05,
+      "loss": 5.3467,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1179193077774424,
+      "grad_norm": 0.1513088952598225,
+      "learning_rate": 2e-05,
+      "loss": 5.5316,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11798638360666734,
+      "grad_norm": 0.1410334433970815,
+      "learning_rate": 2e-05,
+      "loss": 5.3608,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11805345943589228,
+      "grad_norm": 0.14319378127285576,
+      "learning_rate": 2e-05,
+      "loss": 5.5108,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11812053526511722,
+      "grad_norm": 0.1482146029818278,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11818761109434216,
+      "grad_norm": 0.15346423024042347,
+      "learning_rate": 2e-05,
+      "loss": 5.4403,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1182546869235671,
+      "grad_norm": 0.1418742754596509,
+      "learning_rate": 2e-05,
+      "loss": 5.4136,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11832176275279203,
+      "grad_norm": 0.1469556641440429,
+      "learning_rate": 2e-05,
+      "loss": 5.5213,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11838883858201697,
+      "grad_norm": 0.15353016253842008,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11845591441124191,
+      "grad_norm": 0.14449141948659278,
+      "learning_rate": 2e-05,
+      "loss": 5.4051,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11852299024046685,
+      "grad_norm": 0.14507206827471877,
+      "learning_rate": 2e-05,
+      "loss": 5.4058,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11859006606969179,
+      "grad_norm": 0.15068196251027213,
+      "learning_rate": 2e-05,
+      "loss": 5.6398,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11865714189891673,
+      "grad_norm": 0.1477465627195587,
+      "learning_rate": 2e-05,
+      "loss": 5.5693,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11872421772814166,
+      "grad_norm": 0.14848756675269678,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1187912935573666,
+      "grad_norm": 0.14899502324690714,
+      "learning_rate": 2e-05,
+      "loss": 5.5583,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11885836938659154,
+      "grad_norm": 0.15169170494286424,
+      "learning_rate": 2e-05,
+      "loss": 5.4646,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11892544521581648,
+      "grad_norm": 0.14124669424193217,
+      "learning_rate": 2e-05,
+      "loss": 5.5158,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11899252104504142,
+      "grad_norm": 0.15739325085314027,
+      "learning_rate": 2e-05,
+      "loss": 5.4314,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11905959687426636,
+      "grad_norm": 0.15627831597940647,
+      "learning_rate": 2e-05,
+      "loss": 5.5581,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1191266727034913,
+      "grad_norm": 0.1521970702616037,
+      "learning_rate": 2e-05,
+      "loss": 5.3976,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11919374853271623,
+      "grad_norm": 0.1455824628738951,
+      "learning_rate": 2e-05,
+      "loss": 5.3878,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11926082436194117,
+      "grad_norm": 0.15565406336028872,
+      "learning_rate": 2e-05,
+      "loss": 5.6286,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11932790019116611,
+      "grad_norm": 0.1505384288164595,
+      "learning_rate": 2e-05,
+      "loss": 5.3285,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11939497602039105,
+      "grad_norm": 0.13975783166486241,
+      "learning_rate": 2e-05,
+      "loss": 5.4006,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11946205184961599,
+      "grad_norm": 0.14950799239837476,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11952912767884093,
+      "grad_norm": 0.1426771400723545,
+      "learning_rate": 2e-05,
+      "loss": 5.4882,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11959620350806587,
+      "grad_norm": 0.13996186209425482,
+      "learning_rate": 2e-05,
+      "loss": 5.5467,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1196632793372908,
+      "grad_norm": 0.1485821056601315,
+      "learning_rate": 2e-05,
+      "loss": 5.6052,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11973035516651574,
+      "grad_norm": 0.14619350655453636,
+      "learning_rate": 2e-05,
+      "loss": 5.5082,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11979743099574068,
+      "grad_norm": 0.15114660555487425,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11986450682496562,
+      "grad_norm": 0.14429797285235818,
+      "learning_rate": 2e-05,
+      "loss": 5.3817,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11993158265419056,
+      "grad_norm": 0.14385447232376583,
+      "learning_rate": 2e-05,
+      "loss": 5.4046,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1199986584834155,
+      "grad_norm": 0.1449642799706084,
+      "learning_rate": 2e-05,
+      "loss": 5.4803,
+      "step": 1789
+    },
+    {
+      "epoch": 0.12006573431264043,
+      "grad_norm": 0.14700692297466353,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12013281014186537,
+      "grad_norm": 0.14104283101532653,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 1791
+    },
+    {
+      "epoch": 0.12019988597109031,
+      "grad_norm": 0.14367192373260235,
+      "learning_rate": 2e-05,
+      "loss": 5.5578,
+      "step": 1792
+    },
+    {
+      "epoch": 0.12026696180031525,
+      "grad_norm": 0.14372272432025077,
+      "learning_rate": 2e-05,
+      "loss": 5.5144,
+      "step": 1793
+    },
+    {
+      "epoch": 0.12033403762954019,
+      "grad_norm": 0.14473254828566526,
+      "learning_rate": 2e-05,
+      "loss": 5.517,
+      "step": 1794
+    },
+    {
+      "epoch": 0.12040111345876513,
+      "grad_norm": 0.14677646426660745,
+      "learning_rate": 2e-05,
+      "loss": 5.5758,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12046818928799008,
+      "grad_norm": 0.14434640604466636,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12053526511721502,
+      "grad_norm": 0.13746608510826885,
+      "learning_rate": 2e-05,
+      "loss": 5.5242,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12060234094643996,
+      "grad_norm": 0.13860857264334797,
+      "learning_rate": 2e-05,
+      "loss": 5.543,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1206694167756649,
+      "grad_norm": 0.14167408581935662,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12073649260488983,
+      "grad_norm": 0.14362706352981583,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12080356843411477,
+      "grad_norm": 0.14019906744931399,
+      "learning_rate": 2e-05,
+      "loss": 5.3003,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12087064426333971,
+      "grad_norm": 0.14997627086751703,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12093772009256465,
+      "grad_norm": 0.1411805325612246,
+      "learning_rate": 2e-05,
+      "loss": 5.4087,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12100479592178959,
+      "grad_norm": 0.14429349603104452,
+      "learning_rate": 2e-05,
+      "loss": 5.5002,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12107187175101453,
+      "grad_norm": 0.14134480976193212,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12113894758023946,
+      "grad_norm": 0.14400373537430505,
+      "learning_rate": 2e-05,
+      "loss": 5.4729,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1212060234094644,
+      "grad_norm": 0.13692019638223304,
+      "learning_rate": 2e-05,
+      "loss": 5.5117,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12127309923868934,
+      "grad_norm": 0.14467803600626575,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12134017506791428,
+      "grad_norm": 0.14163202456653476,
+      "learning_rate": 2e-05,
+      "loss": 5.28,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12140725089713922,
+      "grad_norm": 0.14784062979483414,
+      "learning_rate": 2e-05,
+      "loss": 5.4748,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12147432672636416,
+      "grad_norm": 0.15147882453855518,
+      "learning_rate": 2e-05,
+      "loss": 5.3962,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1215414025555891,
+      "grad_norm": 0.1457760362401531,
+      "learning_rate": 2e-05,
+      "loss": 5.4332,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12160847838481403,
+      "grad_norm": 0.15215096305764989,
+      "learning_rate": 2e-05,
+      "loss": 5.4669,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12167555421403897,
+      "grad_norm": 0.146954077302628,
+      "learning_rate": 2e-05,
+      "loss": 5.4683,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12174263004326391,
+      "grad_norm": 0.1473960805957019,
+      "learning_rate": 2e-05,
+      "loss": 5.5006,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12180970587248885,
+      "grad_norm": 0.14899294016149073,
+      "learning_rate": 2e-05,
+      "loss": 5.4919,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12187678170171379,
+      "grad_norm": 0.15536722103927272,
+      "learning_rate": 2e-05,
+      "loss": 5.4115,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12194385753093873,
+      "grad_norm": 0.1457999949778778,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12201093336016366,
+      "grad_norm": 0.16759430394248373,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1220780091893886,
+      "grad_norm": 0.14509672532067855,
+      "learning_rate": 2e-05,
+      "loss": 5.3685,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12214508501861354,
+      "grad_norm": 0.14006795050158746,
+      "learning_rate": 2e-05,
+      "loss": 5.5244,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12221216084783848,
+      "grad_norm": 0.14884161729066828,
+      "learning_rate": 2e-05,
+      "loss": 5.521,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12227923667706342,
+      "grad_norm": 0.1660048239175091,
+      "learning_rate": 2e-05,
+      "loss": 5.5241,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12234631250628836,
+      "grad_norm": 0.1451615853965432,
+      "learning_rate": 2e-05,
+      "loss": 5.2139,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1224133883355133,
+      "grad_norm": 0.14097635186848775,
+      "learning_rate": 2e-05,
+      "loss": 5.4372,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12248046416473823,
+      "grad_norm": 0.14200340832336866,
+      "learning_rate": 2e-05,
+      "loss": 5.3357,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12254753999396317,
+      "grad_norm": 0.14119726168484806,
+      "learning_rate": 2e-05,
+      "loss": 5.3946,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12261461582318811,
+      "grad_norm": 0.14972409064749148,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12268169165241305,
+      "grad_norm": 0.1444109028534444,
+      "learning_rate": 2e-05,
+      "loss": 5.4474,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12274876748163799,
+      "grad_norm": 0.14667671317738692,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12281584331086293,
+      "grad_norm": 0.15319668421207489,
+      "learning_rate": 2e-05,
+      "loss": 5.4554,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12288291914008787,
+      "grad_norm": 0.14859146718439736,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1229499949693128,
+      "grad_norm": 0.15259370193275856,
+      "learning_rate": 2e-05,
+      "loss": 5.4208,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12301707079853774,
+      "grad_norm": 0.16701384899575464,
+      "learning_rate": 2e-05,
+      "loss": 5.3841,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12308414662776268,
+      "grad_norm": 0.15466878829413863,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12315122245698762,
+      "grad_norm": 0.1412058327952156,
+      "learning_rate": 2e-05,
+      "loss": 5.4583,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12321829828621256,
+      "grad_norm": 0.1596037769834483,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1232853741154375,
+      "grad_norm": 0.14664410190725172,
+      "learning_rate": 2e-05,
+      "loss": 5.3442,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12335244994466243,
+      "grad_norm": 0.15262053663577485,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12341952577388737,
+      "grad_norm": 0.14092589111297743,
+      "learning_rate": 2e-05,
+      "loss": 5.5207,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12348660160311233,
+      "grad_norm": 0.15231160188741388,
+      "learning_rate": 2e-05,
+      "loss": 5.4059,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12355367743233726,
+      "grad_norm": 0.14521952531422674,
+      "learning_rate": 2e-05,
+      "loss": 5.4591,
+      "step": 1842
+    },
+    {
+      "epoch": 0.1236207532615622,
+      "grad_norm": 0.1509501358463259,
+      "learning_rate": 2e-05,
+      "loss": 5.5053,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12368782909078714,
+      "grad_norm": 0.14549316231166712,
+      "learning_rate": 2e-05,
+      "loss": 5.3897,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12375490492001208,
+      "grad_norm": 0.14614218362454823,
+      "learning_rate": 2e-05,
+      "loss": 5.4183,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12382198074923702,
+      "grad_norm": 0.14321150185029133,
+      "learning_rate": 2e-05,
+      "loss": 5.2831,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12388905657846196,
+      "grad_norm": 0.1401953172109717,
+      "learning_rate": 2e-05,
+      "loss": 5.5607,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1239561324076869,
+      "grad_norm": 0.15283473959572416,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12402320823691183,
+      "grad_norm": 0.14542488368869816,
+      "learning_rate": 2e-05,
+      "loss": 5.4721,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12409028406613677,
+      "grad_norm": 0.142788269645899,
+      "learning_rate": 2e-05,
+      "loss": 5.3877,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12415735989536171,
+      "grad_norm": 0.1596319510109966,
+      "learning_rate": 2e-05,
+      "loss": 5.3418,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12422443572458665,
+      "grad_norm": 0.14236119882169496,
+      "learning_rate": 2e-05,
+      "loss": 5.5382,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12429151155381159,
+      "grad_norm": 0.1450573742205188,
+      "learning_rate": 2e-05,
+      "loss": 5.4858,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12435858738303653,
+      "grad_norm": 0.1439699942350802,
+      "learning_rate": 2e-05,
+      "loss": 5.3036,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12442566321226146,
+      "grad_norm": 0.14623640847729258,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1244927390414864,
+      "grad_norm": 0.14485557228166146,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12455981487071134,
+      "grad_norm": 0.13982738752351312,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12462689069993628,
+      "grad_norm": 0.14719835269097867,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12469396652916122,
+      "grad_norm": 0.14078426163277716,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12476104235838616,
+      "grad_norm": 0.14297307890878072,
+      "learning_rate": 2e-05,
+      "loss": 5.3307,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1248281181876111,
+      "grad_norm": 0.139571070028144,
+      "learning_rate": 2e-05,
+      "loss": 5.336,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12489519401683603,
+      "grad_norm": 0.14754626481044328,
+      "learning_rate": 2e-05,
+      "loss": 5.5069,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12496226984606097,
+      "grad_norm": 0.1476192120179093,
+      "learning_rate": 2e-05,
+      "loss": 5.4071,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12502934567528592,
+      "grad_norm": 0.13944848725842776,
+      "learning_rate": 2e-05,
+      "loss": 5.4003,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12509642150451086,
+      "grad_norm": 0.13628749788420894,
+      "learning_rate": 2e-05,
+      "loss": 5.5264,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1251634973337358,
+      "grad_norm": 0.1406361742064489,
+      "learning_rate": 2e-05,
+      "loss": 5.3506,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12523057316296074,
+      "grad_norm": 0.15379868404412883,
+      "learning_rate": 2e-05,
+      "loss": 5.5913,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12529764899218568,
+      "grad_norm": 0.14041029202813018,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12536472482141062,
+      "grad_norm": 0.1375451448225345,
+      "learning_rate": 2e-05,
+      "loss": 5.4259,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12543180065063556,
+      "grad_norm": 0.14495931203673493,
+      "learning_rate": 2e-05,
+      "loss": 5.3572,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1254988764798605,
+      "grad_norm": 0.13947340564711863,
+      "learning_rate": 2e-05,
+      "loss": 5.6391,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12556595230908543,
+      "grad_norm": 0.14001171728829823,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12563302813831037,
+      "grad_norm": 0.13996946780171685,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1257001039675353,
+      "grad_norm": 0.15940514558606575,
+      "learning_rate": 2e-05,
+      "loss": 5.5038,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12576717979676025,
+      "grad_norm": 0.13998073533499764,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1258342556259852,
+      "grad_norm": 0.14507551405040323,
+      "learning_rate": 2e-05,
+      "loss": 5.5977,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12590133145521012,
+      "grad_norm": 0.14912971550955875,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12596840728443506,
+      "grad_norm": 0.1436517082454124,
+      "learning_rate": 2e-05,
+      "loss": 5.4746,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12603548311366,
+      "grad_norm": 0.14712974202744067,
+      "learning_rate": 2e-05,
+      "loss": 5.4649,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12610255894288494,
+      "grad_norm": 0.14958263073342865,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12616963477210988,
+      "grad_norm": 0.13893508607020036,
+      "learning_rate": 2e-05,
+      "loss": 5.5317,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12623671060133482,
+      "grad_norm": 0.14249023315739473,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12630378643055976,
+      "grad_norm": 0.1390799480003924,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1263708622597847,
+      "grad_norm": 0.13853824550256427,
+      "learning_rate": 2e-05,
+      "loss": 5.7263,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12643793808900963,
+      "grad_norm": 0.1429390863825981,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12650501391823457,
+      "grad_norm": 0.1358428277225231,
+      "learning_rate": 2e-05,
+      "loss": 5.4367,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1265720897474595,
+      "grad_norm": 0.13757380572600764,
+      "learning_rate": 2e-05,
+      "loss": 5.5579,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12663916557668445,
+      "grad_norm": 0.13883640479380413,
+      "learning_rate": 2e-05,
+      "loss": 5.5154,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1267062414059094,
+      "grad_norm": 0.14680222118438638,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12677331723513433,
+      "grad_norm": 0.14522425770104957,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12684039306435926,
+      "grad_norm": 0.1410063239950081,
+      "learning_rate": 2e-05,
+      "loss": 5.4814,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1269074688935842,
+      "grad_norm": 0.14273449715245337,
+      "learning_rate": 2e-05,
+      "loss": 5.3577,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12697454472280914,
+      "grad_norm": 0.14514641394216807,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12704162055203408,
+      "grad_norm": 0.15444539024524154,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12710869638125902,
+      "grad_norm": 0.1418333521815973,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12717577221048396,
+      "grad_norm": 0.14020006919043362,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1272428480397089,
+      "grad_norm": 0.14581041269686396,
+      "learning_rate": 2e-05,
+      "loss": 5.4439,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12730992386893383,
+      "grad_norm": 0.1494569391065688,
+      "learning_rate": 2e-05,
+      "loss": 5.4719,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12737699969815877,
+      "grad_norm": 0.14118852772518073,
+      "learning_rate": 2e-05,
+      "loss": 5.291,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1274440755273837,
+      "grad_norm": 0.1516878162001335,
+      "learning_rate": 2e-05,
+      "loss": 5.4671,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12751115135660865,
+      "grad_norm": 0.14412963651074975,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1275782271858336,
+      "grad_norm": 0.15173290793838062,
+      "learning_rate": 2e-05,
+      "loss": 5.4234,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12764530301505853,
+      "grad_norm": 0.15498519094169122,
+      "learning_rate": 2e-05,
+      "loss": 5.6382,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12771237884428346,
+      "grad_norm": 0.14258458657910297,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1277794546735084,
+      "grad_norm": 0.14485376639806907,
+      "learning_rate": 2e-05,
+      "loss": 5.6236,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12784653050273334,
+      "grad_norm": 0.1396741327894716,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12791360633195828,
+      "grad_norm": 0.14389001065350449,
+      "learning_rate": 2e-05,
+      "loss": 5.5392,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12798068216118322,
+      "grad_norm": 0.1397066841512997,
+      "learning_rate": 2e-05,
+      "loss": 5.4658,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12804775799040816,
+      "grad_norm": 0.14517243869301452,
+      "learning_rate": 2e-05,
+      "loss": 5.4767,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1281148338196331,
+      "grad_norm": 0.1388457872976633,
+      "learning_rate": 2e-05,
+      "loss": 5.5761,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12818190964885803,
+      "grad_norm": 0.14866331702902877,
+      "learning_rate": 2e-05,
+      "loss": 5.5116,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12824898547808297,
+      "grad_norm": 0.14075411505799154,
+      "learning_rate": 2e-05,
+      "loss": 5.4761,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1283160613073079,
+      "grad_norm": 0.1432050551367564,
+      "learning_rate": 2e-05,
+      "loss": 5.5468,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12838313713653285,
+      "grad_norm": 0.13859470417222766,
+      "learning_rate": 2e-05,
+      "loss": 5.5306,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1284502129657578,
+      "grad_norm": 0.14500776706881147,
+      "learning_rate": 2e-05,
+      "loss": 5.3401,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12851728879498273,
+      "grad_norm": 0.1418737488216362,
+      "learning_rate": 2e-05,
+      "loss": 5.5752,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12858436462420766,
+      "grad_norm": 0.137981971877405,
+      "learning_rate": 2e-05,
+      "loss": 5.5088,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1286514404534326,
+      "grad_norm": 0.1462245767545126,
+      "learning_rate": 2e-05,
+      "loss": 5.4777,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12871851628265754,
+      "grad_norm": 0.1356981712344521,
+      "learning_rate": 2e-05,
+      "loss": 5.5119,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12878559211188248,
+      "grad_norm": 0.14204418232973992,
+      "learning_rate": 2e-05,
+      "loss": 5.6162,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12885266794110742,
+      "grad_norm": 0.15423285890345337,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12891974377033236,
+      "grad_norm": 0.1424266148329285,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1289868195995573,
+      "grad_norm": 0.1448449818573184,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12905389542878223,
+      "grad_norm": 0.14961858589307625,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12912097125800717,
+      "grad_norm": 0.14587724104963767,
+      "learning_rate": 2e-05,
+      "loss": 5.3631,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1291880470872321,
+      "grad_norm": 0.1413589834181679,
+      "learning_rate": 2e-05,
+      "loss": 5.4963,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12925512291645705,
+      "grad_norm": 0.14994918456690476,
+      "learning_rate": 2e-05,
+      "loss": 5.639,
+      "step": 1927
+    },
+    {
+      "epoch": 0.129322198745682,
+      "grad_norm": 0.1514793497508335,
+      "learning_rate": 2e-05,
+      "loss": 5.44,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12938927457490693,
+      "grad_norm": 0.13849240355975054,
+      "learning_rate": 2e-05,
+      "loss": 5.4278,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12945635040413186,
+      "grad_norm": 0.14574276114739315,
+      "learning_rate": 2e-05,
+      "loss": 5.3476,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1295234262333568,
+      "grad_norm": 0.14827970077252967,
+      "learning_rate": 2e-05,
+      "loss": 5.3418,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12959050206258174,
+      "grad_norm": 0.1504019455227756,
+      "learning_rate": 2e-05,
+      "loss": 5.4283,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12965757789180668,
+      "grad_norm": 0.14784610859339736,
+      "learning_rate": 2e-05,
+      "loss": 5.5211,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12972465372103162,
+      "grad_norm": 0.14782922214675503,
+      "learning_rate": 2e-05,
+      "loss": 5.4848,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12979172955025656,
+      "grad_norm": 0.14866727881423464,
+      "learning_rate": 2e-05,
+      "loss": 5.5655,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1298588053794815,
+      "grad_norm": 0.14702017211930346,
+      "learning_rate": 2e-05,
+      "loss": 5.4899,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12992588120870643,
+      "grad_norm": 0.14490956006807051,
+      "learning_rate": 2e-05,
+      "loss": 5.4319,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12999295703793137,
+      "grad_norm": 0.14604952746485914,
+      "learning_rate": 2e-05,
+      "loss": 5.3937,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1300600328671563,
+      "grad_norm": 0.14661868018948063,
+      "learning_rate": 2e-05,
+      "loss": 5.5462,
+      "step": 1939
+    },
+    {
+      "epoch": 0.13012710869638125,
+      "grad_norm": 0.1412940904409667,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1301941845256062,
+      "grad_norm": 0.1399941198613891,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 1941
+    },
+    {
+      "epoch": 0.13026126035483113,
+      "grad_norm": 0.14209111357235482,
+      "learning_rate": 2e-05,
+      "loss": 5.413,
+      "step": 1942
+    },
+    {
+      "epoch": 0.13032833618405607,
+      "grad_norm": 0.14345695700888858,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 1943
+    },
+    {
+      "epoch": 0.130395412013281,
+      "grad_norm": 0.140390032319596,
+      "learning_rate": 2e-05,
+      "loss": 5.4833,
+      "step": 1944
+    },
+    {
+      "epoch": 0.13046248784250594,
+      "grad_norm": 0.14957044459065189,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 1945
+    },
+    {
+      "epoch": 0.13052956367173088,
+      "grad_norm": 0.15047774100179956,
+      "learning_rate": 2e-05,
+      "loss": 5.4188,
+      "step": 1946
+    },
+    {
+      "epoch": 0.13059663950095582,
+      "grad_norm": 0.13661743945238827,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 1947
+    },
+    {
+      "epoch": 0.13066371533018076,
+      "grad_norm": 0.1479856596852377,
+      "learning_rate": 2e-05,
+      "loss": 5.4485,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1307307911594057,
+      "grad_norm": 0.14849125874761002,
+      "learning_rate": 2e-05,
+      "loss": 5.3778,
+      "step": 1949
+    },
+    {
+      "epoch": 0.13079786698863063,
+      "grad_norm": 0.138677944949231,
+      "learning_rate": 2e-05,
+      "loss": 5.5215,
+      "step": 1950
+    },
+    {
+      "epoch": 0.13086494281785557,
+      "grad_norm": 0.14113947834098162,
+      "learning_rate": 2e-05,
+      "loss": 5.401,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1309320186470805,
+      "grad_norm": 0.1394489516464427,
+      "learning_rate": 2e-05,
+      "loss": 5.5037,
+      "step": 1952
+    },
+    {
+      "epoch": 0.13099909447630545,
+      "grad_norm": 0.14306128135125692,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1310661703055304,
+      "grad_norm": 0.13735841387593792,
+      "learning_rate": 2e-05,
+      "loss": 5.2974,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13113324613475535,
+      "grad_norm": 0.13723900171513598,
+      "learning_rate": 2e-05,
+      "loss": 5.5029,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1312003219639803,
+      "grad_norm": 0.1465056126284347,
+      "learning_rate": 2e-05,
+      "loss": 5.5137,
+      "step": 1956
+    },
+    {
+      "epoch": 0.13126739779320523,
+      "grad_norm": 0.14029783326024664,
+      "learning_rate": 2e-05,
+      "loss": 5.7241,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13133447362243017,
+      "grad_norm": 0.13784125081570311,
+      "learning_rate": 2e-05,
+      "loss": 5.3664,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1314015494516551,
+      "grad_norm": 0.1409874070215786,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 1959
+    },
+    {
+      "epoch": 0.13146862528088005,
+      "grad_norm": 0.14270340617250896,
+      "learning_rate": 2e-05,
+      "loss": 5.5299,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13153570111010499,
+      "grad_norm": 0.14912783971889138,
+      "learning_rate": 2e-05,
+      "loss": 5.4865,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13160277693932992,
+      "grad_norm": 0.1422516542171241,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13166985276855486,
+      "grad_norm": 0.1453387064797123,
+      "learning_rate": 2e-05,
+      "loss": 5.3643,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1317369285977798,
+      "grad_norm": 0.14949059271812726,
+      "learning_rate": 2e-05,
+      "loss": 5.6873,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13180400442700474,
+      "grad_norm": 0.1404432059411551,
+      "learning_rate": 2e-05,
+      "loss": 5.5055,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13187108025622968,
+      "grad_norm": 0.1432290805989564,
+      "learning_rate": 2e-05,
+      "loss": 5.4086,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13193815608545462,
+      "grad_norm": 0.14123093013397076,
+      "learning_rate": 2e-05,
+      "loss": 5.4156,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13200523191467955,
+      "grad_norm": 0.14214029368687892,
+      "learning_rate": 2e-05,
+      "loss": 5.3892,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1320723077439045,
+      "grad_norm": 0.135519582306652,
+      "learning_rate": 2e-05,
+      "loss": 5.4769,
+      "step": 1969
+    },
+    {
+      "epoch": 0.13213938357312943,
+      "grad_norm": 0.14237597309070363,
+      "learning_rate": 2e-05,
+      "loss": 5.5432,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13220645940235437,
+      "grad_norm": 0.13881577946563872,
+      "learning_rate": 2e-05,
+      "loss": 5.4897,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1322735352315793,
+      "grad_norm": 0.13825644313865254,
+      "learning_rate": 2e-05,
+      "loss": 5.4958,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13234061106080425,
+      "grad_norm": 0.14815681523144525,
+      "learning_rate": 2e-05,
+      "loss": 5.449,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13240768689002919,
+      "grad_norm": 0.14751084474456727,
+      "learning_rate": 2e-05,
+      "loss": 5.4038,
+      "step": 1974
+    },
+    {
+      "epoch": 0.13247476271925412,
+      "grad_norm": 0.1414761258267842,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13254183854847906,
+      "grad_norm": 0.14716181431497516,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 1976
+    },
+    {
+      "epoch": 0.132608914377704,
+      "grad_norm": 0.13806666511571658,
+      "learning_rate": 2e-05,
+      "loss": 5.5455,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13267599020692894,
+      "grad_norm": 0.13922469720910405,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13274306603615388,
+      "grad_norm": 0.14865891412056084,
+      "learning_rate": 2e-05,
+      "loss": 5.3761,
+      "step": 1979
+    },
+    {
+      "epoch": 0.13281014186537882,
+      "grad_norm": 0.1424763076427973,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13287721769460376,
+      "grad_norm": 0.1454696590822448,
+      "learning_rate": 2e-05,
+      "loss": 5.4709,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1329442935238287,
+      "grad_norm": 0.13865041053389177,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13301136935305363,
+      "grad_norm": 0.14551415534948073,
+      "learning_rate": 2e-05,
+      "loss": 5.5256,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13307844518227857,
+      "grad_norm": 0.1474735984356603,
+      "learning_rate": 2e-05,
+      "loss": 5.4556,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1331455210115035,
+      "grad_norm": 0.14394084568015036,
+      "learning_rate": 2e-05,
+      "loss": 5.3924,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13321259684072845,
+      "grad_norm": 0.1461027510380752,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1332796726699534,
+      "grad_norm": 0.1488483071654375,
+      "learning_rate": 2e-05,
+      "loss": 5.5081,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13334674849917832,
+      "grad_norm": 0.1396847864715195,
+      "learning_rate": 2e-05,
+      "loss": 5.369,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13341382432840326,
+      "grad_norm": 0.14002972632189548,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1334809001576282,
+      "grad_norm": 0.14073485541851943,
+      "learning_rate": 2e-05,
+      "loss": 5.4307,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13354797598685314,
+      "grad_norm": 0.14217330375588472,
+      "learning_rate": 2e-05,
+      "loss": 5.4872,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13361505181607808,
+      "grad_norm": 0.14052184462782014,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13368212764530302,
+      "grad_norm": 0.14813376420323252,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13374920347452796,
+      "grad_norm": 0.13646098275699603,
+      "learning_rate": 2e-05,
+      "loss": 5.6466,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1338162793037529,
+      "grad_norm": 0.13957628925526722,
+      "learning_rate": 2e-05,
+      "loss": 5.5453,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13388335513297783,
+      "grad_norm": 0.14932722857328756,
+      "learning_rate": 2e-05,
+      "loss": 5.3958,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13395043096220277,
+      "grad_norm": 0.15151667494292811,
+      "learning_rate": 2e-05,
+      "loss": 5.4424,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1340175067914277,
+      "grad_norm": 0.14609554326218485,
+      "learning_rate": 2e-05,
+      "loss": 5.3698,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13408458262065265,
+      "grad_norm": 0.14181228128269063,
+      "learning_rate": 2e-05,
+      "loss": 5.4951,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1341516584498776,
+      "grad_norm": 0.1494612490698301,
+      "learning_rate": 2e-05,
+      "loss": 5.5776,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13421873427910253,
+      "grad_norm": 0.1447508913663154,
+      "learning_rate": 2e-05,
+      "loss": 5.4347,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13428581010832746,
+      "grad_norm": 0.135354379403607,
+      "learning_rate": 2e-05,
+      "loss": 5.4453,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1343528859375524,
+      "grad_norm": 0.1374306260589667,
+      "learning_rate": 2e-05,
+      "loss": 5.4898,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13441996176677734,
+      "grad_norm": 0.13969995883588932,
+      "learning_rate": 2e-05,
+      "loss": 5.4929,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13448703759600228,
+      "grad_norm": 0.139586502232014,
+      "learning_rate": 2e-05,
+      "loss": 5.483,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13455411342522722,
+      "grad_norm": 0.1387304998321867,
+      "learning_rate": 2e-05,
+      "loss": 5.4568,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13462118925445216,
+      "grad_norm": 0.13693910305412965,
+      "learning_rate": 2e-05,
+      "loss": 5.4467,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1346882650836771,
+      "grad_norm": 0.1368695243524809,
+      "learning_rate": 2e-05,
+      "loss": 5.5192,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13475534091290203,
+      "grad_norm": 0.1438899579255941,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13482241674212697,
+      "grad_norm": 0.14654282517301112,
+      "learning_rate": 2e-05,
+      "loss": 5.4737,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1348894925713519,
+      "grad_norm": 0.13720505402364921,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13495656840057685,
+      "grad_norm": 0.1450592621289548,
+      "learning_rate": 2e-05,
+      "loss": 5.2893,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1350236442298018,
+      "grad_norm": 0.13857029719389655,
+      "learning_rate": 2e-05,
+      "loss": 5.3594,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13509072005902673,
+      "grad_norm": 0.147941972618289,
+      "learning_rate": 2e-05,
+      "loss": 5.3475,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13515779588825166,
+      "grad_norm": 0.14386229383674137,
+      "learning_rate": 2e-05,
+      "loss": 5.5088,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1352248717174766,
+      "grad_norm": 0.13832508383023687,
+      "learning_rate": 2e-05,
+      "loss": 5.4,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13529194754670154,
+      "grad_norm": 0.1410064007792989,
+      "learning_rate": 2e-05,
+      "loss": 5.4877,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13535902337592648,
+      "grad_norm": 0.1457767986959116,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13542609920515142,
+      "grad_norm": 0.1432904923965809,
+      "learning_rate": 2e-05,
+      "loss": 5.5244,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13549317503437636,
+      "grad_norm": 0.13967503937241313,
+      "learning_rate": 2e-05,
+      "loss": 5.4017,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1355602508636013,
+      "grad_norm": 0.14837688305869343,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13562732669282623,
+      "grad_norm": 0.13737611932726357,
+      "learning_rate": 2e-05,
+      "loss": 5.3823,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13569440252205117,
+      "grad_norm": 0.14123374240617587,
+      "learning_rate": 2e-05,
+      "loss": 5.3857,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1357614783512761,
+      "grad_norm": 0.16900805642757474,
+      "learning_rate": 2e-05,
+      "loss": 5.4784,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13582855418050105,
+      "grad_norm": 0.14684777583518316,
+      "learning_rate": 2e-05,
+      "loss": 5.334,
+      "step": 2025
+    },
+    {
+      "epoch": 0.135895630009726,
+      "grad_norm": 0.1446995972305585,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13596270583895093,
+      "grad_norm": 0.14118202053839035,
+      "learning_rate": 2e-05,
+      "loss": 5.446,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13602978166817586,
+      "grad_norm": 0.14555215160068108,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1360968574974008,
+      "grad_norm": 0.1401068715846395,
+      "learning_rate": 2e-05,
+      "loss": 5.3414,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13616393332662574,
+      "grad_norm": 0.14626878551584535,
+      "learning_rate": 2e-05,
+      "loss": 5.3572,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13623100915585068,
+      "grad_norm": 0.1423787095430719,
+      "learning_rate": 2e-05,
+      "loss": 5.3844,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13629808498507562,
+      "grad_norm": 0.14412293339296053,
+      "learning_rate": 2e-05,
+      "loss": 5.4014,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13636516081430056,
+      "grad_norm": 0.15175324440485932,
+      "learning_rate": 2e-05,
+      "loss": 5.389,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1364322366435255,
+      "grad_norm": 0.1397606777273193,
+      "learning_rate": 2e-05,
+      "loss": 5.4963,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13649931247275043,
+      "grad_norm": 0.14674533195408315,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13656638830197537,
+      "grad_norm": 0.1447364283306658,
+      "learning_rate": 2e-05,
+      "loss": 5.2774,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1366334641312003,
+      "grad_norm": 0.14365806038075374,
+      "learning_rate": 2e-05,
+      "loss": 5.5085,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13670053996042525,
+      "grad_norm": 0.14616284778443692,
+      "learning_rate": 2e-05,
+      "loss": 5.4005,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1367676157896502,
+      "grad_norm": 0.149914301050013,
+      "learning_rate": 2e-05,
+      "loss": 5.5219,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13683469161887513,
+      "grad_norm": 0.15234679438903173,
+      "learning_rate": 2e-05,
+      "loss": 5.4839,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13690176744810006,
+      "grad_norm": 0.14143879390156558,
+      "learning_rate": 2e-05,
+      "loss": 5.3685,
+      "step": 2041
+    },
+    {
+      "epoch": 0.136968843277325,
+      "grad_norm": 0.14990220753598701,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13703591910654994,
+      "grad_norm": 0.1449612956720609,
+      "learning_rate": 2e-05,
+      "loss": 5.5107,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13710299493577488,
+      "grad_norm": 0.16479745823575775,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13717007076499982,
+      "grad_norm": 0.1451770064850456,
+      "learning_rate": 2e-05,
+      "loss": 5.458,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13723714659422478,
+      "grad_norm": 0.14789898391209222,
+      "learning_rate": 2e-05,
+      "loss": 5.4245,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13730422242344972,
+      "grad_norm": 0.14693746529462,
+      "learning_rate": 2e-05,
+      "loss": 5.508,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13737129825267466,
+      "grad_norm": 0.14860532131671506,
+      "learning_rate": 2e-05,
+      "loss": 5.5726,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1374383740818996,
+      "grad_norm": 0.14814345311288496,
+      "learning_rate": 2e-05,
+      "loss": 5.4744,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13750544991112454,
+      "grad_norm": 0.14666424999576003,
+      "learning_rate": 2e-05,
+      "loss": 5.6521,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13757252574034948,
+      "grad_norm": 0.15168902537562473,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13763960156957442,
+      "grad_norm": 0.14328150444215826,
+      "learning_rate": 2e-05,
+      "loss": 5.5479,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13770667739879935,
+      "grad_norm": 0.1425490431984674,
+      "learning_rate": 2e-05,
+      "loss": 5.612,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1377737532280243,
+      "grad_norm": 0.14956636970194537,
+      "learning_rate": 2e-05,
+      "loss": 5.5478,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13784082905724923,
+      "grad_norm": 0.14180943967009133,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13790790488647417,
+      "grad_norm": 0.13570420702588037,
+      "learning_rate": 2e-05,
+      "loss": 5.5137,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1379749807156991,
+      "grad_norm": 0.1517526719084816,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13804205654492405,
+      "grad_norm": 0.14640368761339154,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13810913237414899,
+      "grad_norm": 0.1395307926984185,
+      "learning_rate": 2e-05,
+      "loss": 5.4405,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13817620820337392,
+      "grad_norm": 0.14198355180662323,
+      "learning_rate": 2e-05,
+      "loss": 5.4186,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13824328403259886,
+      "grad_norm": 0.1614154130269461,
+      "learning_rate": 2e-05,
+      "loss": 5.3232,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1383103598618238,
+      "grad_norm": 0.15129914459303087,
+      "learning_rate": 2e-05,
+      "loss": 5.5696,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13837743569104874,
+      "grad_norm": 0.13555279176585308,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13844451152027368,
+      "grad_norm": 0.14959272535112905,
+      "learning_rate": 2e-05,
+      "loss": 5.4603,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13851158734949862,
+      "grad_norm": 0.1413531061216604,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13857866317872355,
+      "grad_norm": 0.15615542950039502,
+      "learning_rate": 2e-05,
+      "loss": 5.4343,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1386457390079485,
+      "grad_norm": 0.14526522337447637,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13871281483717343,
+      "grad_norm": 0.14011432675400964,
+      "learning_rate": 2e-05,
+      "loss": 5.4477,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13877989066639837,
+      "grad_norm": 0.1448029966268039,
+      "learning_rate": 2e-05,
+      "loss": 5.4613,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1388469664956233,
+      "grad_norm": 0.1497322643153707,
+      "learning_rate": 2e-05,
+      "loss": 5.2901,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13891404232484825,
+      "grad_norm": 0.14241235246061373,
+      "learning_rate": 2e-05,
+      "loss": 5.4274,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13898111815407319,
+      "grad_norm": 0.14007582098754184,
+      "learning_rate": 2e-05,
+      "loss": 5.2934,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13904819398329812,
+      "grad_norm": 0.1454765670991779,
+      "learning_rate": 2e-05,
+      "loss": 5.582,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13911526981252306,
+      "grad_norm": 0.14230513638145892,
+      "learning_rate": 2e-05,
+      "loss": 5.4706,
+      "step": 2074
+    },
+    {
+      "epoch": 0.139182345641748,
+      "grad_norm": 0.14434933079648948,
+      "learning_rate": 2e-05,
+      "loss": 5.4723,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13924942147097294,
+      "grad_norm": 0.147421816105899,
+      "learning_rate": 2e-05,
+      "loss": 5.3296,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13931649730019788,
+      "grad_norm": 0.13653657684269718,
+      "learning_rate": 2e-05,
+      "loss": 5.516,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13938357312942282,
+      "grad_norm": 0.1374611697153059,
+      "learning_rate": 2e-05,
+      "loss": 5.506,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13945064895864775,
+      "grad_norm": 0.14054374810427406,
+      "learning_rate": 2e-05,
+      "loss": 5.3506,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1395177247878727,
+      "grad_norm": 0.14588591518712912,
+      "learning_rate": 2e-05,
+      "loss": 5.402,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13958480061709763,
+      "grad_norm": 0.14541747549107972,
+      "learning_rate": 2e-05,
+      "loss": 5.4473,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13965187644632257,
+      "grad_norm": 0.14735580456517158,
+      "learning_rate": 2e-05,
+      "loss": 5.4775,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1397189522755475,
+      "grad_norm": 0.14913060319894267,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13978602810477245,
+      "grad_norm": 0.147982923460394,
+      "learning_rate": 2e-05,
+      "loss": 5.577,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13985310393399739,
+      "grad_norm": 0.14612828677525555,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13992017976322232,
+      "grad_norm": 0.14380085827701397,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13998725559244726,
+      "grad_norm": 0.14734830397347093,
+      "learning_rate": 2e-05,
+      "loss": 5.3922,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1400543314216722,
+      "grad_norm": 0.14435415307227997,
+      "learning_rate": 2e-05,
+      "loss": 5.566,
+      "step": 2088
+    },
+    {
+      "epoch": 0.14012140725089714,
+      "grad_norm": 0.13674176147553868,
+      "learning_rate": 2e-05,
+      "loss": 5.5672,
+      "step": 2089
+    },
+    {
+      "epoch": 0.14018848308012208,
+      "grad_norm": 0.1434648950600529,
+      "learning_rate": 2e-05,
+      "loss": 5.5865,
+      "step": 2090
+    },
+    {
+      "epoch": 0.14025555890934702,
+      "grad_norm": 0.1433419822775209,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 2091
+    },
+    {
+      "epoch": 0.14032263473857196,
+      "grad_norm": 0.145537044047533,
+      "learning_rate": 2e-05,
+      "loss": 5.3192,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1403897105677969,
+      "grad_norm": 0.14688936720125742,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 2093
+    },
+    {
+      "epoch": 0.14045678639702183,
+      "grad_norm": 0.1516534831742794,
+      "learning_rate": 2e-05,
+      "loss": 5.5181,
+      "step": 2094
+    },
+    {
+      "epoch": 0.14052386222624677,
+      "grad_norm": 0.14629229619602857,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1405909380554717,
+      "grad_norm": 0.1401692963256046,
+      "learning_rate": 2e-05,
+      "loss": 5.3799,
+      "step": 2096
+    },
+    {
+      "epoch": 0.14065801388469665,
+      "grad_norm": 0.1539715245059927,
+      "learning_rate": 2e-05,
+      "loss": 5.534,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1407250897139216,
+      "grad_norm": 0.1415371240069618,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 2098
+    },
+    {
+      "epoch": 0.14079216554314652,
+      "grad_norm": 0.14804672453359422,
+      "learning_rate": 2e-05,
+      "loss": 5.5047,
+      "step": 2099
+    },
+    {
+      "epoch": 0.14085924137237146,
+      "grad_norm": 0.14262560952410744,
+      "learning_rate": 2e-05,
+      "loss": 5.4216,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1409263172015964,
+      "grad_norm": 0.14208317465400025,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 2101
+    },
+    {
+      "epoch": 0.14099339303082134,
+      "grad_norm": 0.14946715975716882,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 2102
+    },
+    {
+      "epoch": 0.14106046886004628,
+      "grad_norm": 0.14069190633653889,
+      "learning_rate": 2e-05,
+      "loss": 5.4868,
+      "step": 2103
+    },
+    {
+      "epoch": 0.14112754468927122,
+      "grad_norm": 0.13571094887960813,
+      "learning_rate": 2e-05,
+      "loss": 5.3288,
+      "step": 2104
+    },
+    {
+      "epoch": 0.14119462051849616,
+      "grad_norm": 0.14455144686027474,
+      "learning_rate": 2e-05,
+      "loss": 5.4554,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1412616963477211,
+      "grad_norm": 0.1459431448506834,
+      "learning_rate": 2e-05,
+      "loss": 5.3115,
+      "step": 2106
+    },
+    {
+      "epoch": 0.14132877217694603,
+      "grad_norm": 0.1384018680219067,
+      "learning_rate": 2e-05,
+      "loss": 5.5212,
+      "step": 2107
+    },
+    {
+      "epoch": 0.14139584800617097,
+      "grad_norm": 0.14976967076771533,
+      "learning_rate": 2e-05,
+      "loss": 5.3876,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1414629238353959,
+      "grad_norm": 0.14454852192919818,
+      "learning_rate": 2e-05,
+      "loss": 5.6078,
+      "step": 2109
+    },
+    {
+      "epoch": 0.14152999966462085,
+      "grad_norm": 0.14163177259582685,
+      "learning_rate": 2e-05,
+      "loss": 5.4388,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1415970754938458,
+      "grad_norm": 0.14883238154839137,
+      "learning_rate": 2e-05,
+      "loss": 5.4799,
+      "step": 2111
+    },
+    {
+      "epoch": 0.14166415132307072,
+      "grad_norm": 0.14911614115275804,
+      "learning_rate": 2e-05,
+      "loss": 5.5243,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14173122715229566,
+      "grad_norm": 0.13433864203980456,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1417983029815206,
+      "grad_norm": 0.1465446417328623,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14186537881074554,
+      "grad_norm": 0.14600045400918793,
+      "learning_rate": 2e-05,
+      "loss": 5.3702,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14193245463997048,
+      "grad_norm": 0.14406164889968967,
+      "learning_rate": 2e-05,
+      "loss": 5.4832,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14199953046919542,
+      "grad_norm": 0.14090250065905935,
+      "learning_rate": 2e-05,
+      "loss": 5.4051,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14206660629842036,
+      "grad_norm": 0.1430461520860423,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1421336821276453,
+      "grad_norm": 0.14040505064416364,
+      "learning_rate": 2e-05,
+      "loss": 5.5316,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14220075795687023,
+      "grad_norm": 0.13844218925268514,
+      "learning_rate": 2e-05,
+      "loss": 5.6115,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14226783378609517,
+      "grad_norm": 0.15981607081188468,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1423349096153201,
+      "grad_norm": 0.14497336791094118,
+      "learning_rate": 2e-05,
+      "loss": 5.4611,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14240198544454505,
+      "grad_norm": 0.14537369836705907,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14246906127377,
+      "grad_norm": 0.15327627284900278,
+      "learning_rate": 2e-05,
+      "loss": 5.5293,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14253613710299493,
+      "grad_norm": 0.16332699981039317,
+      "learning_rate": 2e-05,
+      "loss": 5.4914,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14260321293221986,
+      "grad_norm": 0.14181143872765845,
+      "learning_rate": 2e-05,
+      "loss": 5.5039,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1426702887614448,
+      "grad_norm": 0.16042452925068149,
+      "learning_rate": 2e-05,
+      "loss": 5.3038,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14273736459066974,
+      "grad_norm": 0.15474470048925182,
+      "learning_rate": 2e-05,
+      "loss": 5.6319,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14280444041989468,
+      "grad_norm": 0.1473012960551814,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14287151624911962,
+      "grad_norm": 0.14819107898481138,
+      "learning_rate": 2e-05,
+      "loss": 5.5904,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14293859207834456,
+      "grad_norm": 0.15625828449722654,
+      "learning_rate": 2e-05,
+      "loss": 5.3722,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1430056679075695,
+      "grad_norm": 0.1494386459669284,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14307274373679443,
+      "grad_norm": 0.1425782423003123,
+      "learning_rate": 2e-05,
+      "loss": 5.4499,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14313981956601937,
+      "grad_norm": 0.15249503770800232,
+      "learning_rate": 2e-05,
+      "loss": 5.5472,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1432068953952443,
+      "grad_norm": 0.1520570401289044,
+      "learning_rate": 2e-05,
+      "loss": 5.4849,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14327397122446925,
+      "grad_norm": 0.14510983100042368,
+      "learning_rate": 2e-05,
+      "loss": 5.4872,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14334104705369421,
+      "grad_norm": 0.14342754556248508,
+      "learning_rate": 2e-05,
+      "loss": 5.5008,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14340812288291915,
+      "grad_norm": 0.1530994065118648,
+      "learning_rate": 2e-05,
+      "loss": 5.2658,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1434751987121441,
+      "grad_norm": 0.14587688386255726,
+      "learning_rate": 2e-05,
+      "loss": 5.5043,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14354227454136903,
+      "grad_norm": 0.14396024841486615,
+      "learning_rate": 2e-05,
+      "loss": 5.5005,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14360935037059397,
+      "grad_norm": 0.1497870886897319,
+      "learning_rate": 2e-05,
+      "loss": 5.5405,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1436764261998189,
+      "grad_norm": 0.15436407238643793,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14374350202904385,
+      "grad_norm": 0.15210275575774748,
+      "learning_rate": 2e-05,
+      "loss": 5.6512,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14381057785826878,
+      "grad_norm": 0.14589209349364704,
+      "learning_rate": 2e-05,
+      "loss": 5.4501,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14387765368749372,
+      "grad_norm": 0.15336538657993987,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14394472951671866,
+      "grad_norm": 0.14618768853695918,
+      "learning_rate": 2e-05,
+      "loss": 5.4508,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1440118053459436,
+      "grad_norm": 0.1420582185925039,
+      "learning_rate": 2e-05,
+      "loss": 5.3978,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14407888117516854,
+      "grad_norm": 0.14480051300594715,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14414595700439348,
+      "grad_norm": 0.14757642621031064,
+      "learning_rate": 2e-05,
+      "loss": 5.3278,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14421303283361842,
+      "grad_norm": 0.14622554225796436,
+      "learning_rate": 2e-05,
+      "loss": 5.4272,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14428010866284335,
+      "grad_norm": 0.14664582831387782,
+      "learning_rate": 2e-05,
+      "loss": 5.5526,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1443471844920683,
+      "grad_norm": 0.14587066778078878,
+      "learning_rate": 2e-05,
+      "loss": 5.4913,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14441426032129323,
+      "grad_norm": 0.14987580868355235,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14448133615051817,
+      "grad_norm": 0.14278796309540473,
+      "learning_rate": 2e-05,
+      "loss": 5.5058,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1445484119797431,
+      "grad_norm": 0.14535872377461978,
+      "learning_rate": 2e-05,
+      "loss": 5.5851,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14461548780896805,
+      "grad_norm": 0.15244676984563427,
+      "learning_rate": 2e-05,
+      "loss": 5.3563,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14468256363819298,
+      "grad_norm": 0.14305299912458483,
+      "learning_rate": 2e-05,
+      "loss": 5.5083,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14474963946741792,
+      "grad_norm": 0.14831822003093237,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14481671529664286,
+      "grad_norm": 0.14068281987142375,
+      "learning_rate": 2e-05,
+      "loss": 5.6524,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1448837911258678,
+      "grad_norm": 0.136001486156199,
+      "learning_rate": 2e-05,
+      "loss": 5.4686,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14495086695509274,
+      "grad_norm": 0.13927899315634554,
+      "learning_rate": 2e-05,
+      "loss": 5.5185,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14501794278431768,
+      "grad_norm": 0.14010794400150614,
+      "learning_rate": 2e-05,
+      "loss": 5.5248,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14508501861354262,
+      "grad_norm": 0.14554846241013303,
+      "learning_rate": 2e-05,
+      "loss": 5.4205,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14515209444276755,
+      "grad_norm": 0.1401670456559598,
+      "learning_rate": 2e-05,
+      "loss": 5.3287,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1452191702719925,
+      "grad_norm": 0.13695529979492535,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14528624610121743,
+      "grad_norm": 0.13870956115657285,
+      "learning_rate": 2e-05,
+      "loss": 5.5754,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14535332193044237,
+      "grad_norm": 0.14419526180863027,
+      "learning_rate": 2e-05,
+      "loss": 5.3541,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1454203977596673,
+      "grad_norm": 0.14658600936191776,
+      "learning_rate": 2e-05,
+      "loss": 5.475,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14548747358889225,
+      "grad_norm": 0.14590146222600692,
+      "learning_rate": 2e-05,
+      "loss": 5.4832,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14555454941811719,
+      "grad_norm": 0.1439962269381841,
+      "learning_rate": 2e-05,
+      "loss": 5.3578,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14562162524734212,
+      "grad_norm": 0.14580884668233787,
+      "learning_rate": 2e-05,
+      "loss": 5.4147,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14568870107656706,
+      "grad_norm": 0.14242865571192384,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 2172
+    },
+    {
+      "epoch": 0.145755776905792,
+      "grad_norm": 0.14735232844407223,
+      "learning_rate": 2e-05,
+      "loss": 5.4937,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14582285273501694,
+      "grad_norm": 0.1388656679383615,
+      "learning_rate": 2e-05,
+      "loss": 5.5155,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14588992856424188,
+      "grad_norm": 0.13938224549764658,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14595700439346682,
+      "grad_norm": 0.14709734331036806,
+      "learning_rate": 2e-05,
+      "loss": 5.6044,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14602408022269175,
+      "grad_norm": 0.14439610456307594,
+      "learning_rate": 2e-05,
+      "loss": 5.5195,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1460911560519167,
+      "grad_norm": 0.13945714518718524,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14615823188114163,
+      "grad_norm": 0.14918330422225048,
+      "learning_rate": 2e-05,
+      "loss": 5.5542,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14622530771036657,
+      "grad_norm": 0.14199905843355493,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1462923835395915,
+      "grad_norm": 0.13958120837397936,
+      "learning_rate": 2e-05,
+      "loss": 5.4251,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14635945936881645,
+      "grad_norm": 0.1435304665307187,
+      "learning_rate": 2e-05,
+      "loss": 5.5456,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14642653519804139,
+      "grad_norm": 0.143505956709722,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14649361102726632,
+      "grad_norm": 0.14519884727835403,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14656068685649126,
+      "grad_norm": 0.1493951824550538,
+      "learning_rate": 2e-05,
+      "loss": 5.5119,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1466277626857162,
+      "grad_norm": 0.14894069244209124,
+      "learning_rate": 2e-05,
+      "loss": 5.5139,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14669483851494114,
+      "grad_norm": 0.14261482756850957,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14676191434416608,
+      "grad_norm": 0.1405120190329673,
+      "learning_rate": 2e-05,
+      "loss": 5.3772,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14682899017339102,
+      "grad_norm": 0.15127998187901046,
+      "learning_rate": 2e-05,
+      "loss": 5.3881,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14689606600261595,
+      "grad_norm": 0.14465456808578442,
+      "learning_rate": 2e-05,
+      "loss": 5.5839,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1469631418318409,
+      "grad_norm": 0.14546482835722385,
+      "learning_rate": 2e-05,
+      "loss": 5.437,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14703021766106583,
+      "grad_norm": 0.15266618479632738,
+      "learning_rate": 2e-05,
+      "loss": 5.5806,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14709729349029077,
+      "grad_norm": 0.14043600101504575,
+      "learning_rate": 2e-05,
+      "loss": 5.5099,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1471643693195157,
+      "grad_norm": 0.1392318886014651,
+      "learning_rate": 2e-05,
+      "loss": 5.4123,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14723144514874065,
+      "grad_norm": 0.14988425749621023,
+      "learning_rate": 2e-05,
+      "loss": 5.4001,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14729852097796559,
+      "grad_norm": 0.14841366435953765,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14736559680719052,
+      "grad_norm": 0.14146947574584215,
+      "learning_rate": 2e-05,
+      "loss": 5.556,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14743267263641546,
+      "grad_norm": 0.15176055246465603,
+      "learning_rate": 2e-05,
+      "loss": 5.3864,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1474997484656404,
+      "grad_norm": 0.15154767872970293,
+      "learning_rate": 2e-05,
+      "loss": 5.3568,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14756682429486534,
+      "grad_norm": 0.1356737335629346,
+      "learning_rate": 2e-05,
+      "loss": 5.4971,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14763390012409028,
+      "grad_norm": 0.14602027840593182,
+      "learning_rate": 2e-05,
+      "loss": 5.5186,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14770097595331522,
+      "grad_norm": 0.14476867460642287,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14776805178254016,
+      "grad_norm": 0.14090129458077094,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1478351276117651,
+      "grad_norm": 0.14932640146795353,
+      "learning_rate": 2e-05,
+      "loss": 5.4009,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14790220344099003,
+      "grad_norm": 0.1431406445598946,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14796927927021497,
+      "grad_norm": 0.14582358722961314,
+      "learning_rate": 2e-05,
+      "loss": 5.3745,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1480363550994399,
+      "grad_norm": 0.1390614303786016,
+      "learning_rate": 2e-05,
+      "loss": 5.6172,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14810343092866485,
+      "grad_norm": 0.1400337280454805,
+      "learning_rate": 2e-05,
+      "loss": 5.4746,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14817050675788979,
+      "grad_norm": 0.14745711725274677,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14823758258711472,
+      "grad_norm": 0.14250627037811864,
+      "learning_rate": 2e-05,
+      "loss": 5.4953,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14830465841633966,
+      "grad_norm": 0.14556539121841414,
+      "learning_rate": 2e-05,
+      "loss": 5.4818,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1483717342455646,
+      "grad_norm": 0.14712378306325027,
+      "learning_rate": 2e-05,
+      "loss": 5.6014,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14843881007478954,
+      "grad_norm": 0.1406234374043565,
+      "learning_rate": 2e-05,
+      "loss": 5.4194,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14850588590401448,
+      "grad_norm": 0.14274074721247146,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14857296173323942,
+      "grad_norm": 0.14979598280292283,
+      "learning_rate": 2e-05,
+      "loss": 5.4035,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14864003756246436,
+      "grad_norm": 0.1453821618882125,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1487071133916893,
+      "grad_norm": 0.14069108566684205,
+      "learning_rate": 2e-05,
+      "loss": 5.4717,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14877418922091423,
+      "grad_norm": 0.14182602725321994,
+      "learning_rate": 2e-05,
+      "loss": 5.3915,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14884126505013917,
+      "grad_norm": 0.14916687410903165,
+      "learning_rate": 2e-05,
+      "loss": 5.5567,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1489083408793641,
+      "grad_norm": 0.14884857125911571,
+      "learning_rate": 2e-05,
+      "loss": 5.4354,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14897541670858905,
+      "grad_norm": 0.14257903090820245,
+      "learning_rate": 2e-05,
+      "loss": 5.4405,
+      "step": 2221
+    },
+    {
+      "epoch": 0.149042492537814,
+      "grad_norm": 0.15517829393693822,
+      "learning_rate": 2e-05,
+      "loss": 5.3558,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14910956836703892,
+      "grad_norm": 0.1451589564604694,
+      "learning_rate": 2e-05,
+      "loss": 5.4937,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14917664419626386,
+      "grad_norm": 0.14736625165276526,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1492437200254888,
+      "grad_norm": 0.16117011334112802,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14931079585471374,
+      "grad_norm": 0.1527624923482027,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14937787168393868,
+      "grad_norm": 0.14502147403392052,
+      "learning_rate": 2e-05,
+      "loss": 5.5065,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14944494751316365,
+      "grad_norm": 0.15810757819337676,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14951202334238858,
+      "grad_norm": 0.1571042219769501,
+      "learning_rate": 2e-05,
+      "loss": 5.5172,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14957909917161352,
+      "grad_norm": 0.14277814223132904,
+      "learning_rate": 2e-05,
+      "loss": 5.4234,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14964617500083846,
+      "grad_norm": 0.1488686070843933,
+      "learning_rate": 2e-05,
+      "loss": 5.4678,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1497132508300634,
+      "grad_norm": 0.15864424669060126,
+      "learning_rate": 2e-05,
+      "loss": 5.5101,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14978032665928834,
+      "grad_norm": 0.14838415723065443,
+      "learning_rate": 2e-05,
+      "loss": 5.4721,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14984740248851328,
+      "grad_norm": 0.14078026528329415,
+      "learning_rate": 2e-05,
+      "loss": 5.5701,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14991447831773821,
+      "grad_norm": 0.15356249959265666,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14998155414696315,
+      "grad_norm": 0.1465234196894582,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1500486299761881,
+      "grad_norm": 0.14998744357722843,
+      "learning_rate": 2e-05,
+      "loss": 5.4431,
+      "step": 2237
+    },
+    {
+      "epoch": 0.15011570580541303,
+      "grad_norm": 0.14743260824118767,
+      "learning_rate": 2e-05,
+      "loss": 5.6873,
+      "step": 2238
+    },
+    {
+      "epoch": 0.15018278163463797,
+      "grad_norm": 0.14209126148813192,
+      "learning_rate": 2e-05,
+      "loss": 5.4788,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1502498574638629,
+      "grad_norm": 0.1524338052146644,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 2240
+    },
+    {
+      "epoch": 0.15031693329308785,
+      "grad_norm": 0.14701611645049883,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 2241
+    },
+    {
+      "epoch": 0.15038400912231278,
+      "grad_norm": 0.1423477713808431,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 2242
+    },
+    {
+      "epoch": 0.15045108495153772,
+      "grad_norm": 0.1412729839765959,
+      "learning_rate": 2e-05,
+      "loss": 5.5248,
+      "step": 2243
+    },
+    {
+      "epoch": 0.15051816078076266,
+      "grad_norm": 0.15299669197471893,
+      "learning_rate": 2e-05,
+      "loss": 5.5234,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1505852366099876,
+      "grad_norm": 0.15097127108213343,
+      "learning_rate": 2e-05,
+      "loss": 5.5145,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15065231243921254,
+      "grad_norm": 0.15650572664124027,
+      "learning_rate": 2e-05,
+      "loss": 5.4188,
+      "step": 2246
+    },
+    {
+      "epoch": 0.15071938826843748,
+      "grad_norm": 0.14165577413322508,
+      "learning_rate": 2e-05,
+      "loss": 5.6206,
+      "step": 2247
+    },
+    {
+      "epoch": 0.15078646409766241,
+      "grad_norm": 0.14550932225739427,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 2248
+    },
+    {
+      "epoch": 0.15085353992688735,
+      "grad_norm": 0.15161439769008564,
+      "learning_rate": 2e-05,
+      "loss": 5.5719,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1509206157561123,
+      "grad_norm": 0.1416963164733599,
+      "learning_rate": 2e-05,
+      "loss": 5.459,
+      "step": 2250
+    },
+    {
+      "epoch": 0.15098769158533723,
+      "grad_norm": 0.1400122232410672,
+      "learning_rate": 2e-05,
+      "loss": 5.2783,
+      "step": 2251
+    },
+    {
+      "epoch": 0.15105476741456217,
+      "grad_norm": 0.1429691044726434,
+      "learning_rate": 2e-05,
+      "loss": 5.5537,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1511218432437871,
+      "grad_norm": 0.1460582380069382,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 2253
+    },
+    {
+      "epoch": 0.15118891907301205,
+      "grad_norm": 0.15753842120203146,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 2254
+    },
+    {
+      "epoch": 0.15125599490223698,
+      "grad_norm": 0.14509997776657477,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15132307073146192,
+      "grad_norm": 0.1415921386618362,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 2256
+    },
+    {
+      "epoch": 0.15139014656068686,
+      "grad_norm": 0.14785146233164903,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1514572223899118,
+      "grad_norm": 0.15086071810392968,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 2258
+    },
+    {
+      "epoch": 0.15152429821913674,
+      "grad_norm": 0.14799069108598728,
+      "learning_rate": 2e-05,
+      "loss": 5.5839,
+      "step": 2259
+    },
+    {
+      "epoch": 0.15159137404836168,
+      "grad_norm": 0.14899993294556388,
+      "learning_rate": 2e-05,
+      "loss": 5.4858,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15165844987758662,
+      "grad_norm": 0.16238576602059426,
+      "learning_rate": 2e-05,
+      "loss": 5.377,
+      "step": 2261
+    },
+    {
+      "epoch": 0.15172552570681155,
+      "grad_norm": 0.14208003644090292,
+      "learning_rate": 2e-05,
+      "loss": 5.4137,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1517926015360365,
+      "grad_norm": 0.14011384053976045,
+      "learning_rate": 2e-05,
+      "loss": 5.4898,
+      "step": 2263
+    },
+    {
+      "epoch": 0.15185967736526143,
+      "grad_norm": 0.14519204184100665,
+      "learning_rate": 2e-05,
+      "loss": 5.4516,
+      "step": 2264
+    },
+    {
+      "epoch": 0.15192675319448637,
+      "grad_norm": 0.16102552409919463,
+      "learning_rate": 2e-05,
+      "loss": 5.3099,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1519938290237113,
+      "grad_norm": 0.1410919397614373,
+      "learning_rate": 2e-05,
+      "loss": 5.5681,
+      "step": 2266
+    },
+    {
+      "epoch": 0.15206090485293625,
+      "grad_norm": 0.1502947123798291,
+      "learning_rate": 2e-05,
+      "loss": 5.3604,
+      "step": 2267
+    },
+    {
+      "epoch": 0.15212798068216118,
+      "grad_norm": 0.14630220817539488,
+      "learning_rate": 2e-05,
+      "loss": 5.4826,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15219505651138612,
+      "grad_norm": 0.1538987882435936,
+      "learning_rate": 2e-05,
+      "loss": 5.3781,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15226213234061106,
+      "grad_norm": 0.15178125495389094,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 2270
+    },
+    {
+      "epoch": 0.152329208169836,
+      "grad_norm": 0.13989645217075813,
+      "learning_rate": 2e-05,
+      "loss": 5.5333,
+      "step": 2271
+    },
+    {
+      "epoch": 0.15239628399906094,
+      "grad_norm": 0.14339244285451885,
+      "learning_rate": 2e-05,
+      "loss": 5.5345,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15246335982828588,
+      "grad_norm": 0.1581101726326712,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15253043565751082,
+      "grad_norm": 0.14044416841294247,
+      "learning_rate": 2e-05,
+      "loss": 5.4124,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15259751148673575,
+      "grad_norm": 0.15354680019139158,
+      "learning_rate": 2e-05,
+      "loss": 5.4627,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1526645873159607,
+      "grad_norm": 0.14965510974736163,
+      "learning_rate": 2e-05,
+      "loss": 5.6755,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15273166314518563,
+      "grad_norm": 0.14054227982067247,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15279873897441057,
+      "grad_norm": 0.14763564405088236,
+      "learning_rate": 2e-05,
+      "loss": 5.5406,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1528658148036355,
+      "grad_norm": 0.14781745796692083,
+      "learning_rate": 2e-05,
+      "loss": 5.252,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15293289063286045,
+      "grad_norm": 0.13788657848427882,
+      "learning_rate": 2e-05,
+      "loss": 5.4278,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15299996646208538,
+      "grad_norm": 0.14235988307616068,
+      "learning_rate": 2e-05,
+      "loss": 5.398,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15306704229131032,
+      "grad_norm": 0.14761321434016464,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15313411812053526,
+      "grad_norm": 0.1469045034281214,
+      "learning_rate": 2e-05,
+      "loss": 5.4509,
+      "step": 2283
+    },
+    {
+      "epoch": 0.1532011939497602,
+      "grad_norm": 0.1395772060824934,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 2284
+    },
+    {
+      "epoch": 0.15326826977898514,
+      "grad_norm": 0.14269735970812533,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15333534560821008,
+      "grad_norm": 0.15165497629321506,
+      "learning_rate": 2e-05,
+      "loss": 5.5399,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15340242143743502,
+      "grad_norm": 0.14593553524768385,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15346949726665995,
+      "grad_norm": 0.1495158109531524,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1535365730958849,
+      "grad_norm": 0.14427889031991184,
+      "learning_rate": 2e-05,
+      "loss": 5.5328,
+      "step": 2289
+    },
+    {
+      "epoch": 0.15360364892510983,
+      "grad_norm": 0.14293444331385835,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15367072475433477,
+      "grad_norm": 0.14300788897448516,
+      "learning_rate": 2e-05,
+      "loss": 5.5457,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1537378005835597,
+      "grad_norm": 0.15839380993173519,
+      "learning_rate": 2e-05,
+      "loss": 5.4744,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15380487641278465,
+      "grad_norm": 0.14287159479701322,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 2293
+    },
+    {
+      "epoch": 0.15387195224200959,
+      "grad_norm": 0.14656089975723957,
+      "learning_rate": 2e-05,
+      "loss": 5.3293,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15393902807123452,
+      "grad_norm": 0.14490025836650466,
+      "learning_rate": 2e-05,
+      "loss": 5.3642,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15400610390045946,
+      "grad_norm": 0.1443542456273993,
+      "learning_rate": 2e-05,
+      "loss": 5.5044,
+      "step": 2296
+    },
+    {
+      "epoch": 0.1540731797296844,
+      "grad_norm": 0.1346579724718058,
+      "learning_rate": 2e-05,
+      "loss": 5.5314,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15414025555890934,
+      "grad_norm": 0.13990973010521732,
+      "learning_rate": 2e-05,
+      "loss": 5.3165,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15420733138813428,
+      "grad_norm": 0.1474743572771329,
+      "learning_rate": 2e-05,
+      "loss": 5.555,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15427440721735922,
+      "grad_norm": 0.13783941455984747,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15434148304658415,
+      "grad_norm": 0.14385386809671397,
+      "learning_rate": 2e-05,
+      "loss": 5.3911,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1544085588758091,
+      "grad_norm": 0.15509563998690715,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 2302
+    },
+    {
+      "epoch": 0.15447563470503403,
+      "grad_norm": 0.13849701367142336,
+      "learning_rate": 2e-05,
+      "loss": 5.4053,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15454271053425897,
+      "grad_norm": 0.1379753445525031,
+      "learning_rate": 2e-05,
+      "loss": 5.5365,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1546097863634839,
+      "grad_norm": 0.1463537822310483,
+      "learning_rate": 2e-05,
+      "loss": 5.6788,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15467686219270885,
+      "grad_norm": 0.14593362015315364,
+      "learning_rate": 2e-05,
+      "loss": 5.4789,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15474393802193379,
+      "grad_norm": 0.14128212457608336,
+      "learning_rate": 2e-05,
+      "loss": 5.3883,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15481101385115872,
+      "grad_norm": 0.14518735415621012,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15487808968038366,
+      "grad_norm": 0.15297932827664257,
+      "learning_rate": 2e-05,
+      "loss": 5.4036,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1549451655096086,
+      "grad_norm": 0.14250968749806642,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15501224133883354,
+      "grad_norm": 0.14176724327749027,
+      "learning_rate": 2e-05,
+      "loss": 5.5528,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15507931716805848,
+      "grad_norm": 0.15678985461689396,
+      "learning_rate": 2e-05,
+      "loss": 5.5408,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15514639299728342,
+      "grad_norm": 0.1438157660426479,
+      "learning_rate": 2e-05,
+      "loss": 5.3095,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15521346882650836,
+      "grad_norm": 0.14569620654338855,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1552805446557333,
+      "grad_norm": 0.15447518775506502,
+      "learning_rate": 2e-05,
+      "loss": 5.3291,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15534762048495823,
+      "grad_norm": 0.14461796037760485,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15541469631418317,
+      "grad_norm": 0.1460319735692185,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1554817721434081,
+      "grad_norm": 0.15427368979106088,
+      "learning_rate": 2e-05,
+      "loss": 5.601,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15554884797263308,
+      "grad_norm": 0.14232126372826354,
+      "learning_rate": 2e-05,
+      "loss": 5.3739,
+      "step": 2319
+    },
+    {
+      "epoch": 0.155615923801858,
+      "grad_norm": 0.14131855537227234,
+      "learning_rate": 2e-05,
+      "loss": 5.4119,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15568299963108295,
+      "grad_norm": 0.15281059650498724,
+      "learning_rate": 2e-05,
+      "loss": 5.4194,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1557500754603079,
+      "grad_norm": 0.1492611365731564,
+      "learning_rate": 2e-05,
+      "loss": 5.5631,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15581715128953283,
+      "grad_norm": 0.14218122528193228,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15588422711875777,
+      "grad_norm": 0.14632708519101542,
+      "learning_rate": 2e-05,
+      "loss": 5.5426,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1559513029479827,
+      "grad_norm": 0.14857366712729925,
+      "learning_rate": 2e-05,
+      "loss": 5.3962,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15601837877720764,
+      "grad_norm": 0.1381742955875007,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15608545460643258,
+      "grad_norm": 0.14093020594417033,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15615253043565752,
+      "grad_norm": 0.15163084726916026,
+      "learning_rate": 2e-05,
+      "loss": 5.4639,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15621960626488246,
+      "grad_norm": 0.14052181429635524,
+      "learning_rate": 2e-05,
+      "loss": 5.3433,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1562866820941074,
+      "grad_norm": 0.14076437535567424,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15635375792333234,
+      "grad_norm": 0.15994112976457536,
+      "learning_rate": 2e-05,
+      "loss": 5.5112,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15642083375255728,
+      "grad_norm": 0.14787053249484572,
+      "learning_rate": 2e-05,
+      "loss": 5.4783,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15648790958178221,
+      "grad_norm": 0.14379822738763856,
+      "learning_rate": 2e-05,
+      "loss": 5.4488,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15655498541100715,
+      "grad_norm": 0.14466737255612103,
+      "learning_rate": 2e-05,
+      "loss": 5.3279,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1566220612402321,
+      "grad_norm": 0.14234589272269482,
+      "learning_rate": 2e-05,
+      "loss": 5.337,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15668913706945703,
+      "grad_norm": 0.1474708564561655,
+      "learning_rate": 2e-05,
+      "loss": 5.4071,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15675621289868197,
+      "grad_norm": 0.15872848546892468,
+      "learning_rate": 2e-05,
+      "loss": 5.3319,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1568232887279069,
+      "grad_norm": 0.13906693290851507,
+      "learning_rate": 2e-05,
+      "loss": 5.4662,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15689036455713185,
+      "grad_norm": 0.146758494209616,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15695744038635678,
+      "grad_norm": 0.14466956002255082,
+      "learning_rate": 2e-05,
+      "loss": 5.6068,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15702451621558172,
+      "grad_norm": 0.15101040715007039,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15709159204480666,
+      "grad_norm": 0.1383889866808492,
+      "learning_rate": 2e-05,
+      "loss": 5.4893,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1571586678740316,
+      "grad_norm": 0.14576142055712432,
+      "learning_rate": 2e-05,
+      "loss": 5.4446,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15722574370325654,
+      "grad_norm": 0.14264499200962064,
+      "learning_rate": 2e-05,
+      "loss": 5.4257,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15729281953248148,
+      "grad_norm": 0.14047120332190827,
+      "learning_rate": 2e-05,
+      "loss": 5.423,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15735989536170641,
+      "grad_norm": 0.14000617612931074,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15742697119093135,
+      "grad_norm": 0.13846066797179535,
+      "learning_rate": 2e-05,
+      "loss": 5.4984,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1574940470201563,
+      "grad_norm": 0.1452805748940401,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15756112284938123,
+      "grad_norm": 0.14804840319473359,
+      "learning_rate": 2e-05,
+      "loss": 5.6158,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15762819867860617,
+      "grad_norm": 0.14148547182330085,
+      "learning_rate": 2e-05,
+      "loss": 5.5848,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1576952745078311,
+      "grad_norm": 0.15533229017682926,
+      "learning_rate": 2e-05,
+      "loss": 5.3411,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15776235033705605,
+      "grad_norm": 0.14671536178520045,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15782942616628098,
+      "grad_norm": 0.14266904759102816,
+      "learning_rate": 2e-05,
+      "loss": 5.4372,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15789650199550592,
+      "grad_norm": 0.14255542211768477,
+      "learning_rate": 2e-05,
+      "loss": 5.3497,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15796357782473086,
+      "grad_norm": 0.14507958252532455,
+      "learning_rate": 2e-05,
+      "loss": 5.5019,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1580306536539558,
+      "grad_norm": 0.148113391721394,
+      "learning_rate": 2e-05,
+      "loss": 5.5866,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15809772948318074,
+      "grad_norm": 0.15192451042737692,
+      "learning_rate": 2e-05,
+      "loss": 5.2997,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15816480531240568,
+      "grad_norm": 0.14638840157250887,
+      "learning_rate": 2e-05,
+      "loss": 5.3291,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15823188114163061,
+      "grad_norm": 0.14401264352904009,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15829895697085555,
+      "grad_norm": 0.14054968350110053,
+      "learning_rate": 2e-05,
+      "loss": 5.2867,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1583660328000805,
+      "grad_norm": 0.1442489339767662,
+      "learning_rate": 2e-05,
+      "loss": 5.5353,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15843310862930543,
+      "grad_norm": 0.14579198364257284,
+      "learning_rate": 2e-05,
+      "loss": 5.5005,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15850018445853037,
+      "grad_norm": 0.1447535931455502,
+      "learning_rate": 2e-05,
+      "loss": 5.4216,
+      "step": 2363
+    },
+    {
+      "epoch": 0.1585672602877553,
+      "grad_norm": 0.14231432042940004,
+      "learning_rate": 2e-05,
+      "loss": 5.5694,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15863433611698025,
+      "grad_norm": 0.13863330570613475,
+      "learning_rate": 2e-05,
+      "loss": 5.5704,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15870141194620518,
+      "grad_norm": 0.14107277331954826,
+      "learning_rate": 2e-05,
+      "loss": 5.3608,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15876848777543012,
+      "grad_norm": 0.14620546600502357,
+      "learning_rate": 2e-05,
+      "loss": 5.4877,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15883556360465506,
+      "grad_norm": 0.1443168444763351,
+      "learning_rate": 2e-05,
+      "loss": 5.4214,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15890263943388,
+      "grad_norm": 0.15068525676391123,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15896971526310494,
+      "grad_norm": 0.1451898049614518,
+      "learning_rate": 2e-05,
+      "loss": 5.4835,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15903679109232988,
+      "grad_norm": 0.14123767103545656,
+      "learning_rate": 2e-05,
+      "loss": 5.3891,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15910386692155482,
+      "grad_norm": 0.1418601006909018,
+      "learning_rate": 2e-05,
+      "loss": 5.5425,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15917094275077975,
+      "grad_norm": 0.1405122826624735,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1592380185800047,
+      "grad_norm": 0.14601902513594311,
+      "learning_rate": 2e-05,
+      "loss": 5.3881,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15930509440922963,
+      "grad_norm": 0.14050065107798837,
+      "learning_rate": 2e-05,
+      "loss": 5.5875,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15937217023845457,
+      "grad_norm": 0.1472672005177831,
+      "learning_rate": 2e-05,
+      "loss": 5.556,
+      "step": 2376
+    },
+    {
+      "epoch": 0.1594392460676795,
+      "grad_norm": 0.14121171956892806,
+      "learning_rate": 2e-05,
+      "loss": 5.466,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15950632189690445,
+      "grad_norm": 0.14198356157085337,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15957339772612938,
+      "grad_norm": 0.14283160145559864,
+      "learning_rate": 2e-05,
+      "loss": 5.4047,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15964047355535432,
+      "grad_norm": 0.14424342998396564,
+      "learning_rate": 2e-05,
+      "loss": 5.3986,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15970754938457926,
+      "grad_norm": 0.14275374736912896,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1597746252138042,
+      "grad_norm": 0.14055076114585294,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15984170104302914,
+      "grad_norm": 0.14430590939194896,
+      "learning_rate": 2e-05,
+      "loss": 5.5807,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15990877687225408,
+      "grad_norm": 0.14568661492738738,
+      "learning_rate": 2e-05,
+      "loss": 5.4035,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15997585270147902,
+      "grad_norm": 0.14533866084154484,
+      "learning_rate": 2e-05,
+      "loss": 5.4999,
+      "step": 2385
+    },
+    {
+      "epoch": 0.16004292853070395,
+      "grad_norm": 0.14165591007247472,
+      "learning_rate": 2e-05,
+      "loss": 5.4439,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1601100043599289,
+      "grad_norm": 0.15333755943418206,
+      "learning_rate": 2e-05,
+      "loss": 5.4862,
+      "step": 2387
+    },
+    {
+      "epoch": 0.16017708018915383,
+      "grad_norm": 0.15097394370361303,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 2388
+    },
+    {
+      "epoch": 0.16024415601837877,
+      "grad_norm": 0.1437002248399569,
+      "learning_rate": 2e-05,
+      "loss": 5.3247,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1603112318476037,
+      "grad_norm": 0.15252247747533082,
+      "learning_rate": 2e-05,
+      "loss": 5.5615,
+      "step": 2390
+    },
+    {
+      "epoch": 0.16037830767682865,
+      "grad_norm": 0.14718975005068968,
+      "learning_rate": 2e-05,
+      "loss": 5.4205,
+      "step": 2391
+    },
+    {
+      "epoch": 0.16044538350605358,
+      "grad_norm": 0.1392312283211577,
+      "learning_rate": 2e-05,
+      "loss": 5.623,
+      "step": 2392
+    },
+    {
+      "epoch": 0.16051245933527852,
+      "grad_norm": 0.1440021197958591,
+      "learning_rate": 2e-05,
+      "loss": 5.491,
+      "step": 2393
+    },
+    {
+      "epoch": 0.16057953516450346,
+      "grad_norm": 0.14931590207611323,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1606466109937284,
+      "grad_norm": 0.13793300310197534,
+      "learning_rate": 2e-05,
+      "loss": 5.5663,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16071368682295334,
+      "grad_norm": 0.14491600666368062,
+      "learning_rate": 2e-05,
+      "loss": 5.3338,
+      "step": 2396
+    },
+    {
+      "epoch": 0.16078076265217828,
+      "grad_norm": 0.14624699069343033,
+      "learning_rate": 2e-05,
+      "loss": 5.4416,
+      "step": 2397
+    },
+    {
+      "epoch": 0.16084783848140322,
+      "grad_norm": 0.14060469449896057,
+      "learning_rate": 2e-05,
+      "loss": 5.509,
+      "step": 2398
+    },
+    {
+      "epoch": 0.16091491431062815,
+      "grad_norm": 0.14333667352401153,
+      "learning_rate": 2e-05,
+      "loss": 5.453,
+      "step": 2399
+    },
+    {
+      "epoch": 0.1609819901398531,
+      "grad_norm": 0.1409207379905979,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 2400
+    },
+    {
+      "epoch": 0.16104906596907803,
+      "grad_norm": 0.14727288087553977,
+      "learning_rate": 2e-05,
+      "loss": 5.3683,
+      "step": 2401
+    },
+    {
+      "epoch": 0.16111614179830297,
+      "grad_norm": 0.1444922269327931,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 2402
+    },
+    {
+      "epoch": 0.1611832176275279,
+      "grad_norm": 0.14470715103006052,
+      "learning_rate": 2e-05,
+      "loss": 5.3595,
+      "step": 2403
+    },
+    {
+      "epoch": 0.16125029345675285,
+      "grad_norm": 0.13843112904215,
+      "learning_rate": 2e-05,
+      "loss": 5.5286,
+      "step": 2404
+    },
+    {
+      "epoch": 0.16131736928597779,
+      "grad_norm": 0.13596651182104355,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 2405
+    },
+    {
+      "epoch": 0.16138444511520272,
+      "grad_norm": 0.1388494231211852,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 2406
+    },
+    {
+      "epoch": 0.16145152094442766,
+      "grad_norm": 0.1405488705765877,
+      "learning_rate": 2e-05,
+      "loss": 5.5408,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1615185967736526,
+      "grad_norm": 0.14393091950556156,
+      "learning_rate": 2e-05,
+      "loss": 5.3754,
+      "step": 2408
+    },
+    {
+      "epoch": 0.16158567260287757,
+      "grad_norm": 0.1384426150831119,
+      "learning_rate": 2e-05,
+      "loss": 5.3615,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1616527484321025,
+      "grad_norm": 0.14487927622349348,
+      "learning_rate": 2e-05,
+      "loss": 5.4452,
+      "step": 2410
+    },
+    {
+      "epoch": 0.16171982426132744,
+      "grad_norm": 0.14137470834390445,
+      "learning_rate": 2e-05,
+      "loss": 5.4825,
+      "step": 2411
+    },
+    {
+      "epoch": 0.16178690009055238,
+      "grad_norm": 0.14263974961942927,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 2412
+    },
+    {
+      "epoch": 0.16185397591977732,
+      "grad_norm": 0.14753571888811567,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 2413
+    },
+    {
+      "epoch": 0.16192105174900226,
+      "grad_norm": 0.14181372368889528,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1619881275782272,
+      "grad_norm": 0.14260603994467655,
+      "learning_rate": 2e-05,
+      "loss": 5.4608,
+      "step": 2415
+    },
+    {
+      "epoch": 0.16205520340745214,
+      "grad_norm": 0.14392079443855477,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 2416
+    },
+    {
+      "epoch": 0.16212227923667707,
+      "grad_norm": 0.14384362961327227,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 2417
+    },
+    {
+      "epoch": 0.162189355065902,
+      "grad_norm": 0.13620456338263684,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 2418
+    },
+    {
+      "epoch": 0.16225643089512695,
+      "grad_norm": 0.14453631292157756,
+      "learning_rate": 2e-05,
+      "loss": 5.3456,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1623235067243519,
+      "grad_norm": 0.148943054681391,
+      "learning_rate": 2e-05,
+      "loss": 5.4943,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16239058255357683,
+      "grad_norm": 0.14697834827808967,
+      "learning_rate": 2e-05,
+      "loss": 5.3806,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16245765838280177,
+      "grad_norm": 0.14614976762829437,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 2422
+    },
+    {
+      "epoch": 0.1625247342120267,
+      "grad_norm": 0.1495011278735686,
+      "learning_rate": 2e-05,
+      "loss": 5.4211,
+      "step": 2423
+    },
+    {
+      "epoch": 0.16259181004125164,
+      "grad_norm": 0.15181933473717937,
+      "learning_rate": 2e-05,
+      "loss": 5.5551,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16265888587047658,
+      "grad_norm": 0.1489569692650287,
+      "learning_rate": 2e-05,
+      "loss": 5.4797,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16272596169970152,
+      "grad_norm": 0.14704931376380545,
+      "learning_rate": 2e-05,
+      "loss": 5.4029,
+      "step": 2426
+    },
+    {
+      "epoch": 0.16279303752892646,
+      "grad_norm": 0.15638987284030623,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1628601133581514,
+      "grad_norm": 0.1415923895528077,
+      "learning_rate": 2e-05,
+      "loss": 5.5635,
+      "step": 2428
+    },
+    {
+      "epoch": 0.16292718918737634,
+      "grad_norm": 0.14864587033960683,
+      "learning_rate": 2e-05,
+      "loss": 5.3037,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16299426501660128,
+      "grad_norm": 0.1540098858896406,
+      "learning_rate": 2e-05,
+      "loss": 5.3839,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1630613408458262,
+      "grad_norm": 0.14910669579702795,
+      "learning_rate": 2e-05,
+      "loss": 5.4609,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16312841667505115,
+      "grad_norm": 0.1537364439712487,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 2432
+    },
+    {
+      "epoch": 0.1631954925042761,
+      "grad_norm": 0.14738337576787472,
+      "learning_rate": 2e-05,
+      "loss": 5.3937,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16326256833350103,
+      "grad_norm": 0.14695192696232734,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16332964416272597,
+      "grad_norm": 0.13994657459932627,
+      "learning_rate": 2e-05,
+      "loss": 5.555,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1633967199919509,
+      "grad_norm": 0.15573085673056739,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16346379582117584,
+      "grad_norm": 0.15569600994706778,
+      "learning_rate": 2e-05,
+      "loss": 5.6321,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16353087165040078,
+      "grad_norm": 0.14659400525148494,
+      "learning_rate": 2e-05,
+      "loss": 5.5168,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16359794747962572,
+      "grad_norm": 0.14710385566080444,
+      "learning_rate": 2e-05,
+      "loss": 5.6336,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16366502330885066,
+      "grad_norm": 0.14565712105286344,
+      "learning_rate": 2e-05,
+      "loss": 5.3387,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1637320991380756,
+      "grad_norm": 0.14829389956014913,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16379917496730054,
+      "grad_norm": 0.1483173395084134,
+      "learning_rate": 2e-05,
+      "loss": 5.3844,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16386625079652548,
+      "grad_norm": 0.14052437686159897,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 2443
+    },
+    {
+      "epoch": 0.16393332662575041,
+      "grad_norm": 0.14391721692042328,
+      "learning_rate": 2e-05,
+      "loss": 5.5697,
+      "step": 2444
+    },
+    {
+      "epoch": 0.16400040245497535,
+      "grad_norm": 0.1529104101336224,
+      "learning_rate": 2e-05,
+      "loss": 5.3822,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1640674782842003,
+      "grad_norm": 0.14719125296781704,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16413455411342523,
+      "grad_norm": 0.1374033884970539,
+      "learning_rate": 2e-05,
+      "loss": 5.518,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16420162994265017,
+      "grad_norm": 0.14612561144026992,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 2448
+    },
+    {
+      "epoch": 0.1642687057718751,
+      "grad_norm": 0.1546508760281814,
+      "learning_rate": 2e-05,
+      "loss": 5.3576,
+      "step": 2449
+    },
+    {
+      "epoch": 0.16433578160110004,
+      "grad_norm": 0.14718804437613583,
+      "learning_rate": 2e-05,
+      "loss": 5.3804,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16440285743032498,
+      "grad_norm": 0.14864599892015126,
+      "learning_rate": 2e-05,
+      "loss": 5.5916,
+      "step": 2451
+    },
+    {
+      "epoch": 0.16446993325954992,
+      "grad_norm": 0.14704502735602804,
+      "learning_rate": 2e-05,
+      "loss": 5.4664,
+      "step": 2452
+    },
+    {
+      "epoch": 0.16453700908877486,
+      "grad_norm": 0.14069335525647098,
+      "learning_rate": 2e-05,
+      "loss": 5.4435,
+      "step": 2453
+    },
+    {
+      "epoch": 0.1646040849179998,
+      "grad_norm": 0.146616862770691,
+      "learning_rate": 2e-05,
+      "loss": 5.3489,
+      "step": 2454
+    },
+    {
+      "epoch": 0.16467116074722474,
+      "grad_norm": 0.14104082784698949,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16473823657644968,
+      "grad_norm": 0.15559415172745028,
+      "learning_rate": 2e-05,
+      "loss": 5.5461,
+      "step": 2456
+    },
+    {
+      "epoch": 0.16480531240567461,
+      "grad_norm": 0.14475506512930691,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 2457
+    },
+    {
+      "epoch": 0.16487238823489955,
+      "grad_norm": 0.1395780744887203,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1649394640641245,
+      "grad_norm": 0.1444394343684067,
+      "learning_rate": 2e-05,
+      "loss": 5.5522,
+      "step": 2459
+    },
+    {
+      "epoch": 0.16500653989334943,
+      "grad_norm": 0.15230689604385875,
+      "learning_rate": 2e-05,
+      "loss": 5.5231,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16507361572257437,
+      "grad_norm": 0.14203627698267512,
+      "learning_rate": 2e-05,
+      "loss": 5.4083,
+      "step": 2461
+    },
+    {
+      "epoch": 0.1651406915517993,
+      "grad_norm": 0.14622774936503205,
+      "learning_rate": 2e-05,
+      "loss": 5.5527,
+      "step": 2462
+    },
+    {
+      "epoch": 0.16520776738102425,
+      "grad_norm": 0.13926649462720872,
+      "learning_rate": 2e-05,
+      "loss": 5.3688,
+      "step": 2463
+    },
+    {
+      "epoch": 0.16527484321024918,
+      "grad_norm": 0.1468571309121751,
+      "learning_rate": 2e-05,
+      "loss": 5.5108,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16534191903947412,
+      "grad_norm": 0.1472747397156153,
+      "learning_rate": 2e-05,
+      "loss": 5.5592,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16540899486869906,
+      "grad_norm": 0.1479018245579689,
+      "learning_rate": 2e-05,
+      "loss": 5.4527,
+      "step": 2466
+    },
+    {
+      "epoch": 0.165476070697924,
+      "grad_norm": 0.1435352654371062,
+      "learning_rate": 2e-05,
+      "loss": 5.4107,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16554314652714894,
+      "grad_norm": 0.14711805060894925,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16561022235637388,
+      "grad_norm": 0.14226674667382594,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16567729818559881,
+      "grad_norm": 0.14123014462152864,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16574437401482375,
+      "grad_norm": 0.14037503317714725,
+      "learning_rate": 2e-05,
+      "loss": 5.5148,
+      "step": 2471
+    },
+    {
+      "epoch": 0.1658114498440487,
+      "grad_norm": 0.14691746862422148,
+      "learning_rate": 2e-05,
+      "loss": 5.4775,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16587852567327363,
+      "grad_norm": 0.14011376358581162,
+      "learning_rate": 2e-05,
+      "loss": 5.5388,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16594560150249857,
+      "grad_norm": 0.14474927034015614,
+      "learning_rate": 2e-05,
+      "loss": 5.5722,
+      "step": 2474
+    },
+    {
+      "epoch": 0.1660126773317235,
+      "grad_norm": 0.14983012045777636,
+      "learning_rate": 2e-05,
+      "loss": 5.4251,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16607975316094845,
+      "grad_norm": 0.14596881905416767,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16614682899017338,
+      "grad_norm": 0.1423328884456758,
+      "learning_rate": 2e-05,
+      "loss": 5.523,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16621390481939832,
+      "grad_norm": 0.14970430880858662,
+      "learning_rate": 2e-05,
+      "loss": 5.3283,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16628098064862326,
+      "grad_norm": 0.14071292896562543,
+      "learning_rate": 2e-05,
+      "loss": 5.4286,
+      "step": 2479
+    },
+    {
+      "epoch": 0.1663480564778482,
+      "grad_norm": 0.1403332150501888,
+      "learning_rate": 2e-05,
+      "loss": 5.5011,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16641513230707314,
+      "grad_norm": 0.1488165086606482,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16648220813629808,
+      "grad_norm": 0.14863697784443577,
+      "learning_rate": 2e-05,
+      "loss": 5.3631,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16654928396552302,
+      "grad_norm": 0.1582226178701242,
+      "learning_rate": 2e-05,
+      "loss": 5.5396,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16661635979474795,
+      "grad_norm": 0.146555542941495,
+      "learning_rate": 2e-05,
+      "loss": 5.5681,
+      "step": 2484
+    },
+    {
+      "epoch": 0.1666834356239729,
+      "grad_norm": 0.143759175618965,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16675051145319783,
+      "grad_norm": 0.14591120245358957,
+      "learning_rate": 2e-05,
+      "loss": 5.3344,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16681758728242277,
+      "grad_norm": 0.1443768288311367,
+      "learning_rate": 2e-05,
+      "loss": 5.5123,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1668846631116477,
+      "grad_norm": 0.1517836606210499,
+      "learning_rate": 2e-05,
+      "loss": 5.4953,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16695173894087265,
+      "grad_norm": 0.15018255158157978,
+      "learning_rate": 2e-05,
+      "loss": 5.3909,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16701881477009758,
+      "grad_norm": 0.15408063460111734,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16708589059932252,
+      "grad_norm": 0.15581964816728078,
+      "learning_rate": 2e-05,
+      "loss": 5.3865,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16715296642854746,
+      "grad_norm": 0.14446025666215612,
+      "learning_rate": 2e-05,
+      "loss": 5.5154,
+      "step": 2492
+    },
+    {
+      "epoch": 0.1672200422577724,
+      "grad_norm": 0.15337181690999274,
+      "learning_rate": 2e-05,
+      "loss": 5.4878,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16728711808699734,
+      "grad_norm": 0.15383406617498016,
+      "learning_rate": 2e-05,
+      "loss": 5.4583,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16735419391622228,
+      "grad_norm": 0.14502013863342825,
+      "learning_rate": 2e-05,
+      "loss": 5.2352,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16742126974544722,
+      "grad_norm": 0.15086844840347172,
+      "learning_rate": 2e-05,
+      "loss": 5.3865,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16748834557467215,
+      "grad_norm": 0.15129741960051374,
+      "learning_rate": 2e-05,
+      "loss": 5.6155,
+      "step": 2497
+    },
+    {
+      "epoch": 0.1675554214038971,
+      "grad_norm": 0.14769741057473393,
+      "learning_rate": 2e-05,
+      "loss": 5.3708,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16762249723312203,
+      "grad_norm": 0.1428227216032321,
+      "learning_rate": 2e-05,
+      "loss": 5.5101,
+      "step": 2499
+    },
+    {
+      "epoch": 0.167689573062347,
+      "grad_norm": 0.153125230388444,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16775664889157194,
+      "grad_norm": 0.14334216860909743,
+      "learning_rate": 2e-05,
+      "loss": 5.4267,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16782372472079687,
+      "grad_norm": 0.14051468999010686,
+      "learning_rate": 2e-05,
+      "loss": 5.505,
+      "step": 2502
+    },
+    {
+      "epoch": 0.1678908005500218,
+      "grad_norm": 0.1463822354280018,
+      "learning_rate": 2e-05,
+      "loss": 5.5593,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16795787637924675,
+      "grad_norm": 0.15397087365010878,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 2504
+    },
+    {
+      "epoch": 0.1680249522084717,
+      "grad_norm": 0.1503609189562958,
+      "learning_rate": 2e-05,
+      "loss": 5.489,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16809202803769663,
+      "grad_norm": 0.15221787536572576,
+      "learning_rate": 2e-05,
+      "loss": 5.4132,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16815910386692157,
+      "grad_norm": 0.14121505644195256,
+      "learning_rate": 2e-05,
+      "loss": 5.5514,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1682261796961465,
+      "grad_norm": 0.14247784496313234,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16829325552537144,
+      "grad_norm": 0.13904329871335092,
+      "learning_rate": 2e-05,
+      "loss": 5.5479,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16836033135459638,
+      "grad_norm": 0.14258450757056296,
+      "learning_rate": 2e-05,
+      "loss": 5.5255,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16842740718382132,
+      "grad_norm": 0.14763348743247004,
+      "learning_rate": 2e-05,
+      "loss": 5.5388,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16849448301304626,
+      "grad_norm": 0.14246726380123792,
+      "learning_rate": 2e-05,
+      "loss": 5.4087,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1685615588422712,
+      "grad_norm": 0.14549953143961375,
+      "learning_rate": 2e-05,
+      "loss": 5.5256,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16862863467149614,
+      "grad_norm": 0.15488517306874744,
+      "learning_rate": 2e-05,
+      "loss": 5.4583,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16869571050072107,
+      "grad_norm": 0.14661708893609032,
+      "learning_rate": 2e-05,
+      "loss": 5.5287,
+      "step": 2515
+    },
+    {
+      "epoch": 0.168762786329946,
+      "grad_norm": 0.14498820294071055,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16882986215917095,
+      "grad_norm": 0.15534088412557862,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1688969379883959,
+      "grad_norm": 0.14469022445763455,
+      "learning_rate": 2e-05,
+      "loss": 5.44,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16896401381762083,
+      "grad_norm": 0.1491309692918812,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16903108964684577,
+      "grad_norm": 0.14412939240514347,
+      "learning_rate": 2e-05,
+      "loss": 5.3184,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1690981654760707,
+      "grad_norm": 0.14863584431136675,
+      "learning_rate": 2e-05,
+      "loss": 5.4974,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16916524130529564,
+      "grad_norm": 0.1457113867389181,
+      "learning_rate": 2e-05,
+      "loss": 5.4469,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16923231713452058,
+      "grad_norm": 0.14586032189747342,
+      "learning_rate": 2e-05,
+      "loss": 5.4538,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16929939296374552,
+      "grad_norm": 0.14436916075773082,
+      "learning_rate": 2e-05,
+      "loss": 5.3812,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16936646879297046,
+      "grad_norm": 0.1438460089824862,
+      "learning_rate": 2e-05,
+      "loss": 5.5459,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1694335446221954,
+      "grad_norm": 0.14858133505452512,
+      "learning_rate": 2e-05,
+      "loss": 5.2766,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16950062045142034,
+      "grad_norm": 0.14327387565957866,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16956769628064527,
+      "grad_norm": 0.13766648705931764,
+      "learning_rate": 2e-05,
+      "loss": 5.495,
+      "step": 2528
+    },
+    {
+      "epoch": 0.1696347721098702,
+      "grad_norm": 0.14375311309086336,
+      "learning_rate": 2e-05,
+      "loss": 5.4773,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16970184793909515,
+      "grad_norm": 0.15714557625651787,
+      "learning_rate": 2e-05,
+      "loss": 5.4038,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1697689237683201,
+      "grad_norm": 0.14878397673859023,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16983599959754503,
+      "grad_norm": 0.14568128767722408,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16990307542676997,
+      "grad_norm": 0.14998663230601098,
+      "learning_rate": 2e-05,
+      "loss": 5.5074,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1699701512559949,
+      "grad_norm": 0.1532521644342136,
+      "learning_rate": 2e-05,
+      "loss": 5.603,
+      "step": 2534
+    },
+    {
+      "epoch": 0.17003722708521984,
+      "grad_norm": 0.15190230926023732,
+      "learning_rate": 2e-05,
+      "loss": 5.4622,
+      "step": 2535
+    },
+    {
+      "epoch": 0.17010430291444478,
+      "grad_norm": 0.14891375046742839,
+      "learning_rate": 2e-05,
+      "loss": 5.446,
+      "step": 2536
+    },
+    {
+      "epoch": 0.17017137874366972,
+      "grad_norm": 0.14116277914526928,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 2537
+    },
+    {
+      "epoch": 0.17023845457289466,
+      "grad_norm": 0.1597181887843393,
+      "learning_rate": 2e-05,
+      "loss": 5.5109,
+      "step": 2538
+    },
+    {
+      "epoch": 0.1703055304021196,
+      "grad_norm": 0.14692313895599957,
+      "learning_rate": 2e-05,
+      "loss": 5.626,
+      "step": 2539
+    },
+    {
+      "epoch": 0.17037260623134454,
+      "grad_norm": 0.13903869536577884,
+      "learning_rate": 2e-05,
+      "loss": 5.4246,
+      "step": 2540
+    },
+    {
+      "epoch": 0.17043968206056948,
+      "grad_norm": 0.14981260472766256,
+      "learning_rate": 2e-05,
+      "loss": 5.2405,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1705067578897944,
+      "grad_norm": 0.15502904623142638,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 2542
+    },
+    {
+      "epoch": 0.17057383371901935,
+      "grad_norm": 0.14405677478882048,
+      "learning_rate": 2e-05,
+      "loss": 5.3952,
+      "step": 2543
+    },
+    {
+      "epoch": 0.1706409095482443,
+      "grad_norm": 0.14452201132290635,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 2544
+    },
+    {
+      "epoch": 0.17070798537746923,
+      "grad_norm": 0.15316940108836102,
+      "learning_rate": 2e-05,
+      "loss": 5.4131,
+      "step": 2545
+    },
+    {
+      "epoch": 0.17077506120669417,
+      "grad_norm": 0.14904761356248086,
+      "learning_rate": 2e-05,
+      "loss": 5.5189,
+      "step": 2546
+    },
+    {
+      "epoch": 0.1708421370359191,
+      "grad_norm": 0.14516760933144324,
+      "learning_rate": 2e-05,
+      "loss": 5.4998,
+      "step": 2547
+    },
+    {
+      "epoch": 0.17090921286514404,
+      "grad_norm": 0.15616887055689624,
+      "learning_rate": 2e-05,
+      "loss": 5.4678,
+      "step": 2548
+    },
+    {
+      "epoch": 0.17097628869436898,
+      "grad_norm": 0.15939941201040944,
+      "learning_rate": 2e-05,
+      "loss": 5.2871,
+      "step": 2549
+    },
+    {
+      "epoch": 0.17104336452359392,
+      "grad_norm": 0.13941306936917536,
+      "learning_rate": 2e-05,
+      "loss": 5.4881,
+      "step": 2550
+    },
+    {
+      "epoch": 0.17111044035281886,
+      "grad_norm": 0.14942280679639994,
+      "learning_rate": 2e-05,
+      "loss": 5.6343,
+      "step": 2551
+    },
+    {
+      "epoch": 0.1711775161820438,
+      "grad_norm": 0.14657035317538097,
+      "learning_rate": 2e-05,
+      "loss": 5.338,
+      "step": 2552
+    },
+    {
+      "epoch": 0.17124459201126874,
+      "grad_norm": 0.14399463887162492,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 2553
+    },
+    {
+      "epoch": 0.17131166784049368,
+      "grad_norm": 0.14313600762184503,
+      "learning_rate": 2e-05,
+      "loss": 5.3833,
+      "step": 2554
+    },
+    {
+      "epoch": 0.17137874366971861,
+      "grad_norm": 0.13820762699174657,
+      "learning_rate": 2e-05,
+      "loss": 5.4736,
+      "step": 2555
+    },
+    {
+      "epoch": 0.17144581949894355,
+      "grad_norm": 0.14277506653599614,
+      "learning_rate": 2e-05,
+      "loss": 5.5234,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1715128953281685,
+      "grad_norm": 0.13722279723212297,
+      "learning_rate": 2e-05,
+      "loss": 5.353,
+      "step": 2557
+    },
+    {
+      "epoch": 0.17157997115739343,
+      "grad_norm": 0.14309688166331935,
+      "learning_rate": 2e-05,
+      "loss": 5.5954,
+      "step": 2558
+    },
+    {
+      "epoch": 0.17164704698661837,
+      "grad_norm": 0.1431883071968044,
+      "learning_rate": 2e-05,
+      "loss": 5.3412,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1717141228158433,
+      "grad_norm": 0.1385769118567593,
+      "learning_rate": 2e-05,
+      "loss": 5.4145,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17178119864506824,
+      "grad_norm": 0.13683091479682571,
+      "learning_rate": 2e-05,
+      "loss": 5.3548,
+      "step": 2561
+    },
+    {
+      "epoch": 0.17184827447429318,
+      "grad_norm": 0.1507851029788443,
+      "learning_rate": 2e-05,
+      "loss": 5.5468,
+      "step": 2562
+    },
+    {
+      "epoch": 0.17191535030351812,
+      "grad_norm": 0.149524309959163,
+      "learning_rate": 2e-05,
+      "loss": 5.3498,
+      "step": 2563
+    },
+    {
+      "epoch": 0.17198242613274306,
+      "grad_norm": 0.14025023771948,
+      "learning_rate": 2e-05,
+      "loss": 5.4778,
+      "step": 2564
+    },
+    {
+      "epoch": 0.172049501961968,
+      "grad_norm": 0.14977864327059184,
+      "learning_rate": 2e-05,
+      "loss": 5.3942,
+      "step": 2565
+    },
+    {
+      "epoch": 0.17211657779119294,
+      "grad_norm": 0.14670199837453518,
+      "learning_rate": 2e-05,
+      "loss": 5.5356,
+      "step": 2566
+    },
+    {
+      "epoch": 0.17218365362041788,
+      "grad_norm": 0.14096757608258875,
+      "learning_rate": 2e-05,
+      "loss": 5.3267,
+      "step": 2567
+    },
+    {
+      "epoch": 0.17225072944964281,
+      "grad_norm": 0.13820350684768012,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 2568
+    },
+    {
+      "epoch": 0.17231780527886775,
+      "grad_norm": 0.14178975332184338,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 2569
+    },
+    {
+      "epoch": 0.1723848811080927,
+      "grad_norm": 0.14966146070522898,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17245195693731763,
+      "grad_norm": 0.1447915902058335,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 2571
+    },
+    {
+      "epoch": 0.17251903276654257,
+      "grad_norm": 0.14400967710915882,
+      "learning_rate": 2e-05,
+      "loss": 5.4193,
+      "step": 2572
+    },
+    {
+      "epoch": 0.1725861085957675,
+      "grad_norm": 0.14221123064682598,
+      "learning_rate": 2e-05,
+      "loss": 5.5736,
+      "step": 2573
+    },
+    {
+      "epoch": 0.17265318442499245,
+      "grad_norm": 0.14181898824140626,
+      "learning_rate": 2e-05,
+      "loss": 5.5926,
+      "step": 2574
+    },
+    {
+      "epoch": 0.17272026025421738,
+      "grad_norm": 0.1379653535792752,
+      "learning_rate": 2e-05,
+      "loss": 5.4791,
+      "step": 2575
+    },
+    {
+      "epoch": 0.17278733608344232,
+      "grad_norm": 0.14432015873467263,
+      "learning_rate": 2e-05,
+      "loss": 5.3382,
+      "step": 2576
+    },
+    {
+      "epoch": 0.17285441191266726,
+      "grad_norm": 0.14133352979716302,
+      "learning_rate": 2e-05,
+      "loss": 5.5156,
+      "step": 2577
+    },
+    {
+      "epoch": 0.1729214877418922,
+      "grad_norm": 0.1406533973839638,
+      "learning_rate": 2e-05,
+      "loss": 5.3921,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17298856357111714,
+      "grad_norm": 0.14107192407303223,
+      "learning_rate": 2e-05,
+      "loss": 5.5486,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17305563940034208,
+      "grad_norm": 0.1454510521294858,
+      "learning_rate": 2e-05,
+      "loss": 5.5314,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17312271522956701,
+      "grad_norm": 0.14265554324679652,
+      "learning_rate": 2e-05,
+      "loss": 5.3908,
+      "step": 2581
+    },
+    {
+      "epoch": 0.17318979105879195,
+      "grad_norm": 0.1498309314122728,
+      "learning_rate": 2e-05,
+      "loss": 5.5536,
+      "step": 2582
+    },
+    {
+      "epoch": 0.1732568668880169,
+      "grad_norm": 0.13908207804614398,
+      "learning_rate": 2e-05,
+      "loss": 5.4089,
+      "step": 2583
+    },
+    {
+      "epoch": 0.17332394271724183,
+      "grad_norm": 0.14382631155236927,
+      "learning_rate": 2e-05,
+      "loss": 5.5428,
+      "step": 2584
+    },
+    {
+      "epoch": 0.17339101854646677,
+      "grad_norm": 0.15081781863531626,
+      "learning_rate": 2e-05,
+      "loss": 5.4121,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1734580943756917,
+      "grad_norm": 0.14714714347337973,
+      "learning_rate": 2e-05,
+      "loss": 5.5223,
+      "step": 2586
+    },
+    {
+      "epoch": 0.17352517020491665,
+      "grad_norm": 0.14297866772501977,
+      "learning_rate": 2e-05,
+      "loss": 5.4195,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17359224603414158,
+      "grad_norm": 0.1457451048375711,
+      "learning_rate": 2e-05,
+      "loss": 5.5022,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17365932186336652,
+      "grad_norm": 0.14653831179435475,
+      "learning_rate": 2e-05,
+      "loss": 5.4827,
+      "step": 2589
+    },
+    {
+      "epoch": 0.17372639769259146,
+      "grad_norm": 0.14882892534755696,
+      "learning_rate": 2e-05,
+      "loss": 5.4405,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17379347352181643,
+      "grad_norm": 0.14080708104937065,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17386054935104137,
+      "grad_norm": 0.14368760097482722,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1739276251802663,
+      "grad_norm": 0.14700209159404232,
+      "learning_rate": 2e-05,
+      "loss": 5.5815,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17399470100949124,
+      "grad_norm": 0.14441299524057935,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 2594
+    },
+    {
+      "epoch": 0.17406177683871618,
+      "grad_norm": 0.14195813061211418,
+      "learning_rate": 2e-05,
+      "loss": 5.42,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17412885266794112,
+      "grad_norm": 0.1385742418863064,
+      "learning_rate": 2e-05,
+      "loss": 5.4735,
+      "step": 2596
+    },
+    {
+      "epoch": 0.17419592849716606,
+      "grad_norm": 0.13999079413236934,
+      "learning_rate": 2e-05,
+      "loss": 5.5407,
+      "step": 2597
+    },
+    {
+      "epoch": 0.174263004326391,
+      "grad_norm": 0.1381939618223655,
+      "learning_rate": 2e-05,
+      "loss": 5.2716,
+      "step": 2598
+    },
+    {
+      "epoch": 0.17433008015561594,
+      "grad_norm": 0.1435419586298622,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 2599
+    },
+    {
+      "epoch": 0.17439715598484087,
+      "grad_norm": 0.14055597295911545,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1744642318140658,
+      "grad_norm": 0.14763784106228464,
+      "learning_rate": 2e-05,
+      "loss": 5.5546,
+      "step": 2601
+    },
+    {
+      "epoch": 0.17453130764329075,
+      "grad_norm": 0.14415645761204046,
+      "learning_rate": 2e-05,
+      "loss": 5.4906,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1745983834725157,
+      "grad_norm": 0.14195762727198927,
+      "learning_rate": 2e-05,
+      "loss": 5.5428,
+      "step": 2603
+    },
+    {
+      "epoch": 0.17466545930174063,
+      "grad_norm": 0.1450020273484119,
+      "learning_rate": 2e-05,
+      "loss": 5.5559,
+      "step": 2604
+    },
+    {
+      "epoch": 0.17473253513096557,
+      "grad_norm": 0.15141374075155653,
+      "learning_rate": 2e-05,
+      "loss": 5.4512,
+      "step": 2605
+    },
+    {
+      "epoch": 0.1747996109601905,
+      "grad_norm": 0.14230648253803874,
+      "learning_rate": 2e-05,
+      "loss": 5.3241,
+      "step": 2606
+    },
+    {
+      "epoch": 0.17486668678941544,
+      "grad_norm": 0.14864009509934972,
+      "learning_rate": 2e-05,
+      "loss": 5.3433,
+      "step": 2607
+    },
+    {
+      "epoch": 0.17493376261864038,
+      "grad_norm": 0.1550631145366756,
+      "learning_rate": 2e-05,
+      "loss": 5.5168,
+      "step": 2608
+    },
+    {
+      "epoch": 0.17500083844786532,
+      "grad_norm": 0.1438838288894027,
+      "learning_rate": 2e-05,
+      "loss": 5.5118,
+      "step": 2609
+    },
+    {
+      "epoch": 0.17506791427709026,
+      "grad_norm": 0.14032889457994752,
+      "learning_rate": 2e-05,
+      "loss": 5.3631,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1751349901063152,
+      "grad_norm": 0.15628384122754785,
+      "learning_rate": 2e-05,
+      "loss": 5.5051,
+      "step": 2611
+    },
+    {
+      "epoch": 0.17520206593554014,
+      "grad_norm": 0.15177421551967998,
+      "learning_rate": 2e-05,
+      "loss": 5.3739,
+      "step": 2612
+    },
+    {
+      "epoch": 0.17526914176476507,
+      "grad_norm": 0.148069587949307,
+      "learning_rate": 2e-05,
+      "loss": 5.4031,
+      "step": 2613
+    },
+    {
+      "epoch": 0.17533621759399,
+      "grad_norm": 0.1483591440474578,
+      "learning_rate": 2e-05,
+      "loss": 5.336,
+      "step": 2614
+    },
+    {
+      "epoch": 0.17540329342321495,
+      "grad_norm": 0.14660471756854285,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1754703692524399,
+      "grad_norm": 0.1463103057879146,
+      "learning_rate": 2e-05,
+      "loss": 5.3364,
+      "step": 2616
+    },
+    {
+      "epoch": 0.17553744508166483,
+      "grad_norm": 0.1375757834191774,
+      "learning_rate": 2e-05,
+      "loss": 5.4093,
+      "step": 2617
+    },
+    {
+      "epoch": 0.17560452091088977,
+      "grad_norm": 0.14765199128969533,
+      "learning_rate": 2e-05,
+      "loss": 5.4037,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1756715967401147,
+      "grad_norm": 0.14717473347746984,
+      "learning_rate": 2e-05,
+      "loss": 5.4089,
+      "step": 2619
+    },
+    {
+      "epoch": 0.17573867256933964,
+      "grad_norm": 0.14380617346966226,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17580574839856458,
+      "grad_norm": 0.1421633363916329,
+      "learning_rate": 2e-05,
+      "loss": 5.5549,
+      "step": 2621
+    },
+    {
+      "epoch": 0.17587282422778952,
+      "grad_norm": 0.1442571869768685,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 2622
+    },
+    {
+      "epoch": 0.17593990005701446,
+      "grad_norm": 0.14656364103136127,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 2623
+    },
+    {
+      "epoch": 0.1760069758862394,
+      "grad_norm": 0.1456528183909178,
+      "learning_rate": 2e-05,
+      "loss": 5.4578,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17607405171546434,
+      "grad_norm": 0.15195099716392566,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17614112754468927,
+      "grad_norm": 0.15006684856112534,
+      "learning_rate": 2e-05,
+      "loss": 5.4918,
+      "step": 2626
+    },
+    {
+      "epoch": 0.1762082033739142,
+      "grad_norm": 0.1431274176488639,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17627527920313915,
+      "grad_norm": 0.14942317917910963,
+      "learning_rate": 2e-05,
+      "loss": 5.5498,
+      "step": 2628
+    },
+    {
+      "epoch": 0.1763423550323641,
+      "grad_norm": 0.15221842826883383,
+      "learning_rate": 2e-05,
+      "loss": 5.3539,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17640943086158903,
+      "grad_norm": 0.14446214704974156,
+      "learning_rate": 2e-05,
+      "loss": 5.5342,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17647650669081397,
+      "grad_norm": 0.1414175728282391,
+      "learning_rate": 2e-05,
+      "loss": 5.597,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1765435825200389,
+      "grad_norm": 0.14161112375737342,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17661065834926384,
+      "grad_norm": 0.15247717739353916,
+      "learning_rate": 2e-05,
+      "loss": 5.539,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17667773417848878,
+      "grad_norm": 0.14753757360803704,
+      "learning_rate": 2e-05,
+      "loss": 5.5711,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17674481000771372,
+      "grad_norm": 0.14763146888509768,
+      "learning_rate": 2e-05,
+      "loss": 5.397,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17681188583693866,
+      "grad_norm": 0.14798587618270526,
+      "learning_rate": 2e-05,
+      "loss": 5.4314,
+      "step": 2636
+    },
+    {
+      "epoch": 0.1768789616661636,
+      "grad_norm": 0.1425484358944931,
+      "learning_rate": 2e-05,
+      "loss": 5.2556,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17694603749538854,
+      "grad_norm": 0.14269894422658494,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17701311332461347,
+      "grad_norm": 0.14166497812989312,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 2639
+    },
+    {
+      "epoch": 0.1770801891538384,
+      "grad_norm": 0.1391571415376557,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17714726498306335,
+      "grad_norm": 0.14746512084361776,
+      "learning_rate": 2e-05,
+      "loss": 5.4347,
+      "step": 2641
+    },
+    {
+      "epoch": 0.1772143408122883,
+      "grad_norm": 0.14116341505120372,
+      "learning_rate": 2e-05,
+      "loss": 5.457,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17728141664151323,
+      "grad_norm": 0.1384718884831778,
+      "learning_rate": 2e-05,
+      "loss": 5.5101,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17734849247073817,
+      "grad_norm": 0.1452093019498892,
+      "learning_rate": 2e-05,
+      "loss": 5.5889,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1774155682999631,
+      "grad_norm": 0.14323930671024596,
+      "learning_rate": 2e-05,
+      "loss": 5.4952,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17748264412918804,
+      "grad_norm": 0.1434597925335034,
+      "learning_rate": 2e-05,
+      "loss": 5.5873,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17754971995841298,
+      "grad_norm": 0.1472814790240145,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17761679578763792,
+      "grad_norm": 0.14723157954616875,
+      "learning_rate": 2e-05,
+      "loss": 5.4142,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17768387161686286,
+      "grad_norm": 0.1424632648440163,
+      "learning_rate": 2e-05,
+      "loss": 5.6168,
+      "step": 2649
+    },
+    {
+      "epoch": 0.1777509474460878,
+      "grad_norm": 0.1478986383009081,
+      "learning_rate": 2e-05,
+      "loss": 5.5999,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17781802327531274,
+      "grad_norm": 0.13960371808328045,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17788509910453768,
+      "grad_norm": 0.1525053275354145,
+      "learning_rate": 2e-05,
+      "loss": 5.3542,
+      "step": 2652
+    },
+    {
+      "epoch": 0.1779521749337626,
+      "grad_norm": 0.13802781553167734,
+      "learning_rate": 2e-05,
+      "loss": 5.5051,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17801925076298755,
+      "grad_norm": 0.14738663285153394,
+      "learning_rate": 2e-05,
+      "loss": 5.4788,
+      "step": 2654
+    },
+    {
+      "epoch": 0.1780863265922125,
+      "grad_norm": 0.1479337981843923,
+      "learning_rate": 2e-05,
+      "loss": 5.3711,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17815340242143743,
+      "grad_norm": 0.15134046369634582,
+      "learning_rate": 2e-05,
+      "loss": 5.3958,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17822047825066237,
+      "grad_norm": 0.13970549209905672,
+      "learning_rate": 2e-05,
+      "loss": 5.6274,
+      "step": 2657
+    },
+    {
+      "epoch": 0.1782875540798873,
+      "grad_norm": 0.14311049887555619,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17835462990911224,
+      "grad_norm": 0.15183222374207075,
+      "learning_rate": 2e-05,
+      "loss": 5.5167,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17842170573833718,
+      "grad_norm": 0.15369356492098218,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17848878156756212,
+      "grad_norm": 0.15517384157640332,
+      "learning_rate": 2e-05,
+      "loss": 5.3945,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17855585739678706,
+      "grad_norm": 0.1432847080804353,
+      "learning_rate": 2e-05,
+      "loss": 5.57,
+      "step": 2662
+    },
+    {
+      "epoch": 0.178622933226012,
+      "grad_norm": 0.14467138511265848,
+      "learning_rate": 2e-05,
+      "loss": 5.5256,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17869000905523694,
+      "grad_norm": 0.14325072474701503,
+      "learning_rate": 2e-05,
+      "loss": 5.4998,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17875708488446188,
+      "grad_norm": 0.14083759093790876,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1788241607136868,
+      "grad_norm": 0.1418019764393983,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17889123654291175,
+      "grad_norm": 0.1465686497629299,
+      "learning_rate": 2e-05,
+      "loss": 5.5252,
+      "step": 2667
+    },
+    {
+      "epoch": 0.1789583123721367,
+      "grad_norm": 0.14881814625406883,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17902538820136163,
+      "grad_norm": 0.14785337629111087,
+      "learning_rate": 2e-05,
+      "loss": 5.5233,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17909246403058657,
+      "grad_norm": 0.14261456501281292,
+      "learning_rate": 2e-05,
+      "loss": 5.4047,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1791595398598115,
+      "grad_norm": 0.14375018882959187,
+      "learning_rate": 2e-05,
+      "loss": 5.3515,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17922661568903644,
+      "grad_norm": 0.14038132868925005,
+      "learning_rate": 2e-05,
+      "loss": 5.4422,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17929369151826138,
+      "grad_norm": 0.13987256477233506,
+      "learning_rate": 2e-05,
+      "loss": 5.382,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17936076734748632,
+      "grad_norm": 0.14011649115468783,
+      "learning_rate": 2e-05,
+      "loss": 5.6314,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17942784317671126,
+      "grad_norm": 0.14884877473985555,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1794949190059362,
+      "grad_norm": 0.14546404752440997,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17956199483516114,
+      "grad_norm": 0.13967347764260432,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17962907066438608,
+      "grad_norm": 0.14521520915931624,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17969614649361101,
+      "grad_norm": 0.1447311625517521,
+      "learning_rate": 2e-05,
+      "loss": 5.5465,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17976322232283595,
+      "grad_norm": 0.14143206780291115,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1798302981520609,
+      "grad_norm": 0.14334486130070173,
+      "learning_rate": 2e-05,
+      "loss": 5.3819,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17989737398128586,
+      "grad_norm": 0.15291732014122578,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 2682
+    },
+    {
+      "epoch": 0.1799644498105108,
+      "grad_norm": 0.14466943312902764,
+      "learning_rate": 2e-05,
+      "loss": 5.5716,
+      "step": 2683
+    },
+    {
+      "epoch": 0.18003152563973573,
+      "grad_norm": 0.14620871349027964,
+      "learning_rate": 2e-05,
+      "loss": 5.4146,
+      "step": 2684
+    },
+    {
+      "epoch": 0.18009860146896067,
+      "grad_norm": 0.1451252652445664,
+      "learning_rate": 2e-05,
+      "loss": 5.4678,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1801656772981856,
+      "grad_norm": 0.14460873433451146,
+      "learning_rate": 2e-05,
+      "loss": 5.5468,
+      "step": 2686
+    },
+    {
+      "epoch": 0.18023275312741055,
+      "grad_norm": 0.13963124286347603,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 2687
+    },
+    {
+      "epoch": 0.1802998289566355,
+      "grad_norm": 0.1420901321803039,
+      "learning_rate": 2e-05,
+      "loss": 5.3788,
+      "step": 2688
+    },
+    {
+      "epoch": 0.18036690478586043,
+      "grad_norm": 0.14343231481174518,
+      "learning_rate": 2e-05,
+      "loss": 5.4295,
+      "step": 2689
+    },
+    {
+      "epoch": 0.18043398061508537,
+      "grad_norm": 0.13709606755888748,
+      "learning_rate": 2e-05,
+      "loss": 5.5792,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1805010564443103,
+      "grad_norm": 0.1468829691034455,
+      "learning_rate": 2e-05,
+      "loss": 5.4253,
+      "step": 2691
+    },
+    {
+      "epoch": 0.18056813227353524,
+      "grad_norm": 0.14048648408284678,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 2692
+    },
+    {
+      "epoch": 0.18063520810276018,
+      "grad_norm": 0.14213439859482596,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 2693
+    },
+    {
+      "epoch": 0.18070228393198512,
+      "grad_norm": 0.1479588447526294,
+      "learning_rate": 2e-05,
+      "loss": 5.4792,
+      "step": 2694
+    },
+    {
+      "epoch": 0.18076935976121006,
+      "grad_norm": 0.1446408909061203,
+      "learning_rate": 2e-05,
+      "loss": 5.6064,
+      "step": 2695
+    },
+    {
+      "epoch": 0.180836435590435,
+      "grad_norm": 0.14184480954709416,
+      "learning_rate": 2e-05,
+      "loss": 5.481,
+      "step": 2696
+    },
+    {
+      "epoch": 0.18090351141965993,
+      "grad_norm": 0.14481317391697998,
+      "learning_rate": 2e-05,
+      "loss": 5.5432,
+      "step": 2697
+    },
+    {
+      "epoch": 0.18097058724888487,
+      "grad_norm": 0.14829224119551393,
+      "learning_rate": 2e-05,
+      "loss": 5.5077,
+      "step": 2698
+    },
+    {
+      "epoch": 0.1810376630781098,
+      "grad_norm": 0.13742211598567433,
+      "learning_rate": 2e-05,
+      "loss": 5.3533,
+      "step": 2699
+    },
+    {
+      "epoch": 0.18110473890733475,
+      "grad_norm": 0.13895532949840653,
+      "learning_rate": 2e-05,
+      "loss": 5.455,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1811718147365597,
+      "grad_norm": 0.14259474766112434,
+      "learning_rate": 2e-05,
+      "loss": 5.4728,
+      "step": 2701
+    },
+    {
+      "epoch": 0.18123889056578463,
+      "grad_norm": 0.14779709666127674,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 2702
+    },
+    {
+      "epoch": 0.18130596639500957,
+      "grad_norm": 0.1373766791159465,
+      "learning_rate": 2e-05,
+      "loss": 5.2521,
+      "step": 2703
+    },
+    {
+      "epoch": 0.1813730422242345,
+      "grad_norm": 0.14163590753237487,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 2704
+    },
+    {
+      "epoch": 0.18144011805345944,
+      "grad_norm": 0.14626805488372077,
+      "learning_rate": 2e-05,
+      "loss": 5.2864,
+      "step": 2705
+    },
+    {
+      "epoch": 0.18150719388268438,
+      "grad_norm": 0.1425107728687988,
+      "learning_rate": 2e-05,
+      "loss": 5.542,
+      "step": 2706
+    },
+    {
+      "epoch": 0.18157426971190932,
+      "grad_norm": 0.144817730661276,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 2707
+    },
+    {
+      "epoch": 0.18164134554113426,
+      "grad_norm": 0.1493922397984341,
+      "learning_rate": 2e-05,
+      "loss": 5.5852,
+      "step": 2708
+    },
+    {
+      "epoch": 0.1817084213703592,
+      "grad_norm": 0.14772091061664142,
+      "learning_rate": 2e-05,
+      "loss": 5.515,
+      "step": 2709
+    },
+    {
+      "epoch": 0.18177549719958414,
+      "grad_norm": 0.14612150406129878,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18184257302880907,
+      "grad_norm": 0.14354653922212127,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 2711
+    },
+    {
+      "epoch": 0.181909648858034,
+      "grad_norm": 0.161969393548271,
+      "learning_rate": 2e-05,
+      "loss": 5.3918,
+      "step": 2712
+    },
+    {
+      "epoch": 0.18197672468725895,
+      "grad_norm": 0.1425930264558271,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 2713
+    },
+    {
+      "epoch": 0.1820438005164839,
+      "grad_norm": 0.15342648392973557,
+      "learning_rate": 2e-05,
+      "loss": 5.6016,
+      "step": 2714
+    },
+    {
+      "epoch": 0.18211087634570883,
+      "grad_norm": 0.14690007618081394,
+      "learning_rate": 2e-05,
+      "loss": 5.5322,
+      "step": 2715
+    },
+    {
+      "epoch": 0.18217795217493377,
+      "grad_norm": 0.1526073653053425,
+      "learning_rate": 2e-05,
+      "loss": 5.4781,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1822450280041587,
+      "grad_norm": 0.14258527717093078,
+      "learning_rate": 2e-05,
+      "loss": 5.3875,
+      "step": 2717
+    },
+    {
+      "epoch": 0.18231210383338364,
+      "grad_norm": 0.15120540034054075,
+      "learning_rate": 2e-05,
+      "loss": 5.4578,
+      "step": 2718
+    },
+    {
+      "epoch": 0.18237917966260858,
+      "grad_norm": 0.15403161016922193,
+      "learning_rate": 2e-05,
+      "loss": 5.3631,
+      "step": 2719
+    },
+    {
+      "epoch": 0.18244625549183352,
+      "grad_norm": 0.149735981742783,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 2720
+    },
+    {
+      "epoch": 0.18251333132105846,
+      "grad_norm": 0.1442048921818772,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 2721
+    },
+    {
+      "epoch": 0.1825804071502834,
+      "grad_norm": 0.1580239085696565,
+      "learning_rate": 2e-05,
+      "loss": 5.5962,
+      "step": 2722
+    },
+    {
+      "epoch": 0.18264748297950834,
+      "grad_norm": 0.15596444762442319,
+      "learning_rate": 2e-05,
+      "loss": 5.3992,
+      "step": 2723
+    },
+    {
+      "epoch": 0.18271455880873327,
+      "grad_norm": 0.15152131974537503,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 2724
+    },
+    {
+      "epoch": 0.1827816346379582,
+      "grad_norm": 0.16651044146488672,
+      "learning_rate": 2e-05,
+      "loss": 5.4338,
+      "step": 2725
+    },
+    {
+      "epoch": 0.18284871046718315,
+      "grad_norm": 0.1566294463142734,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 2726
+    },
+    {
+      "epoch": 0.1829157862964081,
+      "grad_norm": 0.15208251439897374,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 2727
+    },
+    {
+      "epoch": 0.18298286212563303,
+      "grad_norm": 0.16130038926673448,
+      "learning_rate": 2e-05,
+      "loss": 5.5043,
+      "step": 2728
+    },
+    {
+      "epoch": 0.18304993795485797,
+      "grad_norm": 0.15103364020010096,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 2729
+    },
+    {
+      "epoch": 0.1831170137840829,
+      "grad_norm": 0.1407859004150815,
+      "learning_rate": 2e-05,
+      "loss": 5.3311,
+      "step": 2730
+    },
+    {
+      "epoch": 0.18318408961330784,
+      "grad_norm": 0.15245416632274317,
+      "learning_rate": 2e-05,
+      "loss": 5.5932,
+      "step": 2731
+    },
+    {
+      "epoch": 0.18325116544253278,
+      "grad_norm": 0.15435921761370103,
+      "learning_rate": 2e-05,
+      "loss": 5.3567,
+      "step": 2732
+    },
+    {
+      "epoch": 0.18331824127175772,
+      "grad_norm": 0.14207736295263176,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 2733
+    },
+    {
+      "epoch": 0.18338531710098266,
+      "grad_norm": 0.1541551064521531,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 2734
+    },
+    {
+      "epoch": 0.1834523929302076,
+      "grad_norm": 0.15438840326322978,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18351946875943254,
+      "grad_norm": 0.1508957888806164,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 2736
+    },
+    {
+      "epoch": 0.18358654458865747,
+      "grad_norm": 0.15056652095232892,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 2737
+    },
+    {
+      "epoch": 0.1836536204178824,
+      "grad_norm": 0.15683653732639452,
+      "learning_rate": 2e-05,
+      "loss": 5.4238,
+      "step": 2738
+    },
+    {
+      "epoch": 0.18372069624710735,
+      "grad_norm": 0.14136064398771256,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1837877720763323,
+      "grad_norm": 0.14530185931233602,
+      "learning_rate": 2e-05,
+      "loss": 5.4354,
+      "step": 2740
+    },
+    {
+      "epoch": 0.18385484790555723,
+      "grad_norm": 0.1426834525108686,
+      "learning_rate": 2e-05,
+      "loss": 5.4332,
+      "step": 2741
+    },
+    {
+      "epoch": 0.18392192373478217,
+      "grad_norm": 0.1408597167301315,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 2742
+    },
+    {
+      "epoch": 0.1839889995640071,
+      "grad_norm": 0.15610807574360927,
+      "learning_rate": 2e-05,
+      "loss": 5.6129,
+      "step": 2743
+    },
+    {
+      "epoch": 0.18405607539323204,
+      "grad_norm": 0.14602729654530727,
+      "learning_rate": 2e-05,
+      "loss": 5.438,
+      "step": 2744
+    },
+    {
+      "epoch": 0.18412315122245698,
+      "grad_norm": 0.162189654013063,
+      "learning_rate": 2e-05,
+      "loss": 5.5383,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18419022705168192,
+      "grad_norm": 0.14376361851029346,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 2746
+    },
+    {
+      "epoch": 0.18425730288090686,
+      "grad_norm": 0.14820936781411953,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1843243787101318,
+      "grad_norm": 0.1509668959826662,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 2748
+    },
+    {
+      "epoch": 0.18439145453935674,
+      "grad_norm": 0.14415063719527912,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 2749
+    },
+    {
+      "epoch": 0.18445853036858167,
+      "grad_norm": 0.15394337679983203,
+      "learning_rate": 2e-05,
+      "loss": 5.3641,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1845256061978066,
+      "grad_norm": 0.14943830999848118,
+      "learning_rate": 2e-05,
+      "loss": 5.4878,
+      "step": 2751
+    },
+    {
+      "epoch": 0.18459268202703155,
+      "grad_norm": 0.15306978949617306,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 2752
+    },
+    {
+      "epoch": 0.1846597578562565,
+      "grad_norm": 0.14796552718289854,
+      "learning_rate": 2e-05,
+      "loss": 5.3559,
+      "step": 2753
+    },
+    {
+      "epoch": 0.18472683368548143,
+      "grad_norm": 0.16068124344554538,
+      "learning_rate": 2e-05,
+      "loss": 5.5151,
+      "step": 2754
+    },
+    {
+      "epoch": 0.18479390951470637,
+      "grad_norm": 0.14478781246575143,
+      "learning_rate": 2e-05,
+      "loss": 5.4016,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1848609853439313,
+      "grad_norm": 0.14595903385641057,
+      "learning_rate": 2e-05,
+      "loss": 5.5139,
+      "step": 2756
+    },
+    {
+      "epoch": 0.18492806117315624,
+      "grad_norm": 0.14541072156839957,
+      "learning_rate": 2e-05,
+      "loss": 5.5161,
+      "step": 2757
+    },
+    {
+      "epoch": 0.18499513700238118,
+      "grad_norm": 0.1544261597403794,
+      "learning_rate": 2e-05,
+      "loss": 5.4788,
+      "step": 2758
+    },
+    {
+      "epoch": 0.18506221283160612,
+      "grad_norm": 0.14763974846137595,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 2759
+    },
+    {
+      "epoch": 0.18512928866083106,
+      "grad_norm": 0.14375436940507896,
+      "learning_rate": 2e-05,
+      "loss": 5.5288,
+      "step": 2760
+    },
+    {
+      "epoch": 0.185196364490056,
+      "grad_norm": 0.15286340799644232,
+      "learning_rate": 2e-05,
+      "loss": 5.408,
+      "step": 2761
+    },
+    {
+      "epoch": 0.18526344031928094,
+      "grad_norm": 0.14775311585695267,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 2762
+    },
+    {
+      "epoch": 0.18533051614850587,
+      "grad_norm": 0.14522994453796517,
+      "learning_rate": 2e-05,
+      "loss": 5.4787,
+      "step": 2763
+    },
+    {
+      "epoch": 0.1853975919777308,
+      "grad_norm": 0.15468972742722598,
+      "learning_rate": 2e-05,
+      "loss": 5.4885,
+      "step": 2764
+    },
+    {
+      "epoch": 0.18546466780695575,
+      "grad_norm": 0.14762409806535293,
+      "learning_rate": 2e-05,
+      "loss": 5.2785,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1855317436361807,
+      "grad_norm": 0.14602240294651064,
+      "learning_rate": 2e-05,
+      "loss": 5.4999,
+      "step": 2766
+    },
+    {
+      "epoch": 0.18559881946540563,
+      "grad_norm": 0.14988994705552325,
+      "learning_rate": 2e-05,
+      "loss": 5.3385,
+      "step": 2767
+    },
+    {
+      "epoch": 0.18566589529463057,
+      "grad_norm": 0.1417013672882784,
+      "learning_rate": 2e-05,
+      "loss": 5.4869,
+      "step": 2768
+    },
+    {
+      "epoch": 0.1857329711238555,
+      "grad_norm": 0.14465079861219646,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 2769
+    },
+    {
+      "epoch": 0.18580004695308044,
+      "grad_norm": 0.1483081013498772,
+      "learning_rate": 2e-05,
+      "loss": 5.4942,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18586712278230538,
+      "grad_norm": 0.15386824332441423,
+      "learning_rate": 2e-05,
+      "loss": 5.5363,
+      "step": 2771
+    },
+    {
+      "epoch": 0.18593419861153032,
+      "grad_norm": 0.140655535647225,
+      "learning_rate": 2e-05,
+      "loss": 5.5318,
+      "step": 2772
+    },
+    {
+      "epoch": 0.1860012744407553,
+      "grad_norm": 0.14525645416409158,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 2773
+    },
+    {
+      "epoch": 0.18606835026998023,
+      "grad_norm": 0.1453030410120523,
+      "learning_rate": 2e-05,
+      "loss": 5.3575,
+      "step": 2774
+    },
+    {
+      "epoch": 0.18613542609920516,
+      "grad_norm": 0.15139970626429225,
+      "learning_rate": 2e-05,
+      "loss": 5.5578,
+      "step": 2775
+    },
+    {
+      "epoch": 0.1862025019284301,
+      "grad_norm": 0.1380328181502607,
+      "learning_rate": 2e-05,
+      "loss": 5.3625,
+      "step": 2776
+    },
+    {
+      "epoch": 0.18626957775765504,
+      "grad_norm": 0.14222643249375677,
+      "learning_rate": 2e-05,
+      "loss": 5.4745,
+      "step": 2777
+    },
+    {
+      "epoch": 0.18633665358687998,
+      "grad_norm": 0.1518684449334853,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 2778
+    },
+    {
+      "epoch": 0.18640372941610492,
+      "grad_norm": 0.15455306309884742,
+      "learning_rate": 2e-05,
+      "loss": 5.3815,
+      "step": 2779
+    },
+    {
+      "epoch": 0.18647080524532986,
+      "grad_norm": 0.1368926311066186,
+      "learning_rate": 2e-05,
+      "loss": 5.487,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1865378810745548,
+      "grad_norm": 0.14749508507398076,
+      "learning_rate": 2e-05,
+      "loss": 5.6635,
+      "step": 2781
+    },
+    {
+      "epoch": 0.18660495690377973,
+      "grad_norm": 0.1554580748416227,
+      "learning_rate": 2e-05,
+      "loss": 5.444,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18667203273300467,
+      "grad_norm": 0.1445286448880499,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1867391085622296,
+      "grad_norm": 0.1570790952450586,
+      "learning_rate": 2e-05,
+      "loss": 5.549,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18680618439145455,
+      "grad_norm": 0.1406678087234311,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1868732602206795,
+      "grad_norm": 0.14130881816708843,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18694033604990443,
+      "grad_norm": 0.14912546607224975,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18700741187912936,
+      "grad_norm": 0.14221206355037808,
+      "learning_rate": 2e-05,
+      "loss": 5.5247,
+      "step": 2788
+    },
+    {
+      "epoch": 0.1870744877083543,
+      "grad_norm": 0.13620575288463382,
+      "learning_rate": 2e-05,
+      "loss": 5.3195,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18714156353757924,
+      "grad_norm": 0.13942055899930447,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18720863936680418,
+      "grad_norm": 0.13762686775756489,
+      "learning_rate": 2e-05,
+      "loss": 5.3106,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18727571519602912,
+      "grad_norm": 0.14104708237963642,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18734279102525406,
+      "grad_norm": 0.14387836108780838,
+      "learning_rate": 2e-05,
+      "loss": 5.4443,
+      "step": 2793
+    },
+    {
+      "epoch": 0.187409866854479,
+      "grad_norm": 0.14852322677390425,
+      "learning_rate": 2e-05,
+      "loss": 5.6148,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18747694268370393,
+      "grad_norm": 0.1373032761849103,
+      "learning_rate": 2e-05,
+      "loss": 5.4469,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18754401851292887,
+      "grad_norm": 0.14733484001729938,
+      "learning_rate": 2e-05,
+      "loss": 5.4145,
+      "step": 2796
+    },
+    {
+      "epoch": 0.1876110943421538,
+      "grad_norm": 0.1445911783593561,
+      "learning_rate": 2e-05,
+      "loss": 5.4494,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18767817017137875,
+      "grad_norm": 0.1504190609642747,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1877452460006037,
+      "grad_norm": 0.1426187510550691,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18781232182982863,
+      "grad_norm": 0.15500711248605933,
+      "learning_rate": 2e-05,
+      "loss": 5.4148,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18787939765905357,
+      "grad_norm": 0.1441784902913559,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1879464734882785,
+      "grad_norm": 0.13962365403417273,
+      "learning_rate": 2e-05,
+      "loss": 5.302,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18801354931750344,
+      "grad_norm": 0.14860217834896863,
+      "learning_rate": 2e-05,
+      "loss": 5.4326,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18808062514672838,
+      "grad_norm": 0.14716173046589545,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18814770097595332,
+      "grad_norm": 0.14649588594771532,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18821477680517826,
+      "grad_norm": 0.15367918569550446,
+      "learning_rate": 2e-05,
+      "loss": 5.4133,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1882818526344032,
+      "grad_norm": 0.15158986858518406,
+      "learning_rate": 2e-05,
+      "loss": 5.5573,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18834892846362813,
+      "grad_norm": 0.1431643320024629,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18841600429285307,
+      "grad_norm": 0.15153110795528937,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 2809
+    },
+    {
+      "epoch": 0.188483080122078,
+      "grad_norm": 0.14914957267185086,
+      "learning_rate": 2e-05,
+      "loss": 5.6928,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18855015595130295,
+      "grad_norm": 0.14804035514882688,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 2811
+    },
+    {
+      "epoch": 0.1886172317805279,
+      "grad_norm": 0.1446475806562322,
+      "learning_rate": 2e-05,
+      "loss": 5.559,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18868430760975283,
+      "grad_norm": 0.14098515559135927,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18875138343897777,
+      "grad_norm": 0.14644313606939166,
+      "learning_rate": 2e-05,
+      "loss": 5.5955,
+      "step": 2814
+    },
+    {
+      "epoch": 0.1888184592682027,
+      "grad_norm": 0.15042458276289053,
+      "learning_rate": 2e-05,
+      "loss": 5.5086,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18888553509742764,
+      "grad_norm": 0.14558034065585312,
+      "learning_rate": 2e-05,
+      "loss": 5.2867,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18895261092665258,
+      "grad_norm": 0.13490574351485027,
+      "learning_rate": 2e-05,
+      "loss": 5.4483,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18901968675587752,
+      "grad_norm": 0.14399322630548952,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18908676258510246,
+      "grad_norm": 0.14873050911469793,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 2819
+    },
+    {
+      "epoch": 0.1891538384143274,
+      "grad_norm": 0.14062363108697062,
+      "learning_rate": 2e-05,
+      "loss": 5.3979,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18922091424355234,
+      "grad_norm": 0.14625551336910175,
+      "learning_rate": 2e-05,
+      "loss": 5.5137,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18928799007277727,
+      "grad_norm": 0.14391127016380653,
+      "learning_rate": 2e-05,
+      "loss": 5.4028,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1893550659020022,
+      "grad_norm": 0.1460262753577528,
+      "learning_rate": 2e-05,
+      "loss": 5.5043,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18942214173122715,
+      "grad_norm": 0.1495346580748495,
+      "learning_rate": 2e-05,
+      "loss": 5.5205,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1894892175604521,
+      "grad_norm": 0.14365089700990966,
+      "learning_rate": 2e-05,
+      "loss": 5.4019,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18955629338967703,
+      "grad_norm": 0.1480401307073672,
+      "learning_rate": 2e-05,
+      "loss": 5.5275,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18962336921890197,
+      "grad_norm": 0.1432371967272493,
+      "learning_rate": 2e-05,
+      "loss": 5.2553,
+      "step": 2827
+    },
+    {
+      "epoch": 0.1896904450481269,
+      "grad_norm": 0.14682541935574384,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18975752087735184,
+      "grad_norm": 0.15127493187002242,
+      "learning_rate": 2e-05,
+      "loss": 5.5113,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18982459670657678,
+      "grad_norm": 0.14193655844692712,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18989167253580172,
+      "grad_norm": 0.149888673996713,
+      "learning_rate": 2e-05,
+      "loss": 5.4137,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18995874836502666,
+      "grad_norm": 0.15261384859298927,
+      "learning_rate": 2e-05,
+      "loss": 5.6057,
+      "step": 2832
+    },
+    {
+      "epoch": 0.1900258241942516,
+      "grad_norm": 0.14016270777498008,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 2833
+    },
+    {
+      "epoch": 0.19009290002347654,
+      "grad_norm": 0.13795216577659897,
+      "learning_rate": 2e-05,
+      "loss": 5.5151,
+      "step": 2834
+    },
+    {
+      "epoch": 0.19015997585270147,
+      "grad_norm": 0.14961518059548296,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 2835
+    },
+    {
+      "epoch": 0.1902270516819264,
+      "grad_norm": 0.1503754769406176,
+      "learning_rate": 2e-05,
+      "loss": 5.5119,
+      "step": 2836
+    },
+    {
+      "epoch": 0.19029412751115135,
+      "grad_norm": 0.1457004641868794,
+      "learning_rate": 2e-05,
+      "loss": 5.3622,
+      "step": 2837
+    },
+    {
+      "epoch": 0.1903612033403763,
+      "grad_norm": 0.1468702368491807,
+      "learning_rate": 2e-05,
+      "loss": 5.4572,
+      "step": 2838
+    },
+    {
+      "epoch": 0.19042827916960123,
+      "grad_norm": 0.14609720986799696,
+      "learning_rate": 2e-05,
+      "loss": 5.4388,
+      "step": 2839
+    },
+    {
+      "epoch": 0.19049535499882617,
+      "grad_norm": 0.14414027482299493,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1905624308280511,
+      "grad_norm": 0.14670241542858473,
+      "learning_rate": 2e-05,
+      "loss": 5.2781,
+      "step": 2841
+    },
+    {
+      "epoch": 0.19062950665727604,
+      "grad_norm": 0.14318237467451572,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 2842
+    },
+    {
+      "epoch": 0.19069658248650098,
+      "grad_norm": 0.15338943257226637,
+      "learning_rate": 2e-05,
+      "loss": 5.3478,
+      "step": 2843
+    },
+    {
+      "epoch": 0.19076365831572592,
+      "grad_norm": 0.14589096640450133,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 2844
+    },
+    {
+      "epoch": 0.19083073414495086,
+      "grad_norm": 0.13895507997822087,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1908978099741758,
+      "grad_norm": 0.14306723954204897,
+      "learning_rate": 2e-05,
+      "loss": 5.4905,
+      "step": 2846
+    },
+    {
+      "epoch": 0.19096488580340074,
+      "grad_norm": 0.14519262260760044,
+      "learning_rate": 2e-05,
+      "loss": 5.3107,
+      "step": 2847
+    },
+    {
+      "epoch": 0.19103196163262567,
+      "grad_norm": 0.1536035240889774,
+      "learning_rate": 2e-05,
+      "loss": 5.5027,
+      "step": 2848
+    },
+    {
+      "epoch": 0.1910990374618506,
+      "grad_norm": 0.1461179990146056,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 2849
+    },
+    {
+      "epoch": 0.19116611329107555,
+      "grad_norm": 0.1397937145348719,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1912331891203005,
+      "grad_norm": 0.16158993138502456,
+      "learning_rate": 2e-05,
+      "loss": 5.4313,
+      "step": 2851
+    },
+    {
+      "epoch": 0.19130026494952543,
+      "grad_norm": 0.15848794293139673,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 2852
+    },
+    {
+      "epoch": 0.19136734077875037,
+      "grad_norm": 0.14298786795795257,
+      "learning_rate": 2e-05,
+      "loss": 5.3269,
+      "step": 2853
+    },
+    {
+      "epoch": 0.1914344166079753,
+      "grad_norm": 0.15084784600941908,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 2854
+    },
+    {
+      "epoch": 0.19150149243720024,
+      "grad_norm": 0.1497944368921153,
+      "learning_rate": 2e-05,
+      "loss": 5.447,
+      "step": 2855
+    },
+    {
+      "epoch": 0.19156856826642518,
+      "grad_norm": 0.14554790478013255,
+      "learning_rate": 2e-05,
+      "loss": 5.6201,
+      "step": 2856
+    },
+    {
+      "epoch": 0.19163564409565012,
+      "grad_norm": 0.14519836170893075,
+      "learning_rate": 2e-05,
+      "loss": 5.34,
+      "step": 2857
+    },
+    {
+      "epoch": 0.19170271992487506,
+      "grad_norm": 0.1541332672491981,
+      "learning_rate": 2e-05,
+      "loss": 5.6442,
+      "step": 2858
+    },
+    {
+      "epoch": 0.1917697957541,
+      "grad_norm": 0.143269012402466,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 2859
+    },
+    {
+      "epoch": 0.19183687158332494,
+      "grad_norm": 0.1465888609781174,
+      "learning_rate": 2e-05,
+      "loss": 5.4873,
+      "step": 2860
+    },
+    {
+      "epoch": 0.19190394741254987,
+      "grad_norm": 0.14463491684166604,
+      "learning_rate": 2e-05,
+      "loss": 5.5835,
+      "step": 2861
+    },
+    {
+      "epoch": 0.1919710232417748,
+      "grad_norm": 0.15009420783234587,
+      "learning_rate": 2e-05,
+      "loss": 5.5894,
+      "step": 2862
+    },
+    {
+      "epoch": 0.19203809907099975,
+      "grad_norm": 0.1432231677263341,
+      "learning_rate": 2e-05,
+      "loss": 5.5369,
+      "step": 2863
+    },
+    {
+      "epoch": 0.19210517490022472,
+      "grad_norm": 0.14165033841451444,
+      "learning_rate": 2e-05,
+      "loss": 5.5392,
+      "step": 2864
+    },
+    {
+      "epoch": 0.19217225072944966,
+      "grad_norm": 0.14803025931130132,
+      "learning_rate": 2e-05,
+      "loss": 5.5598,
+      "step": 2865
+    },
+    {
+      "epoch": 0.1922393265586746,
+      "grad_norm": 0.15647560470016336,
+      "learning_rate": 2e-05,
+      "loss": 5.293,
+      "step": 2866
+    },
+    {
+      "epoch": 0.19230640238789953,
+      "grad_norm": 0.14825904475899848,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 2867
+    },
+    {
+      "epoch": 0.19237347821712447,
+      "grad_norm": 0.15001004503178583,
+      "learning_rate": 2e-05,
+      "loss": 5.2631,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1924405540463494,
+      "grad_norm": 0.1391735848468873,
+      "learning_rate": 2e-05,
+      "loss": 5.5774,
+      "step": 2869
+    },
+    {
+      "epoch": 0.19250762987557435,
+      "grad_norm": 0.14532375646203816,
+      "learning_rate": 2e-05,
+      "loss": 5.3736,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1925747057047993,
+      "grad_norm": 0.1468656291839792,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 2871
+    },
+    {
+      "epoch": 0.19264178153402423,
+      "grad_norm": 0.1416055489028985,
+      "learning_rate": 2e-05,
+      "loss": 5.4278,
+      "step": 2872
+    },
+    {
+      "epoch": 0.19270885736324916,
+      "grad_norm": 0.14627645099368602,
+      "learning_rate": 2e-05,
+      "loss": 5.3814,
+      "step": 2873
+    },
+    {
+      "epoch": 0.1927759331924741,
+      "grad_norm": 0.13947088066865052,
+      "learning_rate": 2e-05,
+      "loss": 5.4594,
+      "step": 2874
+    },
+    {
+      "epoch": 0.19284300902169904,
+      "grad_norm": 0.14330913600085518,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 2875
+    },
+    {
+      "epoch": 0.19291008485092398,
+      "grad_norm": 0.14235222976570247,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 2876
+    },
+    {
+      "epoch": 0.19297716068014892,
+      "grad_norm": 0.1402779631021663,
+      "learning_rate": 2e-05,
+      "loss": 5.553,
+      "step": 2877
+    },
+    {
+      "epoch": 0.19304423650937386,
+      "grad_norm": 0.16429993451400132,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 2878
+    },
+    {
+      "epoch": 0.1931113123385988,
+      "grad_norm": 0.14157038787220258,
+      "learning_rate": 2e-05,
+      "loss": 5.421,
+      "step": 2879
+    },
+    {
+      "epoch": 0.19317838816782373,
+      "grad_norm": 0.14648374932740207,
+      "learning_rate": 2e-05,
+      "loss": 5.3988,
+      "step": 2880
+    },
+    {
+      "epoch": 0.19324546399704867,
+      "grad_norm": 0.14154215326838476,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 2881
+    },
+    {
+      "epoch": 0.1933125398262736,
+      "grad_norm": 0.1424851202592815,
+      "learning_rate": 2e-05,
+      "loss": 5.366,
+      "step": 2882
+    },
+    {
+      "epoch": 0.19337961565549855,
+      "grad_norm": 0.14059870584448472,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 2883
+    },
+    {
+      "epoch": 0.1934466914847235,
+      "grad_norm": 0.14935950219158964,
+      "learning_rate": 2e-05,
+      "loss": 5.5553,
+      "step": 2884
+    },
+    {
+      "epoch": 0.19351376731394843,
+      "grad_norm": 0.14275499011538914,
+      "learning_rate": 2e-05,
+      "loss": 5.4438,
+      "step": 2885
+    },
+    {
+      "epoch": 0.19358084314317336,
+      "grad_norm": 0.14400243706243512,
+      "learning_rate": 2e-05,
+      "loss": 5.5797,
+      "step": 2886
+    },
+    {
+      "epoch": 0.1936479189723983,
+      "grad_norm": 0.14670462175667487,
+      "learning_rate": 2e-05,
+      "loss": 5.6267,
+      "step": 2887
+    },
+    {
+      "epoch": 0.19371499480162324,
+      "grad_norm": 0.1454304716232998,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 2888
+    },
+    {
+      "epoch": 0.19378207063084818,
+      "grad_norm": 0.1409886488750574,
+      "learning_rate": 2e-05,
+      "loss": 5.3999,
+      "step": 2889
+    },
+    {
+      "epoch": 0.19384914646007312,
+      "grad_norm": 0.14510571418571616,
+      "learning_rate": 2e-05,
+      "loss": 5.3707,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19391622228929806,
+      "grad_norm": 0.15303720996987785,
+      "learning_rate": 2e-05,
+      "loss": 5.5571,
+      "step": 2891
+    },
+    {
+      "epoch": 0.193983298118523,
+      "grad_norm": 0.15239792851986214,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 2892
+    },
+    {
+      "epoch": 0.19405037394774793,
+      "grad_norm": 0.14084363883606532,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 2893
+    },
+    {
+      "epoch": 0.19411744977697287,
+      "grad_norm": 0.14285824642692535,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 2894
+    },
+    {
+      "epoch": 0.1941845256061978,
+      "grad_norm": 0.1516624702427743,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19425160143542275,
+      "grad_norm": 0.14257036787884608,
+      "learning_rate": 2e-05,
+      "loss": 5.4641,
+      "step": 2896
+    },
+    {
+      "epoch": 0.1943186772646477,
+      "grad_norm": 0.14697329044831953,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 2897
+    },
+    {
+      "epoch": 0.19438575309387263,
+      "grad_norm": 0.1507798655710949,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 2898
+    },
+    {
+      "epoch": 0.19445282892309756,
+      "grad_norm": 0.14839607566750623,
+      "learning_rate": 2e-05,
+      "loss": 5.5276,
+      "step": 2899
+    },
+    {
+      "epoch": 0.1945199047523225,
+      "grad_norm": 0.14504630689286738,
+      "learning_rate": 2e-05,
+      "loss": 5.3915,
+      "step": 2900
+    },
+    {
+      "epoch": 0.19458698058154744,
+      "grad_norm": 0.1450465230432479,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 2901
+    },
+    {
+      "epoch": 0.19465405641077238,
+      "grad_norm": 0.147474843975319,
+      "learning_rate": 2e-05,
+      "loss": 5.5343,
+      "step": 2902
+    },
+    {
+      "epoch": 0.19472113223999732,
+      "grad_norm": 0.14323788470312038,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 2903
+    },
+    {
+      "epoch": 0.19478820806922226,
+      "grad_norm": 0.1416027150700193,
+      "learning_rate": 2e-05,
+      "loss": 5.5662,
+      "step": 2904
+    },
+    {
+      "epoch": 0.1948552838984472,
+      "grad_norm": 0.1469141777885197,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19492235972767213,
+      "grad_norm": 0.1393415885910724,
+      "learning_rate": 2e-05,
+      "loss": 5.5315,
+      "step": 2906
+    },
+    {
+      "epoch": 0.19498943555689707,
+      "grad_norm": 0.13803519236915313,
+      "learning_rate": 2e-05,
+      "loss": 5.3177,
+      "step": 2907
+    },
+    {
+      "epoch": 0.195056511386122,
+      "grad_norm": 0.15044764516446688,
+      "learning_rate": 2e-05,
+      "loss": 5.3923,
+      "step": 2908
+    },
+    {
+      "epoch": 0.19512358721534695,
+      "grad_norm": 0.1465902076227777,
+      "learning_rate": 2e-05,
+      "loss": 5.5504,
+      "step": 2909
+    },
+    {
+      "epoch": 0.1951906630445719,
+      "grad_norm": 0.14476497371196184,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19525773887379683,
+      "grad_norm": 0.14067391054563042,
+      "learning_rate": 2e-05,
+      "loss": 5.551,
+      "step": 2911
+    },
+    {
+      "epoch": 0.19532481470302177,
+      "grad_norm": 0.14310475035001582,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 2912
+    },
+    {
+      "epoch": 0.1953918905322467,
+      "grad_norm": 0.14137499452177452,
+      "learning_rate": 2e-05,
+      "loss": 5.5288,
+      "step": 2913
+    },
+    {
+      "epoch": 0.19545896636147164,
+      "grad_norm": 0.1422984321620052,
+      "learning_rate": 2e-05,
+      "loss": 5.3457,
+      "step": 2914
+    },
+    {
+      "epoch": 0.19552604219069658,
+      "grad_norm": 0.14875434772696083,
+      "learning_rate": 2e-05,
+      "loss": 5.5561,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19559311801992152,
+      "grad_norm": 0.1494127224488448,
+      "learning_rate": 2e-05,
+      "loss": 5.5104,
+      "step": 2916
+    },
+    {
+      "epoch": 0.19566019384914646,
+      "grad_norm": 0.15162618474343506,
+      "learning_rate": 2e-05,
+      "loss": 5.4282,
+      "step": 2917
+    },
+    {
+      "epoch": 0.1957272696783714,
+      "grad_norm": 0.14019621118932557,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 2918
+    },
+    {
+      "epoch": 0.19579434550759633,
+      "grad_norm": 0.14691067164752378,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 2919
+    },
+    {
+      "epoch": 0.19586142133682127,
+      "grad_norm": 0.14490590862901187,
+      "learning_rate": 2e-05,
+      "loss": 5.4807,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1959284971660462,
+      "grad_norm": 0.14644723159612702,
+      "learning_rate": 2e-05,
+      "loss": 5.6153,
+      "step": 2921
+    },
+    {
+      "epoch": 0.19599557299527115,
+      "grad_norm": 0.14171362047899058,
+      "learning_rate": 2e-05,
+      "loss": 5.3515,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1960626488244961,
+      "grad_norm": 0.15254886209315727,
+      "learning_rate": 2e-05,
+      "loss": 5.3314,
+      "step": 2923
+    },
+    {
+      "epoch": 0.19612972465372103,
+      "grad_norm": 0.14397449922930072,
+      "learning_rate": 2e-05,
+      "loss": 5.4136,
+      "step": 2924
+    },
+    {
+      "epoch": 0.19619680048294597,
+      "grad_norm": 0.14217260222584494,
+      "learning_rate": 2e-05,
+      "loss": 5.3712,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1962638763121709,
+      "grad_norm": 0.14396192623322784,
+      "learning_rate": 2e-05,
+      "loss": 5.3989,
+      "step": 2926
+    },
+    {
+      "epoch": 0.19633095214139584,
+      "grad_norm": 0.14798556040184313,
+      "learning_rate": 2e-05,
+      "loss": 5.4361,
+      "step": 2927
+    },
+    {
+      "epoch": 0.19639802797062078,
+      "grad_norm": 0.1596110501625947,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 2928
+    },
+    {
+      "epoch": 0.19646510379984572,
+      "grad_norm": 0.13849286762911275,
+      "learning_rate": 2e-05,
+      "loss": 5.466,
+      "step": 2929
+    },
+    {
+      "epoch": 0.19653217962907066,
+      "grad_norm": 0.14550873180838436,
+      "learning_rate": 2e-05,
+      "loss": 5.2974,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1965992554582956,
+      "grad_norm": 0.1452155635316248,
+      "learning_rate": 2e-05,
+      "loss": 5.3564,
+      "step": 2931
+    },
+    {
+      "epoch": 0.19666633128752053,
+      "grad_norm": 0.14506895506157305,
+      "learning_rate": 2e-05,
+      "loss": 5.5959,
+      "step": 2932
+    },
+    {
+      "epoch": 0.19673340711674547,
+      "grad_norm": 0.14291070006779216,
+      "learning_rate": 2e-05,
+      "loss": 5.4041,
+      "step": 2933
+    },
+    {
+      "epoch": 0.1968004829459704,
+      "grad_norm": 0.1496124887368297,
+      "learning_rate": 2e-05,
+      "loss": 5.3981,
+      "step": 2934
+    },
+    {
+      "epoch": 0.19686755877519535,
+      "grad_norm": 0.1477471892679978,
+      "learning_rate": 2e-05,
+      "loss": 5.5308,
+      "step": 2935
+    },
+    {
+      "epoch": 0.1969346346044203,
+      "grad_norm": 0.1454020489397554,
+      "learning_rate": 2e-05,
+      "loss": 5.3662,
+      "step": 2936
+    },
+    {
+      "epoch": 0.19700171043364523,
+      "grad_norm": 0.1440655944722557,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 2937
+    },
+    {
+      "epoch": 0.19706878626287017,
+      "grad_norm": 0.14048519518114946,
+      "learning_rate": 2e-05,
+      "loss": 5.2793,
+      "step": 2938
+    },
+    {
+      "epoch": 0.1971358620920951,
+      "grad_norm": 0.147431055952558,
+      "learning_rate": 2e-05,
+      "loss": 5.2396,
+      "step": 2939
+    },
+    {
+      "epoch": 0.19720293792132004,
+      "grad_norm": 0.14135486565868316,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19727001375054498,
+      "grad_norm": 0.1457632065633095,
+      "learning_rate": 2e-05,
+      "loss": 5.4476,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19733708957976992,
+      "grad_norm": 0.14423474478854234,
+      "learning_rate": 2e-05,
+      "loss": 5.3503,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19740416540899486,
+      "grad_norm": 0.13646219392576245,
+      "learning_rate": 2e-05,
+      "loss": 5.3473,
+      "step": 2943
+    },
+    {
+      "epoch": 0.1974712412382198,
+      "grad_norm": 0.1452072373436208,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19753831706744474,
+      "grad_norm": 0.13886950230872394,
+      "learning_rate": 2e-05,
+      "loss": 5.53,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19760539289666967,
+      "grad_norm": 0.15021817086696096,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 2946
+    },
+    {
+      "epoch": 0.1976724687258946,
+      "grad_norm": 0.14358627487370337,
+      "learning_rate": 2e-05,
+      "loss": 5.5446,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19773954455511955,
+      "grad_norm": 0.14006671558069284,
+      "learning_rate": 2e-05,
+      "loss": 5.3814,
+      "step": 2948
+    },
+    {
+      "epoch": 0.1978066203843445,
+      "grad_norm": 0.14617541309244586,
+      "learning_rate": 2e-05,
+      "loss": 5.3645,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19787369621356943,
+      "grad_norm": 0.1408974284803663,
+      "learning_rate": 2e-05,
+      "loss": 5.4826,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19794077204279437,
+      "grad_norm": 0.14432837367587809,
+      "learning_rate": 2e-05,
+      "loss": 5.4906,
+      "step": 2951
+    },
+    {
+      "epoch": 0.1980078478720193,
+      "grad_norm": 0.1419883461572055,
+      "learning_rate": 2e-05,
+      "loss": 5.4803,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19807492370124424,
+      "grad_norm": 0.13786954744113436,
+      "learning_rate": 2e-05,
+      "loss": 5.4306,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19814199953046918,
+      "grad_norm": 0.14977241792447413,
+      "learning_rate": 2e-05,
+      "loss": 5.3918,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19820907535969415,
+      "grad_norm": 0.1488793118886182,
+      "learning_rate": 2e-05,
+      "loss": 5.4931,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1982761511889191,
+      "grad_norm": 0.15066805631126257,
+      "learning_rate": 2e-05,
+      "loss": 5.3643,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19834322701814402,
+      "grad_norm": 0.14039294609730038,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19841030284736896,
+      "grad_norm": 0.1491944935638163,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1984773786765939,
+      "grad_norm": 0.1385960263272716,
+      "learning_rate": 2e-05,
+      "loss": 5.4046,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19854445450581884,
+      "grad_norm": 0.14011673784966122,
+      "learning_rate": 2e-05,
+      "loss": 5.5052,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19861153033504378,
+      "grad_norm": 0.14373254507816535,
+      "learning_rate": 2e-05,
+      "loss": 5.5055,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19867860616426872,
+      "grad_norm": 0.14907275217826996,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19874568199349366,
+      "grad_norm": 0.14554415970519327,
+      "learning_rate": 2e-05,
+      "loss": 5.4974,
+      "step": 2963
+    },
+    {
+      "epoch": 0.1988127578227186,
+      "grad_norm": 0.1412335780950371,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19887983365194353,
+      "grad_norm": 0.1453733363268682,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19894690948116847,
+      "grad_norm": 0.150013769308575,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1990139853103934,
+      "grad_norm": 0.1447414116162861,
+      "learning_rate": 2e-05,
+      "loss": 5.439,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19908106113961835,
+      "grad_norm": 0.14440229339971944,
+      "learning_rate": 2e-05,
+      "loss": 5.3161,
+      "step": 2968
+    },
+    {
+      "epoch": 0.1991481369688433,
+      "grad_norm": 0.14168709151750497,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19921521279806823,
+      "grad_norm": 0.14581846115145042,
+      "learning_rate": 2e-05,
+      "loss": 5.508,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19928228862729316,
+      "grad_norm": 0.1553515211920712,
+      "learning_rate": 2e-05,
+      "loss": 5.4904,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1993493644565181,
+      "grad_norm": 0.14930143558295958,
+      "learning_rate": 2e-05,
+      "loss": 5.4535,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19941644028574304,
+      "grad_norm": 0.14424912912144242,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19948351611496798,
+      "grad_norm": 0.14928587791927395,
+      "learning_rate": 2e-05,
+      "loss": 5.4346,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19955059194419292,
+      "grad_norm": 0.1501448890260835,
+      "learning_rate": 2e-05,
+      "loss": 5.549,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19961766777341786,
+      "grad_norm": 0.15285281060663317,
+      "learning_rate": 2e-05,
+      "loss": 5.4986,
+      "step": 2976
+    },
+    {
+      "epoch": 0.1996847436026428,
+      "grad_norm": 0.14894586563965428,
+      "learning_rate": 2e-05,
+      "loss": 5.4183,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19975181943186773,
+      "grad_norm": 0.14506213434268828,
+      "learning_rate": 2e-05,
+      "loss": 5.591,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19981889526109267,
+      "grad_norm": 0.1398618418000234,
+      "learning_rate": 2e-05,
+      "loss": 5.5728,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1998859710903176,
+      "grad_norm": 0.14548107968677967,
+      "learning_rate": 2e-05,
+      "loss": 5.361,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19995304691954255,
+      "grad_norm": 0.1527923048202449,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2000201227487675,
+      "grad_norm": 0.13865889975026682,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 2982
+    },
+    {
+      "epoch": 0.20008719857799243,
+      "grad_norm": 0.14406150591607975,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 2983
+    },
+    {
+      "epoch": 0.20015427440721736,
+      "grad_norm": 0.15273271941215824,
+      "learning_rate": 2e-05,
+      "loss": 5.4474,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2002213502364423,
+      "grad_norm": 0.15415823641395715,
+      "learning_rate": 2e-05,
+      "loss": 5.6198,
+      "step": 2985
+    },
+    {
+      "epoch": 0.20028842606566724,
+      "grad_norm": 0.1425073688530836,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 2986
+    },
+    {
+      "epoch": 0.20035550189489218,
+      "grad_norm": 0.14155754497040868,
+      "learning_rate": 2e-05,
+      "loss": 5.4924,
+      "step": 2987
+    },
+    {
+      "epoch": 0.20042257772411712,
+      "grad_norm": 0.1514580387316104,
+      "learning_rate": 2e-05,
+      "loss": 5.4208,
+      "step": 2988
+    },
+    {
+      "epoch": 0.20048965355334206,
+      "grad_norm": 0.1579241935874637,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 2989
+    },
+    {
+      "epoch": 0.200556729382567,
+      "grad_norm": 0.15378931524097514,
+      "learning_rate": 2e-05,
+      "loss": 5.5408,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20062380521179193,
+      "grad_norm": 0.15092259684605497,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 2991
+    },
+    {
+      "epoch": 0.20069088104101687,
+      "grad_norm": 0.14950536600708964,
+      "learning_rate": 2e-05,
+      "loss": 5.5204,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2007579568702418,
+      "grad_norm": 0.15792384311486418,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 2993
+    },
+    {
+      "epoch": 0.20082503269946675,
+      "grad_norm": 0.1503961952033972,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2008921085286917,
+      "grad_norm": 0.1470019024708196,
+      "learning_rate": 2e-05,
+      "loss": 5.4011,
+      "step": 2995
+    },
+    {
+      "epoch": 0.20095918435791663,
+      "grad_norm": 0.15425820093858025,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 2996
+    },
+    {
+      "epoch": 0.20102626018714156,
+      "grad_norm": 0.15902190633585256,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2010933360163665,
+      "grad_norm": 0.1491845283942766,
+      "learning_rate": 2e-05,
+      "loss": 5.5094,
+      "step": 2998
+    },
+    {
+      "epoch": 0.20116041184559144,
+      "grad_norm": 0.1449410759603749,
+      "learning_rate": 2e-05,
+      "loss": 5.4203,
+      "step": 2999
+    },
+    {
+      "epoch": 0.20122748767481638,
+      "grad_norm": 0.15198180333884964,
+      "learning_rate": 2e-05,
+      "loss": 5.3732,
+      "step": 3000
+    },
+    {
+      "epoch": 0.20129456350404132,
+      "grad_norm": 0.14307529685683662,
+      "learning_rate": 2e-05,
+      "loss": 5.3358,
+      "step": 3001
+    },
+    {
+      "epoch": 0.20136163933326626,
+      "grad_norm": 0.14712041125910671,
+      "learning_rate": 2e-05,
+      "loss": 5.4459,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2014287151624912,
+      "grad_norm": 0.14934833205345135,
+      "learning_rate": 2e-05,
+      "loss": 5.3753,
+      "step": 3003
+    },
+    {
+      "epoch": 0.20149579099171613,
+      "grad_norm": 0.14952202147035834,
+      "learning_rate": 2e-05,
+      "loss": 5.6429,
+      "step": 3004
+    },
+    {
+      "epoch": 0.20156286682094107,
+      "grad_norm": 0.15032636294942786,
+      "learning_rate": 2e-05,
+      "loss": 5.3222,
+      "step": 3005
+    },
+    {
+      "epoch": 0.201629942650166,
+      "grad_norm": 0.14471347402146728,
+      "learning_rate": 2e-05,
+      "loss": 5.5036,
+      "step": 3006
+    },
+    {
+      "epoch": 0.20169701847939095,
+      "grad_norm": 0.14708795060029764,
+      "learning_rate": 2e-05,
+      "loss": 5.6083,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2017640943086159,
+      "grad_norm": 0.1547541836407129,
+      "learning_rate": 2e-05,
+      "loss": 5.3391,
+      "step": 3008
+    },
+    {
+      "epoch": 0.20183117013784083,
+      "grad_norm": 0.1439416517725702,
+      "learning_rate": 2e-05,
+      "loss": 5.49,
+      "step": 3009
+    },
+    {
+      "epoch": 0.20189824596706576,
+      "grad_norm": 0.14598872997078965,
+      "learning_rate": 2e-05,
+      "loss": 5.5713,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2019653217962907,
+      "grad_norm": 0.1406840286073035,
+      "learning_rate": 2e-05,
+      "loss": 5.5344,
+      "step": 3011
+    },
+    {
+      "epoch": 0.20203239762551564,
+      "grad_norm": 0.14309123967604118,
+      "learning_rate": 2e-05,
+      "loss": 5.4291,
+      "step": 3012
+    },
+    {
+      "epoch": 0.20209947345474058,
+      "grad_norm": 0.14753820072330392,
+      "learning_rate": 2e-05,
+      "loss": 5.3468,
+      "step": 3013
+    },
+    {
+      "epoch": 0.20216654928396552,
+      "grad_norm": 0.14611787469283935,
+      "learning_rate": 2e-05,
+      "loss": 5.4188,
+      "step": 3014
+    },
+    {
+      "epoch": 0.20223362511319046,
+      "grad_norm": 0.14751560672488104,
+      "learning_rate": 2e-05,
+      "loss": 5.4619,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2023007009424154,
+      "grad_norm": 0.1516185010812857,
+      "learning_rate": 2e-05,
+      "loss": 5.5085,
+      "step": 3016
+    },
+    {
+      "epoch": 0.20236777677164033,
+      "grad_norm": 0.1406281677341113,
+      "learning_rate": 2e-05,
+      "loss": 5.3114,
+      "step": 3017
+    },
+    {
+      "epoch": 0.20243485260086527,
+      "grad_norm": 0.1447394514764819,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2025019284300902,
+      "grad_norm": 0.15826957399022767,
+      "learning_rate": 2e-05,
+      "loss": 5.3987,
+      "step": 3019
+    },
+    {
+      "epoch": 0.20256900425931515,
+      "grad_norm": 0.14759104499065448,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2026360800885401,
+      "grad_norm": 0.1463592915398451,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 3021
+    },
+    {
+      "epoch": 0.20270315591776503,
+      "grad_norm": 0.15318663395955756,
+      "learning_rate": 2e-05,
+      "loss": 5.4005,
+      "step": 3022
+    },
+    {
+      "epoch": 0.20277023174698997,
+      "grad_norm": 0.15285848570593338,
+      "learning_rate": 2e-05,
+      "loss": 5.602,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2028373075762149,
+      "grad_norm": 0.1444383521060343,
+      "learning_rate": 2e-05,
+      "loss": 5.4171,
+      "step": 3024
+    },
+    {
+      "epoch": 0.20290438340543984,
+      "grad_norm": 0.15722692752490042,
+      "learning_rate": 2e-05,
+      "loss": 5.6045,
+      "step": 3025
+    },
+    {
+      "epoch": 0.20297145923466478,
+      "grad_norm": 0.14455820122339122,
+      "learning_rate": 2e-05,
+      "loss": 5.4946,
+      "step": 3026
+    },
+    {
+      "epoch": 0.20303853506388972,
+      "grad_norm": 0.1378644971484586,
+      "learning_rate": 2e-05,
+      "loss": 5.4347,
+      "step": 3027
+    },
+    {
+      "epoch": 0.20310561089311466,
+      "grad_norm": 0.15561304764081713,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2031726867223396,
+      "grad_norm": 0.1550419748531997,
+      "learning_rate": 2e-05,
+      "loss": 5.5441,
+      "step": 3029
+    },
+    {
+      "epoch": 0.20323976255156453,
+      "grad_norm": 0.1469918152921543,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 3030
+    },
+    {
+      "epoch": 0.20330683838078947,
+      "grad_norm": 0.14925571325921638,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2033739142100144,
+      "grad_norm": 0.15580832083467977,
+      "learning_rate": 2e-05,
+      "loss": 5.5111,
+      "step": 3032
+    },
+    {
+      "epoch": 0.20344099003923935,
+      "grad_norm": 0.1469633740790967,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 3033
+    },
+    {
+      "epoch": 0.2035080658684643,
+      "grad_norm": 0.14541796164468618,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 3034
+    },
+    {
+      "epoch": 0.20357514169768923,
+      "grad_norm": 0.1457480369224864,
+      "learning_rate": 2e-05,
+      "loss": 5.4848,
+      "step": 3035
+    },
+    {
+      "epoch": 0.20364221752691417,
+      "grad_norm": 0.15139681659143542,
+      "learning_rate": 2e-05,
+      "loss": 5.3304,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2037092933561391,
+      "grad_norm": 0.14609842605462234,
+      "learning_rate": 2e-05,
+      "loss": 5.348,
+      "step": 3037
+    },
+    {
+      "epoch": 0.20377636918536404,
+      "grad_norm": 0.1491384865468703,
+      "learning_rate": 2e-05,
+      "loss": 5.5582,
+      "step": 3038
+    },
+    {
+      "epoch": 0.20384344501458898,
+      "grad_norm": 0.14477399941425576,
+      "learning_rate": 2e-05,
+      "loss": 5.4143,
+      "step": 3039
+    },
+    {
+      "epoch": 0.20391052084381392,
+      "grad_norm": 0.14526841609338192,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 3040
+    },
+    {
+      "epoch": 0.20397759667303886,
+      "grad_norm": 0.14665081328067248,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2040446725022638,
+      "grad_norm": 0.14088013445111372,
+      "learning_rate": 2e-05,
+      "loss": 5.4019,
+      "step": 3042
+    },
+    {
+      "epoch": 0.20411174833148873,
+      "grad_norm": 0.14067080565106252,
+      "learning_rate": 2e-05,
+      "loss": 5.461,
+      "step": 3043
+    },
+    {
+      "epoch": 0.20417882416071367,
+      "grad_norm": 0.14203718642181404,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2042458999899386,
+      "grad_norm": 0.1443076929440761,
+      "learning_rate": 2e-05,
+      "loss": 5.4189,
+      "step": 3045
+    },
+    {
+      "epoch": 0.20431297581916358,
+      "grad_norm": 0.15702968043159682,
+      "learning_rate": 2e-05,
+      "loss": 5.524,
+      "step": 3046
+    },
+    {
+      "epoch": 0.20438005164838852,
+      "grad_norm": 0.15510329913132856,
+      "learning_rate": 2e-05,
+      "loss": 5.5507,
+      "step": 3047
+    },
+    {
+      "epoch": 0.20444712747761346,
+      "grad_norm": 0.15240369296619588,
+      "learning_rate": 2e-05,
+      "loss": 5.5217,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2045142033068384,
+      "grad_norm": 0.14794097646818508,
+      "learning_rate": 2e-05,
+      "loss": 5.4078,
+      "step": 3049
+    },
+    {
+      "epoch": 0.20458127913606333,
+      "grad_norm": 0.14471544849454027,
+      "learning_rate": 2e-05,
+      "loss": 5.4842,
+      "step": 3050
+    },
+    {
+      "epoch": 0.20464835496528827,
+      "grad_norm": 0.1477870596710543,
+      "learning_rate": 2e-05,
+      "loss": 5.3182,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2047154307945132,
+      "grad_norm": 0.1498538820148795,
+      "learning_rate": 2e-05,
+      "loss": 5.518,
+      "step": 3052
+    },
+    {
+      "epoch": 0.20478250662373815,
+      "grad_norm": 0.14533719782339297,
+      "learning_rate": 2e-05,
+      "loss": 5.6222,
+      "step": 3053
+    },
+    {
+      "epoch": 0.20484958245296309,
+      "grad_norm": 0.14485387796394072,
+      "learning_rate": 2e-05,
+      "loss": 5.4144,
+      "step": 3054
+    },
+    {
+      "epoch": 0.20491665828218802,
+      "grad_norm": 0.14890287660451448,
+      "learning_rate": 2e-05,
+      "loss": 5.5016,
+      "step": 3055
+    },
+    {
+      "epoch": 0.20498373411141296,
+      "grad_norm": 0.15389066140015165,
+      "learning_rate": 2e-05,
+      "loss": 5.3707,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2050508099406379,
+      "grad_norm": 0.13853450760140834,
+      "learning_rate": 2e-05,
+      "loss": 5.4202,
+      "step": 3057
+    },
+    {
+      "epoch": 0.20511788576986284,
+      "grad_norm": 0.14586949257984608,
+      "learning_rate": 2e-05,
+      "loss": 5.5217,
+      "step": 3058
+    },
+    {
+      "epoch": 0.20518496159908778,
+      "grad_norm": 0.1447376551167985,
+      "learning_rate": 2e-05,
+      "loss": 5.467,
+      "step": 3059
+    },
+    {
+      "epoch": 0.20525203742831272,
+      "grad_norm": 0.14572140049185617,
+      "learning_rate": 2e-05,
+      "loss": 5.499,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20531911325753766,
+      "grad_norm": 0.1411607688031229,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2053861890867626,
+      "grad_norm": 0.15111612573315109,
+      "learning_rate": 2e-05,
+      "loss": 5.4026,
+      "step": 3062
+    },
+    {
+      "epoch": 0.20545326491598753,
+      "grad_norm": 0.1530701160224052,
+      "learning_rate": 2e-05,
+      "loss": 5.6171,
+      "step": 3063
+    },
+    {
+      "epoch": 0.20552034074521247,
+      "grad_norm": 0.14180896945216864,
+      "learning_rate": 2e-05,
+      "loss": 5.5197,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2055874165744374,
+      "grad_norm": 0.150533677399395,
+      "learning_rate": 2e-05,
+      "loss": 5.6637,
+      "step": 3065
+    },
+    {
+      "epoch": 0.20565449240366235,
+      "grad_norm": 0.16139257674162552,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2057215682328873,
+      "grad_norm": 0.14819002880831822,
+      "learning_rate": 2e-05,
+      "loss": 5.3762,
+      "step": 3067
+    },
+    {
+      "epoch": 0.20578864406211222,
+      "grad_norm": 0.14312912764192268,
+      "learning_rate": 2e-05,
+      "loss": 5.5971,
+      "step": 3068
+    },
+    {
+      "epoch": 0.20585571989133716,
+      "grad_norm": 0.15011488469917977,
+      "learning_rate": 2e-05,
+      "loss": 5.5412,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2059227957205621,
+      "grad_norm": 0.1522955329626157,
+      "learning_rate": 2e-05,
+      "loss": 5.4292,
+      "step": 3070
+    },
+    {
+      "epoch": 0.20598987154978704,
+      "grad_norm": 0.13825495118739617,
+      "learning_rate": 2e-05,
+      "loss": 5.4571,
+      "step": 3071
+    },
+    {
+      "epoch": 0.20605694737901198,
+      "grad_norm": 0.1515500318919278,
+      "learning_rate": 2e-05,
+      "loss": 5.5153,
+      "step": 3072
+    },
+    {
+      "epoch": 0.20612402320823692,
+      "grad_norm": 0.1512379040154233,
+      "learning_rate": 2e-05,
+      "loss": 5.4477,
+      "step": 3073
+    },
+    {
+      "epoch": 0.20619109903746186,
+      "grad_norm": 0.14639059241416527,
+      "learning_rate": 2e-05,
+      "loss": 5.4339,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2062581748666868,
+      "grad_norm": 0.14300149376879287,
+      "learning_rate": 2e-05,
+      "loss": 5.3653,
+      "step": 3075
+    },
+    {
+      "epoch": 0.20632525069591173,
+      "grad_norm": 0.13824641512010052,
+      "learning_rate": 2e-05,
+      "loss": 5.5345,
+      "step": 3076
+    },
+    {
+      "epoch": 0.20639232652513667,
+      "grad_norm": 0.14349130958336473,
+      "learning_rate": 2e-05,
+      "loss": 5.4653,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2064594023543616,
+      "grad_norm": 0.14720363780776052,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 3078
+    },
+    {
+      "epoch": 0.20652647818358655,
+      "grad_norm": 0.14495890879856804,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2065935540128115,
+      "grad_norm": 0.14999245671107905,
+      "learning_rate": 2e-05,
+      "loss": 5.4082,
+      "step": 3080
+    },
+    {
+      "epoch": 0.20666062984203643,
+      "grad_norm": 0.1423471871868468,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 3081
+    },
+    {
+      "epoch": 0.20672770567126136,
+      "grad_norm": 0.14458321749817846,
+      "learning_rate": 2e-05,
+      "loss": 5.3577,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2067947815004863,
+      "grad_norm": 0.1487991137110896,
+      "learning_rate": 2e-05,
+      "loss": 5.504,
+      "step": 3083
+    },
+    {
+      "epoch": 0.20686185732971124,
+      "grad_norm": 0.1489922212059875,
+      "learning_rate": 2e-05,
+      "loss": 5.5994,
+      "step": 3084
+    },
+    {
+      "epoch": 0.20692893315893618,
+      "grad_norm": 0.14316286435275205,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 3085
+    },
+    {
+      "epoch": 0.20699600898816112,
+      "grad_norm": 0.1437785535676628,
+      "learning_rate": 2e-05,
+      "loss": 5.5321,
+      "step": 3086
+    },
+    {
+      "epoch": 0.20706308481738606,
+      "grad_norm": 0.14419866075194698,
+      "learning_rate": 2e-05,
+      "loss": 5.4459,
+      "step": 3087
+    },
+    {
+      "epoch": 0.207130160646611,
+      "grad_norm": 0.14433626923338122,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 3088
+    },
+    {
+      "epoch": 0.20719723647583593,
+      "grad_norm": 0.14101152975084455,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 3089
+    },
+    {
+      "epoch": 0.20726431230506087,
+      "grad_norm": 0.14785642561144374,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2073313881342858,
+      "grad_norm": 0.1485231640271306,
+      "learning_rate": 2e-05,
+      "loss": 5.4409,
+      "step": 3091
+    },
+    {
+      "epoch": 0.20739846396351075,
+      "grad_norm": 0.1427861150447786,
+      "learning_rate": 2e-05,
+      "loss": 5.3397,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2074655397927357,
+      "grad_norm": 0.1455448676900824,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 3093
+    },
+    {
+      "epoch": 0.20753261562196063,
+      "grad_norm": 0.14138365405725,
+      "learning_rate": 2e-05,
+      "loss": 5.5633,
+      "step": 3094
+    },
+    {
+      "epoch": 0.20759969145118556,
+      "grad_norm": 0.14179023979294056,
+      "learning_rate": 2e-05,
+      "loss": 5.4528,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2076667672804105,
+      "grad_norm": 0.1438254618429699,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 3096
+    },
+    {
+      "epoch": 0.20773384310963544,
+      "grad_norm": 0.1423015275787241,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 3097
+    },
+    {
+      "epoch": 0.20780091893886038,
+      "grad_norm": 0.14643741836625412,
+      "learning_rate": 2e-05,
+      "loss": 5.4445,
+      "step": 3098
+    },
+    {
+      "epoch": 0.20786799476808532,
+      "grad_norm": 0.14535342766017958,
+      "learning_rate": 2e-05,
+      "loss": 5.513,
+      "step": 3099
+    },
+    {
+      "epoch": 0.20793507059731026,
+      "grad_norm": 0.14292422123627008,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2080021464265352,
+      "grad_norm": 0.14652997007372942,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 3101
+    },
+    {
+      "epoch": 0.20806922225576013,
+      "grad_norm": 0.14108371346331539,
+      "learning_rate": 2e-05,
+      "loss": 5.4824,
+      "step": 3102
+    },
+    {
+      "epoch": 0.20813629808498507,
+      "grad_norm": 0.15138135369770928,
+      "learning_rate": 2e-05,
+      "loss": 5.5671,
+      "step": 3103
+    },
+    {
+      "epoch": 0.20820337391421,
+      "grad_norm": 0.15446227763580825,
+      "learning_rate": 2e-05,
+      "loss": 5.5476,
+      "step": 3104
+    },
+    {
+      "epoch": 0.20827044974343495,
+      "grad_norm": 0.14455946589848603,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2083375255726599,
+      "grad_norm": 0.15556147259144065,
+      "learning_rate": 2e-05,
+      "loss": 5.3669,
+      "step": 3106
+    },
+    {
+      "epoch": 0.20840460140188483,
+      "grad_norm": 0.1473600234054407,
+      "learning_rate": 2e-05,
+      "loss": 5.495,
+      "step": 3107
+    },
+    {
+      "epoch": 0.20847167723110976,
+      "grad_norm": 0.14546566514039044,
+      "learning_rate": 2e-05,
+      "loss": 5.5236,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2085387530603347,
+      "grad_norm": 0.15207936101904435,
+      "learning_rate": 2e-05,
+      "loss": 5.6357,
+      "step": 3109
+    },
+    {
+      "epoch": 0.20860582888955964,
+      "grad_norm": 0.1683997431935039,
+      "learning_rate": 2e-05,
+      "loss": 5.4837,
+      "step": 3110
+    },
+    {
+      "epoch": 0.20867290471878458,
+      "grad_norm": 0.14979660783303064,
+      "learning_rate": 2e-05,
+      "loss": 5.4416,
+      "step": 3111
+    },
+    {
+      "epoch": 0.20873998054800952,
+      "grad_norm": 0.1397880593419318,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 3112
+    },
+    {
+      "epoch": 0.20880705637723446,
+      "grad_norm": 0.14062837375466844,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2088741322064594,
+      "grad_norm": 0.15061629227849166,
+      "learning_rate": 2e-05,
+      "loss": 5.516,
+      "step": 3114
+    },
+    {
+      "epoch": 0.20894120803568433,
+      "grad_norm": 0.14680081895498812,
+      "learning_rate": 2e-05,
+      "loss": 5.3401,
+      "step": 3115
+    },
+    {
+      "epoch": 0.20900828386490927,
+      "grad_norm": 0.14102022693714888,
+      "learning_rate": 2e-05,
+      "loss": 5.4155,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2090753596941342,
+      "grad_norm": 0.1399473133115436,
+      "learning_rate": 2e-05,
+      "loss": 5.3414,
+      "step": 3117
+    },
+    {
+      "epoch": 0.20914243552335915,
+      "grad_norm": 0.1439648553301222,
+      "learning_rate": 2e-05,
+      "loss": 5.4817,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2092095113525841,
+      "grad_norm": 0.14056909046096888,
+      "learning_rate": 2e-05,
+      "loss": 5.403,
+      "step": 3119
+    },
+    {
+      "epoch": 0.20927658718180903,
+      "grad_norm": 0.14105973042678108,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20934366301103396,
+      "grad_norm": 0.1397450244096343,
+      "learning_rate": 2e-05,
+      "loss": 5.3693,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2094107388402589,
+      "grad_norm": 0.13733889778034322,
+      "learning_rate": 2e-05,
+      "loss": 5.6639,
+      "step": 3122
+    },
+    {
+      "epoch": 0.20947781466948384,
+      "grad_norm": 0.15010955109128885,
+      "learning_rate": 2e-05,
+      "loss": 5.6688,
+      "step": 3123
+    },
+    {
+      "epoch": 0.20954489049870878,
+      "grad_norm": 0.13893557795618847,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 3124
+    },
+    {
+      "epoch": 0.20961196632793372,
+      "grad_norm": 0.14121483235751464,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 3125
+    },
+    {
+      "epoch": 0.20967904215715866,
+      "grad_norm": 0.15358336776355982,
+      "learning_rate": 2e-05,
+      "loss": 5.3989,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2097461179863836,
+      "grad_norm": 0.14406810623991959,
+      "learning_rate": 2e-05,
+      "loss": 5.5001,
+      "step": 3127
+    },
+    {
+      "epoch": 0.20981319381560853,
+      "grad_norm": 0.14197236869185342,
+      "learning_rate": 2e-05,
+      "loss": 5.5416,
+      "step": 3128
+    },
+    {
+      "epoch": 0.20988026964483347,
+      "grad_norm": 0.1418805774394157,
+      "learning_rate": 2e-05,
+      "loss": 5.4431,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2099473454740584,
+      "grad_norm": 0.14411777774640497,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 3130
+    },
+    {
+      "epoch": 0.21001442130328335,
+      "grad_norm": 0.1453863508997706,
+      "learning_rate": 2e-05,
+      "loss": 5.3509,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2100814971325083,
+      "grad_norm": 0.13690146130641365,
+      "learning_rate": 2e-05,
+      "loss": 5.4378,
+      "step": 3132
+    },
+    {
+      "epoch": 0.21014857296173323,
+      "grad_norm": 0.14395788264883455,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 3133
+    },
+    {
+      "epoch": 0.21021564879095817,
+      "grad_norm": 0.1468773331653128,
+      "learning_rate": 2e-05,
+      "loss": 5.3514,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2102827246201831,
+      "grad_norm": 0.13928174181323075,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 3135
+    },
+    {
+      "epoch": 0.21034980044940804,
+      "grad_norm": 0.13813346516889752,
+      "learning_rate": 2e-05,
+      "loss": 5.4438,
+      "step": 3136
+    },
+    {
+      "epoch": 0.210416876278633,
+      "grad_norm": 0.14097473210244377,
+      "learning_rate": 2e-05,
+      "loss": 5.6157,
+      "step": 3137
+    },
+    {
+      "epoch": 0.21048395210785795,
+      "grad_norm": 0.1420229371175282,
+      "learning_rate": 2e-05,
+      "loss": 5.6865,
+      "step": 3138
+    },
+    {
+      "epoch": 0.21055102793708289,
+      "grad_norm": 0.14541383268996122,
+      "learning_rate": 2e-05,
+      "loss": 5.3944,
+      "step": 3139
+    },
+    {
+      "epoch": 0.21061810376630782,
+      "grad_norm": 0.14447513673296128,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 3140
+    },
+    {
+      "epoch": 0.21068517959553276,
+      "grad_norm": 0.13951692048735007,
+      "learning_rate": 2e-05,
+      "loss": 5.44,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2107522554247577,
+      "grad_norm": 0.14601729616907377,
+      "learning_rate": 2e-05,
+      "loss": 5.3517,
+      "step": 3142
+    },
+    {
+      "epoch": 0.21081933125398264,
+      "grad_norm": 0.14288211221162064,
+      "learning_rate": 2e-05,
+      "loss": 5.3477,
+      "step": 3143
+    },
+    {
+      "epoch": 0.21088640708320758,
+      "grad_norm": 0.14340310440472676,
+      "learning_rate": 2e-05,
+      "loss": 5.5222,
+      "step": 3144
+    },
+    {
+      "epoch": 0.21095348291243252,
+      "grad_norm": 0.15689416530214037,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21102055874165745,
+      "grad_norm": 0.14978960066678307,
+      "learning_rate": 2e-05,
+      "loss": 5.5155,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2110876345708824,
+      "grad_norm": 0.14333937157852977,
+      "learning_rate": 2e-05,
+      "loss": 5.5935,
+      "step": 3147
+    },
+    {
+      "epoch": 0.21115471040010733,
+      "grad_norm": 0.15003934278966846,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 3148
+    },
+    {
+      "epoch": 0.21122178622933227,
+      "grad_norm": 0.1636410947478519,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2112888620585572,
+      "grad_norm": 0.14607752283981512,
+      "learning_rate": 2e-05,
+      "loss": 5.422,
+      "step": 3150
+    },
+    {
+      "epoch": 0.21135593788778215,
+      "grad_norm": 0.14513506994288355,
+      "learning_rate": 2e-05,
+      "loss": 5.472,
+      "step": 3151
+    },
+    {
+      "epoch": 0.21142301371700709,
+      "grad_norm": 0.14452170337977938,
+      "learning_rate": 2e-05,
+      "loss": 5.5079,
+      "step": 3152
+    },
+    {
+      "epoch": 0.21149008954623202,
+      "grad_norm": 0.13936625305994524,
+      "learning_rate": 2e-05,
+      "loss": 5.5671,
+      "step": 3153
+    },
+    {
+      "epoch": 0.21155716537545696,
+      "grad_norm": 0.14761432001487632,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2116242412046819,
+      "grad_norm": 0.14248112726470508,
+      "learning_rate": 2e-05,
+      "loss": 5.4651,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21169131703390684,
+      "grad_norm": 0.15024303380893234,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 3156
+    },
+    {
+      "epoch": 0.21175839286313178,
+      "grad_norm": 0.13688357045495048,
+      "learning_rate": 2e-05,
+      "loss": 5.5145,
+      "step": 3157
+    },
+    {
+      "epoch": 0.21182546869235672,
+      "grad_norm": 0.14113721288626668,
+      "learning_rate": 2e-05,
+      "loss": 5.5873,
+      "step": 3158
+    },
+    {
+      "epoch": 0.21189254452158165,
+      "grad_norm": 0.15612672163791408,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2119596203508066,
+      "grad_norm": 0.13889565587375519,
+      "learning_rate": 2e-05,
+      "loss": 5.3583,
+      "step": 3160
+    },
+    {
+      "epoch": 0.21202669618003153,
+      "grad_norm": 0.14595303473167015,
+      "learning_rate": 2e-05,
+      "loss": 5.3332,
+      "step": 3161
+    },
+    {
+      "epoch": 0.21209377200925647,
+      "grad_norm": 0.1557610289199363,
+      "learning_rate": 2e-05,
+      "loss": 5.3329,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2121608478384814,
+      "grad_norm": 0.1445801329059681,
+      "learning_rate": 2e-05,
+      "loss": 5.3537,
+      "step": 3163
+    },
+    {
+      "epoch": 0.21222792366770635,
+      "grad_norm": 0.16015550030013584,
+      "learning_rate": 2e-05,
+      "loss": 5.5481,
+      "step": 3164
+    },
+    {
+      "epoch": 0.21229499949693129,
+      "grad_norm": 0.14403213981203153,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 3165
+    },
+    {
+      "epoch": 0.21236207532615622,
+      "grad_norm": 0.1468230070825452,
+      "learning_rate": 2e-05,
+      "loss": 5.2819,
+      "step": 3166
+    },
+    {
+      "epoch": 0.21242915115538116,
+      "grad_norm": 0.1403777006966255,
+      "learning_rate": 2e-05,
+      "loss": 5.5679,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2124962269846061,
+      "grad_norm": 0.1551011245255942,
+      "learning_rate": 2e-05,
+      "loss": 5.562,
+      "step": 3168
+    },
+    {
+      "epoch": 0.21256330281383104,
+      "grad_norm": 0.14890717538912843,
+      "learning_rate": 2e-05,
+      "loss": 5.4026,
+      "step": 3169
+    },
+    {
+      "epoch": 0.21263037864305598,
+      "grad_norm": 0.14826186593089416,
+      "learning_rate": 2e-05,
+      "loss": 5.5025,
+      "step": 3170
+    },
+    {
+      "epoch": 0.21269745447228092,
+      "grad_norm": 0.1446312561245324,
+      "learning_rate": 2e-05,
+      "loss": 5.2921,
+      "step": 3171
+    },
+    {
+      "epoch": 0.21276453030150586,
+      "grad_norm": 0.1527081329037847,
+      "learning_rate": 2e-05,
+      "loss": 5.3111,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2128316061307308,
+      "grad_norm": 0.13974109410024763,
+      "learning_rate": 2e-05,
+      "loss": 5.4236,
+      "step": 3173
+    },
+    {
+      "epoch": 0.21289868195995573,
+      "grad_norm": 0.14551930917950906,
+      "learning_rate": 2e-05,
+      "loss": 5.5298,
+      "step": 3174
+    },
+    {
+      "epoch": 0.21296575778918067,
+      "grad_norm": 0.137250810046067,
+      "learning_rate": 2e-05,
+      "loss": 5.4215,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2130328336184056,
+      "grad_norm": 0.1407280997901509,
+      "learning_rate": 2e-05,
+      "loss": 5.5511,
+      "step": 3176
+    },
+    {
+      "epoch": 0.21309990944763055,
+      "grad_norm": 0.1371370898914445,
+      "learning_rate": 2e-05,
+      "loss": 5.4351,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2131669852768555,
+      "grad_norm": 0.14419278137887057,
+      "learning_rate": 2e-05,
+      "loss": 5.4751,
+      "step": 3178
+    },
+    {
+      "epoch": 0.21323406110608042,
+      "grad_norm": 0.1433636171299948,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 3179
+    },
+    {
+      "epoch": 0.21330113693530536,
+      "grad_norm": 0.14472148554097586,
+      "learning_rate": 2e-05,
+      "loss": 5.2808,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2133682127645303,
+      "grad_norm": 0.14728076298555853,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 3181
+    },
+    {
+      "epoch": 0.21343528859375524,
+      "grad_norm": 0.14789719435084364,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 3182
+    },
+    {
+      "epoch": 0.21350236442298018,
+      "grad_norm": 0.14191005317421204,
+      "learning_rate": 2e-05,
+      "loss": 5.5851,
+      "step": 3183
+    },
+    {
+      "epoch": 0.21356944025220512,
+      "grad_norm": 0.1549257820440864,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 3184
+    },
+    {
+      "epoch": 0.21363651608143006,
+      "grad_norm": 0.1521058709323602,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 3185
+    },
+    {
+      "epoch": 0.213703591910655,
+      "grad_norm": 0.1416611609507571,
+      "learning_rate": 2e-05,
+      "loss": 5.454,
+      "step": 3186
+    },
+    {
+      "epoch": 0.21377066773987993,
+      "grad_norm": 0.14213355592181234,
+      "learning_rate": 2e-05,
+      "loss": 5.3991,
+      "step": 3187
+    },
+    {
+      "epoch": 0.21383774356910487,
+      "grad_norm": 0.14764065001647575,
+      "learning_rate": 2e-05,
+      "loss": 5.5522,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2139048193983298,
+      "grad_norm": 0.14588074924878125,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 3189
+    },
+    {
+      "epoch": 0.21397189522755475,
+      "grad_norm": 0.14130905345234476,
+      "learning_rate": 2e-05,
+      "loss": 5.3555,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2140389710567797,
+      "grad_norm": 0.1467764646808757,
+      "learning_rate": 2e-05,
+      "loss": 5.5283,
+      "step": 3191
+    },
+    {
+      "epoch": 0.21410604688600463,
+      "grad_norm": 0.14902790681130731,
+      "learning_rate": 2e-05,
+      "loss": 5.4571,
+      "step": 3192
+    },
+    {
+      "epoch": 0.21417312271522956,
+      "grad_norm": 0.1481406915322595,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2142401985444545,
+      "grad_norm": 0.14196425274046848,
+      "learning_rate": 2e-05,
+      "loss": 5.3809,
+      "step": 3194
+    },
+    {
+      "epoch": 0.21430727437367944,
+      "grad_norm": 0.14749043729935088,
+      "learning_rate": 2e-05,
+      "loss": 5.5198,
+      "step": 3195
+    },
+    {
+      "epoch": 0.21437435020290438,
+      "grad_norm": 0.14972169286494344,
+      "learning_rate": 2e-05,
+      "loss": 5.4541,
+      "step": 3196
+    },
+    {
+      "epoch": 0.21444142603212932,
+      "grad_norm": 0.14476836684585181,
+      "learning_rate": 2e-05,
+      "loss": 5.591,
+      "step": 3197
+    },
+    {
+      "epoch": 0.21450850186135426,
+      "grad_norm": 0.14204390727496452,
+      "learning_rate": 2e-05,
+      "loss": 5.4689,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2145755776905792,
+      "grad_norm": 0.15151050088300744,
+      "learning_rate": 2e-05,
+      "loss": 5.5278,
+      "step": 3199
+    },
+    {
+      "epoch": 0.21464265351980413,
+      "grad_norm": 0.14227925152320525,
+      "learning_rate": 2e-05,
+      "loss": 5.387,
+      "step": 3200
+    },
+    {
+      "epoch": 0.21470972934902907,
+      "grad_norm": 0.14558011965356218,
+      "learning_rate": 2e-05,
+      "loss": 5.5078,
+      "step": 3201
+    },
+    {
+      "epoch": 0.214776805178254,
+      "grad_norm": 0.14756292718284436,
+      "learning_rate": 2e-05,
+      "loss": 5.4132,
+      "step": 3202
+    },
+    {
+      "epoch": 0.21484388100747895,
+      "grad_norm": 0.14235864038439133,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2149109568367039,
+      "grad_norm": 0.15624519886200725,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 3204
+    },
+    {
+      "epoch": 0.21497803266592883,
+      "grad_norm": 0.15070229734644655,
+      "learning_rate": 2e-05,
+      "loss": 5.2519,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21504510849515376,
+      "grad_norm": 0.1450976489631152,
+      "learning_rate": 2e-05,
+      "loss": 5.5052,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2151121843243787,
+      "grad_norm": 0.1531313471246393,
+      "learning_rate": 2e-05,
+      "loss": 5.5375,
+      "step": 3207
+    },
+    {
+      "epoch": 0.21517926015360364,
+      "grad_norm": 0.14150071520545918,
+      "learning_rate": 2e-05,
+      "loss": 5.3538,
+      "step": 3208
+    },
+    {
+      "epoch": 0.21524633598282858,
+      "grad_norm": 0.14109334493976483,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 3209
+    },
+    {
+      "epoch": 0.21531341181205352,
+      "grad_norm": 0.14654135730583567,
+      "learning_rate": 2e-05,
+      "loss": 5.4281,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21538048764127846,
+      "grad_norm": 0.1448929443754271,
+      "learning_rate": 2e-05,
+      "loss": 5.5809,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2154475634705034,
+      "grad_norm": 0.14222667671408426,
+      "learning_rate": 2e-05,
+      "loss": 5.6381,
+      "step": 3212
+    },
+    {
+      "epoch": 0.21551463929972833,
+      "grad_norm": 0.14910683795322652,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 3213
+    },
+    {
+      "epoch": 0.21558171512895327,
+      "grad_norm": 0.15093734562031114,
+      "learning_rate": 2e-05,
+      "loss": 5.3076,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2156487909581782,
+      "grad_norm": 0.1483919168014747,
+      "learning_rate": 2e-05,
+      "loss": 5.5474,
+      "step": 3215
+    },
+    {
+      "epoch": 0.21571586678740315,
+      "grad_norm": 0.15535868975876266,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2157829426166281,
+      "grad_norm": 0.14198821066288736,
+      "learning_rate": 2e-05,
+      "loss": 5.5111,
+      "step": 3217
+    },
+    {
+      "epoch": 0.21585001844585303,
+      "grad_norm": 0.1447599838450322,
+      "learning_rate": 2e-05,
+      "loss": 5.4439,
+      "step": 3218
+    },
+    {
+      "epoch": 0.21591709427507796,
+      "grad_norm": 0.15128707909766362,
+      "learning_rate": 2e-05,
+      "loss": 5.4865,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2159841701043029,
+      "grad_norm": 0.14264956653343508,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21605124593352784,
+      "grad_norm": 0.1403301628615722,
+      "learning_rate": 2e-05,
+      "loss": 5.6286,
+      "step": 3221
+    },
+    {
+      "epoch": 0.21611832176275278,
+      "grad_norm": 0.15528770396372726,
+      "learning_rate": 2e-05,
+      "loss": 5.5553,
+      "step": 3222
+    },
+    {
+      "epoch": 0.21618539759197772,
+      "grad_norm": 0.14319346434747807,
+      "learning_rate": 2e-05,
+      "loss": 5.3659,
+      "step": 3223
+    },
+    {
+      "epoch": 0.21625247342120266,
+      "grad_norm": 0.14091543787734176,
+      "learning_rate": 2e-05,
+      "loss": 5.3949,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2163195492504276,
+      "grad_norm": 0.1486425088198627,
+      "learning_rate": 2e-05,
+      "loss": 5.4431,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21638662507965253,
+      "grad_norm": 0.14397993746299345,
+      "learning_rate": 2e-05,
+      "loss": 5.4264,
+      "step": 3226
+    },
+    {
+      "epoch": 0.21645370090887747,
+      "grad_norm": 0.1469011599629992,
+      "learning_rate": 2e-05,
+      "loss": 5.5253,
+      "step": 3227
+    },
+    {
+      "epoch": 0.21652077673810244,
+      "grad_norm": 0.1531627977947837,
+      "learning_rate": 2e-05,
+      "loss": 5.3494,
+      "step": 3228
+    },
+    {
+      "epoch": 0.21658785256732738,
+      "grad_norm": 0.14520325677009419,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 3229
+    },
+    {
+      "epoch": 0.21665492839655232,
+      "grad_norm": 0.1407434117043489,
+      "learning_rate": 2e-05,
+      "loss": 5.4839,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21672200422577725,
+      "grad_norm": 0.14700419185765715,
+      "learning_rate": 2e-05,
+      "loss": 5.4021,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2167890800550022,
+      "grad_norm": 0.15434291636388722,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 3232
+    },
+    {
+      "epoch": 0.21685615588422713,
+      "grad_norm": 0.14570134096286255,
+      "learning_rate": 2e-05,
+      "loss": 5.421,
+      "step": 3233
+    },
+    {
+      "epoch": 0.21692323171345207,
+      "grad_norm": 0.1472617270373705,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 3234
+    },
+    {
+      "epoch": 0.216990307542677,
+      "grad_norm": 0.14876830020010962,
+      "learning_rate": 2e-05,
+      "loss": 5.5196,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21705738337190195,
+      "grad_norm": 0.1534012752778758,
+      "learning_rate": 2e-05,
+      "loss": 5.3637,
+      "step": 3236
+    },
+    {
+      "epoch": 0.21712445920112688,
+      "grad_norm": 0.1419744824114444,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 3237
+    },
+    {
+      "epoch": 0.21719153503035182,
+      "grad_norm": 0.1498624838089467,
+      "learning_rate": 2e-05,
+      "loss": 5.6543,
+      "step": 3238
+    },
+    {
+      "epoch": 0.21725861085957676,
+      "grad_norm": 0.15403356299412144,
+      "learning_rate": 2e-05,
+      "loss": 5.5744,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2173256866888017,
+      "grad_norm": 0.1433927376672018,
+      "learning_rate": 2e-05,
+      "loss": 5.5316,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21739276251802664,
+      "grad_norm": 0.14889277686589178,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 3241
+    },
+    {
+      "epoch": 0.21745983834725158,
+      "grad_norm": 0.14310320150254194,
+      "learning_rate": 2e-05,
+      "loss": 5.595,
+      "step": 3242
+    },
+    {
+      "epoch": 0.21752691417647652,
+      "grad_norm": 0.14054839129450325,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 3243
+    },
+    {
+      "epoch": 0.21759399000570145,
+      "grad_norm": 0.14145450877128427,
+      "learning_rate": 2e-05,
+      "loss": 5.5252,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2176610658349264,
+      "grad_norm": 0.14823888110315733,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21772814166415133,
+      "grad_norm": 0.14593728728926186,
+      "learning_rate": 2e-05,
+      "loss": 5.4083,
+      "step": 3246
+    },
+    {
+      "epoch": 0.21779521749337627,
+      "grad_norm": 0.14625440803653394,
+      "learning_rate": 2e-05,
+      "loss": 5.5574,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2178622933226012,
+      "grad_norm": 0.14707612878696016,
+      "learning_rate": 2e-05,
+      "loss": 5.4345,
+      "step": 3248
+    },
+    {
+      "epoch": 0.21792936915182615,
+      "grad_norm": 0.14025601940512689,
+      "learning_rate": 2e-05,
+      "loss": 5.537,
+      "step": 3249
+    },
+    {
+      "epoch": 0.21799644498105109,
+      "grad_norm": 0.15137618728006727,
+      "learning_rate": 2e-05,
+      "loss": 5.5326,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21806352081027602,
+      "grad_norm": 0.1480765163879402,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 3251
+    },
+    {
+      "epoch": 0.21813059663950096,
+      "grad_norm": 0.14618901214492772,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2181976724687259,
+      "grad_norm": 0.14908071561772773,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 3253
+    },
+    {
+      "epoch": 0.21826474829795084,
+      "grad_norm": 0.15103693742322666,
+      "learning_rate": 2e-05,
+      "loss": 5.4662,
+      "step": 3254
+    },
+    {
+      "epoch": 0.21833182412717578,
+      "grad_norm": 0.13993391759080342,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21839889995640072,
+      "grad_norm": 0.14469112313820162,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 3256
+    },
+    {
+      "epoch": 0.21846597578562565,
+      "grad_norm": 0.15549967815390786,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2185330516148506,
+      "grad_norm": 0.15013196977659907,
+      "learning_rate": 2e-05,
+      "loss": 5.4123,
+      "step": 3258
+    },
+    {
+      "epoch": 0.21860012744407553,
+      "grad_norm": 0.15391835666807346,
+      "learning_rate": 2e-05,
+      "loss": 5.424,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21866720327330047,
+      "grad_norm": 0.15189234727666506,
+      "learning_rate": 2e-05,
+      "loss": 5.3853,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2187342791025254,
+      "grad_norm": 0.14316351491160523,
+      "learning_rate": 2e-05,
+      "loss": 5.522,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21880135493175035,
+      "grad_norm": 0.15071205847410488,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21886843076097529,
+      "grad_norm": 0.14677867840987388,
+      "learning_rate": 2e-05,
+      "loss": 5.31,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21893550659020022,
+      "grad_norm": 0.14316265616788784,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21900258241942516,
+      "grad_norm": 0.14514472590572142,
+      "learning_rate": 2e-05,
+      "loss": 5.5157,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2190696582486501,
+      "grad_norm": 0.1436745732882428,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21913673407787504,
+      "grad_norm": 0.14903705531823794,
+      "learning_rate": 2e-05,
+      "loss": 5.558,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21920380990709998,
+      "grad_norm": 0.14453176518241367,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21927088573632492,
+      "grad_norm": 0.14253449044735628,
+      "learning_rate": 2e-05,
+      "loss": 5.6043,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21933796156554985,
+      "grad_norm": 0.1479703266613821,
+      "learning_rate": 2e-05,
+      "loss": 5.3884,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2194050373947748,
+      "grad_norm": 0.14400393451620722,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21947211322399973,
+      "grad_norm": 0.1486525039917543,
+      "learning_rate": 2e-05,
+      "loss": 5.5454,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21953918905322467,
+      "grad_norm": 0.15274187946424048,
+      "learning_rate": 2e-05,
+      "loss": 5.5541,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2196062648824496,
+      "grad_norm": 0.1535018462194799,
+      "learning_rate": 2e-05,
+      "loss": 5.3389,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21967334071167455,
+      "grad_norm": 0.1420593685377505,
+      "learning_rate": 2e-05,
+      "loss": 5.5557,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21974041654089949,
+      "grad_norm": 0.1487048801258307,
+      "learning_rate": 2e-05,
+      "loss": 5.5296,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21980749237012442,
+      "grad_norm": 0.14747191990978456,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21987456819934936,
+      "grad_norm": 0.14125494959602256,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2199416440285743,
+      "grad_norm": 0.14698309356986805,
+      "learning_rate": 2e-05,
+      "loss": 5.3914,
+      "step": 3279
+    },
+    {
+      "epoch": 0.22000871985779924,
+      "grad_norm": 0.15091536411562856,
+      "learning_rate": 2e-05,
+      "loss": 5.5673,
+      "step": 3280
+    },
+    {
+      "epoch": 0.22007579568702418,
+      "grad_norm": 0.14345997483532777,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 3281
+    },
+    {
+      "epoch": 0.22014287151624912,
+      "grad_norm": 0.14016199289496584,
+      "learning_rate": 2e-05,
+      "loss": 5.5704,
+      "step": 3282
+    },
+    {
+      "epoch": 0.22020994734547406,
+      "grad_norm": 0.14522198421909455,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 3283
+    },
+    {
+      "epoch": 0.220277023174699,
+      "grad_norm": 0.14663207333966252,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 3284
+    },
+    {
+      "epoch": 0.22034409900392393,
+      "grad_norm": 0.14839668137633485,
+      "learning_rate": 2e-05,
+      "loss": 5.4721,
+      "step": 3285
+    },
+    {
+      "epoch": 0.22041117483314887,
+      "grad_norm": 0.14384895434823158,
+      "learning_rate": 2e-05,
+      "loss": 5.2787,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2204782506623738,
+      "grad_norm": 0.14323077201248935,
+      "learning_rate": 2e-05,
+      "loss": 5.3615,
+      "step": 3287
+    },
+    {
+      "epoch": 0.22054532649159875,
+      "grad_norm": 0.15661144852426015,
+      "learning_rate": 2e-05,
+      "loss": 5.3908,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2206124023208237,
+      "grad_norm": 0.14969866329865555,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 3289
+    },
+    {
+      "epoch": 0.22067947815004862,
+      "grad_norm": 0.14150137018848877,
+      "learning_rate": 2e-05,
+      "loss": 5.369,
+      "step": 3290
+    },
+    {
+      "epoch": 0.22074655397927356,
+      "grad_norm": 0.14733168570098099,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2208136298084985,
+      "grad_norm": 0.14074849988433014,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 3292
+    },
+    {
+      "epoch": 0.22088070563772344,
+      "grad_norm": 0.15360931466649588,
+      "learning_rate": 2e-05,
+      "loss": 5.4907,
+      "step": 3293
+    },
+    {
+      "epoch": 0.22094778146694838,
+      "grad_norm": 0.1441457513637673,
+      "learning_rate": 2e-05,
+      "loss": 5.3295,
+      "step": 3294
+    },
+    {
+      "epoch": 0.22101485729617332,
+      "grad_norm": 0.1454151361769673,
+      "learning_rate": 2e-05,
+      "loss": 5.4096,
+      "step": 3295
+    },
+    {
+      "epoch": 0.22108193312539826,
+      "grad_norm": 0.1482526917231663,
+      "learning_rate": 2e-05,
+      "loss": 5.4018,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2211490089546232,
+      "grad_norm": 0.14779623260860175,
+      "learning_rate": 2e-05,
+      "loss": 5.4151,
+      "step": 3297
+    },
+    {
+      "epoch": 0.22121608478384813,
+      "grad_norm": 0.14494605532190408,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 3298
+    },
+    {
+      "epoch": 0.22128316061307307,
+      "grad_norm": 0.14773534375713246,
+      "learning_rate": 2e-05,
+      "loss": 5.3223,
+      "step": 3299
+    },
+    {
+      "epoch": 0.221350236442298,
+      "grad_norm": 0.15442321120473268,
+      "learning_rate": 2e-05,
+      "loss": 5.403,
+      "step": 3300
+    },
+    {
+      "epoch": 0.22141731227152295,
+      "grad_norm": 0.13754664208473438,
+      "learning_rate": 2e-05,
+      "loss": 5.5776,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2214843881007479,
+      "grad_norm": 0.1412273666297994,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 3302
+    },
+    {
+      "epoch": 0.22155146392997283,
+      "grad_norm": 0.14711497291752107,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 3303
+    },
+    {
+      "epoch": 0.22161853975919776,
+      "grad_norm": 0.14899014956174633,
+      "learning_rate": 2e-05,
+      "loss": 5.418,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2216856155884227,
+      "grad_norm": 0.14121523671287264,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 3305
+    },
+    {
+      "epoch": 0.22175269141764764,
+      "grad_norm": 0.15424168549465048,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 3306
+    },
+    {
+      "epoch": 0.22181976724687258,
+      "grad_norm": 0.14163486181898038,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 3307
+    },
+    {
+      "epoch": 0.22188684307609752,
+      "grad_norm": 0.14515407993835638,
+      "learning_rate": 2e-05,
+      "loss": 5.4077,
+      "step": 3308
+    },
+    {
+      "epoch": 0.22195391890532246,
+      "grad_norm": 0.14571876901783734,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2220209947345474,
+      "grad_norm": 0.14402978354917284,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 3310
+    },
+    {
+      "epoch": 0.22208807056377233,
+      "grad_norm": 0.1417472126820956,
+      "learning_rate": 2e-05,
+      "loss": 5.4274,
+      "step": 3311
+    },
+    {
+      "epoch": 0.22215514639299727,
+      "grad_norm": 0.1450794222434863,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.1431991683922062,
+      "learning_rate": 2e-05,
+      "loss": 5.3527,
+      "step": 3313
+    },
+    {
+      "epoch": 0.22228929805144715,
+      "grad_norm": 0.13859018370680973,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2223563738806721,
+      "grad_norm": 0.14407800033283474,
+      "learning_rate": 2e-05,
+      "loss": 5.5719,
+      "step": 3315
+    },
+    {
+      "epoch": 0.22242344970989703,
+      "grad_norm": 0.15198461766798824,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 3316
+    },
+    {
+      "epoch": 0.22249052553912196,
+      "grad_norm": 0.14830205883938186,
+      "learning_rate": 2e-05,
+      "loss": 5.5057,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2225576013683469,
+      "grad_norm": 0.1456247687418642,
+      "learning_rate": 2e-05,
+      "loss": 5.563,
+      "step": 3318
+    },
+    {
+      "epoch": 0.22262467719757187,
+      "grad_norm": 0.148248463875476,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2226917530267968,
+      "grad_norm": 0.14897145918519974,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 3320
+    },
+    {
+      "epoch": 0.22275882885602175,
+      "grad_norm": 0.13984777476904017,
+      "learning_rate": 2e-05,
+      "loss": 5.5934,
+      "step": 3321
+    },
+    {
+      "epoch": 0.22282590468524668,
+      "grad_norm": 0.13851013485340352,
+      "learning_rate": 2e-05,
+      "loss": 5.409,
+      "step": 3322
+    },
+    {
+      "epoch": 0.22289298051447162,
+      "grad_norm": 0.14887849420642546,
+      "learning_rate": 2e-05,
+      "loss": 5.3563,
+      "step": 3323
+    },
+    {
+      "epoch": 0.22296005634369656,
+      "grad_norm": 0.1399338920570312,
+      "learning_rate": 2e-05,
+      "loss": 5.3355,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2230271321729215,
+      "grad_norm": 0.1437955602929798,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 3325
+    },
+    {
+      "epoch": 0.22309420800214644,
+      "grad_norm": 0.14574781976642454,
+      "learning_rate": 2e-05,
+      "loss": 5.4015,
+      "step": 3326
+    },
+    {
+      "epoch": 0.22316128383137138,
+      "grad_norm": 0.1452574798881683,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 3327
+    },
+    {
+      "epoch": 0.22322835966059631,
+      "grad_norm": 0.14171525233435625,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 3328
+    },
+    {
+      "epoch": 0.22329543548982125,
+      "grad_norm": 0.143561487256503,
+      "learning_rate": 2e-05,
+      "loss": 5.394,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2233625113190462,
+      "grad_norm": 0.14548669086288804,
+      "learning_rate": 2e-05,
+      "loss": 5.358,
+      "step": 3330
+    },
+    {
+      "epoch": 0.22342958714827113,
+      "grad_norm": 0.1391528585107554,
+      "learning_rate": 2e-05,
+      "loss": 5.5248,
+      "step": 3331
+    },
+    {
+      "epoch": 0.22349666297749607,
+      "grad_norm": 0.14379887452054305,
+      "learning_rate": 2e-05,
+      "loss": 5.3513,
+      "step": 3332
+    },
+    {
+      "epoch": 0.223563738806721,
+      "grad_norm": 0.14445073766755653,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 3333
+    },
+    {
+      "epoch": 0.22363081463594595,
+      "grad_norm": 0.14474146647118105,
+      "learning_rate": 2e-05,
+      "loss": 5.4753,
+      "step": 3334
+    },
+    {
+      "epoch": 0.22369789046517088,
+      "grad_norm": 0.14160179207495563,
+      "learning_rate": 2e-05,
+      "loss": 5.63,
+      "step": 3335
+    },
+    {
+      "epoch": 0.22376496629439582,
+      "grad_norm": 0.1359150134910531,
+      "learning_rate": 2e-05,
+      "loss": 5.3557,
+      "step": 3336
+    },
+    {
+      "epoch": 0.22383204212362076,
+      "grad_norm": 0.13853732856451736,
+      "learning_rate": 2e-05,
+      "loss": 5.315,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2238991179528457,
+      "grad_norm": 0.1466469528710574,
+      "learning_rate": 2e-05,
+      "loss": 5.4195,
+      "step": 3338
+    },
+    {
+      "epoch": 0.22396619378207064,
+      "grad_norm": 0.145500337621402,
+      "learning_rate": 2e-05,
+      "loss": 5.384,
+      "step": 3339
+    },
+    {
+      "epoch": 0.22403326961129558,
+      "grad_norm": 0.14527607811757642,
+      "learning_rate": 2e-05,
+      "loss": 5.4743,
+      "step": 3340
+    },
+    {
+      "epoch": 0.22410034544052052,
+      "grad_norm": 0.14417387607357432,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 3341
+    },
+    {
+      "epoch": 0.22416742126974545,
+      "grad_norm": 0.14642908162842402,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2242344970989704,
+      "grad_norm": 0.14648563762939465,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 3343
+    },
+    {
+      "epoch": 0.22430157292819533,
+      "grad_norm": 0.14969092754863134,
+      "learning_rate": 2e-05,
+      "loss": 5.3663,
+      "step": 3344
+    },
+    {
+      "epoch": 0.22436864875742027,
+      "grad_norm": 0.1453288398702135,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2244357245866452,
+      "grad_norm": 0.1501449053626544,
+      "learning_rate": 2e-05,
+      "loss": 5.3492,
+      "step": 3346
+    },
+    {
+      "epoch": 0.22450280041587015,
+      "grad_norm": 0.14118306940853848,
+      "learning_rate": 2e-05,
+      "loss": 5.3102,
+      "step": 3347
+    },
+    {
+      "epoch": 0.22456987624509508,
+      "grad_norm": 0.1452119444311658,
+      "learning_rate": 2e-05,
+      "loss": 5.4148,
+      "step": 3348
+    },
+    {
+      "epoch": 0.22463695207432002,
+      "grad_norm": 0.1462631968904061,
+      "learning_rate": 2e-05,
+      "loss": 5.5368,
+      "step": 3349
+    },
+    {
+      "epoch": 0.22470402790354496,
+      "grad_norm": 0.14258471998355282,
+      "learning_rate": 2e-05,
+      "loss": 5.4189,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2247711037327699,
+      "grad_norm": 0.14053407163490214,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 3351
+    },
+    {
+      "epoch": 0.22483817956199484,
+      "grad_norm": 0.14868010203872675,
+      "learning_rate": 2e-05,
+      "loss": 5.4498,
+      "step": 3352
+    },
+    {
+      "epoch": 0.22490525539121978,
+      "grad_norm": 0.15479655157381114,
+      "learning_rate": 2e-05,
+      "loss": 5.5889,
+      "step": 3353
+    },
+    {
+      "epoch": 0.22497233122044472,
+      "grad_norm": 0.14789543006263284,
+      "learning_rate": 2e-05,
+      "loss": 5.4545,
+      "step": 3354
+    },
+    {
+      "epoch": 0.22503940704966965,
+      "grad_norm": 0.14540681134346461,
+      "learning_rate": 2e-05,
+      "loss": 5.4728,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2251064828788946,
+      "grad_norm": 0.1418552082768162,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 3356
+    },
+    {
+      "epoch": 0.22517355870811953,
+      "grad_norm": 0.14996099255898973,
+      "learning_rate": 2e-05,
+      "loss": 5.4556,
+      "step": 3357
+    },
+    {
+      "epoch": 0.22524063453734447,
+      "grad_norm": 0.1466803115494233,
+      "learning_rate": 2e-05,
+      "loss": 5.3674,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2253077103665694,
+      "grad_norm": 0.14245217870405116,
+      "learning_rate": 2e-05,
+      "loss": 5.3492,
+      "step": 3359
+    },
+    {
+      "epoch": 0.22537478619579435,
+      "grad_norm": 0.14662140655960706,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22544186202501929,
+      "grad_norm": 0.14690895233302784,
+      "learning_rate": 2e-05,
+      "loss": 5.3943,
+      "step": 3361
+    },
+    {
+      "epoch": 0.22550893785424422,
+      "grad_norm": 0.14813828259902137,
+      "learning_rate": 2e-05,
+      "loss": 5.4247,
+      "step": 3362
+    },
+    {
+      "epoch": 0.22557601368346916,
+      "grad_norm": 0.14560137512926544,
+      "learning_rate": 2e-05,
+      "loss": 5.4723,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2256430895126941,
+      "grad_norm": 0.13979113657917733,
+      "learning_rate": 2e-05,
+      "loss": 5.3748,
+      "step": 3364
+    },
+    {
+      "epoch": 0.22571016534191904,
+      "grad_norm": 0.14708609837116485,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22577724117114398,
+      "grad_norm": 0.14614842506070927,
+      "learning_rate": 2e-05,
+      "loss": 5.5455,
+      "step": 3366
+    },
+    {
+      "epoch": 0.22584431700036892,
+      "grad_norm": 0.1474168667901843,
+      "learning_rate": 2e-05,
+      "loss": 5.4803,
+      "step": 3367
+    },
+    {
+      "epoch": 0.22591139282959385,
+      "grad_norm": 0.15021074426409145,
+      "learning_rate": 2e-05,
+      "loss": 5.363,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2259784686588188,
+      "grad_norm": 0.14144366793432883,
+      "learning_rate": 2e-05,
+      "loss": 5.5611,
+      "step": 3369
+    },
+    {
+      "epoch": 0.22604554448804373,
+      "grad_norm": 0.142355204104326,
+      "learning_rate": 2e-05,
+      "loss": 5.5282,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22611262031726867,
+      "grad_norm": 0.15224386056668945,
+      "learning_rate": 2e-05,
+      "loss": 5.4439,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2261796961464936,
+      "grad_norm": 0.14253957973109355,
+      "learning_rate": 2e-05,
+      "loss": 5.5983,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22624677197571855,
+      "grad_norm": 0.1469803824604715,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 3373
+    },
+    {
+      "epoch": 0.22631384780494349,
+      "grad_norm": 0.14141834465343275,
+      "learning_rate": 2e-05,
+      "loss": 5.5093,
+      "step": 3374
+    },
+    {
+      "epoch": 0.22638092363416842,
+      "grad_norm": 0.14481688844242438,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22644799946339336,
+      "grad_norm": 0.15328909068071533,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2265150752926183,
+      "grad_norm": 0.14125314777478323,
+      "learning_rate": 2e-05,
+      "loss": 5.4363,
+      "step": 3377
+    },
+    {
+      "epoch": 0.22658215112184324,
+      "grad_norm": 0.14248935886426195,
+      "learning_rate": 2e-05,
+      "loss": 5.3558,
+      "step": 3378
+    },
+    {
+      "epoch": 0.22664922695106818,
+      "grad_norm": 0.14040882023903561,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 3379
+    },
+    {
+      "epoch": 0.22671630278029312,
+      "grad_norm": 0.14171605958872224,
+      "learning_rate": 2e-05,
+      "loss": 5.4552,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22678337860951805,
+      "grad_norm": 0.14645327192167784,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 3381
+    },
+    {
+      "epoch": 0.226850454438743,
+      "grad_norm": 0.14040358696436314,
+      "learning_rate": 2e-05,
+      "loss": 5.3237,
+      "step": 3382
+    },
+    {
+      "epoch": 0.22691753026796793,
+      "grad_norm": 0.1384265205732937,
+      "learning_rate": 2e-05,
+      "loss": 5.338,
+      "step": 3383
+    },
+    {
+      "epoch": 0.22698460609719287,
+      "grad_norm": 0.14383877630117523,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2270516819264178,
+      "grad_norm": 0.15100340295462444,
+      "learning_rate": 2e-05,
+      "loss": 5.3593,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22711875775564275,
+      "grad_norm": 0.14130549310510448,
+      "learning_rate": 2e-05,
+      "loss": 5.505,
+      "step": 3386
+    },
+    {
+      "epoch": 0.22718583358486769,
+      "grad_norm": 0.14256663011015833,
+      "learning_rate": 2e-05,
+      "loss": 5.5269,
+      "step": 3387
+    },
+    {
+      "epoch": 0.22725290941409262,
+      "grad_norm": 0.15085827804340995,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 3388
+    },
+    {
+      "epoch": 0.22731998524331756,
+      "grad_norm": 0.14828126460966765,
+      "learning_rate": 2e-05,
+      "loss": 5.3036,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2273870610725425,
+      "grad_norm": 0.14498907927164847,
+      "learning_rate": 2e-05,
+      "loss": 5.447,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22745413690176744,
+      "grad_norm": 0.14413043156888655,
+      "learning_rate": 2e-05,
+      "loss": 5.6121,
+      "step": 3391
+    },
+    {
+      "epoch": 0.22752121273099238,
+      "grad_norm": 0.14234778631638192,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 3392
+    },
+    {
+      "epoch": 0.22758828856021732,
+      "grad_norm": 0.13753813639303925,
+      "learning_rate": 2e-05,
+      "loss": 5.2939,
+      "step": 3393
+    },
+    {
+      "epoch": 0.22765536438944226,
+      "grad_norm": 0.1528255449976704,
+      "learning_rate": 2e-05,
+      "loss": 5.2939,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2277224402186672,
+      "grad_norm": 0.14421302592863017,
+      "learning_rate": 2e-05,
+      "loss": 5.3352,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22778951604789213,
+      "grad_norm": 0.15010412878968277,
+      "learning_rate": 2e-05,
+      "loss": 5.5838,
+      "step": 3396
+    },
+    {
+      "epoch": 0.22785659187711707,
+      "grad_norm": 0.15464374175148057,
+      "learning_rate": 2e-05,
+      "loss": 5.4792,
+      "step": 3397
+    },
+    {
+      "epoch": 0.227923667706342,
+      "grad_norm": 0.14797515090400598,
+      "learning_rate": 2e-05,
+      "loss": 5.4047,
+      "step": 3398
+    },
+    {
+      "epoch": 0.22799074353556695,
+      "grad_norm": 0.1425283979324664,
+      "learning_rate": 2e-05,
+      "loss": 5.4424,
+      "step": 3399
+    },
+    {
+      "epoch": 0.22805781936479189,
+      "grad_norm": 0.16006468965061685,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22812489519401682,
+      "grad_norm": 0.14523925385595776,
+      "learning_rate": 2e-05,
+      "loss": 5.5093,
+      "step": 3401
+    },
+    {
+      "epoch": 0.22819197102324176,
+      "grad_norm": 0.14726897613179657,
+      "learning_rate": 2e-05,
+      "loss": 5.4243,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2282590468524667,
+      "grad_norm": 0.14690783609994362,
+      "learning_rate": 2e-05,
+      "loss": 5.4418,
+      "step": 3403
+    },
+    {
+      "epoch": 0.22832612268169164,
+      "grad_norm": 0.1565688716198834,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 3404
+    },
+    {
+      "epoch": 0.22839319851091658,
+      "grad_norm": 0.14404877109519776,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22846027434014152,
+      "grad_norm": 0.15211231795216418,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 3406
+    },
+    {
+      "epoch": 0.22852735016936646,
+      "grad_norm": 0.15012913020479116,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2285944259985914,
+      "grad_norm": 0.14492099842965925,
+      "learning_rate": 2e-05,
+      "loss": 5.407,
+      "step": 3408
+    },
+    {
+      "epoch": 0.22866150182781636,
+      "grad_norm": 0.1468269201156155,
+      "learning_rate": 2e-05,
+      "loss": 5.416,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2287285776570413,
+      "grad_norm": 0.1464244650819511,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22879565348626624,
+      "grad_norm": 0.14878296563401044,
+      "learning_rate": 2e-05,
+      "loss": 5.5534,
+      "step": 3411
+    },
+    {
+      "epoch": 0.22886272931549118,
+      "grad_norm": 0.13959208750680374,
+      "learning_rate": 2e-05,
+      "loss": 5.4725,
+      "step": 3412
+    },
+    {
+      "epoch": 0.22892980514471611,
+      "grad_norm": 0.15171536164152227,
+      "learning_rate": 2e-05,
+      "loss": 5.3108,
+      "step": 3413
+    },
+    {
+      "epoch": 0.22899688097394105,
+      "grad_norm": 0.14533678396583066,
+      "learning_rate": 2e-05,
+      "loss": 5.3943,
+      "step": 3414
+    },
+    {
+      "epoch": 0.229063956803166,
+      "grad_norm": 0.1450557298121004,
+      "learning_rate": 2e-05,
+      "loss": 5.5631,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22913103263239093,
+      "grad_norm": 0.1518773227471714,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 3416
+    },
+    {
+      "epoch": 0.22919810846161587,
+      "grad_norm": 0.14291958947556643,
+      "learning_rate": 2e-05,
+      "loss": 5.4769,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2292651842908408,
+      "grad_norm": 0.15401083614957103,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22933226012006575,
+      "grad_norm": 0.1571937733112811,
+      "learning_rate": 2e-05,
+      "loss": 5.507,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22939933594929068,
+      "grad_norm": 0.15682683820937696,
+      "learning_rate": 2e-05,
+      "loss": 5.6094,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22946641177851562,
+      "grad_norm": 0.14697385852311382,
+      "learning_rate": 2e-05,
+      "loss": 5.4774,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22953348760774056,
+      "grad_norm": 0.1622536003713561,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2296005634369655,
+      "grad_norm": 0.14725023302156642,
+      "learning_rate": 2e-05,
+      "loss": 5.493,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22966763926619044,
+      "grad_norm": 0.14245512755067502,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22973471509541538,
+      "grad_norm": 0.14632260849898043,
+      "learning_rate": 2e-05,
+      "loss": 5.4943,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22980179092464031,
+      "grad_norm": 0.163994804967215,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22986886675386525,
+      "grad_norm": 0.1465502317429114,
+      "learning_rate": 2e-05,
+      "loss": 5.4666,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2299359425830902,
+      "grad_norm": 0.15241126838888572,
+      "learning_rate": 2e-05,
+      "loss": 5.5447,
+      "step": 3428
+    },
+    {
+      "epoch": 0.23000301841231513,
+      "grad_norm": 0.1514294029175094,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 3429
+    },
+    {
+      "epoch": 0.23007009424154007,
+      "grad_norm": 0.15463798545433138,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 3430
+    },
+    {
+      "epoch": 0.230137170070765,
+      "grad_norm": 0.14434279594774765,
+      "learning_rate": 2e-05,
+      "loss": 5.3553,
+      "step": 3431
+    },
+    {
+      "epoch": 0.23020424589998995,
+      "grad_norm": 0.15159219476294827,
+      "learning_rate": 2e-05,
+      "loss": 5.5237,
+      "step": 3432
+    },
+    {
+      "epoch": 0.23027132172921488,
+      "grad_norm": 0.14740066096012147,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 3433
+    },
+    {
+      "epoch": 0.23033839755843982,
+      "grad_norm": 0.1439765915280828,
+      "learning_rate": 2e-05,
+      "loss": 5.2935,
+      "step": 3434
+    },
+    {
+      "epoch": 0.23040547338766476,
+      "grad_norm": 0.1471021682882794,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2304725492168897,
+      "grad_norm": 0.14297794671159805,
+      "learning_rate": 2e-05,
+      "loss": 5.4506,
+      "step": 3436
+    },
+    {
+      "epoch": 0.23053962504611464,
+      "grad_norm": 0.14308505249073547,
+      "learning_rate": 2e-05,
+      "loss": 5.313,
+      "step": 3437
+    },
+    {
+      "epoch": 0.23060670087533958,
+      "grad_norm": 0.1477452365182314,
+      "learning_rate": 2e-05,
+      "loss": 5.4021,
+      "step": 3438
+    },
+    {
+      "epoch": 0.23067377670456451,
+      "grad_norm": 0.1407774898466553,
+      "learning_rate": 2e-05,
+      "loss": 5.4568,
+      "step": 3439
+    },
+    {
+      "epoch": 0.23074085253378945,
+      "grad_norm": 0.1458398085465949,
+      "learning_rate": 2e-05,
+      "loss": 5.4968,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2308079283630144,
+      "grad_norm": 0.14672830791462918,
+      "learning_rate": 2e-05,
+      "loss": 5.6024,
+      "step": 3441
+    },
+    {
+      "epoch": 0.23087500419223933,
+      "grad_norm": 0.1407893521094788,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 3442
+    },
+    {
+      "epoch": 0.23094208002146427,
+      "grad_norm": 0.14272943978723976,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2310091558506892,
+      "grad_norm": 0.15382186391746266,
+      "learning_rate": 2e-05,
+      "loss": 5.4774,
+      "step": 3444
+    },
+    {
+      "epoch": 0.23107623167991415,
+      "grad_norm": 0.1457638253653574,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 3445
+    },
+    {
+      "epoch": 0.23114330750913908,
+      "grad_norm": 0.14927022932748973,
+      "learning_rate": 2e-05,
+      "loss": 5.37,
+      "step": 3446
+    },
+    {
+      "epoch": 0.23121038333836402,
+      "grad_norm": 0.13740387995035644,
+      "learning_rate": 2e-05,
+      "loss": 5.4411,
+      "step": 3447
+    },
+    {
+      "epoch": 0.23127745916758896,
+      "grad_norm": 0.14247515026610486,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2313445349968139,
+      "grad_norm": 0.1469279389738208,
+      "learning_rate": 2e-05,
+      "loss": 5.5468,
+      "step": 3449
+    },
+    {
+      "epoch": 0.23141161082603884,
+      "grad_norm": 0.14226201634245214,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 3450
+    },
+    {
+      "epoch": 0.23147868665526378,
+      "grad_norm": 0.14075024937949654,
+      "learning_rate": 2e-05,
+      "loss": 5.3123,
+      "step": 3451
+    },
+    {
+      "epoch": 0.23154576248448872,
+      "grad_norm": 0.14596852944026004,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 3452
+    },
+    {
+      "epoch": 0.23161283831371365,
+      "grad_norm": 0.1554609594807938,
+      "learning_rate": 2e-05,
+      "loss": 5.5661,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2316799141429386,
+      "grad_norm": 0.14429336364512899,
+      "learning_rate": 2e-05,
+      "loss": 5.4758,
+      "step": 3454
+    },
+    {
+      "epoch": 0.23174698997216353,
+      "grad_norm": 0.14286153756938957,
+      "learning_rate": 2e-05,
+      "loss": 5.3602,
+      "step": 3455
+    },
+    {
+      "epoch": 0.23181406580138847,
+      "grad_norm": 0.14364683734649591,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2318811416306134,
+      "grad_norm": 0.1423188118060661,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 3457
+    },
+    {
+      "epoch": 0.23194821745983835,
+      "grad_norm": 0.15434914701733232,
+      "learning_rate": 2e-05,
+      "loss": 5.4899,
+      "step": 3458
+    },
+    {
+      "epoch": 0.23201529328906328,
+      "grad_norm": 0.15044817775695768,
+      "learning_rate": 2e-05,
+      "loss": 5.377,
+      "step": 3459
+    },
+    {
+      "epoch": 0.23208236911828822,
+      "grad_norm": 0.1447980728131735,
+      "learning_rate": 2e-05,
+      "loss": 5.5556,
+      "step": 3460
+    },
+    {
+      "epoch": 0.23214944494751316,
+      "grad_norm": 0.1473657646073293,
+      "learning_rate": 2e-05,
+      "loss": 5.513,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2322165207767381,
+      "grad_norm": 0.1540195375523279,
+      "learning_rate": 2e-05,
+      "loss": 5.3536,
+      "step": 3462
+    },
+    {
+      "epoch": 0.23228359660596304,
+      "grad_norm": 0.15140455484196938,
+      "learning_rate": 2e-05,
+      "loss": 5.4865,
+      "step": 3463
+    },
+    {
+      "epoch": 0.23235067243518798,
+      "grad_norm": 0.1425535170030364,
+      "learning_rate": 2e-05,
+      "loss": 5.3209,
+      "step": 3464
+    },
+    {
+      "epoch": 0.23241774826441292,
+      "grad_norm": 0.14989365672785368,
+      "learning_rate": 2e-05,
+      "loss": 5.3769,
+      "step": 3465
+    },
+    {
+      "epoch": 0.23248482409363785,
+      "grad_norm": 0.14819540122780983,
+      "learning_rate": 2e-05,
+      "loss": 5.2552,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2325518999228628,
+      "grad_norm": 0.14893145733154475,
+      "learning_rate": 2e-05,
+      "loss": 5.4109,
+      "step": 3467
+    },
+    {
+      "epoch": 0.23261897575208773,
+      "grad_norm": 0.14504790903785594,
+      "learning_rate": 2e-05,
+      "loss": 5.4506,
+      "step": 3468
+    },
+    {
+      "epoch": 0.23268605158131267,
+      "grad_norm": 0.15100237555233534,
+      "learning_rate": 2e-05,
+      "loss": 5.5081,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2327531274105376,
+      "grad_norm": 0.14828666784357752,
+      "learning_rate": 2e-05,
+      "loss": 5.4153,
+      "step": 3470
+    },
+    {
+      "epoch": 0.23282020323976255,
+      "grad_norm": 0.14958997152927891,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 3471
+    },
+    {
+      "epoch": 0.23288727906898748,
+      "grad_norm": 0.15029568920565664,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 3472
+    },
+    {
+      "epoch": 0.23295435489821242,
+      "grad_norm": 0.15398294644031,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 3473
+    },
+    {
+      "epoch": 0.23302143072743736,
+      "grad_norm": 0.13548851722105862,
+      "learning_rate": 2e-05,
+      "loss": 5.5548,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2330885065566623,
+      "grad_norm": 0.15430529302511342,
+      "learning_rate": 2e-05,
+      "loss": 5.3476,
+      "step": 3475
+    },
+    {
+      "epoch": 0.23315558238588724,
+      "grad_norm": 0.14153151682452225,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 3476
+    },
+    {
+      "epoch": 0.23322265821511218,
+      "grad_norm": 0.14232073483353452,
+      "learning_rate": 2e-05,
+      "loss": 5.5149,
+      "step": 3477
+    },
+    {
+      "epoch": 0.23328973404433712,
+      "grad_norm": 0.1440756610156961,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 3478
+    },
+    {
+      "epoch": 0.23335680987356205,
+      "grad_norm": 0.14911651047203564,
+      "learning_rate": 2e-05,
+      "loss": 5.4429,
+      "step": 3479
+    },
+    {
+      "epoch": 0.233423885702787,
+      "grad_norm": 0.15000868980779755,
+      "learning_rate": 2e-05,
+      "loss": 5.4138,
+      "step": 3480
+    },
+    {
+      "epoch": 0.23349096153201193,
+      "grad_norm": 0.14457725499352092,
+      "learning_rate": 2e-05,
+      "loss": 5.3754,
+      "step": 3481
+    },
+    {
+      "epoch": 0.23355803736123687,
+      "grad_norm": 0.14157268385249638,
+      "learning_rate": 2e-05,
+      "loss": 5.3373,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2336251131904618,
+      "grad_norm": 0.15794916005534493,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 3483
+    },
+    {
+      "epoch": 0.23369218901968675,
+      "grad_norm": 0.1503698387383471,
+      "learning_rate": 2e-05,
+      "loss": 5.5343,
+      "step": 3484
+    },
+    {
+      "epoch": 0.23375926484891169,
+      "grad_norm": 0.13872005015624286,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 3485
+    },
+    {
+      "epoch": 0.23382634067813662,
+      "grad_norm": 0.14663081469979158,
+      "learning_rate": 2e-05,
+      "loss": 5.3847,
+      "step": 3486
+    },
+    {
+      "epoch": 0.23389341650736156,
+      "grad_norm": 0.14664795209165352,
+      "learning_rate": 2e-05,
+      "loss": 5.4122,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2339604923365865,
+      "grad_norm": 0.13615857855746596,
+      "learning_rate": 2e-05,
+      "loss": 5.3387,
+      "step": 3488
+    },
+    {
+      "epoch": 0.23402756816581144,
+      "grad_norm": 0.14138756527291793,
+      "learning_rate": 2e-05,
+      "loss": 5.3293,
+      "step": 3489
+    },
+    {
+      "epoch": 0.23409464399503638,
+      "grad_norm": 0.1473045378295767,
+      "learning_rate": 2e-05,
+      "loss": 5.443,
+      "step": 3490
+    },
+    {
+      "epoch": 0.23416171982426132,
+      "grad_norm": 0.1493431720275265,
+      "learning_rate": 2e-05,
+      "loss": 5.557,
+      "step": 3491
+    },
+    {
+      "epoch": 0.23422879565348625,
+      "grad_norm": 0.1431735933763373,
+      "learning_rate": 2e-05,
+      "loss": 5.3499,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2342958714827112,
+      "grad_norm": 0.14248237496004026,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 3493
+    },
+    {
+      "epoch": 0.23436294731193613,
+      "grad_norm": 0.15630906030636985,
+      "learning_rate": 2e-05,
+      "loss": 5.3829,
+      "step": 3494
+    },
+    {
+      "epoch": 0.23443002314116107,
+      "grad_norm": 0.1539356198777102,
+      "learning_rate": 2e-05,
+      "loss": 5.5334,
+      "step": 3495
+    },
+    {
+      "epoch": 0.234497098970386,
+      "grad_norm": 0.14489593999433611,
+      "learning_rate": 2e-05,
+      "loss": 5.4182,
+      "step": 3496
+    },
+    {
+      "epoch": 0.23456417479961095,
+      "grad_norm": 0.15279631393833318,
+      "learning_rate": 2e-05,
+      "loss": 5.5939,
+      "step": 3497
+    },
+    {
+      "epoch": 0.23463125062883589,
+      "grad_norm": 0.15048770041683063,
+      "learning_rate": 2e-05,
+      "loss": 5.4113,
+      "step": 3498
+    },
+    {
+      "epoch": 0.23469832645806082,
+      "grad_norm": 0.14277888894777233,
+      "learning_rate": 2e-05,
+      "loss": 5.4244,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2347654022872858,
+      "grad_norm": 0.14752158291177558,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 3500
+    },
+    {
+      "epoch": 0.23483247811651073,
+      "grad_norm": 0.1475969242931102,
+      "learning_rate": 2e-05,
+      "loss": 5.389,
+      "step": 3501
+    },
+    {
+      "epoch": 0.23489955394573567,
+      "grad_norm": 0.14429252207405874,
+      "learning_rate": 2e-05,
+      "loss": 5.5832,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2349666297749606,
+      "grad_norm": 0.1510452579535065,
+      "learning_rate": 2e-05,
+      "loss": 5.3752,
+      "step": 3503
+    },
+    {
+      "epoch": 0.23503370560418554,
+      "grad_norm": 0.14280845807719614,
+      "learning_rate": 2e-05,
+      "loss": 5.528,
+      "step": 3504
+    },
+    {
+      "epoch": 0.23510078143341048,
+      "grad_norm": 0.13904514598864562,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 3505
+    },
+    {
+      "epoch": 0.23516785726263542,
+      "grad_norm": 0.14999033017336383,
+      "learning_rate": 2e-05,
+      "loss": 5.495,
+      "step": 3506
+    },
+    {
+      "epoch": 0.23523493309186036,
+      "grad_norm": 0.14992065335260224,
+      "learning_rate": 2e-05,
+      "loss": 5.5619,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2353020089210853,
+      "grad_norm": 0.14270677132538787,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 3508
+    },
+    {
+      "epoch": 0.23536908475031024,
+      "grad_norm": 0.14028672998724132,
+      "learning_rate": 2e-05,
+      "loss": 5.4181,
+      "step": 3509
+    },
+    {
+      "epoch": 0.23543616057953518,
+      "grad_norm": 0.1438872156184744,
+      "learning_rate": 2e-05,
+      "loss": 5.509,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2355032364087601,
+      "grad_norm": 0.13445183921744125,
+      "learning_rate": 2e-05,
+      "loss": 5.3785,
+      "step": 3511
+    },
+    {
+      "epoch": 0.23557031223798505,
+      "grad_norm": 0.14351326430539219,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 3512
+    },
+    {
+      "epoch": 0.23563738806721,
+      "grad_norm": 0.1385464283831032,
+      "learning_rate": 2e-05,
+      "loss": 5.3699,
+      "step": 3513
+    },
+    {
+      "epoch": 0.23570446389643493,
+      "grad_norm": 0.1423727695269095,
+      "learning_rate": 2e-05,
+      "loss": 5.5019,
+      "step": 3514
+    },
+    {
+      "epoch": 0.23577153972565987,
+      "grad_norm": 0.14893616145305844,
+      "learning_rate": 2e-05,
+      "loss": 5.476,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2358386155548848,
+      "grad_norm": 0.14503258207101175,
+      "learning_rate": 2e-05,
+      "loss": 5.5196,
+      "step": 3516
+    },
+    {
+      "epoch": 0.23590569138410974,
+      "grad_norm": 0.144262530989285,
+      "learning_rate": 2e-05,
+      "loss": 5.2734,
+      "step": 3517
+    },
+    {
+      "epoch": 0.23597276721333468,
+      "grad_norm": 0.15243440179725587,
+      "learning_rate": 2e-05,
+      "loss": 5.6379,
+      "step": 3518
+    },
+    {
+      "epoch": 0.23603984304255962,
+      "grad_norm": 0.14688829945791226,
+      "learning_rate": 2e-05,
+      "loss": 5.3312,
+      "step": 3519
+    },
+    {
+      "epoch": 0.23610691887178456,
+      "grad_norm": 0.1408167230208686,
+      "learning_rate": 2e-05,
+      "loss": 5.5007,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2361739947010095,
+      "grad_norm": 0.14811484390425758,
+      "learning_rate": 2e-05,
+      "loss": 5.362,
+      "step": 3521
+    },
+    {
+      "epoch": 0.23624107053023444,
+      "grad_norm": 0.14487484193208983,
+      "learning_rate": 2e-05,
+      "loss": 5.2847,
+      "step": 3522
+    },
+    {
+      "epoch": 0.23630814635945938,
+      "grad_norm": 0.14225623478135174,
+      "learning_rate": 2e-05,
+      "loss": 5.524,
+      "step": 3523
+    },
+    {
+      "epoch": 0.23637522218868431,
+      "grad_norm": 0.14417705268103714,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 3524
+    },
+    {
+      "epoch": 0.23644229801790925,
+      "grad_norm": 0.15439279877724937,
+      "learning_rate": 2e-05,
+      "loss": 5.3358,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2365093738471342,
+      "grad_norm": 0.1520473111980663,
+      "learning_rate": 2e-05,
+      "loss": 5.3886,
+      "step": 3526
+    },
+    {
+      "epoch": 0.23657644967635913,
+      "grad_norm": 0.1477856534585931,
+      "learning_rate": 2e-05,
+      "loss": 5.5556,
+      "step": 3527
+    },
+    {
+      "epoch": 0.23664352550558407,
+      "grad_norm": 0.14843615579474798,
+      "learning_rate": 2e-05,
+      "loss": 5.4057,
+      "step": 3528
+    },
+    {
+      "epoch": 0.236710601334809,
+      "grad_norm": 0.1415872776402702,
+      "learning_rate": 2e-05,
+      "loss": 5.4979,
+      "step": 3529
+    },
+    {
+      "epoch": 0.23677767716403395,
+      "grad_norm": 0.14756182423521544,
+      "learning_rate": 2e-05,
+      "loss": 5.4503,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23684475299325888,
+      "grad_norm": 0.15226614404675812,
+      "learning_rate": 2e-05,
+      "loss": 5.408,
+      "step": 3531
+    },
+    {
+      "epoch": 0.23691182882248382,
+      "grad_norm": 0.14807275574227344,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 3532
+    },
+    {
+      "epoch": 0.23697890465170876,
+      "grad_norm": 0.1486057618220397,
+      "learning_rate": 2e-05,
+      "loss": 5.2641,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2370459804809337,
+      "grad_norm": 0.1436224862867708,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 3534
+    },
+    {
+      "epoch": 0.23711305631015864,
+      "grad_norm": 0.14886049371535018,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23718013213938358,
+      "grad_norm": 0.14543112434322694,
+      "learning_rate": 2e-05,
+      "loss": 5.4001,
+      "step": 3536
+    },
+    {
+      "epoch": 0.23724720796860851,
+      "grad_norm": 0.14177993838503603,
+      "learning_rate": 2e-05,
+      "loss": 5.5375,
+      "step": 3537
+    },
+    {
+      "epoch": 0.23731428379783345,
+      "grad_norm": 0.1429583396398804,
+      "learning_rate": 2e-05,
+      "loss": 5.4175,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2373813596270584,
+      "grad_norm": 0.14872975530012988,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 3539
+    },
+    {
+      "epoch": 0.23744843545628333,
+      "grad_norm": 0.1459126885075444,
+      "learning_rate": 2e-05,
+      "loss": 5.4484,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23751551128550827,
+      "grad_norm": 0.14597136972819438,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2375825871147332,
+      "grad_norm": 0.1432984832948743,
+      "learning_rate": 2e-05,
+      "loss": 5.3988,
+      "step": 3542
+    },
+    {
+      "epoch": 0.23764966294395815,
+      "grad_norm": 0.14140264500951494,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 3543
+    },
+    {
+      "epoch": 0.23771673877318308,
+      "grad_norm": 0.1431937122299652,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 3544
+    },
+    {
+      "epoch": 0.23778381460240802,
+      "grad_norm": 0.14400014551122753,
+      "learning_rate": 2e-05,
+      "loss": 5.5884,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23785089043163296,
+      "grad_norm": 0.14337834691045587,
+      "learning_rate": 2e-05,
+      "loss": 5.4924,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2379179662608579,
+      "grad_norm": 0.14423321464853942,
+      "learning_rate": 2e-05,
+      "loss": 5.2118,
+      "step": 3547
+    },
+    {
+      "epoch": 0.23798504209008284,
+      "grad_norm": 0.1494603828070573,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 3548
+    },
+    {
+      "epoch": 0.23805211791930778,
+      "grad_norm": 0.1468641727126983,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 3549
+    },
+    {
+      "epoch": 0.23811919374853271,
+      "grad_norm": 0.14535791353807256,
+      "learning_rate": 2e-05,
+      "loss": 5.35,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23818626957775765,
+      "grad_norm": 0.14726389618685923,
+      "learning_rate": 2e-05,
+      "loss": 5.4449,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2382533454069826,
+      "grad_norm": 0.14886412298754007,
+      "learning_rate": 2e-05,
+      "loss": 5.3301,
+      "step": 3552
+    },
+    {
+      "epoch": 0.23832042123620753,
+      "grad_norm": 0.1447393389411653,
+      "learning_rate": 2e-05,
+      "loss": 5.446,
+      "step": 3553
+    },
+    {
+      "epoch": 0.23838749706543247,
+      "grad_norm": 0.15149331545970063,
+      "learning_rate": 2e-05,
+      "loss": 5.398,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2384545728946574,
+      "grad_norm": 0.14306428571320842,
+      "learning_rate": 2e-05,
+      "loss": 5.4073,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23852164872388235,
+      "grad_norm": 0.14645303842261187,
+      "learning_rate": 2e-05,
+      "loss": 5.4945,
+      "step": 3556
+    },
+    {
+      "epoch": 0.23858872455310728,
+      "grad_norm": 0.15418742552398584,
+      "learning_rate": 2e-05,
+      "loss": 5.5669,
+      "step": 3557
+    },
+    {
+      "epoch": 0.23865580038233222,
+      "grad_norm": 0.15006469042220186,
+      "learning_rate": 2e-05,
+      "loss": 5.3745,
+      "step": 3558
+    },
+    {
+      "epoch": 0.23872287621155716,
+      "grad_norm": 0.1490717846176477,
+      "learning_rate": 2e-05,
+      "loss": 5.5104,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2387899520407821,
+      "grad_norm": 0.14782202119172577,
+      "learning_rate": 2e-05,
+      "loss": 5.6135,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23885702787000704,
+      "grad_norm": 0.13979660984578765,
+      "learning_rate": 2e-05,
+      "loss": 5.5384,
+      "step": 3561
+    },
+    {
+      "epoch": 0.23892410369923198,
+      "grad_norm": 0.15194676048056097,
+      "learning_rate": 2e-05,
+      "loss": 5.5315,
+      "step": 3562
+    },
+    {
+      "epoch": 0.23899117952845692,
+      "grad_norm": 0.15701889055919505,
+      "learning_rate": 2e-05,
+      "loss": 5.5611,
+      "step": 3563
+    },
+    {
+      "epoch": 0.23905825535768185,
+      "grad_norm": 0.14897405207653808,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2391253311869068,
+      "grad_norm": 0.14635998946054196,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23919240701613173,
+      "grad_norm": 0.14590361784241354,
+      "learning_rate": 2e-05,
+      "loss": 5.4474,
+      "step": 3566
+    },
+    {
+      "epoch": 0.23925948284535667,
+      "grad_norm": 0.1498224270900654,
+      "learning_rate": 2e-05,
+      "loss": 5.3986,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2393265586745816,
+      "grad_norm": 0.1407735369937029,
+      "learning_rate": 2e-05,
+      "loss": 5.4607,
+      "step": 3568
+    },
+    {
+      "epoch": 0.23939363450380655,
+      "grad_norm": 0.1483895539546799,
+      "learning_rate": 2e-05,
+      "loss": 5.4781,
+      "step": 3569
+    },
+    {
+      "epoch": 0.23946071033303148,
+      "grad_norm": 0.14651194215700625,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23952778616225642,
+      "grad_norm": 0.14731860878398914,
+      "learning_rate": 2e-05,
+      "loss": 5.3801,
+      "step": 3571
+    },
+    {
+      "epoch": 0.23959486199148136,
+      "grad_norm": 0.14426093008705018,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2396619378207063,
+      "grad_norm": 0.14651187314024125,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 3573
+    },
+    {
+      "epoch": 0.23972901364993124,
+      "grad_norm": 0.1376884545853337,
+      "learning_rate": 2e-05,
+      "loss": 5.3039,
+      "step": 3574
+    },
+    {
+      "epoch": 0.23979608947915618,
+      "grad_norm": 0.14828809284676006,
+      "learning_rate": 2e-05,
+      "loss": 5.557,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23986316530838112,
+      "grad_norm": 0.1508494931122979,
+      "learning_rate": 2e-05,
+      "loss": 5.4242,
+      "step": 3576
+    },
+    {
+      "epoch": 0.23993024113760605,
+      "grad_norm": 0.14495512147559056,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 3577
+    },
+    {
+      "epoch": 0.239997316966831,
+      "grad_norm": 0.13877851207772546,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 3578
+    },
+    {
+      "epoch": 0.24006439279605593,
+      "grad_norm": 0.14814119844113935,
+      "learning_rate": 2e-05,
+      "loss": 5.4427,
+      "step": 3579
+    },
+    {
+      "epoch": 0.24013146862528087,
+      "grad_norm": 0.15009500487518862,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2401985444545058,
+      "grad_norm": 0.14627368534510096,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 3581
+    },
+    {
+      "epoch": 0.24026562028373075,
+      "grad_norm": 0.1474676260741314,
+      "learning_rate": 2e-05,
+      "loss": 5.4589,
+      "step": 3582
+    },
+    {
+      "epoch": 0.24033269611295568,
+      "grad_norm": 0.14760527699882298,
+      "learning_rate": 2e-05,
+      "loss": 5.651,
+      "step": 3583
+    },
+    {
+      "epoch": 0.24039977194218062,
+      "grad_norm": 0.13822640472121228,
+      "learning_rate": 2e-05,
+      "loss": 5.5158,
+      "step": 3584
+    },
+    {
+      "epoch": 0.24046684777140556,
+      "grad_norm": 0.1476073540813708,
+      "learning_rate": 2e-05,
+      "loss": 5.5566,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2405339236006305,
+      "grad_norm": 0.15050068481690607,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 3586
+    },
+    {
+      "epoch": 0.24060099942985544,
+      "grad_norm": 0.15515377109165715,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 3587
+    },
+    {
+      "epoch": 0.24066807525908038,
+      "grad_norm": 0.14670988659683643,
+      "learning_rate": 2e-05,
+      "loss": 5.3674,
+      "step": 3588
+    },
+    {
+      "epoch": 0.24073515108830532,
+      "grad_norm": 0.14360318996635782,
+      "learning_rate": 2e-05,
+      "loss": 5.5265,
+      "step": 3589
+    },
+    {
+      "epoch": 0.24080222691753025,
+      "grad_norm": 0.14920994153156833,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 3590
+    },
+    {
+      "epoch": 0.24086930274675522,
+      "grad_norm": 0.15357917390173467,
+      "learning_rate": 2e-05,
+      "loss": 5.487,
+      "step": 3591
+    },
+    {
+      "epoch": 0.24093637857598016,
+      "grad_norm": 0.13866932594504833,
+      "learning_rate": 2e-05,
+      "loss": 5.394,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2410034544052051,
+      "grad_norm": 0.13928740049290605,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 3593
+    },
+    {
+      "epoch": 0.24107053023443004,
+      "grad_norm": 0.1559794287200337,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 3594
+    },
+    {
+      "epoch": 0.24113760606365497,
+      "grad_norm": 0.15810769698501753,
+      "learning_rate": 2e-05,
+      "loss": 5.4507,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2412046818928799,
+      "grad_norm": 0.14015399892917801,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 3596
+    },
+    {
+      "epoch": 0.24127175772210485,
+      "grad_norm": 0.14935159513319188,
+      "learning_rate": 2e-05,
+      "loss": 5.6543,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2413388335513298,
+      "grad_norm": 0.15259754328296427,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 3598
+    },
+    {
+      "epoch": 0.24140590938055473,
+      "grad_norm": 0.14248153915199116,
+      "learning_rate": 2e-05,
+      "loss": 5.5191,
+      "step": 3599
+    },
+    {
+      "epoch": 0.24147298520977967,
+      "grad_norm": 0.14719228140804352,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2415400610390046,
+      "grad_norm": 0.150872142997005,
+      "learning_rate": 2e-05,
+      "loss": 5.1958,
+      "step": 3601
+    },
+    {
+      "epoch": 0.24160713686822954,
+      "grad_norm": 0.15848387675462616,
+      "learning_rate": 2e-05,
+      "loss": 5.6504,
+      "step": 3602
+    },
+    {
+      "epoch": 0.24167421269745448,
+      "grad_norm": 0.1535823111291965,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 3603
+    },
+    {
+      "epoch": 0.24174128852667942,
+      "grad_norm": 0.15431040139636454,
+      "learning_rate": 2e-05,
+      "loss": 5.4069,
+      "step": 3604
+    },
+    {
+      "epoch": 0.24180836435590436,
+      "grad_norm": 0.14845076148491315,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2418754401851293,
+      "grad_norm": 0.1439580280563471,
+      "learning_rate": 2e-05,
+      "loss": 5.4982,
+      "step": 3606
+    },
+    {
+      "epoch": 0.24194251601435424,
+      "grad_norm": 0.15751179764710962,
+      "learning_rate": 2e-05,
+      "loss": 5.5652,
+      "step": 3607
+    },
+    {
+      "epoch": 0.24200959184357917,
+      "grad_norm": 0.16556794501088962,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2420766676728041,
+      "grad_norm": 0.1467700363359307,
+      "learning_rate": 2e-05,
+      "loss": 5.4524,
+      "step": 3609
+    },
+    {
+      "epoch": 0.24214374350202905,
+      "grad_norm": 0.14497654148201797,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 3610
+    },
+    {
+      "epoch": 0.242210819331254,
+      "grad_norm": 0.1544073697127881,
+      "learning_rate": 2e-05,
+      "loss": 5.5565,
+      "step": 3611
+    },
+    {
+      "epoch": 0.24227789516047893,
+      "grad_norm": 0.15412135702011834,
+      "learning_rate": 2e-05,
+      "loss": 5.5116,
+      "step": 3612
+    },
+    {
+      "epoch": 0.24234497098970387,
+      "grad_norm": 0.15051854975464043,
+      "learning_rate": 2e-05,
+      "loss": 5.5159,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2424120468189288,
+      "grad_norm": 0.1526509858189395,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 3614
+    },
+    {
+      "epoch": 0.24247912264815374,
+      "grad_norm": 0.15302755061037368,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 3615
+    },
+    {
+      "epoch": 0.24254619847737868,
+      "grad_norm": 0.15757455898302683,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 3616
+    },
+    {
+      "epoch": 0.24261327430660362,
+      "grad_norm": 0.14166734259692065,
+      "learning_rate": 2e-05,
+      "loss": 5.332,
+      "step": 3617
+    },
+    {
+      "epoch": 0.24268035013582856,
+      "grad_norm": 0.1460331856112316,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2427474259650535,
+      "grad_norm": 0.14436487902974549,
+      "learning_rate": 2e-05,
+      "loss": 5.3426,
+      "step": 3619
+    },
+    {
+      "epoch": 0.24281450179427844,
+      "grad_norm": 0.14817355376869543,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 3620
+    },
+    {
+      "epoch": 0.24288157762350338,
+      "grad_norm": 0.14938676055263458,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2429486534527283,
+      "grad_norm": 0.14366101523528207,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 3622
+    },
+    {
+      "epoch": 0.24301572928195325,
+      "grad_norm": 0.14507404526858197,
+      "learning_rate": 2e-05,
+      "loss": 5.3432,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2430828051111782,
+      "grad_norm": 0.14366351234079802,
+      "learning_rate": 2e-05,
+      "loss": 5.236,
+      "step": 3624
+    },
+    {
+      "epoch": 0.24314988094040313,
+      "grad_norm": 0.14065453590913157,
+      "learning_rate": 2e-05,
+      "loss": 5.3985,
+      "step": 3625
+    },
+    {
+      "epoch": 0.24321695676962807,
+      "grad_norm": 0.14310340429876883,
+      "learning_rate": 2e-05,
+      "loss": 5.5357,
+      "step": 3626
+    },
+    {
+      "epoch": 0.243284032598853,
+      "grad_norm": 0.14561869461475388,
+      "learning_rate": 2e-05,
+      "loss": 5.4521,
+      "step": 3627
+    },
+    {
+      "epoch": 0.24335110842807794,
+      "grad_norm": 0.14718491287197175,
+      "learning_rate": 2e-05,
+      "loss": 5.541,
+      "step": 3628
+    },
+    {
+      "epoch": 0.24341818425730288,
+      "grad_norm": 0.15811075863944035,
+      "learning_rate": 2e-05,
+      "loss": 5.5721,
+      "step": 3629
+    },
+    {
+      "epoch": 0.24348526008652782,
+      "grad_norm": 0.1437612483477177,
+      "learning_rate": 2e-05,
+      "loss": 5.4801,
+      "step": 3630
+    },
+    {
+      "epoch": 0.24355233591575276,
+      "grad_norm": 0.14965258242721233,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2436194117449777,
+      "grad_norm": 0.14638016257053468,
+      "learning_rate": 2e-05,
+      "loss": 5.4041,
+      "step": 3632
+    },
+    {
+      "epoch": 0.24368648757420264,
+      "grad_norm": 0.14798879041810833,
+      "learning_rate": 2e-05,
+      "loss": 5.5178,
+      "step": 3633
+    },
+    {
+      "epoch": 0.24375356340342758,
+      "grad_norm": 0.1456325835833653,
+      "learning_rate": 2e-05,
+      "loss": 5.4517,
+      "step": 3634
+    },
+    {
+      "epoch": 0.24382063923265251,
+      "grad_norm": 0.1512598891344156,
+      "learning_rate": 2e-05,
+      "loss": 5.5248,
+      "step": 3635
+    },
+    {
+      "epoch": 0.24388771506187745,
+      "grad_norm": 0.14296517565879271,
+      "learning_rate": 2e-05,
+      "loss": 5.2985,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2439547908911024,
+      "grad_norm": 0.14712023750578207,
+      "learning_rate": 2e-05,
+      "loss": 5.4917,
+      "step": 3637
+    },
+    {
+      "epoch": 0.24402186672032733,
+      "grad_norm": 0.14756282521171962,
+      "learning_rate": 2e-05,
+      "loss": 5.4272,
+      "step": 3638
+    },
+    {
+      "epoch": 0.24408894254955227,
+      "grad_norm": 0.14730008794004007,
+      "learning_rate": 2e-05,
+      "loss": 5.5766,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2441560183787772,
+      "grad_norm": 0.1396090166995398,
+      "learning_rate": 2e-05,
+      "loss": 5.4968,
+      "step": 3640
+    },
+    {
+      "epoch": 0.24422309420800214,
+      "grad_norm": 0.14251020257537156,
+      "learning_rate": 2e-05,
+      "loss": 5.5977,
+      "step": 3641
+    },
+    {
+      "epoch": 0.24429017003722708,
+      "grad_norm": 0.1464377714773839,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 3642
+    },
+    {
+      "epoch": 0.24435724586645202,
+      "grad_norm": 0.14620689406947984,
+      "learning_rate": 2e-05,
+      "loss": 5.3781,
+      "step": 3643
+    },
+    {
+      "epoch": 0.24442432169567696,
+      "grad_norm": 0.13980799334282612,
+      "learning_rate": 2e-05,
+      "loss": 5.4723,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2444913975249019,
+      "grad_norm": 0.15254023659266766,
+      "learning_rate": 2e-05,
+      "loss": 5.5534,
+      "step": 3645
+    },
+    {
+      "epoch": 0.24455847335412684,
+      "grad_norm": 0.14857320394254767,
+      "learning_rate": 2e-05,
+      "loss": 5.5112,
+      "step": 3646
+    },
+    {
+      "epoch": 0.24462554918335178,
+      "grad_norm": 0.14440664136098097,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 3647
+    },
+    {
+      "epoch": 0.24469262501257671,
+      "grad_norm": 0.14974671831282674,
+      "learning_rate": 2e-05,
+      "loss": 5.3042,
+      "step": 3648
+    },
+    {
+      "epoch": 0.24475970084180165,
+      "grad_norm": 0.149235735910126,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2448267766710266,
+      "grad_norm": 0.14814151198799683,
+      "learning_rate": 2e-05,
+      "loss": 5.4355,
+      "step": 3650
+    },
+    {
+      "epoch": 0.24489385250025153,
+      "grad_norm": 0.14496377003461322,
+      "learning_rate": 2e-05,
+      "loss": 5.4834,
+      "step": 3651
+    },
+    {
+      "epoch": 0.24496092832947647,
+      "grad_norm": 0.16068906904224686,
+      "learning_rate": 2e-05,
+      "loss": 5.5942,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2450280041587014,
+      "grad_norm": 0.14587251339580518,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 3653
+    },
+    {
+      "epoch": 0.24509507998792635,
+      "grad_norm": 0.14508859271347674,
+      "learning_rate": 2e-05,
+      "loss": 5.5362,
+      "step": 3654
+    },
+    {
+      "epoch": 0.24516215581715128,
+      "grad_norm": 0.14269287334823746,
+      "learning_rate": 2e-05,
+      "loss": 5.4556,
+      "step": 3655
+    },
+    {
+      "epoch": 0.24522923164637622,
+      "grad_norm": 0.14282063322235541,
+      "learning_rate": 2e-05,
+      "loss": 5.3507,
+      "step": 3656
+    },
+    {
+      "epoch": 0.24529630747560116,
+      "grad_norm": 0.14213918266620293,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2453633833048261,
+      "grad_norm": 0.1414017873713863,
+      "learning_rate": 2e-05,
+      "loss": 5.358,
+      "step": 3658
+    },
+    {
+      "epoch": 0.24543045913405104,
+      "grad_norm": 0.14491051464040294,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 3659
+    },
+    {
+      "epoch": 0.24549753496327598,
+      "grad_norm": 0.16307578897831285,
+      "learning_rate": 2e-05,
+      "loss": 5.332,
+      "step": 3660
+    },
+    {
+      "epoch": 0.24556461079250091,
+      "grad_norm": 0.15323985260479694,
+      "learning_rate": 2e-05,
+      "loss": 5.645,
+      "step": 3661
+    },
+    {
+      "epoch": 0.24563168662172585,
+      "grad_norm": 0.14618345101920524,
+      "learning_rate": 2e-05,
+      "loss": 5.5975,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2456987624509508,
+      "grad_norm": 0.15171265184919502,
+      "learning_rate": 2e-05,
+      "loss": 5.516,
+      "step": 3663
+    },
+    {
+      "epoch": 0.24576583828017573,
+      "grad_norm": 0.1459931451054656,
+      "learning_rate": 2e-05,
+      "loss": 5.4694,
+      "step": 3664
+    },
+    {
+      "epoch": 0.24583291410940067,
+      "grad_norm": 0.14177041467419058,
+      "learning_rate": 2e-05,
+      "loss": 5.5929,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2458999899386256,
+      "grad_norm": 0.14672452588147475,
+      "learning_rate": 2e-05,
+      "loss": 5.6084,
+      "step": 3666
+    },
+    {
+      "epoch": 0.24596706576785055,
+      "grad_norm": 0.14533926188747898,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 3667
+    },
+    {
+      "epoch": 0.24603414159707548,
+      "grad_norm": 0.14981871517150505,
+      "learning_rate": 2e-05,
+      "loss": 5.3662,
+      "step": 3668
+    },
+    {
+      "epoch": 0.24610121742630042,
+      "grad_norm": 0.141950703064853,
+      "learning_rate": 2e-05,
+      "loss": 5.5151,
+      "step": 3669
+    },
+    {
+      "epoch": 0.24616829325552536,
+      "grad_norm": 0.14813405072550195,
+      "learning_rate": 2e-05,
+      "loss": 5.407,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2462353690847503,
+      "grad_norm": 0.14502555308141019,
+      "learning_rate": 2e-05,
+      "loss": 5.4116,
+      "step": 3671
+    },
+    {
+      "epoch": 0.24630244491397524,
+      "grad_norm": 0.13977201120380367,
+      "learning_rate": 2e-05,
+      "loss": 5.3089,
+      "step": 3672
+    },
+    {
+      "epoch": 0.24636952074320018,
+      "grad_norm": 0.14439439972962667,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 3673
+    },
+    {
+      "epoch": 0.24643659657242512,
+      "grad_norm": 0.155415093590654,
+      "learning_rate": 2e-05,
+      "loss": 5.52,
+      "step": 3674
+    },
+    {
+      "epoch": 0.24650367240165005,
+      "grad_norm": 0.1443454422547581,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 3675
+    },
+    {
+      "epoch": 0.246570748230875,
+      "grad_norm": 0.14617138806972998,
+      "learning_rate": 2e-05,
+      "loss": 5.3552,
+      "step": 3676
+    },
+    {
+      "epoch": 0.24663782406009993,
+      "grad_norm": 0.1473838118538251,
+      "learning_rate": 2e-05,
+      "loss": 5.5768,
+      "step": 3677
+    },
+    {
+      "epoch": 0.24670489988932487,
+      "grad_norm": 0.14007642205061163,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2467719757185498,
+      "grad_norm": 0.14288371376793463,
+      "learning_rate": 2e-05,
+      "loss": 5.3388,
+      "step": 3679
+    },
+    {
+      "epoch": 0.24683905154777475,
+      "grad_norm": 0.15348546011205694,
+      "learning_rate": 2e-05,
+      "loss": 5.546,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24690612737699968,
+      "grad_norm": 0.1487894866190546,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 3681
+    },
+    {
+      "epoch": 0.24697320320622465,
+      "grad_norm": 0.15238782699231776,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2470402790354496,
+      "grad_norm": 0.15212350324453344,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 3683
+    },
+    {
+      "epoch": 0.24710735486467453,
+      "grad_norm": 0.1490143392798747,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 3684
+    },
+    {
+      "epoch": 0.24717443069389947,
+      "grad_norm": 0.1492434181391777,
+      "learning_rate": 2e-05,
+      "loss": 5.3927,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2472415065231244,
+      "grad_norm": 0.14856207578599978,
+      "learning_rate": 2e-05,
+      "loss": 5.4936,
+      "step": 3686
+    },
+    {
+      "epoch": 0.24730858235234934,
+      "grad_norm": 0.1433958767245302,
+      "learning_rate": 2e-05,
+      "loss": 5.3249,
+      "step": 3687
+    },
+    {
+      "epoch": 0.24737565818157428,
+      "grad_norm": 0.1459598563999363,
+      "learning_rate": 2e-05,
+      "loss": 5.4155,
+      "step": 3688
+    },
+    {
+      "epoch": 0.24744273401079922,
+      "grad_norm": 0.14893763962528928,
+      "learning_rate": 2e-05,
+      "loss": 5.3354,
+      "step": 3689
+    },
+    {
+      "epoch": 0.24750980984002416,
+      "grad_norm": 0.14281587981729854,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2475768856692491,
+      "grad_norm": 0.1454576184429072,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 3691
+    },
+    {
+      "epoch": 0.24764396149847404,
+      "grad_norm": 0.14992064152839124,
+      "learning_rate": 2e-05,
+      "loss": 5.4173,
+      "step": 3692
+    },
+    {
+      "epoch": 0.24771103732769897,
+      "grad_norm": 0.1457027203661565,
+      "learning_rate": 2e-05,
+      "loss": 5.4855,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2477781131569239,
+      "grad_norm": 0.13712480928021217,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24784518898614885,
+      "grad_norm": 0.1506826082913375,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2479122648153738,
+      "grad_norm": 0.14475299215034776,
+      "learning_rate": 2e-05,
+      "loss": 5.3777,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24797934064459873,
+      "grad_norm": 0.14241727883424193,
+      "learning_rate": 2e-05,
+      "loss": 5.4091,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24804641647382367,
+      "grad_norm": 0.15242285404685815,
+      "learning_rate": 2e-05,
+      "loss": 5.5338,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2481134923030486,
+      "grad_norm": 0.14355888793153426,
+      "learning_rate": 2e-05,
+      "loss": 5.3647,
+      "step": 3699
+    },
+    {
+      "epoch": 0.24818056813227354,
+      "grad_norm": 0.14509671238841956,
+      "learning_rate": 2e-05,
+      "loss": 5.5661,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24824764396149848,
+      "grad_norm": 0.15105971295539494,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 3701
+    },
+    {
+      "epoch": 0.24831471979072342,
+      "grad_norm": 0.14776018167086183,
+      "learning_rate": 2e-05,
+      "loss": 5.3371,
+      "step": 3702
+    },
+    {
+      "epoch": 0.24838179561994836,
+      "grad_norm": 0.14908279479261377,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2484488714491733,
+      "grad_norm": 0.14701820043350675,
+      "learning_rate": 2e-05,
+      "loss": 5.2194,
+      "step": 3704
+    },
+    {
+      "epoch": 0.24851594727839824,
+      "grad_norm": 0.14137802709287903,
+      "learning_rate": 2e-05,
+      "loss": 5.3789,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24858302310762317,
+      "grad_norm": 0.14528206783243755,
+      "learning_rate": 2e-05,
+      "loss": 5.5396,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2486500989368481,
+      "grad_norm": 0.1530842988147788,
+      "learning_rate": 2e-05,
+      "loss": 5.3475,
+      "step": 3707
+    },
+    {
+      "epoch": 0.24871717476607305,
+      "grad_norm": 0.14914422394800178,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 3708
+    },
+    {
+      "epoch": 0.248784250595298,
+      "grad_norm": 0.14537689798870795,
+      "learning_rate": 2e-05,
+      "loss": 5.4776,
+      "step": 3709
+    },
+    {
+      "epoch": 0.24885132642452293,
+      "grad_norm": 0.14557369011991786,
+      "learning_rate": 2e-05,
+      "loss": 5.4662,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24891840225374787,
+      "grad_norm": 0.15725903389508947,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2489854780829728,
+      "grad_norm": 0.14321110469564327,
+      "learning_rate": 2e-05,
+      "loss": 5.3905,
+      "step": 3712
+    },
+    {
+      "epoch": 0.24905255391219774,
+      "grad_norm": 0.14172155653544785,
+      "learning_rate": 2e-05,
+      "loss": 5.5528,
+      "step": 3713
+    },
+    {
+      "epoch": 0.24911962974142268,
+      "grad_norm": 0.14778521722156662,
+      "learning_rate": 2e-05,
+      "loss": 5.4289,
+      "step": 3714
+    },
+    {
+      "epoch": 0.24918670557064762,
+      "grad_norm": 0.14685067760891327,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24925378139987256,
+      "grad_norm": 0.14686093946460982,
+      "learning_rate": 2e-05,
+      "loss": 5.5111,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2493208572290975,
+      "grad_norm": 0.14921292697583893,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 3717
+    },
+    {
+      "epoch": 0.24938793305832244,
+      "grad_norm": 0.14574756178581724,
+      "learning_rate": 2e-05,
+      "loss": 5.4568,
+      "step": 3718
+    },
+    {
+      "epoch": 0.24945500888754737,
+      "grad_norm": 0.14294330856094223,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2495220847167723,
+      "grad_norm": 0.14732073835469947,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24958916054599725,
+      "grad_norm": 0.14214136761395885,
+      "learning_rate": 2e-05,
+      "loss": 5.5093,
+      "step": 3721
+    },
+    {
+      "epoch": 0.2496562363752222,
+      "grad_norm": 0.15016760130954998,
+      "learning_rate": 2e-05,
+      "loss": 5.4977,
+      "step": 3722
+    },
+    {
+      "epoch": 0.24972331220444713,
+      "grad_norm": 0.1449032185486146,
+      "learning_rate": 2e-05,
+      "loss": 5.2996,
+      "step": 3723
+    },
+    {
+      "epoch": 0.24979038803367207,
+      "grad_norm": 0.14698770217095639,
+      "learning_rate": 2e-05,
+      "loss": 5.4727,
+      "step": 3724
+    },
+    {
+      "epoch": 0.249857463862897,
+      "grad_norm": 0.14735703440715242,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 3725
+    },
+    {
+      "epoch": 0.24992453969212194,
+      "grad_norm": 0.14584714101193985,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 3726
+    },
+    {
+      "epoch": 0.24999161552134688,
+      "grad_norm": 0.14714183157184782,
+      "learning_rate": 2e-05,
+      "loss": 5.3559,
+      "step": 3727
+    },
+    {
+      "epoch": 0.25005869135057185,
+      "grad_norm": 0.14185319776475563,
+      "learning_rate": 2e-05,
+      "loss": 5.6197,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2501257671797968,
+      "grad_norm": 0.14445518904789215,
+      "learning_rate": 2e-05,
+      "loss": 5.6117,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2501928430090217,
+      "grad_norm": 0.14256658378959786,
+      "learning_rate": 2e-05,
+      "loss": 5.5514,
+      "step": 3730
+    },
+    {
+      "epoch": 0.25025991883824666,
+      "grad_norm": 0.1456064586389974,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2503269946674716,
+      "grad_norm": 0.1469171011547121,
+      "learning_rate": 2e-05,
+      "loss": 5.4971,
+      "step": 3732
+    },
+    {
+      "epoch": 0.25039407049669654,
+      "grad_norm": 0.15165530625035067,
+      "learning_rate": 2e-05,
+      "loss": 5.3415,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2504611463259215,
+      "grad_norm": 0.14581348188795631,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2505282221551464,
+      "grad_norm": 0.14654471430777896,
+      "learning_rate": 2e-05,
+      "loss": 5.5605,
+      "step": 3735
+    },
+    {
+      "epoch": 0.25059529798437136,
+      "grad_norm": 0.14291705656766163,
+      "learning_rate": 2e-05,
+      "loss": 5.5325,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2506623738135963,
+      "grad_norm": 0.1570339929073719,
+      "learning_rate": 2e-05,
+      "loss": 5.532,
+      "step": 3737
+    },
+    {
+      "epoch": 0.25072944964282123,
+      "grad_norm": 0.1481393537747991,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2507965254720462,
+      "grad_norm": 0.14942812086890742,
+      "learning_rate": 2e-05,
+      "loss": 5.5956,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2508636013012711,
+      "grad_norm": 0.15728490446844373,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 3740
+    },
+    {
+      "epoch": 0.25093067713049605,
+      "grad_norm": 0.14820747106245247,
+      "learning_rate": 2e-05,
+      "loss": 5.5815,
+      "step": 3741
+    },
+    {
+      "epoch": 0.250997752959721,
+      "grad_norm": 0.15508411824280052,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2510648287889459,
+      "grad_norm": 0.1558609173490631,
+      "learning_rate": 2e-05,
+      "loss": 5.3829,
+      "step": 3743
+    },
+    {
+      "epoch": 0.25113190461817086,
+      "grad_norm": 0.15560234145068735,
+      "learning_rate": 2e-05,
+      "loss": 5.3069,
+      "step": 3744
+    },
+    {
+      "epoch": 0.2511989804473958,
+      "grad_norm": 0.1491312291805695,
+      "learning_rate": 2e-05,
+      "loss": 5.3313,
+      "step": 3745
+    },
+    {
+      "epoch": 0.25126605627662074,
+      "grad_norm": 0.1630677209260026,
+      "learning_rate": 2e-05,
+      "loss": 5.4552,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2513331321058457,
+      "grad_norm": 0.1537602319057921,
+      "learning_rate": 2e-05,
+      "loss": 5.4446,
+      "step": 3747
+    },
+    {
+      "epoch": 0.2514002079350706,
+      "grad_norm": 0.14658220678469497,
+      "learning_rate": 2e-05,
+      "loss": 5.3901,
+      "step": 3748
+    },
+    {
+      "epoch": 0.25146728376429556,
+      "grad_norm": 0.14897090348839365,
+      "learning_rate": 2e-05,
+      "loss": 5.3604,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2515343595935205,
+      "grad_norm": 0.16215945255206385,
+      "learning_rate": 2e-05,
+      "loss": 5.4713,
+      "step": 3750
+    },
+    {
+      "epoch": 0.25160143542274543,
+      "grad_norm": 0.1416117342497126,
+      "learning_rate": 2e-05,
+      "loss": 5.2477,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2516685112519704,
+      "grad_norm": 0.14825452416081117,
+      "learning_rate": 2e-05,
+      "loss": 5.5632,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2517355870811953,
+      "grad_norm": 0.1604543401766358,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 3753
+    },
+    {
+      "epoch": 0.25180266291042025,
+      "grad_norm": 0.14660550400353697,
+      "learning_rate": 2e-05,
+      "loss": 5.561,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2518697387396452,
+      "grad_norm": 0.14069657923597087,
+      "learning_rate": 2e-05,
+      "loss": 5.4712,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2519368145688701,
+      "grad_norm": 0.1474505129365506,
+      "learning_rate": 2e-05,
+      "loss": 5.5361,
+      "step": 3756
+    },
+    {
+      "epoch": 0.25200389039809507,
+      "grad_norm": 0.15000483023731537,
+      "learning_rate": 2e-05,
+      "loss": 5.4254,
+      "step": 3757
+    },
+    {
+      "epoch": 0.25207096622732,
+      "grad_norm": 0.1510018599640942,
+      "learning_rate": 2e-05,
+      "loss": 5.3964,
+      "step": 3758
+    },
+    {
+      "epoch": 0.25213804205654494,
+      "grad_norm": 0.14974894808130063,
+      "learning_rate": 2e-05,
+      "loss": 5.3482,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2522051178857699,
+      "grad_norm": 0.14453146224163974,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2522721937149948,
+      "grad_norm": 0.15149553743198957,
+      "learning_rate": 2e-05,
+      "loss": 5.3518,
+      "step": 3761
+    },
+    {
+      "epoch": 0.25233926954421976,
+      "grad_norm": 0.15140957033331257,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2524063453734447,
+      "grad_norm": 0.15756766371317613,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 3763
+    },
+    {
+      "epoch": 0.25247342120266963,
+      "grad_norm": 0.1522980868889702,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 3764
+    },
+    {
+      "epoch": 0.2525404970318946,
+      "grad_norm": 0.14788416174679844,
+      "learning_rate": 2e-05,
+      "loss": 5.4031,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2526075728611195,
+      "grad_norm": 0.14981441306040225,
+      "learning_rate": 2e-05,
+      "loss": 5.6343,
+      "step": 3766
+    },
+    {
+      "epoch": 0.25267464869034445,
+      "grad_norm": 0.16060534549702618,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2527417245195694,
+      "grad_norm": 0.14968009800683948,
+      "learning_rate": 2e-05,
+      "loss": 5.337,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2528088003487943,
+      "grad_norm": 0.1487903883699617,
+      "learning_rate": 2e-05,
+      "loss": 5.5062,
+      "step": 3769
+    },
+    {
+      "epoch": 0.25287587617801927,
+      "grad_norm": 0.15088994900414518,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2529429520072442,
+      "grad_norm": 0.17307729005810243,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 3771
+    },
+    {
+      "epoch": 0.25301002783646914,
+      "grad_norm": 0.14467965448648515,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2530771036656941,
+      "grad_norm": 0.1570833072890613,
+      "learning_rate": 2e-05,
+      "loss": 5.5496,
+      "step": 3773
+    },
+    {
+      "epoch": 0.253144179494919,
+      "grad_norm": 0.15385072835908956,
+      "learning_rate": 2e-05,
+      "loss": 5.4976,
+      "step": 3774
+    },
+    {
+      "epoch": 0.25321125532414396,
+      "grad_norm": 0.15863042549085676,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2532783311533689,
+      "grad_norm": 0.1495700043424086,
+      "learning_rate": 2e-05,
+      "loss": 5.3646,
+      "step": 3776
+    },
+    {
+      "epoch": 0.25334540698259383,
+      "grad_norm": 0.15822225013416757,
+      "learning_rate": 2e-05,
+      "loss": 5.3834,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2534124828118188,
+      "grad_norm": 0.151028509168438,
+      "learning_rate": 2e-05,
+      "loss": 5.5125,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2534795586410437,
+      "grad_norm": 0.1580718401617811,
+      "learning_rate": 2e-05,
+      "loss": 5.4318,
+      "step": 3779
+    },
+    {
+      "epoch": 0.25354663447026865,
+      "grad_norm": 0.15241922542955663,
+      "learning_rate": 2e-05,
+      "loss": 5.5999,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2536137102994936,
+      "grad_norm": 0.15025556110340946,
+      "learning_rate": 2e-05,
+      "loss": 5.4019,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2536807861287185,
+      "grad_norm": 0.1597545137971037,
+      "learning_rate": 2e-05,
+      "loss": 5.4116,
+      "step": 3782
+    },
+    {
+      "epoch": 0.25374786195794347,
+      "grad_norm": 0.14521459928213876,
+      "learning_rate": 2e-05,
+      "loss": 5.3941,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2538149377871684,
+      "grad_norm": 0.152505074989981,
+      "learning_rate": 2e-05,
+      "loss": 5.3728,
+      "step": 3784
+    },
+    {
+      "epoch": 0.25388201361639334,
+      "grad_norm": 0.14091224865268126,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2539490894456183,
+      "grad_norm": 0.14637375470173636,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2540161652748432,
+      "grad_norm": 0.1469739070043092,
+      "learning_rate": 2e-05,
+      "loss": 5.5227,
+      "step": 3787
+    },
+    {
+      "epoch": 0.25408324110406816,
+      "grad_norm": 0.14646275326756972,
+      "learning_rate": 2e-05,
+      "loss": 5.5411,
+      "step": 3788
+    },
+    {
+      "epoch": 0.2541503169332931,
+      "grad_norm": 0.15032197446633735,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 3789
+    },
+    {
+      "epoch": 0.25421739276251804,
+      "grad_norm": 0.14617054790420025,
+      "learning_rate": 2e-05,
+      "loss": 5.3097,
+      "step": 3790
+    },
+    {
+      "epoch": 0.254284468591743,
+      "grad_norm": 0.1461232428379103,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2543515444209679,
+      "grad_norm": 0.14750983523767622,
+      "learning_rate": 2e-05,
+      "loss": 5.3047,
+      "step": 3792
+    },
+    {
+      "epoch": 0.25441862025019285,
+      "grad_norm": 0.14369089058307205,
+      "learning_rate": 2e-05,
+      "loss": 5.4326,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2544856960794178,
+      "grad_norm": 0.14370718070951505,
+      "learning_rate": 2e-05,
+      "loss": 5.4828,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2545527719086427,
+      "grad_norm": 0.15415262987685194,
+      "learning_rate": 2e-05,
+      "loss": 5.4204,
+      "step": 3795
+    },
+    {
+      "epoch": 0.25461984773786767,
+      "grad_norm": 0.1418703521584847,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2546869235670926,
+      "grad_norm": 0.14686268848231318,
+      "learning_rate": 2e-05,
+      "loss": 5.4446,
+      "step": 3797
+    },
+    {
+      "epoch": 0.25475399939631754,
+      "grad_norm": 0.14763250521108806,
+      "learning_rate": 2e-05,
+      "loss": 5.3383,
+      "step": 3798
+    },
+    {
+      "epoch": 0.2548210752255425,
+      "grad_norm": 0.14658773401178146,
+      "learning_rate": 2e-05,
+      "loss": 5.631,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2548881510547674,
+      "grad_norm": 0.13927429931640617,
+      "learning_rate": 2e-05,
+      "loss": 5.4475,
+      "step": 3800
+    },
+    {
+      "epoch": 0.25495522688399236,
+      "grad_norm": 0.15900578915114108,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 3801
+    },
+    {
+      "epoch": 0.2550223027132173,
+      "grad_norm": 0.1440663609132752,
+      "learning_rate": 2e-05,
+      "loss": 5.2744,
+      "step": 3802
+    },
+    {
+      "epoch": 0.25508937854244224,
+      "grad_norm": 0.14750285643092637,
+      "learning_rate": 2e-05,
+      "loss": 5.3756,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2551564543716672,
+      "grad_norm": 0.15487437217489824,
+      "learning_rate": 2e-05,
+      "loss": 5.5675,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2552235302008921,
+      "grad_norm": 0.14333348006163538,
+      "learning_rate": 2e-05,
+      "loss": 5.3834,
+      "step": 3805
+    },
+    {
+      "epoch": 0.25529060603011705,
+      "grad_norm": 0.15086835385834996,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 3806
+    },
+    {
+      "epoch": 0.255357681859342,
+      "grad_norm": 0.15251924229132585,
+      "learning_rate": 2e-05,
+      "loss": 5.4025,
+      "step": 3807
+    },
+    {
+      "epoch": 0.25542475768856693,
+      "grad_norm": 0.1491293072520056,
+      "learning_rate": 2e-05,
+      "loss": 5.3184,
+      "step": 3808
+    },
+    {
+      "epoch": 0.25549183351779187,
+      "grad_norm": 0.15171308682643084,
+      "learning_rate": 2e-05,
+      "loss": 5.4826,
+      "step": 3809
+    },
+    {
+      "epoch": 0.2555589093470168,
+      "grad_norm": 0.14670873216509572,
+      "learning_rate": 2e-05,
+      "loss": 5.6444,
+      "step": 3810
+    },
+    {
+      "epoch": 0.25562598517624174,
+      "grad_norm": 0.1461583902989626,
+      "learning_rate": 2e-05,
+      "loss": 5.5252,
+      "step": 3811
+    },
+    {
+      "epoch": 0.2556930610054667,
+      "grad_norm": 0.1552298614685276,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2557601368346916,
+      "grad_norm": 0.15263088244554182,
+      "learning_rate": 2e-05,
+      "loss": 5.5625,
+      "step": 3813
+    },
+    {
+      "epoch": 0.25582721266391656,
+      "grad_norm": 0.14703879994773886,
+      "learning_rate": 2e-05,
+      "loss": 5.3791,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2558942884931415,
+      "grad_norm": 0.14592848941260114,
+      "learning_rate": 2e-05,
+      "loss": 5.5148,
+      "step": 3815
+    },
+    {
+      "epoch": 0.25596136432236644,
+      "grad_norm": 0.155792664196072,
+      "learning_rate": 2e-05,
+      "loss": 5.4763,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2560284401515914,
+      "grad_norm": 0.14735033733964242,
+      "learning_rate": 2e-05,
+      "loss": 5.6019,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2560955159808163,
+      "grad_norm": 0.13735293431396386,
+      "learning_rate": 2e-05,
+      "loss": 5.4016,
+      "step": 3818
+    },
+    {
+      "epoch": 0.25616259181004125,
+      "grad_norm": 0.15081874571023168,
+      "learning_rate": 2e-05,
+      "loss": 5.5326,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2562296676392662,
+      "grad_norm": 0.1447062918806684,
+      "learning_rate": 2e-05,
+      "loss": 5.6289,
+      "step": 3820
+    },
+    {
+      "epoch": 0.25629674346849113,
+      "grad_norm": 0.14860696124792633,
+      "learning_rate": 2e-05,
+      "loss": 5.6646,
+      "step": 3821
+    },
+    {
+      "epoch": 0.25636381929771607,
+      "grad_norm": 0.14705168204088293,
+      "learning_rate": 2e-05,
+      "loss": 5.3814,
+      "step": 3822
+    },
+    {
+      "epoch": 0.256430895126941,
+      "grad_norm": 0.1504028860838788,
+      "learning_rate": 2e-05,
+      "loss": 5.5603,
+      "step": 3823
+    },
+    {
+      "epoch": 0.25649797095616594,
+      "grad_norm": 0.1506198426992679,
+      "learning_rate": 2e-05,
+      "loss": 5.5622,
+      "step": 3824
+    },
+    {
+      "epoch": 0.2565650467853909,
+      "grad_norm": 0.15015485439058282,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2566321226146158,
+      "grad_norm": 0.14783799152041394,
+      "learning_rate": 2e-05,
+      "loss": 5.4901,
+      "step": 3826
+    },
+    {
+      "epoch": 0.25669919844384076,
+      "grad_norm": 0.14556315217727664,
+      "learning_rate": 2e-05,
+      "loss": 5.5117,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2567662742730657,
+      "grad_norm": 0.14748817661083055,
+      "learning_rate": 2e-05,
+      "loss": 5.3964,
+      "step": 3828
+    },
+    {
+      "epoch": 0.25683335010229064,
+      "grad_norm": 0.1463040750964665,
+      "learning_rate": 2e-05,
+      "loss": 5.4465,
+      "step": 3829
+    },
+    {
+      "epoch": 0.2569004259315156,
+      "grad_norm": 0.15307181328508748,
+      "learning_rate": 2e-05,
+      "loss": 5.4208,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2569675017607405,
+      "grad_norm": 0.1454127697673739,
+      "learning_rate": 2e-05,
+      "loss": 5.3367,
+      "step": 3831
+    },
+    {
+      "epoch": 0.25703457758996545,
+      "grad_norm": 0.14649840510161494,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 3832
+    },
+    {
+      "epoch": 0.2571016534191904,
+      "grad_norm": 0.1548959519444856,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 3833
+    },
+    {
+      "epoch": 0.25716872924841533,
+      "grad_norm": 0.1565744244466912,
+      "learning_rate": 2e-05,
+      "loss": 5.4358,
+      "step": 3834
+    },
+    {
+      "epoch": 0.25723580507764027,
+      "grad_norm": 0.14936729323215645,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2573028809068652,
+      "grad_norm": 0.1491608006987523,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 3836
+    },
+    {
+      "epoch": 0.25736995673609014,
+      "grad_norm": 0.14811164173008498,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2574370325653151,
+      "grad_norm": 0.14763781029501316,
+      "learning_rate": 2e-05,
+      "loss": 5.425,
+      "step": 3838
+    },
+    {
+      "epoch": 0.25750410839454,
+      "grad_norm": 0.14470296608904837,
+      "learning_rate": 2e-05,
+      "loss": 5.4426,
+      "step": 3839
+    },
+    {
+      "epoch": 0.25757118422376496,
+      "grad_norm": 0.1426215585856793,
+      "learning_rate": 2e-05,
+      "loss": 5.4508,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2576382600529899,
+      "grad_norm": 0.15617763765753245,
+      "learning_rate": 2e-05,
+      "loss": 5.4295,
+      "step": 3841
+    },
+    {
+      "epoch": 0.25770533588221484,
+      "grad_norm": 0.1469652035429382,
+      "learning_rate": 2e-05,
+      "loss": 5.4878,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2577724117114398,
+      "grad_norm": 0.14025767234877837,
+      "learning_rate": 2e-05,
+      "loss": 5.4263,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2578394875406647,
+      "grad_norm": 0.14460337641806692,
+      "learning_rate": 2e-05,
+      "loss": 5.5263,
+      "step": 3844
+    },
+    {
+      "epoch": 0.25790656336988965,
+      "grad_norm": 0.14428370616266806,
+      "learning_rate": 2e-05,
+      "loss": 5.6108,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2579736391991146,
+      "grad_norm": 0.14754051544332777,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 3846
+    },
+    {
+      "epoch": 0.25804071502833953,
+      "grad_norm": 0.1417889974685112,
+      "learning_rate": 2e-05,
+      "loss": 5.4731,
+      "step": 3847
+    },
+    {
+      "epoch": 0.25810779085756447,
+      "grad_norm": 0.1405041118169406,
+      "learning_rate": 2e-05,
+      "loss": 5.4093,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2581748666867894,
+      "grad_norm": 0.14858446476047155,
+      "learning_rate": 2e-05,
+      "loss": 5.5142,
+      "step": 3849
+    },
+    {
+      "epoch": 0.25824194251601434,
+      "grad_norm": 0.14073322544613465,
+      "learning_rate": 2e-05,
+      "loss": 5.4722,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2583090183452393,
+      "grad_norm": 0.14826431672421808,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2583760941744642,
+      "grad_norm": 0.14755538902732437,
+      "learning_rate": 2e-05,
+      "loss": 5.5777,
+      "step": 3852
+    },
+    {
+      "epoch": 0.25844317000368916,
+      "grad_norm": 0.14771586463461778,
+      "learning_rate": 2e-05,
+      "loss": 5.5744,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2585102458329141,
+      "grad_norm": 0.14775919043476643,
+      "learning_rate": 2e-05,
+      "loss": 5.5132,
+      "step": 3854
+    },
+    {
+      "epoch": 0.25857732166213904,
+      "grad_norm": 0.1410593424589937,
+      "learning_rate": 2e-05,
+      "loss": 5.5138,
+      "step": 3855
+    },
+    {
+      "epoch": 0.258644397491364,
+      "grad_norm": 0.1505672524848943,
+      "learning_rate": 2e-05,
+      "loss": 5.2929,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2587114733205889,
+      "grad_norm": 0.15266885993994908,
+      "learning_rate": 2e-05,
+      "loss": 5.4722,
+      "step": 3857
+    },
+    {
+      "epoch": 0.25877854914981385,
+      "grad_norm": 0.1595065505652384,
+      "learning_rate": 2e-05,
+      "loss": 5.4876,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2588456249790388,
+      "grad_norm": 0.1475495976306526,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 3859
+    },
+    {
+      "epoch": 0.25891270080826373,
+      "grad_norm": 0.15039833089015914,
+      "learning_rate": 2e-05,
+      "loss": 5.5299,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25897977663748867,
+      "grad_norm": 0.15212458560309677,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2590468524667136,
+      "grad_norm": 0.14830852614970091,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 3862
+    },
+    {
+      "epoch": 0.25911392829593854,
+      "grad_norm": 0.1472576863894037,
+      "learning_rate": 2e-05,
+      "loss": 5.4509,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2591810041251635,
+      "grad_norm": 0.16413465281062362,
+      "learning_rate": 2e-05,
+      "loss": 5.3765,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2592480799543884,
+      "grad_norm": 0.15304029171526892,
+      "learning_rate": 2e-05,
+      "loss": 5.5081,
+      "step": 3865
+    },
+    {
+      "epoch": 0.25931515578361336,
+      "grad_norm": 0.1490416272033554,
+      "learning_rate": 2e-05,
+      "loss": 5.4831,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2593822316128383,
+      "grad_norm": 0.15357769296085355,
+      "learning_rate": 2e-05,
+      "loss": 5.5006,
+      "step": 3867
+    },
+    {
+      "epoch": 0.25944930744206324,
+      "grad_norm": 0.15082440023242194,
+      "learning_rate": 2e-05,
+      "loss": 5.3,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2595163832712882,
+      "grad_norm": 0.1406917571594882,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 3869
+    },
+    {
+      "epoch": 0.2595834591005131,
+      "grad_norm": 0.1417999790171647,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25965053492973805,
+      "grad_norm": 0.1576180265007791,
+      "learning_rate": 2e-05,
+      "loss": 5.3812,
+      "step": 3871
+    },
+    {
+      "epoch": 0.259717610758963,
+      "grad_norm": 0.1435934273525419,
+      "learning_rate": 2e-05,
+      "loss": 5.5301,
+      "step": 3872
+    },
+    {
+      "epoch": 0.25978468658818793,
+      "grad_norm": 0.14153321502770955,
+      "learning_rate": 2e-05,
+      "loss": 5.3181,
+      "step": 3873
+    },
+    {
+      "epoch": 0.25985176241741287,
+      "grad_norm": 0.1484391906080592,
+      "learning_rate": 2e-05,
+      "loss": 5.3863,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2599188382466378,
+      "grad_norm": 0.14866640606694945,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 3875
+    },
+    {
+      "epoch": 0.25998591407586275,
+      "grad_norm": 0.14430413833194114,
+      "learning_rate": 2e-05,
+      "loss": 5.4291,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2600529899050877,
+      "grad_norm": 0.15408431081341858,
+      "learning_rate": 2e-05,
+      "loss": 5.471,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2601200657343126,
+      "grad_norm": 0.156422461148273,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 3878
+    },
+    {
+      "epoch": 0.26018714156353756,
+      "grad_norm": 0.1457714697220853,
+      "learning_rate": 2e-05,
+      "loss": 5.414,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2602542173927625,
+      "grad_norm": 0.1429113153140056,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 3880
+    },
+    {
+      "epoch": 0.26032129322198744,
+      "grad_norm": 0.1520664443222717,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2603883690512124,
+      "grad_norm": 0.14763482164581493,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2604554448804373,
+      "grad_norm": 0.1437934131799041,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 3883
+    },
+    {
+      "epoch": 0.26052252070966225,
+      "grad_norm": 0.14536924910016258,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2605895965388872,
+      "grad_norm": 0.14392646022379038,
+      "learning_rate": 2e-05,
+      "loss": 5.443,
+      "step": 3885
+    },
+    {
+      "epoch": 0.26065667236811213,
+      "grad_norm": 0.1490503659274492,
+      "learning_rate": 2e-05,
+      "loss": 5.4564,
+      "step": 3886
+    },
+    {
+      "epoch": 0.26072374819733707,
+      "grad_norm": 0.15401581298820416,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 3887
+    },
+    {
+      "epoch": 0.260790824026562,
+      "grad_norm": 0.1451792881645716,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 3888
+    },
+    {
+      "epoch": 0.26085789985578695,
+      "grad_norm": 0.14623125135775142,
+      "learning_rate": 2e-05,
+      "loss": 5.5564,
+      "step": 3889
+    },
+    {
+      "epoch": 0.2609249756850119,
+      "grad_norm": 0.16040584855767342,
+      "learning_rate": 2e-05,
+      "loss": 5.4483,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2609920515142368,
+      "grad_norm": 0.14454050225788131,
+      "learning_rate": 2e-05,
+      "loss": 5.5406,
+      "step": 3891
+    },
+    {
+      "epoch": 0.26105912734346176,
+      "grad_norm": 0.15719706931169644,
+      "learning_rate": 2e-05,
+      "loss": 5.5034,
+      "step": 3892
+    },
+    {
+      "epoch": 0.2611262031726867,
+      "grad_norm": 0.1535580874405099,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 3893
+    },
+    {
+      "epoch": 0.26119327900191164,
+      "grad_norm": 0.14711019299757844,
+      "learning_rate": 2e-05,
+      "loss": 5.3776,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2612603548311366,
+      "grad_norm": 0.15194779175487633,
+      "learning_rate": 2e-05,
+      "loss": 5.3385,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2613274306603615,
+      "grad_norm": 0.14930527392892204,
+      "learning_rate": 2e-05,
+      "loss": 5.4857,
+      "step": 3896
+    },
+    {
+      "epoch": 0.26139450648958645,
+      "grad_norm": 0.1413100453519304,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 3897
+    },
+    {
+      "epoch": 0.2614615823188114,
+      "grad_norm": 0.15073416321804295,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 3898
+    },
+    {
+      "epoch": 0.26152865814803633,
+      "grad_norm": 0.1419101693912533,
+      "learning_rate": 2e-05,
+      "loss": 5.3436,
+      "step": 3899
+    },
+    {
+      "epoch": 0.26159573397726127,
+      "grad_norm": 0.1483493834825796,
+      "learning_rate": 2e-05,
+      "loss": 5.4655,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2616628098064862,
+      "grad_norm": 0.1416133825578569,
+      "learning_rate": 2e-05,
+      "loss": 5.3605,
+      "step": 3901
+    },
+    {
+      "epoch": 0.26172988563571115,
+      "grad_norm": 0.14401923491992427,
+      "learning_rate": 2e-05,
+      "loss": 5.5828,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2617969614649361,
+      "grad_norm": 0.15144671099689408,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 3903
+    },
+    {
+      "epoch": 0.261864037294161,
+      "grad_norm": 0.14393306547585147,
+      "learning_rate": 2e-05,
+      "loss": 5.4686,
+      "step": 3904
+    },
+    {
+      "epoch": 0.26193111312338596,
+      "grad_norm": 0.14233292188288868,
+      "learning_rate": 2e-05,
+      "loss": 5.3599,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2619981889526109,
+      "grad_norm": 0.14665231024257255,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 3906
+    },
+    {
+      "epoch": 0.26206526478183584,
+      "grad_norm": 0.15419678874675444,
+      "learning_rate": 2e-05,
+      "loss": 5.4649,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2621323406110608,
+      "grad_norm": 0.1551897951298403,
+      "learning_rate": 2e-05,
+      "loss": 5.4885,
+      "step": 3908
+    },
+    {
+      "epoch": 0.26219941644028577,
+      "grad_norm": 0.1504294017000213,
+      "learning_rate": 2e-05,
+      "loss": 5.5096,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2622664922695107,
+      "grad_norm": 0.1515171634727965,
+      "learning_rate": 2e-05,
+      "loss": 5.4202,
+      "step": 3910
+    },
+    {
+      "epoch": 0.26233356809873565,
+      "grad_norm": 0.1485661533598932,
+      "learning_rate": 2e-05,
+      "loss": 5.3466,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2624006439279606,
+      "grad_norm": 0.15183025941885792,
+      "learning_rate": 2e-05,
+      "loss": 5.5035,
+      "step": 3912
+    },
+    {
+      "epoch": 0.2624677197571855,
+      "grad_norm": 0.14732117833667932,
+      "learning_rate": 2e-05,
+      "loss": 5.1908,
+      "step": 3913
+    },
+    {
+      "epoch": 0.26253479558641046,
+      "grad_norm": 0.14970898733761662,
+      "learning_rate": 2e-05,
+      "loss": 5.4572,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2626018714156354,
+      "grad_norm": 0.1570667863469998,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 3915
+    },
+    {
+      "epoch": 0.26266894724486034,
+      "grad_norm": 0.15037231996967612,
+      "learning_rate": 2e-05,
+      "loss": 5.602,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2627360230740853,
+      "grad_norm": 0.14599062190341205,
+      "learning_rate": 2e-05,
+      "loss": 5.5927,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2628030989033102,
+      "grad_norm": 0.15241655082673747,
+      "learning_rate": 2e-05,
+      "loss": 5.3781,
+      "step": 3918
+    },
+    {
+      "epoch": 0.26287017473253516,
+      "grad_norm": 0.14834906260284883,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 3919
+    },
+    {
+      "epoch": 0.2629372505617601,
+      "grad_norm": 0.1388934989012939,
+      "learning_rate": 2e-05,
+      "loss": 5.4115,
+      "step": 3920
+    },
+    {
+      "epoch": 0.26300432639098503,
+      "grad_norm": 0.14629063040826207,
+      "learning_rate": 2e-05,
+      "loss": 5.3671,
+      "step": 3921
+    },
+    {
+      "epoch": 0.26307140222020997,
+      "grad_norm": 0.15262539854031368,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2631384780494349,
+      "grad_norm": 0.13944660490305602,
+      "learning_rate": 2e-05,
+      "loss": 5.36,
+      "step": 3923
+    },
+    {
+      "epoch": 0.26320555387865985,
+      "grad_norm": 0.14970038808776698,
+      "learning_rate": 2e-05,
+      "loss": 5.3034,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2632726297078848,
+      "grad_norm": 0.14666369191607267,
+      "learning_rate": 2e-05,
+      "loss": 5.358,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2633397055371097,
+      "grad_norm": 0.1464924921731778,
+      "learning_rate": 2e-05,
+      "loss": 5.3519,
+      "step": 3926
+    },
+    {
+      "epoch": 0.26340678136633466,
+      "grad_norm": 0.14295580467681032,
+      "learning_rate": 2e-05,
+      "loss": 5.3948,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2634738571955596,
+      "grad_norm": 0.15315361360115567,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 3928
+    },
+    {
+      "epoch": 0.26354093302478454,
+      "grad_norm": 0.1549928196158761,
+      "learning_rate": 2e-05,
+      "loss": 5.2922,
+      "step": 3929
+    },
+    {
+      "epoch": 0.2636080088540095,
+      "grad_norm": 0.14831235448930838,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2636750846832344,
+      "grad_norm": 0.14631020907619874,
+      "learning_rate": 2e-05,
+      "loss": 5.348,
+      "step": 3931
+    },
+    {
+      "epoch": 0.26374216051245936,
+      "grad_norm": 0.14315035868996318,
+      "learning_rate": 2e-05,
+      "loss": 5.4006,
+      "step": 3932
+    },
+    {
+      "epoch": 0.2638092363416843,
+      "grad_norm": 0.138619142752392,
+      "learning_rate": 2e-05,
+      "loss": 5.339,
+      "step": 3933
+    },
+    {
+      "epoch": 0.26387631217090923,
+      "grad_norm": 0.14090254676230443,
+      "learning_rate": 2e-05,
+      "loss": 5.3838,
+      "step": 3934
+    },
+    {
+      "epoch": 0.26394338800013417,
+      "grad_norm": 0.14146528358016539,
+      "learning_rate": 2e-05,
+      "loss": 5.5542,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2640104638293591,
+      "grad_norm": 0.15280248826125098,
+      "learning_rate": 2e-05,
+      "loss": 5.3512,
+      "step": 3936
+    },
+    {
+      "epoch": 0.26407753965858405,
+      "grad_norm": 0.1365271342293026,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 3937
+    },
+    {
+      "epoch": 0.264144615487809,
+      "grad_norm": 0.14160049577973796,
+      "learning_rate": 2e-05,
+      "loss": 5.6048,
+      "step": 3938
+    },
+    {
+      "epoch": 0.2642116913170339,
+      "grad_norm": 0.14497702634602677,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 3939
+    },
+    {
+      "epoch": 0.26427876714625886,
+      "grad_norm": 0.14378048039287575,
+      "learning_rate": 2e-05,
+      "loss": 5.5025,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2643458429754838,
+      "grad_norm": 0.15170355920336792,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 3941
+    },
+    {
+      "epoch": 0.26441291880470874,
+      "grad_norm": 0.1491013017415287,
+      "learning_rate": 2e-05,
+      "loss": 5.3491,
+      "step": 3942
+    },
+    {
+      "epoch": 0.2644799946339337,
+      "grad_norm": 0.15300722296795793,
+      "learning_rate": 2e-05,
+      "loss": 5.5461,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2645470704631586,
+      "grad_norm": 0.1422665084623843,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 3944
+    },
+    {
+      "epoch": 0.26461414629238356,
+      "grad_norm": 0.14086613750178717,
+      "learning_rate": 2e-05,
+      "loss": 5.4694,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2646812221216085,
+      "grad_norm": 0.14246314511349153,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 3946
+    },
+    {
+      "epoch": 0.26474829795083343,
+      "grad_norm": 0.1478285621139698,
+      "learning_rate": 2e-05,
+      "loss": 5.4613,
+      "step": 3947
+    },
+    {
+      "epoch": 0.26481537378005837,
+      "grad_norm": 0.1381168605582325,
+      "learning_rate": 2e-05,
+      "loss": 5.4834,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2648824496092833,
+      "grad_norm": 0.14302195792472977,
+      "learning_rate": 2e-05,
+      "loss": 5.4121,
+      "step": 3949
+    },
+    {
+      "epoch": 0.26494952543850825,
+      "grad_norm": 0.14662659215734974,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2650166012677332,
+      "grad_norm": 0.14353767664312628,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 3951
+    },
+    {
+      "epoch": 0.2650836770969581,
+      "grad_norm": 0.15503948120396469,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 3952
+    },
+    {
+      "epoch": 0.26515075292618306,
+      "grad_norm": 0.14106078704891806,
+      "learning_rate": 2e-05,
+      "loss": 5.4382,
+      "step": 3953
+    },
+    {
+      "epoch": 0.265217828755408,
+      "grad_norm": 0.14534439590715068,
+      "learning_rate": 2e-05,
+      "loss": 5.5083,
+      "step": 3954
+    },
+    {
+      "epoch": 0.26528490458463294,
+      "grad_norm": 0.14851161744696703,
+      "learning_rate": 2e-05,
+      "loss": 5.4736,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2653519804138579,
+      "grad_norm": 0.1409302765135972,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 3956
+    },
+    {
+      "epoch": 0.2654190562430828,
+      "grad_norm": 0.1560010311825181,
+      "learning_rate": 2e-05,
+      "loss": 5.5792,
+      "step": 3957
+    },
+    {
+      "epoch": 0.26548613207230776,
+      "grad_norm": 0.14469956025173653,
+      "learning_rate": 2e-05,
+      "loss": 5.6303,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2655532079015327,
+      "grad_norm": 0.15427136406346895,
+      "learning_rate": 2e-05,
+      "loss": 5.3819,
+      "step": 3959
+    },
+    {
+      "epoch": 0.26562028373075763,
+      "grad_norm": 0.1419142390436936,
+      "learning_rate": 2e-05,
+      "loss": 5.4751,
+      "step": 3960
+    },
+    {
+      "epoch": 0.26568735955998257,
+      "grad_norm": 0.14024886363030958,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2657544353892075,
+      "grad_norm": 0.14106118556531205,
+      "learning_rate": 2e-05,
+      "loss": 5.4772,
+      "step": 3962
+    },
+    {
+      "epoch": 0.26582151121843245,
+      "grad_norm": 0.14514251672518408,
+      "learning_rate": 2e-05,
+      "loss": 5.393,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2658885870476574,
+      "grad_norm": 0.1442537694326055,
+      "learning_rate": 2e-05,
+      "loss": 5.5645,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2659556628768823,
+      "grad_norm": 0.14791947101068886,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 3965
+    },
+    {
+      "epoch": 0.26602273870610726,
+      "grad_norm": 0.14291208790020654,
+      "learning_rate": 2e-05,
+      "loss": 5.5062,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2660898145353322,
+      "grad_norm": 0.15183833808373706,
+      "learning_rate": 2e-05,
+      "loss": 5.281,
+      "step": 3967
+    },
+    {
+      "epoch": 0.26615689036455714,
+      "grad_norm": 0.15192532819502788,
+      "learning_rate": 2e-05,
+      "loss": 5.3464,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2662239661937821,
+      "grad_norm": 0.1465302899812399,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 3969
+    },
+    {
+      "epoch": 0.266291042023007,
+      "grad_norm": 0.14724375900207798,
+      "learning_rate": 2e-05,
+      "loss": 5.244,
+      "step": 3970
+    },
+    {
+      "epoch": 0.26635811785223196,
+      "grad_norm": 0.1589793481967423,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2664251936814569,
+      "grad_norm": 0.14623702837254354,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 3972
+    },
+    {
+      "epoch": 0.26649226951068183,
+      "grad_norm": 0.149889735227389,
+      "learning_rate": 2e-05,
+      "loss": 5.2071,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2665593453399068,
+      "grad_norm": 0.14827475231902176,
+      "learning_rate": 2e-05,
+      "loss": 5.5338,
+      "step": 3974
+    },
+    {
+      "epoch": 0.2666264211691317,
+      "grad_norm": 0.15200149399781676,
+      "learning_rate": 2e-05,
+      "loss": 5.3729,
+      "step": 3975
+    },
+    {
+      "epoch": 0.26669349699835665,
+      "grad_norm": 0.15610579812519548,
+      "learning_rate": 2e-05,
+      "loss": 5.3062,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2667605728275816,
+      "grad_norm": 0.14780158405639887,
+      "learning_rate": 2e-05,
+      "loss": 5.4289,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2668276486568065,
+      "grad_norm": 0.1485248240817027,
+      "learning_rate": 2e-05,
+      "loss": 5.5061,
+      "step": 3978
+    },
+    {
+      "epoch": 0.26689472448603146,
+      "grad_norm": 0.15110609963248903,
+      "learning_rate": 2e-05,
+      "loss": 5.5296,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2669618003152564,
+      "grad_norm": 0.1546623604405714,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 3980
+    },
+    {
+      "epoch": 0.26702887614448134,
+      "grad_norm": 0.1560529227763332,
+      "learning_rate": 2e-05,
+      "loss": 5.5236,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2670959519737063,
+      "grad_norm": 0.15160115450932635,
+      "learning_rate": 2e-05,
+      "loss": 5.3832,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2671630278029312,
+      "grad_norm": 0.14940625709504377,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 3983
+    },
+    {
+      "epoch": 0.26723010363215616,
+      "grad_norm": 0.15089418420658865,
+      "learning_rate": 2e-05,
+      "loss": 5.4891,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2672971794613811,
+      "grad_norm": 0.15149544081078248,
+      "learning_rate": 2e-05,
+      "loss": 5.4743,
+      "step": 3985
+    },
+    {
+      "epoch": 0.26736425529060603,
+      "grad_norm": 0.1590470388736261,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 3986
+    },
+    {
+      "epoch": 0.267431331119831,
+      "grad_norm": 0.14569233425858072,
+      "learning_rate": 2e-05,
+      "loss": 5.421,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2674984069490559,
+      "grad_norm": 0.15299566876328138,
+      "learning_rate": 2e-05,
+      "loss": 5.3958,
+      "step": 3988
+    },
+    {
+      "epoch": 0.26756548277828085,
+      "grad_norm": 0.1530500652312546,
+      "learning_rate": 2e-05,
+      "loss": 5.4839,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2676325586075058,
+      "grad_norm": 0.1493130861294051,
+      "learning_rate": 2e-05,
+      "loss": 5.433,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2676996344367307,
+      "grad_norm": 0.14637908964705038,
+      "learning_rate": 2e-05,
+      "loss": 5.5441,
+      "step": 3991
+    },
+    {
+      "epoch": 0.26776671026595567,
+      "grad_norm": 0.14461814367437367,
+      "learning_rate": 2e-05,
+      "loss": 5.5222,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2678337860951806,
+      "grad_norm": 0.14941149992524735,
+      "learning_rate": 2e-05,
+      "loss": 5.5347,
+      "step": 3993
+    },
+    {
+      "epoch": 0.26790086192440554,
+      "grad_norm": 0.15257382657383972,
+      "learning_rate": 2e-05,
+      "loss": 5.5434,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2679679377536305,
+      "grad_norm": 0.14264895502712277,
+      "learning_rate": 2e-05,
+      "loss": 5.4963,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2680350135828554,
+      "grad_norm": 0.1529453805687773,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 3996
+    },
+    {
+      "epoch": 0.26810208941208036,
+      "grad_norm": 0.14387444150300663,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 3997
+    },
+    {
+      "epoch": 0.2681691652413053,
+      "grad_norm": 0.1528660580646947,
+      "learning_rate": 2e-05,
+      "loss": 5.4257,
+      "step": 3998
+    },
+    {
+      "epoch": 0.26823624107053023,
+      "grad_norm": 0.15016025250543324,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2683033168997552,
+      "grad_norm": 0.1506383656572807,
+      "learning_rate": 2e-05,
+      "loss": 5.4627,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2683703927289801,
+      "grad_norm": 0.1491226530026307,
+      "learning_rate": 2e-05,
+      "loss": 5.3009,
+      "step": 4001
+    },
+    {
+      "epoch": 0.26843746855820505,
+      "grad_norm": 0.1479862197980063,
+      "learning_rate": 2e-05,
+      "loss": 5.3096,
+      "step": 4002
+    },
+    {
+      "epoch": 0.26850454438743,
+      "grad_norm": 0.15844295410647932,
+      "learning_rate": 2e-05,
+      "loss": 5.3801,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2685716202166549,
+      "grad_norm": 0.14948410190571407,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 4004
+    },
+    {
+      "epoch": 0.26863869604587987,
+      "grad_norm": 0.15637300431842427,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2687057718751048,
+      "grad_norm": 0.1651002846933649,
+      "learning_rate": 2e-05,
+      "loss": 5.3597,
+      "step": 4006
+    },
+    {
+      "epoch": 0.26877284770432974,
+      "grad_norm": 0.16170135210059888,
+      "learning_rate": 2e-05,
+      "loss": 5.6067,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2688399235335547,
+      "grad_norm": 0.1533657588586712,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2689069993627796,
+      "grad_norm": 0.15426636658665674,
+      "learning_rate": 2e-05,
+      "loss": 5.3782,
+      "step": 4009
+    },
+    {
+      "epoch": 0.26897407519200456,
+      "grad_norm": 0.15897235335286908,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2690411510212295,
+      "grad_norm": 0.1501870374393895,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 4011
+    },
+    {
+      "epoch": 0.26910822685045444,
+      "grad_norm": 0.14628722061409957,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2691753026796794,
+      "grad_norm": 0.15770345305024938,
+      "learning_rate": 2e-05,
+      "loss": 5.5962,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2692423785089043,
+      "grad_norm": 0.15924108180482627,
+      "learning_rate": 2e-05,
+      "loss": 5.4535,
+      "step": 4014
+    },
+    {
+      "epoch": 0.26930945433812925,
+      "grad_norm": 0.14698488926723588,
+      "learning_rate": 2e-05,
+      "loss": 5.5448,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2693765301673542,
+      "grad_norm": 0.15015298849145567,
+      "learning_rate": 2e-05,
+      "loss": 5.3155,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2694436059965791,
+      "grad_norm": 0.15647140162074277,
+      "learning_rate": 2e-05,
+      "loss": 5.3404,
+      "step": 4017
+    },
+    {
+      "epoch": 0.26951068182580407,
+      "grad_norm": 0.14859888982245759,
+      "learning_rate": 2e-05,
+      "loss": 5.4315,
+      "step": 4018
+    },
+    {
+      "epoch": 0.269577757655029,
+      "grad_norm": 0.1436600048186052,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 4019
+    },
+    {
+      "epoch": 0.26964483348425394,
+      "grad_norm": 0.15560455067486184,
+      "learning_rate": 2e-05,
+      "loss": 5.5367,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2697119093134789,
+      "grad_norm": 0.15844949643341005,
+      "learning_rate": 2e-05,
+      "loss": 5.3786,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2697789851427038,
+      "grad_norm": 0.15122823814188055,
+      "learning_rate": 2e-05,
+      "loss": 5.4912,
+      "step": 4022
+    },
+    {
+      "epoch": 0.26984606097192876,
+      "grad_norm": 0.15103928086612023,
+      "learning_rate": 2e-05,
+      "loss": 5.5401,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2699131368011537,
+      "grad_norm": 0.156539107872407,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 4024
+    },
+    {
+      "epoch": 0.26998021263037864,
+      "grad_norm": 0.15130195428876642,
+      "learning_rate": 2e-05,
+      "loss": 5.4775,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2700472884596036,
+      "grad_norm": 0.14459042634177868,
+      "learning_rate": 2e-05,
+      "loss": 5.2902,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2701143642888285,
+      "grad_norm": 0.1491607686163095,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 4027
+    },
+    {
+      "epoch": 0.27018144011805345,
+      "grad_norm": 0.1440402534633965,
+      "learning_rate": 2e-05,
+      "loss": 5.3895,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2702485159472784,
+      "grad_norm": 0.15550186736623625,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 4029
+    },
+    {
+      "epoch": 0.27031559177650333,
+      "grad_norm": 0.1395105477728193,
+      "learning_rate": 2e-05,
+      "loss": 5.61,
+      "step": 4030
+    },
+    {
+      "epoch": 0.27038266760572827,
+      "grad_norm": 0.14352264799104159,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2704497434349532,
+      "grad_norm": 0.15306248515528173,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 4032
+    },
+    {
+      "epoch": 0.27051681926417814,
+      "grad_norm": 0.14294084866275805,
+      "learning_rate": 2e-05,
+      "loss": 5.5591,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2705838950934031,
+      "grad_norm": 0.1510892460330588,
+      "learning_rate": 2e-05,
+      "loss": 5.4882,
+      "step": 4034
+    },
+    {
+      "epoch": 0.270650970922628,
+      "grad_norm": 0.15467593086176307,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 4035
+    },
+    {
+      "epoch": 0.27071804675185296,
+      "grad_norm": 0.1443802659692695,
+      "learning_rate": 2e-05,
+      "loss": 5.4275,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2707851225810779,
+      "grad_norm": 0.14109784694533045,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 4037
+    },
+    {
+      "epoch": 0.27085219841030284,
+      "grad_norm": 0.1506743384497489,
+      "learning_rate": 2e-05,
+      "loss": 5.3074,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2709192742395278,
+      "grad_norm": 0.14412564676548184,
+      "learning_rate": 2e-05,
+      "loss": 5.5208,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2709863500687527,
+      "grad_norm": 0.14359260235394306,
+      "learning_rate": 2e-05,
+      "loss": 5.5417,
+      "step": 4040
+    },
+    {
+      "epoch": 0.27105342589797765,
+      "grad_norm": 0.152271357844385,
+      "learning_rate": 2e-05,
+      "loss": 5.3863,
+      "step": 4041
+    },
+    {
+      "epoch": 0.2711205017272026,
+      "grad_norm": 0.14231721872606426,
+      "learning_rate": 2e-05,
+      "loss": 5.4674,
+      "step": 4042
+    },
+    {
+      "epoch": 0.27118757755642753,
+      "grad_norm": 0.14783462599573524,
+      "learning_rate": 2e-05,
+      "loss": 5.4661,
+      "step": 4043
+    },
+    {
+      "epoch": 0.27125465338565247,
+      "grad_norm": 0.15470527384936894,
+      "learning_rate": 2e-05,
+      "loss": 5.5303,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2713217292148774,
+      "grad_norm": 0.14160153972229034,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 4045
+    },
+    {
+      "epoch": 0.27138880504410234,
+      "grad_norm": 0.1458510318409033,
+      "learning_rate": 2e-05,
+      "loss": 5.3225,
+      "step": 4046
+    },
+    {
+      "epoch": 0.2714558808733273,
+      "grad_norm": 0.14609365537346464,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2715229567025522,
+      "grad_norm": 0.14356178478901552,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 4048
+    },
+    {
+      "epoch": 0.27159003253177716,
+      "grad_norm": 0.144997692370976,
+      "learning_rate": 2e-05,
+      "loss": 5.4863,
+      "step": 4049
+    },
+    {
+      "epoch": 0.2716571083610021,
+      "grad_norm": 0.14748758711363538,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 4050
+    },
+    {
+      "epoch": 0.27172418419022704,
+      "grad_norm": 0.1491829699718845,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 4051
+    },
+    {
+      "epoch": 0.271791260019452,
+      "grad_norm": 0.14877011809843038,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 4052
+    },
+    {
+      "epoch": 0.2718583358486769,
+      "grad_norm": 0.14957604341918115,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 4053
+    },
+    {
+      "epoch": 0.27192541167790185,
+      "grad_norm": 0.1552661196361312,
+      "learning_rate": 2e-05,
+      "loss": 5.5378,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2719924875071268,
+      "grad_norm": 0.15146516915767905,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 4055
+    },
+    {
+      "epoch": 0.27205956333635173,
+      "grad_norm": 0.16191816221876068,
+      "learning_rate": 2e-05,
+      "loss": 5.5532,
+      "step": 4056
+    },
+    {
+      "epoch": 0.27212663916557667,
+      "grad_norm": 0.1514337350992612,
+      "learning_rate": 2e-05,
+      "loss": 5.45,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2721937149948016,
+      "grad_norm": 0.16639592704904713,
+      "learning_rate": 2e-05,
+      "loss": 5.4002,
+      "step": 4058
+    },
+    {
+      "epoch": 0.27226079082402654,
+      "grad_norm": 0.14756464890440504,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 4059
+    },
+    {
+      "epoch": 0.2723278666532515,
+      "grad_norm": 0.15539069524109989,
+      "learning_rate": 2e-05,
+      "loss": 5.3893,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2723949424824764,
+      "grad_norm": 0.15018571363125266,
+      "learning_rate": 2e-05,
+      "loss": 5.3747,
+      "step": 4061
+    },
+    {
+      "epoch": 0.27246201831170136,
+      "grad_norm": 0.16201294288413687,
+      "learning_rate": 2e-05,
+      "loss": 5.5854,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2725290941409263,
+      "grad_norm": 0.14649527033602472,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 4063
+    },
+    {
+      "epoch": 0.27259616997015124,
+      "grad_norm": 0.1486480958896805,
+      "learning_rate": 2e-05,
+      "loss": 5.4531,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2726632457993762,
+      "grad_norm": 0.1522739555437533,
+      "learning_rate": 2e-05,
+      "loss": 5.4016,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2727303216286011,
+      "grad_norm": 0.15556159222965216,
+      "learning_rate": 2e-05,
+      "loss": 5.4615,
+      "step": 4066
+    },
+    {
+      "epoch": 0.27279739745782605,
+      "grad_norm": 0.14766064252694125,
+      "learning_rate": 2e-05,
+      "loss": 5.4462,
+      "step": 4067
+    },
+    {
+      "epoch": 0.272864473287051,
+      "grad_norm": 0.15345247497740444,
+      "learning_rate": 2e-05,
+      "loss": 5.3887,
+      "step": 4068
+    },
+    {
+      "epoch": 0.27293154911627593,
+      "grad_norm": 0.16445379288701506,
+      "learning_rate": 2e-05,
+      "loss": 5.3135,
+      "step": 4069
+    },
+    {
+      "epoch": 0.27299862494550087,
+      "grad_norm": 0.14867166578157512,
+      "learning_rate": 2e-05,
+      "loss": 5.4107,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2730657007747258,
+      "grad_norm": 0.14348702191429824,
+      "learning_rate": 2e-05,
+      "loss": 5.4058,
+      "step": 4071
+    },
+    {
+      "epoch": 0.27313277660395074,
+      "grad_norm": 0.16109184780567992,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2731998524331757,
+      "grad_norm": 0.15874077859733723,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2732669282624006,
+      "grad_norm": 0.14926414454299927,
+      "learning_rate": 2e-05,
+      "loss": 5.6299,
+      "step": 4074
+    },
+    {
+      "epoch": 0.27333400409162556,
+      "grad_norm": 0.15514703189121695,
+      "learning_rate": 2e-05,
+      "loss": 5.4423,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2734010799208505,
+      "grad_norm": 0.15289669340540735,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 4076
+    },
+    {
+      "epoch": 0.27346815575007544,
+      "grad_norm": 0.14784179912154802,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 4077
+    },
+    {
+      "epoch": 0.2735352315793004,
+      "grad_norm": 0.15309393763101015,
+      "learning_rate": 2e-05,
+      "loss": 5.584,
+      "step": 4078
+    },
+    {
+      "epoch": 0.2736023074085253,
+      "grad_norm": 0.15323639274083803,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 4079
+    },
+    {
+      "epoch": 0.27366938323775025,
+      "grad_norm": 0.14694123643484733,
+      "learning_rate": 2e-05,
+      "loss": 5.4059,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2737364590669752,
+      "grad_norm": 0.14819336260368848,
+      "learning_rate": 2e-05,
+      "loss": 5.5959,
+      "step": 4081
+    },
+    {
+      "epoch": 0.27380353489620013,
+      "grad_norm": 0.1428748049394505,
+      "learning_rate": 2e-05,
+      "loss": 5.3837,
+      "step": 4082
+    },
+    {
+      "epoch": 0.27387061072542507,
+      "grad_norm": 0.15379129208287598,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 4083
+    },
+    {
+      "epoch": 0.27393768655465,
+      "grad_norm": 0.14993411784787034,
+      "learning_rate": 2e-05,
+      "loss": 5.632,
+      "step": 4084
+    },
+    {
+      "epoch": 0.27400476238387494,
+      "grad_norm": 0.14324043086721272,
+      "learning_rate": 2e-05,
+      "loss": 5.4773,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2740718382130999,
+      "grad_norm": 0.15412434909886946,
+      "learning_rate": 2e-05,
+      "loss": 5.5022,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2741389140423248,
+      "grad_norm": 0.14464645574447263,
+      "learning_rate": 2e-05,
+      "loss": 5.5289,
+      "step": 4087
+    },
+    {
+      "epoch": 0.27420598987154976,
+      "grad_norm": 0.1525400868062185,
+      "learning_rate": 2e-05,
+      "loss": 5.3105,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2742730657007747,
+      "grad_norm": 0.1439768415401476,
+      "learning_rate": 2e-05,
+      "loss": 5.5135,
+      "step": 4089
+    },
+    {
+      "epoch": 0.27434014152999964,
+      "grad_norm": 0.15081843220636237,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 4090
+    },
+    {
+      "epoch": 0.27440721735922463,
+      "grad_norm": 0.14582595422154468,
+      "learning_rate": 2e-05,
+      "loss": 5.4494,
+      "step": 4091
+    },
+    {
+      "epoch": 0.27447429318844957,
+      "grad_norm": 0.1520888535031454,
+      "learning_rate": 2e-05,
+      "loss": 5.4683,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2745413690176745,
+      "grad_norm": 0.1443313101156938,
+      "learning_rate": 2e-05,
+      "loss": 5.5678,
+      "step": 4093
+    },
+    {
+      "epoch": 0.27460844484689945,
+      "grad_norm": 0.14869574062537894,
+      "learning_rate": 2e-05,
+      "loss": 5.537,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2746755206761244,
+      "grad_norm": 0.15064387718796454,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2747425965053493,
+      "grad_norm": 0.1456645371340093,
+      "learning_rate": 2e-05,
+      "loss": 5.4509,
+      "step": 4096
+    },
+    {
+      "epoch": 0.27480967233457426,
+      "grad_norm": 0.1486928354832347,
+      "learning_rate": 2e-05,
+      "loss": 5.4929,
+      "step": 4097
+    },
+    {
+      "epoch": 0.2748767481637992,
+      "grad_norm": 0.1522355285410282,
+      "learning_rate": 2e-05,
+      "loss": 5.314,
+      "step": 4098
+    },
+    {
+      "epoch": 0.27494382399302414,
+      "grad_norm": 0.15467220232952622,
+      "learning_rate": 2e-05,
+      "loss": 5.3424,
+      "step": 4099
+    },
+    {
+      "epoch": 0.2750108998222491,
+      "grad_norm": 0.1501567214299044,
+      "learning_rate": 2e-05,
+      "loss": 5.455,
+      "step": 4100
+    },
+    {
+      "epoch": 0.275077975651474,
+      "grad_norm": 0.15733114066295525,
+      "learning_rate": 2e-05,
+      "loss": 5.6115,
+      "step": 4101
+    },
+    {
+      "epoch": 0.27514505148069895,
+      "grad_norm": 0.15180588695206082,
+      "learning_rate": 2e-05,
+      "loss": 5.2306,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2752121273099239,
+      "grad_norm": 0.15727496676538766,
+      "learning_rate": 2e-05,
+      "loss": 5.5279,
+      "step": 4103
+    },
+    {
+      "epoch": 0.27527920313914883,
+      "grad_norm": 0.1544517405112958,
+      "learning_rate": 2e-05,
+      "loss": 5.4355,
+      "step": 4104
+    },
+    {
+      "epoch": 0.27534627896837377,
+      "grad_norm": 0.1598482361541539,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2754133547975987,
+      "grad_norm": 0.15582340920433344,
+      "learning_rate": 2e-05,
+      "loss": 5.369,
+      "step": 4106
+    },
+    {
+      "epoch": 0.27548043062682365,
+      "grad_norm": 0.1475449891490308,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2755475064560486,
+      "grad_norm": 0.16543471690675432,
+      "learning_rate": 2e-05,
+      "loss": 5.5459,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2756145822852735,
+      "grad_norm": 0.16244288708880583,
+      "learning_rate": 2e-05,
+      "loss": 5.3778,
+      "step": 4109
+    },
+    {
+      "epoch": 0.27568165811449846,
+      "grad_norm": 0.15532901434566232,
+      "learning_rate": 2e-05,
+      "loss": 5.541,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2757487339437234,
+      "grad_norm": 0.15138743205016916,
+      "learning_rate": 2e-05,
+      "loss": 5.5883,
+      "step": 4111
+    },
+    {
+      "epoch": 0.27581580977294834,
+      "grad_norm": 0.15812695668384347,
+      "learning_rate": 2e-05,
+      "loss": 5.3698,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2758828856021733,
+      "grad_norm": 0.15152618403766843,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2759499614313982,
+      "grad_norm": 0.1589672420581012,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 4114
+    },
+    {
+      "epoch": 0.27601703726062315,
+      "grad_norm": 0.16761580538270693,
+      "learning_rate": 2e-05,
+      "loss": 5.4443,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2760841130898481,
+      "grad_norm": 0.15982646374919612,
+      "learning_rate": 2e-05,
+      "loss": 5.624,
+      "step": 4116
+    },
+    {
+      "epoch": 0.27615118891907303,
+      "grad_norm": 0.14601158824396357,
+      "learning_rate": 2e-05,
+      "loss": 5.4237,
+      "step": 4117
+    },
+    {
+      "epoch": 0.27621826474829797,
+      "grad_norm": 0.15759836363620652,
+      "learning_rate": 2e-05,
+      "loss": 5.4242,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2762853405775229,
+      "grad_norm": 0.1628020328155809,
+      "learning_rate": 2e-05,
+      "loss": 5.3557,
+      "step": 4119
+    },
+    {
+      "epoch": 0.27635241640674785,
+      "grad_norm": 0.1536032641970959,
+      "learning_rate": 2e-05,
+      "loss": 5.3027,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2764194922359728,
+      "grad_norm": 0.15605052472157843,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2764865680651977,
+      "grad_norm": 0.1820969902861478,
+      "learning_rate": 2e-05,
+      "loss": 5.4095,
+      "step": 4122
+    },
+    {
+      "epoch": 0.27655364389442266,
+      "grad_norm": 0.16397926952136221,
+      "learning_rate": 2e-05,
+      "loss": 5.5056,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2766207197236476,
+      "grad_norm": 0.14673198283871172,
+      "learning_rate": 2e-05,
+      "loss": 5.5883,
+      "step": 4124
+    },
+    {
+      "epoch": 0.27668779555287254,
+      "grad_norm": 0.1640938474772526,
+      "learning_rate": 2e-05,
+      "loss": 5.5054,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2767548713820975,
+      "grad_norm": 0.16061381092949495,
+      "learning_rate": 2e-05,
+      "loss": 5.4159,
+      "step": 4126
+    },
+    {
+      "epoch": 0.2768219472113224,
+      "grad_norm": 0.14342164030566945,
+      "learning_rate": 2e-05,
+      "loss": 5.4554,
+      "step": 4127
+    },
+    {
+      "epoch": 0.27688902304054736,
+      "grad_norm": 0.1559434792775877,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2769560988697723,
+      "grad_norm": 0.15539676962788176,
+      "learning_rate": 2e-05,
+      "loss": 5.3616,
+      "step": 4129
+    },
+    {
+      "epoch": 0.27702317469899723,
+      "grad_norm": 0.15433447065500333,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 4130
+    },
+    {
+      "epoch": 0.27709025052822217,
+      "grad_norm": 0.14622631409558381,
+      "learning_rate": 2e-05,
+      "loss": 5.5679,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2771573263574471,
+      "grad_norm": 0.15956853347736275,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 4132
+    },
+    {
+      "epoch": 0.27722440218667205,
+      "grad_norm": 0.1615426320250839,
+      "learning_rate": 2e-05,
+      "loss": 5.3634,
+      "step": 4133
+    },
+    {
+      "epoch": 0.277291478015897,
+      "grad_norm": 0.15208850531064724,
+      "learning_rate": 2e-05,
+      "loss": 5.4281,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2773585538451219,
+      "grad_norm": 0.14830541811032244,
+      "learning_rate": 2e-05,
+      "loss": 5.3906,
+      "step": 4135
+    },
+    {
+      "epoch": 0.27742562967434686,
+      "grad_norm": 0.15949689886700133,
+      "learning_rate": 2e-05,
+      "loss": 5.3592,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2774927055035718,
+      "grad_norm": 0.14721716640531549,
+      "learning_rate": 2e-05,
+      "loss": 5.4507,
+      "step": 4137
+    },
+    {
+      "epoch": 0.27755978133279674,
+      "grad_norm": 0.1477403648455687,
+      "learning_rate": 2e-05,
+      "loss": 5.34,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2776268571620217,
+      "grad_norm": 0.1539807589677248,
+      "learning_rate": 2e-05,
+      "loss": 5.5452,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2776939329912466,
+      "grad_norm": 0.16099008450939825,
+      "learning_rate": 2e-05,
+      "loss": 5.6494,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27776100882047156,
+      "grad_norm": 0.14834040413891822,
+      "learning_rate": 2e-05,
+      "loss": 5.5821,
+      "step": 4141
+    },
+    {
+      "epoch": 0.2778280846496965,
+      "grad_norm": 0.1545181555443981,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 4142
+    },
+    {
+      "epoch": 0.27789516047892143,
+      "grad_norm": 0.14912546129805215,
+      "learning_rate": 2e-05,
+      "loss": 5.5659,
+      "step": 4143
+    },
+    {
+      "epoch": 0.27796223630814637,
+      "grad_norm": 0.14288639505888356,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2780293121373713,
+      "grad_norm": 0.15450336006430415,
+      "learning_rate": 2e-05,
+      "loss": 5.3666,
+      "step": 4145
+    },
+    {
+      "epoch": 0.27809638796659625,
+      "grad_norm": 0.14996483340793465,
+      "learning_rate": 2e-05,
+      "loss": 5.4091,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2781634637958212,
+      "grad_norm": 0.14761261997075417,
+      "learning_rate": 2e-05,
+      "loss": 5.4636,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2782305396250461,
+      "grad_norm": 0.14381956413504757,
+      "learning_rate": 2e-05,
+      "loss": 5.4845,
+      "step": 4148
+    },
+    {
+      "epoch": 0.27829761545427106,
+      "grad_norm": 0.14300703638995454,
+      "learning_rate": 2e-05,
+      "loss": 5.416,
+      "step": 4149
+    },
+    {
+      "epoch": 0.278364691283496,
+      "grad_norm": 0.14622953558546262,
+      "learning_rate": 2e-05,
+      "loss": 5.5719,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27843176711272094,
+      "grad_norm": 0.14806251125801995,
+      "learning_rate": 2e-05,
+      "loss": 5.5814,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2784988429419459,
+      "grad_norm": 0.14259532621084495,
+      "learning_rate": 2e-05,
+      "loss": 5.4264,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2785659187711708,
+      "grad_norm": 0.14235747638672733,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 4153
+    },
+    {
+      "epoch": 0.27863299460039576,
+      "grad_norm": 0.141832544719108,
+      "learning_rate": 2e-05,
+      "loss": 5.3878,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2787000704296207,
+      "grad_norm": 0.14408348193020845,
+      "learning_rate": 2e-05,
+      "loss": 5.4368,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27876714625884563,
+      "grad_norm": 0.14226990978082277,
+      "learning_rate": 2e-05,
+      "loss": 5.4957,
+      "step": 4156
+    },
+    {
+      "epoch": 0.27883422208807057,
+      "grad_norm": 0.14218932570966877,
+      "learning_rate": 2e-05,
+      "loss": 5.5188,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2789012979172955,
+      "grad_norm": 0.14377661241677933,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 4158
+    },
+    {
+      "epoch": 0.27896837374652045,
+      "grad_norm": 0.1456146730981462,
+      "learning_rate": 2e-05,
+      "loss": 5.4299,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2790354495757454,
+      "grad_norm": 0.14671007785539272,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2791025254049703,
+      "grad_norm": 0.14948237138971202,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 4161
+    },
+    {
+      "epoch": 0.27916960123419526,
+      "grad_norm": 0.1468236459499631,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2792366770634202,
+      "grad_norm": 0.15905936796499737,
+      "learning_rate": 2e-05,
+      "loss": 5.5655,
+      "step": 4163
+    },
+    {
+      "epoch": 0.27930375289264514,
+      "grad_norm": 0.15467165982750114,
+      "learning_rate": 2e-05,
+      "loss": 5.3729,
+      "step": 4164
+    },
+    {
+      "epoch": 0.2793708287218701,
+      "grad_norm": 0.15755806898519642,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 4165
+    },
+    {
+      "epoch": 0.279437904551095,
+      "grad_norm": 0.15428913035791528,
+      "learning_rate": 2e-05,
+      "loss": 5.508,
+      "step": 4166
+    },
+    {
+      "epoch": 0.27950498038031996,
+      "grad_norm": 0.14857360779200088,
+      "learning_rate": 2e-05,
+      "loss": 5.3467,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2795720562095449,
+      "grad_norm": 0.1564175276209204,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 4168
+    },
+    {
+      "epoch": 0.27963913203876983,
+      "grad_norm": 0.15514158477316717,
+      "learning_rate": 2e-05,
+      "loss": 5.3002,
+      "step": 4169
+    },
+    {
+      "epoch": 0.27970620786799477,
+      "grad_norm": 0.14790686312905987,
+      "learning_rate": 2e-05,
+      "loss": 5.4431,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2797732836972197,
+      "grad_norm": 0.15365544574792472,
+      "learning_rate": 2e-05,
+      "loss": 5.3262,
+      "step": 4171
+    },
+    {
+      "epoch": 0.27984035952644465,
+      "grad_norm": 0.15290492268639044,
+      "learning_rate": 2e-05,
+      "loss": 5.4973,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2799074353556696,
+      "grad_norm": 0.14033138523962047,
+      "learning_rate": 2e-05,
+      "loss": 5.3542,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2799745111848945,
+      "grad_norm": 0.15712957055940224,
+      "learning_rate": 2e-05,
+      "loss": 5.3513,
+      "step": 4174
+    },
+    {
+      "epoch": 0.28004158701411946,
+      "grad_norm": 0.14290999720345637,
+      "learning_rate": 2e-05,
+      "loss": 5.3529,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2801086628433444,
+      "grad_norm": 0.1488015984379451,
+      "learning_rate": 2e-05,
+      "loss": 5.606,
+      "step": 4176
+    },
+    {
+      "epoch": 0.28017573867256934,
+      "grad_norm": 0.14622815311354515,
+      "learning_rate": 2e-05,
+      "loss": 5.3576,
+      "step": 4177
+    },
+    {
+      "epoch": 0.2802428145017943,
+      "grad_norm": 0.1568594891759342,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2803098903310192,
+      "grad_norm": 0.14731535138858792,
+      "learning_rate": 2e-05,
+      "loss": 5.397,
+      "step": 4179
+    },
+    {
+      "epoch": 0.28037696616024416,
+      "grad_norm": 0.1467484556578845,
+      "learning_rate": 2e-05,
+      "loss": 5.3449,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2804440419894691,
+      "grad_norm": 0.15275502248270334,
+      "learning_rate": 2e-05,
+      "loss": 5.4738,
+      "step": 4181
+    },
+    {
+      "epoch": 0.28051111781869403,
+      "grad_norm": 0.14334141755914057,
+      "learning_rate": 2e-05,
+      "loss": 5.3698,
+      "step": 4182
+    },
+    {
+      "epoch": 0.28057819364791897,
+      "grad_norm": 0.13882716663401373,
+      "learning_rate": 2e-05,
+      "loss": 5.4051,
+      "step": 4183
+    },
+    {
+      "epoch": 0.2806452694771439,
+      "grad_norm": 0.15564568447183752,
+      "learning_rate": 2e-05,
+      "loss": 5.459,
+      "step": 4184
+    },
+    {
+      "epoch": 0.28071234530636885,
+      "grad_norm": 0.1500014552308296,
+      "learning_rate": 2e-05,
+      "loss": 5.4165,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2807794211355938,
+      "grad_norm": 0.15189539813154015,
+      "learning_rate": 2e-05,
+      "loss": 5.3912,
+      "step": 4186
+    },
+    {
+      "epoch": 0.2808464969648187,
+      "grad_norm": 0.14451676195304103,
+      "learning_rate": 2e-05,
+      "loss": 5.4254,
+      "step": 4187
+    },
+    {
+      "epoch": 0.28091357279404366,
+      "grad_norm": 0.14034127385671746,
+      "learning_rate": 2e-05,
+      "loss": 5.4794,
+      "step": 4188
+    },
+    {
+      "epoch": 0.2809806486232686,
+      "grad_norm": 0.14902262907837083,
+      "learning_rate": 2e-05,
+      "loss": 5.427,
+      "step": 4189
+    },
+    {
+      "epoch": 0.28104772445249354,
+      "grad_norm": 0.14314775836058813,
+      "learning_rate": 2e-05,
+      "loss": 5.4628,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2811148002817185,
+      "grad_norm": 0.15074886905597892,
+      "learning_rate": 2e-05,
+      "loss": 5.535,
+      "step": 4191
+    },
+    {
+      "epoch": 0.2811818761109434,
+      "grad_norm": 0.14782900024066162,
+      "learning_rate": 2e-05,
+      "loss": 5.3076,
+      "step": 4192
+    },
+    {
+      "epoch": 0.28124895194016836,
+      "grad_norm": 0.14696032416589855,
+      "learning_rate": 2e-05,
+      "loss": 5.3914,
+      "step": 4193
+    },
+    {
+      "epoch": 0.2813160277693933,
+      "grad_norm": 0.14215574163411482,
+      "learning_rate": 2e-05,
+      "loss": 5.514,
+      "step": 4194
+    },
+    {
+      "epoch": 0.28138310359861823,
+      "grad_norm": 0.1474656141750625,
+      "learning_rate": 2e-05,
+      "loss": 5.3927,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2814501794278432,
+      "grad_norm": 0.14574274058110845,
+      "learning_rate": 2e-05,
+      "loss": 5.3956,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2815172552570681,
+      "grad_norm": 0.1401248331516252,
+      "learning_rate": 2e-05,
+      "loss": 5.5348,
+      "step": 4197
+    },
+    {
+      "epoch": 0.28158433108629305,
+      "grad_norm": 0.14959442910966878,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 4198
+    },
+    {
+      "epoch": 0.281651406915518,
+      "grad_norm": 0.15403469901247946,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2817184827447429,
+      "grad_norm": 0.14827839395139344,
+      "learning_rate": 2e-05,
+      "loss": 5.3541,
+      "step": 4200
+    },
+    {
+      "epoch": 0.28178555857396786,
+      "grad_norm": 0.1383740848699618,
+      "learning_rate": 2e-05,
+      "loss": 5.3293,
+      "step": 4201
+    },
+    {
+      "epoch": 0.2818526344031928,
+      "grad_norm": 0.1499252854927373,
+      "learning_rate": 2e-05,
+      "loss": 5.4775,
+      "step": 4202
+    },
+    {
+      "epoch": 0.28191971023241774,
+      "grad_norm": 0.1554792443971762,
+      "learning_rate": 2e-05,
+      "loss": 5.3921,
+      "step": 4203
+    },
+    {
+      "epoch": 0.2819867860616427,
+      "grad_norm": 0.14638100891734543,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 4204
+    },
+    {
+      "epoch": 0.2820538618908676,
+      "grad_norm": 0.14666548034922644,
+      "learning_rate": 2e-05,
+      "loss": 5.328,
+      "step": 4205
+    },
+    {
+      "epoch": 0.28212093772009256,
+      "grad_norm": 0.14884032752229057,
+      "learning_rate": 2e-05,
+      "loss": 5.2952,
+      "step": 4206
+    },
+    {
+      "epoch": 0.2821880135493175,
+      "grad_norm": 0.14510826304858915,
+      "learning_rate": 2e-05,
+      "loss": 5.4527,
+      "step": 4207
+    },
+    {
+      "epoch": 0.28225508937854243,
+      "grad_norm": 0.15595811184824168,
+      "learning_rate": 2e-05,
+      "loss": 5.618,
+      "step": 4208
+    },
+    {
+      "epoch": 0.2823221652077674,
+      "grad_norm": 0.16191743973650896,
+      "learning_rate": 2e-05,
+      "loss": 5.3533,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2823892410369923,
+      "grad_norm": 0.15258227415808104,
+      "learning_rate": 2e-05,
+      "loss": 5.4655,
+      "step": 4210
+    },
+    {
+      "epoch": 0.28245631686621725,
+      "grad_norm": 0.14851336250865804,
+      "learning_rate": 2e-05,
+      "loss": 5.3805,
+      "step": 4211
+    },
+    {
+      "epoch": 0.2825233926954422,
+      "grad_norm": 0.1647521698813511,
+      "learning_rate": 2e-05,
+      "loss": 5.4848,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2825904685246671,
+      "grad_norm": 0.1517039055581649,
+      "learning_rate": 2e-05,
+      "loss": 5.3001,
+      "step": 4213
+    },
+    {
+      "epoch": 0.28265754435389207,
+      "grad_norm": 0.1475924345944766,
+      "learning_rate": 2e-05,
+      "loss": 5.4462,
+      "step": 4214
+    },
+    {
+      "epoch": 0.282724620183117,
+      "grad_norm": 0.15120550351234685,
+      "learning_rate": 2e-05,
+      "loss": 5.5405,
+      "step": 4215
+    },
+    {
+      "epoch": 0.28279169601234194,
+      "grad_norm": 0.1690886397091709,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2828587718415669,
+      "grad_norm": 0.14657669637677134,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 4217
+    },
+    {
+      "epoch": 0.2829258476707918,
+      "grad_norm": 0.16140914921294114,
+      "learning_rate": 2e-05,
+      "loss": 5.293,
+      "step": 4218
+    },
+    {
+      "epoch": 0.28299292350001676,
+      "grad_norm": 0.1528805094124935,
+      "learning_rate": 2e-05,
+      "loss": 5.6173,
+      "step": 4219
+    },
+    {
+      "epoch": 0.2830599993292417,
+      "grad_norm": 0.1447067373184129,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 4220
+    },
+    {
+      "epoch": 0.28312707515846663,
+      "grad_norm": 0.15762814207113207,
+      "learning_rate": 2e-05,
+      "loss": 5.2724,
+      "step": 4221
+    },
+    {
+      "epoch": 0.2831941509876916,
+      "grad_norm": 0.1511915831229116,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 4222
+    },
+    {
+      "epoch": 0.2832612268169165,
+      "grad_norm": 0.15398291091200844,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 4223
+    },
+    {
+      "epoch": 0.28332830264614145,
+      "grad_norm": 0.15428083126602,
+      "learning_rate": 2e-05,
+      "loss": 5.3306,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2833953784753664,
+      "grad_norm": 0.15369427482719158,
+      "learning_rate": 2e-05,
+      "loss": 5.4152,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2834624543045913,
+      "grad_norm": 0.15025278925108435,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 4226
+    },
+    {
+      "epoch": 0.28352953013381627,
+      "grad_norm": 0.14402203189121326,
+      "learning_rate": 2e-05,
+      "loss": 5.4837,
+      "step": 4227
+    },
+    {
+      "epoch": 0.2835966059630412,
+      "grad_norm": 0.15624450031401624,
+      "learning_rate": 2e-05,
+      "loss": 5.513,
+      "step": 4228
+    },
+    {
+      "epoch": 0.28366368179226614,
+      "grad_norm": 0.15845192321540835,
+      "learning_rate": 2e-05,
+      "loss": 5.5336,
+      "step": 4229
+    },
+    {
+      "epoch": 0.2837307576214911,
+      "grad_norm": 0.15069380865015877,
+      "learning_rate": 2e-05,
+      "loss": 5.522,
+      "step": 4230
+    },
+    {
+      "epoch": 0.283797833450716,
+      "grad_norm": 0.15741295650484055,
+      "learning_rate": 2e-05,
+      "loss": 5.4564,
+      "step": 4231
+    },
+    {
+      "epoch": 0.28386490927994096,
+      "grad_norm": 0.15550182829337533,
+      "learning_rate": 2e-05,
+      "loss": 5.4777,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2839319851091659,
+      "grad_norm": 0.1462831659853542,
+      "learning_rate": 2e-05,
+      "loss": 5.3358,
+      "step": 4233
+    },
+    {
+      "epoch": 0.28399906093839083,
+      "grad_norm": 0.14953505529372607,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2840661367676158,
+      "grad_norm": 0.14770102032657817,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2841332125968407,
+      "grad_norm": 0.1460947997895811,
+      "learning_rate": 2e-05,
+      "loss": 5.4132,
+      "step": 4236
+    },
+    {
+      "epoch": 0.28420028842606565,
+      "grad_norm": 0.14641119995317137,
+      "learning_rate": 2e-05,
+      "loss": 5.3903,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2842673642552906,
+      "grad_norm": 0.1465546840599459,
+      "learning_rate": 2e-05,
+      "loss": 5.3945,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2843344400845155,
+      "grad_norm": 0.14470492581411526,
+      "learning_rate": 2e-05,
+      "loss": 5.4289,
+      "step": 4239
+    },
+    {
+      "epoch": 0.28440151591374047,
+      "grad_norm": 0.14902268564176963,
+      "learning_rate": 2e-05,
+      "loss": 5.4603,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2844685917429654,
+      "grad_norm": 0.13940256052779812,
+      "learning_rate": 2e-05,
+      "loss": 5.3546,
+      "step": 4241
+    },
+    {
+      "epoch": 0.28453566757219034,
+      "grad_norm": 0.13963924675026573,
+      "learning_rate": 2e-05,
+      "loss": 5.423,
+      "step": 4242
+    },
+    {
+      "epoch": 0.2846027434014153,
+      "grad_norm": 0.1398212103324495,
+      "learning_rate": 2e-05,
+      "loss": 5.4062,
+      "step": 4243
+    },
+    {
+      "epoch": 0.2846698192306402,
+      "grad_norm": 0.1545992844861654,
+      "learning_rate": 2e-05,
+      "loss": 5.5085,
+      "step": 4244
+    },
+    {
+      "epoch": 0.28473689505986516,
+      "grad_norm": 0.1431345672617432,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2848039708890901,
+      "grad_norm": 0.1524796085236813,
+      "learning_rate": 2e-05,
+      "loss": 5.346,
+      "step": 4246
+    },
+    {
+      "epoch": 0.28487104671831504,
+      "grad_norm": 0.15297949604255817,
+      "learning_rate": 2e-05,
+      "loss": 5.5322,
+      "step": 4247
+    },
+    {
+      "epoch": 0.28493812254754,
+      "grad_norm": 0.14092259222754722,
+      "learning_rate": 2e-05,
+      "loss": 5.3634,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2850051983767649,
+      "grad_norm": 0.14555791032461388,
+      "learning_rate": 2e-05,
+      "loss": 5.5858,
+      "step": 4249
+    },
+    {
+      "epoch": 0.28507227420598985,
+      "grad_norm": 0.14154241573589404,
+      "learning_rate": 2e-05,
+      "loss": 5.3594,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2851393500352148,
+      "grad_norm": 0.1466775146157266,
+      "learning_rate": 2e-05,
+      "loss": 5.5421,
+      "step": 4251
+    },
+    {
+      "epoch": 0.2852064258644397,
+      "grad_norm": 0.14014761227278524,
+      "learning_rate": 2e-05,
+      "loss": 5.3744,
+      "step": 4252
+    },
+    {
+      "epoch": 0.28527350169366467,
+      "grad_norm": 0.15173824257328689,
+      "learning_rate": 2e-05,
+      "loss": 5.279,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2853405775228896,
+      "grad_norm": 0.14251731130802656,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 4254
+    },
+    {
+      "epoch": 0.28540765335211454,
+      "grad_norm": 0.1455381096664899,
+      "learning_rate": 2e-05,
+      "loss": 5.5024,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2854747291813395,
+      "grad_norm": 0.1427573540339378,
+      "learning_rate": 2e-05,
+      "loss": 5.4839,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2855418050105644,
+      "grad_norm": 0.15045251391984651,
+      "learning_rate": 2e-05,
+      "loss": 5.4157,
+      "step": 4257
+    },
+    {
+      "epoch": 0.28560888083978936,
+      "grad_norm": 0.14435715512928363,
+      "learning_rate": 2e-05,
+      "loss": 5.4799,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2856759566690143,
+      "grad_norm": 0.1455162445726727,
+      "learning_rate": 2e-05,
+      "loss": 5.4887,
+      "step": 4259
+    },
+    {
+      "epoch": 0.28574303249823924,
+      "grad_norm": 0.1424067586851983,
+      "learning_rate": 2e-05,
+      "loss": 5.4017,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2858101083274642,
+      "grad_norm": 0.15221837115703768,
+      "learning_rate": 2e-05,
+      "loss": 5.4975,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2858771841566891,
+      "grad_norm": 0.14696557163057142,
+      "learning_rate": 2e-05,
+      "loss": 5.4884,
+      "step": 4262
+    },
+    {
+      "epoch": 0.28594425998591405,
+      "grad_norm": 0.14550119046362278,
+      "learning_rate": 2e-05,
+      "loss": 5.4121,
+      "step": 4263
+    },
+    {
+      "epoch": 0.286011335815139,
+      "grad_norm": 0.14298297968219814,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 4264
+    },
+    {
+      "epoch": 0.28607841164436393,
+      "grad_norm": 0.1532524102246561,
+      "learning_rate": 2e-05,
+      "loss": 5.3701,
+      "step": 4265
+    },
+    {
+      "epoch": 0.28614548747358887,
+      "grad_norm": 0.14283410644476482,
+      "learning_rate": 2e-05,
+      "loss": 5.3352,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2862125633028138,
+      "grad_norm": 0.144203414744147,
+      "learning_rate": 2e-05,
+      "loss": 5.4735,
+      "step": 4267
+    },
+    {
+      "epoch": 0.28627963913203874,
+      "grad_norm": 0.14729842141463267,
+      "learning_rate": 2e-05,
+      "loss": 5.4758,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2863467149612637,
+      "grad_norm": 0.14541994368856737,
+      "learning_rate": 2e-05,
+      "loss": 5.5883,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2864137907904886,
+      "grad_norm": 0.1437541539115959,
+      "learning_rate": 2e-05,
+      "loss": 5.2122,
+      "step": 4270
+    },
+    {
+      "epoch": 0.28648086661971356,
+      "grad_norm": 0.14930893513950666,
+      "learning_rate": 2e-05,
+      "loss": 5.3754,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2865479424489385,
+      "grad_norm": 0.1494631918432986,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2866150182781635,
+      "grad_norm": 0.15222266826343825,
+      "learning_rate": 2e-05,
+      "loss": 5.5045,
+      "step": 4273
+    },
+    {
+      "epoch": 0.28668209410738843,
+      "grad_norm": 0.14718894360367596,
+      "learning_rate": 2e-05,
+      "loss": 5.6169,
+      "step": 4274
+    },
+    {
+      "epoch": 0.28674916993661337,
+      "grad_norm": 0.14977024941949604,
+      "learning_rate": 2e-05,
+      "loss": 5.5523,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2868162457658383,
+      "grad_norm": 0.14617614944006752,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 4276
+    },
+    {
+      "epoch": 0.28688332159506325,
+      "grad_norm": 0.1396768816950971,
+      "learning_rate": 2e-05,
+      "loss": 5.3594,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2869503974242882,
+      "grad_norm": 0.1417514557075887,
+      "learning_rate": 2e-05,
+      "loss": 5.3836,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2870174732535131,
+      "grad_norm": 0.15002992399542392,
+      "learning_rate": 2e-05,
+      "loss": 5.4586,
+      "step": 4279
+    },
+    {
+      "epoch": 0.28708454908273806,
+      "grad_norm": 0.1555813753189421,
+      "learning_rate": 2e-05,
+      "loss": 5.562,
+      "step": 4280
+    },
+    {
+      "epoch": 0.287151624911963,
+      "grad_norm": 0.14252335442648353,
+      "learning_rate": 2e-05,
+      "loss": 5.4556,
+      "step": 4281
+    },
+    {
+      "epoch": 0.28721870074118794,
+      "grad_norm": 0.14445060486795183,
+      "learning_rate": 2e-05,
+      "loss": 5.2581,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2872857765704129,
+      "grad_norm": 0.1520681897236542,
+      "learning_rate": 2e-05,
+      "loss": 5.4272,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2873528523996378,
+      "grad_norm": 0.14807937495954765,
+      "learning_rate": 2e-05,
+      "loss": 5.3824,
+      "step": 4284
+    },
+    {
+      "epoch": 0.28741992822886275,
+      "grad_norm": 0.1449371621135098,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2874870040580877,
+      "grad_norm": 0.1428422126491746,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 4286
+    },
+    {
+      "epoch": 0.28755407988731263,
+      "grad_norm": 0.15344917760474994,
+      "learning_rate": 2e-05,
+      "loss": 5.3433,
+      "step": 4287
+    },
+    {
+      "epoch": 0.28762115571653757,
+      "grad_norm": 0.15105761832320116,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2876882315457625,
+      "grad_norm": 0.14724558691720097,
+      "learning_rate": 2e-05,
+      "loss": 5.3819,
+      "step": 4289
+    },
+    {
+      "epoch": 0.28775530737498745,
+      "grad_norm": 0.14816770269809068,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2878223832042124,
+      "grad_norm": 0.15337850201303138,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2878894590334373,
+      "grad_norm": 0.14874407320366576,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 4292
+    },
+    {
+      "epoch": 0.28795653486266226,
+      "grad_norm": 0.15238311601592341,
+      "learning_rate": 2e-05,
+      "loss": 5.4416,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2880236106918872,
+      "grad_norm": 0.1458882285462947,
+      "learning_rate": 2e-05,
+      "loss": 5.552,
+      "step": 4294
+    },
+    {
+      "epoch": 0.28809068652111214,
+      "grad_norm": 0.1642624077157309,
+      "learning_rate": 2e-05,
+      "loss": 5.4825,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2881577623503371,
+      "grad_norm": 0.15529732770382726,
+      "learning_rate": 2e-05,
+      "loss": 5.5949,
+      "step": 4296
+    },
+    {
+      "epoch": 0.288224838179562,
+      "grad_norm": 0.15024288125546403,
+      "learning_rate": 2e-05,
+      "loss": 5.3361,
+      "step": 4297
+    },
+    {
+      "epoch": 0.28829191400878695,
+      "grad_norm": 0.15819722878698475,
+      "learning_rate": 2e-05,
+      "loss": 5.4869,
+      "step": 4298
+    },
+    {
+      "epoch": 0.2883589898380119,
+      "grad_norm": 0.15345583034394156,
+      "learning_rate": 2e-05,
+      "loss": 5.5279,
+      "step": 4299
+    },
+    {
+      "epoch": 0.28842606566723683,
+      "grad_norm": 0.14381233987527614,
+      "learning_rate": 2e-05,
+      "loss": 5.4576,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28849314149646177,
+      "grad_norm": 0.14615054076258546,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2885602173256867,
+      "grad_norm": 0.14183682211588364,
+      "learning_rate": 2e-05,
+      "loss": 5.3402,
+      "step": 4302
+    },
+    {
+      "epoch": 0.28862729315491165,
+      "grad_norm": 0.14717642058018157,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2886943689841366,
+      "grad_norm": 0.14605156989785198,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2887614448133615,
+      "grad_norm": 0.14247841949594403,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 4305
+    },
+    {
+      "epoch": 0.28882852064258646,
+      "grad_norm": 0.15054617023011208,
+      "learning_rate": 2e-05,
+      "loss": 5.358,
+      "step": 4306
+    },
+    {
+      "epoch": 0.2888955964718114,
+      "grad_norm": 0.14439270322498457,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 4307
+    },
+    {
+      "epoch": 0.28896267230103634,
+      "grad_norm": 0.1449675409269972,
+      "learning_rate": 2e-05,
+      "loss": 5.3064,
+      "step": 4308
+    },
+    {
+      "epoch": 0.2890297481302613,
+      "grad_norm": 0.14267879918924425,
+      "learning_rate": 2e-05,
+      "loss": 5.4579,
+      "step": 4309
+    },
+    {
+      "epoch": 0.2890968239594862,
+      "grad_norm": 0.14485250689728393,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 4310
+    },
+    {
+      "epoch": 0.28916389978871115,
+      "grad_norm": 0.1481834686438,
+      "learning_rate": 2e-05,
+      "loss": 5.5298,
+      "step": 4311
+    },
+    {
+      "epoch": 0.2892309756179361,
+      "grad_norm": 0.14747414129523978,
+      "learning_rate": 2e-05,
+      "loss": 5.4527,
+      "step": 4312
+    },
+    {
+      "epoch": 0.28929805144716103,
+      "grad_norm": 0.14801995445269878,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 4313
+    },
+    {
+      "epoch": 0.28936512727638597,
+      "grad_norm": 0.15289396494609864,
+      "learning_rate": 2e-05,
+      "loss": 5.4983,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2894322031056109,
+      "grad_norm": 0.14445784161868983,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28949927893483585,
+      "grad_norm": 0.155576578738933,
+      "learning_rate": 2e-05,
+      "loss": 5.3076,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2895663547640608,
+      "grad_norm": 0.14699644009880616,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2896334305932857,
+      "grad_norm": 0.1488365524057276,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 4318
+    },
+    {
+      "epoch": 0.28970050642251066,
+      "grad_norm": 0.15108991112734016,
+      "learning_rate": 2e-05,
+      "loss": 5.2401,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2897675822517356,
+      "grad_norm": 0.14426307248364983,
+      "learning_rate": 2e-05,
+      "loss": 5.6304,
+      "step": 4320
+    },
+    {
+      "epoch": 0.28983465808096054,
+      "grad_norm": 0.15533068273241857,
+      "learning_rate": 2e-05,
+      "loss": 5.4293,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2899017339101855,
+      "grad_norm": 0.1427738767420857,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 4322
+    },
+    {
+      "epoch": 0.2899688097394104,
+      "grad_norm": 0.14284822360424737,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 4323
+    },
+    {
+      "epoch": 0.29003588556863535,
+      "grad_norm": 0.148833307756587,
+      "learning_rate": 2e-05,
+      "loss": 5.4584,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2901029613978603,
+      "grad_norm": 0.14503416008537315,
+      "learning_rate": 2e-05,
+      "loss": 5.4992,
+      "step": 4325
+    },
+    {
+      "epoch": 0.29017003722708523,
+      "grad_norm": 0.14563257658197198,
+      "learning_rate": 2e-05,
+      "loss": 5.514,
+      "step": 4326
+    },
+    {
+      "epoch": 0.29023711305631017,
+      "grad_norm": 0.14011796899937717,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 4327
+    },
+    {
+      "epoch": 0.2903041888855351,
+      "grad_norm": 0.1488081468341025,
+      "learning_rate": 2e-05,
+      "loss": 5.5469,
+      "step": 4328
+    },
+    {
+      "epoch": 0.29037126471476005,
+      "grad_norm": 0.14467475651196673,
+      "learning_rate": 2e-05,
+      "loss": 5.4835,
+      "step": 4329
+    },
+    {
+      "epoch": 0.290438340543985,
+      "grad_norm": 0.14360370321116983,
+      "learning_rate": 2e-05,
+      "loss": 5.5184,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2905054163732099,
+      "grad_norm": 0.14624865996482275,
+      "learning_rate": 2e-05,
+      "loss": 5.5744,
+      "step": 4331
+    },
+    {
+      "epoch": 0.29057249220243486,
+      "grad_norm": 0.15576076980633363,
+      "learning_rate": 2e-05,
+      "loss": 5.4778,
+      "step": 4332
+    },
+    {
+      "epoch": 0.2906395680316598,
+      "grad_norm": 0.14864447880188478,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 4333
+    },
+    {
+      "epoch": 0.29070664386088474,
+      "grad_norm": 0.14156268680209527,
+      "learning_rate": 2e-05,
+      "loss": 5.3782,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2907737196901097,
+      "grad_norm": 0.1490225963056267,
+      "learning_rate": 2e-05,
+      "loss": 5.3003,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2908407955193346,
+      "grad_norm": 0.1435082157957248,
+      "learning_rate": 2e-05,
+      "loss": 5.36,
+      "step": 4336
+    },
+    {
+      "epoch": 0.29090787134855955,
+      "grad_norm": 0.14857276303867534,
+      "learning_rate": 2e-05,
+      "loss": 5.5576,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2909749471777845,
+      "grad_norm": 0.14476241511603105,
+      "learning_rate": 2e-05,
+      "loss": 5.433,
+      "step": 4338
+    },
+    {
+      "epoch": 0.29104202300700943,
+      "grad_norm": 0.15483432577596914,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 4339
+    },
+    {
+      "epoch": 0.29110909883623437,
+      "grad_norm": 0.14294267323724022,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2911761746654593,
+      "grad_norm": 0.14182168303428214,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 4341
+    },
+    {
+      "epoch": 0.29124325049468425,
+      "grad_norm": 0.14728199619845228,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2913103263239092,
+      "grad_norm": 0.14780555936889528,
+      "learning_rate": 2e-05,
+      "loss": 5.3953,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2913774021531341,
+      "grad_norm": 0.15540606676039548,
+      "learning_rate": 2e-05,
+      "loss": 5.5326,
+      "step": 4344
+    },
+    {
+      "epoch": 0.29144447798235906,
+      "grad_norm": 0.14321255713923634,
+      "learning_rate": 2e-05,
+      "loss": 5.4475,
+      "step": 4345
+    },
+    {
+      "epoch": 0.291511553811584,
+      "grad_norm": 0.16422176130238858,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 4346
+    },
+    {
+      "epoch": 0.29157862964080894,
+      "grad_norm": 0.1666693770309578,
+      "learning_rate": 2e-05,
+      "loss": 5.3495,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2916457054700339,
+      "grad_norm": 0.14132604925109069,
+      "learning_rate": 2e-05,
+      "loss": 5.4086,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2917127812992588,
+      "grad_norm": 0.15400337114408036,
+      "learning_rate": 2e-05,
+      "loss": 5.3539,
+      "step": 4349
+    },
+    {
+      "epoch": 0.29177985712848375,
+      "grad_norm": 0.14979018276038622,
+      "learning_rate": 2e-05,
+      "loss": 5.4969,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2918469329577087,
+      "grad_norm": 0.15776750933257022,
+      "learning_rate": 2e-05,
+      "loss": 5.4548,
+      "step": 4351
+    },
+    {
+      "epoch": 0.29191400878693363,
+      "grad_norm": 0.14681575721973045,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 4352
+    },
+    {
+      "epoch": 0.29198108461615857,
+      "grad_norm": 0.15785570703274107,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 4353
+    },
+    {
+      "epoch": 0.2920481604453835,
+      "grad_norm": 0.14917910401799198,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 4354
+    },
+    {
+      "epoch": 0.29211523627460845,
+      "grad_norm": 0.14274659393309294,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2921823121038334,
+      "grad_norm": 0.15043793573792263,
+      "learning_rate": 2e-05,
+      "loss": 5.4095,
+      "step": 4356
+    },
+    {
+      "epoch": 0.2922493879330583,
+      "grad_norm": 0.14459329603168142,
+      "learning_rate": 2e-05,
+      "loss": 5.4744,
+      "step": 4357
+    },
+    {
+      "epoch": 0.29231646376228326,
+      "grad_norm": 0.1438452005111967,
+      "learning_rate": 2e-05,
+      "loss": 5.5961,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2923835395915082,
+      "grad_norm": 0.15267815843964683,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 4359
+    },
+    {
+      "epoch": 0.29245061542073314,
+      "grad_norm": 0.15317532505140832,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2925176912499581,
+      "grad_norm": 0.14951602274794937,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 4361
+    },
+    {
+      "epoch": 0.292584767079183,
+      "grad_norm": 0.14418839642863918,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 4362
+    },
+    {
+      "epoch": 0.29265184290840796,
+      "grad_norm": 0.14974391073500276,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 4363
+    },
+    {
+      "epoch": 0.2927189187376329,
+      "grad_norm": 0.1505174566015882,
+      "learning_rate": 2e-05,
+      "loss": 5.4077,
+      "step": 4364
+    },
+    {
+      "epoch": 0.29278599456685783,
+      "grad_norm": 0.16085270475530777,
+      "learning_rate": 2e-05,
+      "loss": 5.3752,
+      "step": 4365
+    },
+    {
+      "epoch": 0.29285307039608277,
+      "grad_norm": 0.15495109361103707,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2929201462253077,
+      "grad_norm": 0.14833805759782798,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 4367
+    },
+    {
+      "epoch": 0.29298722205453265,
+      "grad_norm": 0.14115324921952613,
+      "learning_rate": 2e-05,
+      "loss": 5.3789,
+      "step": 4368
+    },
+    {
+      "epoch": 0.2930542978837576,
+      "grad_norm": 0.15822603829061954,
+      "learning_rate": 2e-05,
+      "loss": 5.2587,
+      "step": 4369
+    },
+    {
+      "epoch": 0.2931213737129825,
+      "grad_norm": 0.15678305989512015,
+      "learning_rate": 2e-05,
+      "loss": 5.4746,
+      "step": 4370
+    },
+    {
+      "epoch": 0.29318844954220746,
+      "grad_norm": 0.15166846111621055,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 4371
+    },
+    {
+      "epoch": 0.2932555253714324,
+      "grad_norm": 0.14472523301687565,
+      "learning_rate": 2e-05,
+      "loss": 5.4848,
+      "step": 4372
+    },
+    {
+      "epoch": 0.29332260120065734,
+      "grad_norm": 0.16099566881847294,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2933896770298823,
+      "grad_norm": 0.1401712186724716,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 4374
+    },
+    {
+      "epoch": 0.2934567528591072,
+      "grad_norm": 0.14381719732024206,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 4375
+    },
+    {
+      "epoch": 0.29352382868833216,
+      "grad_norm": 0.1539149795790917,
+      "learning_rate": 2e-05,
+      "loss": 5.3745,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2935909045175571,
+      "grad_norm": 0.14384358779099418,
+      "learning_rate": 2e-05,
+      "loss": 5.4185,
+      "step": 4377
+    },
+    {
+      "epoch": 0.29365798034678203,
+      "grad_norm": 0.14980271897636543,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 4378
+    },
+    {
+      "epoch": 0.29372505617600697,
+      "grad_norm": 0.14751700166753023,
+      "learning_rate": 2e-05,
+      "loss": 5.4583,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2937921320052319,
+      "grad_norm": 0.15025908167934107,
+      "learning_rate": 2e-05,
+      "loss": 5.2662,
+      "step": 4380
+    },
+    {
+      "epoch": 0.29385920783445685,
+      "grad_norm": 0.13611355302971653,
+      "learning_rate": 2e-05,
+      "loss": 5.3662,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2939262836636818,
+      "grad_norm": 0.14494893343278173,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 4382
+    },
+    {
+      "epoch": 0.2939933594929067,
+      "grad_norm": 0.15069572941396875,
+      "learning_rate": 2e-05,
+      "loss": 5.4282,
+      "step": 4383
+    },
+    {
+      "epoch": 0.29406043532213166,
+      "grad_norm": 0.1491662264512494,
+      "learning_rate": 2e-05,
+      "loss": 5.5031,
+      "step": 4384
+    },
+    {
+      "epoch": 0.2941275111513566,
+      "grad_norm": 0.14871270794708297,
+      "learning_rate": 2e-05,
+      "loss": 5.3856,
+      "step": 4385
+    },
+    {
+      "epoch": 0.29419458698058154,
+      "grad_norm": 0.14020941658232328,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 4386
+    },
+    {
+      "epoch": 0.2942616628098065,
+      "grad_norm": 0.14058482186589166,
+      "learning_rate": 2e-05,
+      "loss": 5.6088,
+      "step": 4387
+    },
+    {
+      "epoch": 0.2943287386390314,
+      "grad_norm": 0.1448478448679813,
+      "learning_rate": 2e-05,
+      "loss": 5.3369,
+      "step": 4388
+    },
+    {
+      "epoch": 0.29439581446825636,
+      "grad_norm": 0.13848112939753796,
+      "learning_rate": 2e-05,
+      "loss": 5.3978,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2944628902974813,
+      "grad_norm": 0.15421834380214994,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 4390
+    },
+    {
+      "epoch": 0.29452996612670623,
+      "grad_norm": 0.14490901826190594,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 4391
+    },
+    {
+      "epoch": 0.29459704195593117,
+      "grad_norm": 0.15202333469189552,
+      "learning_rate": 2e-05,
+      "loss": 5.4159,
+      "step": 4392
+    },
+    {
+      "epoch": 0.2946641177851561,
+      "grad_norm": 0.14823784492761957,
+      "learning_rate": 2e-05,
+      "loss": 5.5003,
+      "step": 4393
+    },
+    {
+      "epoch": 0.29473119361438105,
+      "grad_norm": 0.14246537939700812,
+      "learning_rate": 2e-05,
+      "loss": 5.5757,
+      "step": 4394
+    },
+    {
+      "epoch": 0.294798269443606,
+      "grad_norm": 0.14293471968587468,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 4395
+    },
+    {
+      "epoch": 0.2948653452728309,
+      "grad_norm": 0.14882017491081653,
+      "learning_rate": 2e-05,
+      "loss": 5.5194,
+      "step": 4396
+    },
+    {
+      "epoch": 0.29493242110205586,
+      "grad_norm": 0.1508040710568239,
+      "learning_rate": 2e-05,
+      "loss": 5.3642,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2949994969312808,
+      "grad_norm": 0.1427744858949439,
+      "learning_rate": 2e-05,
+      "loss": 5.3706,
+      "step": 4398
+    },
+    {
+      "epoch": 0.29506657276050574,
+      "grad_norm": 0.14986633544094705,
+      "learning_rate": 2e-05,
+      "loss": 5.5787,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2951336485897307,
+      "grad_norm": 0.14473535437439156,
+      "learning_rate": 2e-05,
+      "loss": 5.4453,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2952007244189556,
+      "grad_norm": 0.15058433889888256,
+      "learning_rate": 2e-05,
+      "loss": 5.491,
+      "step": 4401
+    },
+    {
+      "epoch": 0.29526780024818056,
+      "grad_norm": 0.1409289391080362,
+      "learning_rate": 2e-05,
+      "loss": 5.3046,
+      "step": 4402
+    },
+    {
+      "epoch": 0.2953348760774055,
+      "grad_norm": 0.14491970562493814,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 4403
+    },
+    {
+      "epoch": 0.29540195190663043,
+      "grad_norm": 0.14717005845628495,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 4404
+    },
+    {
+      "epoch": 0.29546902773585537,
+      "grad_norm": 0.14729994198806665,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2955361035650803,
+      "grad_norm": 0.14434017108729597,
+      "learning_rate": 2e-05,
+      "loss": 5.4031,
+      "step": 4406
+    },
+    {
+      "epoch": 0.29560317939430525,
+      "grad_norm": 0.14793860476063672,
+      "learning_rate": 2e-05,
+      "loss": 5.4988,
+      "step": 4407
+    },
+    {
+      "epoch": 0.2956702552235302,
+      "grad_norm": 0.14876447857390662,
+      "learning_rate": 2e-05,
+      "loss": 5.412,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2957373310527551,
+      "grad_norm": 0.14569830718837296,
+      "learning_rate": 2e-05,
+      "loss": 5.3668,
+      "step": 4409
+    },
+    {
+      "epoch": 0.29580440688198006,
+      "grad_norm": 0.14632030385002823,
+      "learning_rate": 2e-05,
+      "loss": 5.4831,
+      "step": 4410
+    },
+    {
+      "epoch": 0.295871482711205,
+      "grad_norm": 0.1476153505678602,
+      "learning_rate": 2e-05,
+      "loss": 5.3409,
+      "step": 4411
+    },
+    {
+      "epoch": 0.29593855854042994,
+      "grad_norm": 0.14777398583212123,
+      "learning_rate": 2e-05,
+      "loss": 5.549,
+      "step": 4412
+    },
+    {
+      "epoch": 0.2960056343696549,
+      "grad_norm": 0.1448086725182846,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 4413
+    },
+    {
+      "epoch": 0.2960727101988798,
+      "grad_norm": 0.14777133451904728,
+      "learning_rate": 2e-05,
+      "loss": 5.35,
+      "step": 4414
+    },
+    {
+      "epoch": 0.29613978602810476,
+      "grad_norm": 0.14266071386095056,
+      "learning_rate": 2e-05,
+      "loss": 5.4658,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2962068618573297,
+      "grad_norm": 0.14299321872054224,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 4416
+    },
+    {
+      "epoch": 0.29627393768655463,
+      "grad_norm": 0.1447166550864562,
+      "learning_rate": 2e-05,
+      "loss": 5.4378,
+      "step": 4417
+    },
+    {
+      "epoch": 0.29634101351577957,
+      "grad_norm": 0.14307158506394135,
+      "learning_rate": 2e-05,
+      "loss": 5.6019,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2964080893450045,
+      "grad_norm": 0.15423776942039816,
+      "learning_rate": 2e-05,
+      "loss": 5.3995,
+      "step": 4419
+    },
+    {
+      "epoch": 0.29647516517422945,
+      "grad_norm": 0.14195525026085457,
+      "learning_rate": 2e-05,
+      "loss": 5.3615,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2965422410034544,
+      "grad_norm": 0.14826855831766986,
+      "learning_rate": 2e-05,
+      "loss": 5.3519,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2966093168326793,
+      "grad_norm": 0.14531877147399958,
+      "learning_rate": 2e-05,
+      "loss": 5.4244,
+      "step": 4422
+    },
+    {
+      "epoch": 0.29667639266190426,
+      "grad_norm": 0.14083779448189743,
+      "learning_rate": 2e-05,
+      "loss": 5.2964,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2967434684911292,
+      "grad_norm": 0.1399890268274982,
+      "learning_rate": 2e-05,
+      "loss": 5.4173,
+      "step": 4424
+    },
+    {
+      "epoch": 0.29681054432035414,
+      "grad_norm": 0.14924720137934283,
+      "learning_rate": 2e-05,
+      "loss": 5.362,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2968776201495791,
+      "grad_norm": 0.14965245515966227,
+      "learning_rate": 2e-05,
+      "loss": 5.6122,
+      "step": 4426
+    },
+    {
+      "epoch": 0.296944695978804,
+      "grad_norm": 0.15673724511652062,
+      "learning_rate": 2e-05,
+      "loss": 5.4865,
+      "step": 4427
+    },
+    {
+      "epoch": 0.29701177180802896,
+      "grad_norm": 0.14710387998951466,
+      "learning_rate": 2e-05,
+      "loss": 5.5013,
+      "step": 4428
+    },
+    {
+      "epoch": 0.2970788476372539,
+      "grad_norm": 0.15048683130509707,
+      "learning_rate": 2e-05,
+      "loss": 5.4658,
+      "step": 4429
+    },
+    {
+      "epoch": 0.29714592346647883,
+      "grad_norm": 0.1584824446267396,
+      "learning_rate": 2e-05,
+      "loss": 5.3736,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2972129992957038,
+      "grad_norm": 0.147278775604171,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2972800751249287,
+      "grad_norm": 0.14205231536068444,
+      "learning_rate": 2e-05,
+      "loss": 5.5721,
+      "step": 4432
+    },
+    {
+      "epoch": 0.29734715095415365,
+      "grad_norm": 0.1509313547195019,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 4433
+    },
+    {
+      "epoch": 0.2974142267833786,
+      "grad_norm": 0.15131996830782682,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2974813026126035,
+      "grad_norm": 0.14179402217895634,
+      "learning_rate": 2e-05,
+      "loss": 5.4686,
+      "step": 4435
+    },
+    {
+      "epoch": 0.29754837844182846,
+      "grad_norm": 0.1485118189011419,
+      "learning_rate": 2e-05,
+      "loss": 5.4804,
+      "step": 4436
+    },
+    {
+      "epoch": 0.2976154542710534,
+      "grad_norm": 0.1484314800655757,
+      "learning_rate": 2e-05,
+      "loss": 5.5409,
+      "step": 4437
+    },
+    {
+      "epoch": 0.29768253010027834,
+      "grad_norm": 0.14255539655996666,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 4438
+    },
+    {
+      "epoch": 0.2977496059295033,
+      "grad_norm": 0.1499389444958537,
+      "learning_rate": 2e-05,
+      "loss": 5.3837,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2978166817587282,
+      "grad_norm": 0.15009202824377305,
+      "learning_rate": 2e-05,
+      "loss": 5.3945,
+      "step": 4440
+    },
+    {
+      "epoch": 0.29788375758795316,
+      "grad_norm": 0.14726187033022906,
+      "learning_rate": 2e-05,
+      "loss": 5.4124,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2979508334171781,
+      "grad_norm": 0.1526875965152604,
+      "learning_rate": 2e-05,
+      "loss": 5.6277,
+      "step": 4442
+    },
+    {
+      "epoch": 0.29801790924640303,
+      "grad_norm": 0.1559547808493094,
+      "learning_rate": 2e-05,
+      "loss": 5.2882,
+      "step": 4443
+    },
+    {
+      "epoch": 0.298084985075628,
+      "grad_norm": 0.150176601761596,
+      "learning_rate": 2e-05,
+      "loss": 5.4001,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2981520609048529,
+      "grad_norm": 0.1511907626233967,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 4445
+    },
+    {
+      "epoch": 0.29821913673407785,
+      "grad_norm": 0.14704256311815742,
+      "learning_rate": 2e-05,
+      "loss": 5.3864,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2982862125633028,
+      "grad_norm": 0.1564172063278242,
+      "learning_rate": 2e-05,
+      "loss": 5.475,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2983532883925277,
+      "grad_norm": 0.14431487494891698,
+      "learning_rate": 2e-05,
+      "loss": 5.3474,
+      "step": 4448
+    },
+    {
+      "epoch": 0.29842036422175267,
+      "grad_norm": 0.15261969486484003,
+      "learning_rate": 2e-05,
+      "loss": 5.3892,
+      "step": 4449
+    },
+    {
+      "epoch": 0.2984874400509776,
+      "grad_norm": 0.1572944377686291,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 4450
+    },
+    {
+      "epoch": 0.29855451588020254,
+      "grad_norm": 0.1486465518305097,
+      "learning_rate": 2e-05,
+      "loss": 5.6022,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2986215917094275,
+      "grad_norm": 0.14185341573847346,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 4452
+    },
+    {
+      "epoch": 0.2986886675386524,
+      "grad_norm": 0.15008265859735462,
+      "learning_rate": 2e-05,
+      "loss": 5.3657,
+      "step": 4453
+    },
+    {
+      "epoch": 0.29875574336787736,
+      "grad_norm": 0.15177590647888936,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 4454
+    },
+    {
+      "epoch": 0.29882281919710235,
+      "grad_norm": 0.145296358062232,
+      "learning_rate": 2e-05,
+      "loss": 5.5344,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2988898950263273,
+      "grad_norm": 0.14393210131626882,
+      "learning_rate": 2e-05,
+      "loss": 5.3791,
+      "step": 4456
+    },
+    {
+      "epoch": 0.29895697085555223,
+      "grad_norm": 0.15042910440636892,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 4457
+    },
+    {
+      "epoch": 0.29902404668477717,
+      "grad_norm": 0.15163348213761865,
+      "learning_rate": 2e-05,
+      "loss": 5.3517,
+      "step": 4458
+    },
+    {
+      "epoch": 0.2990911225140021,
+      "grad_norm": 0.14320934245938174,
+      "learning_rate": 2e-05,
+      "loss": 5.575,
+      "step": 4459
+    },
+    {
+      "epoch": 0.29915819834322704,
+      "grad_norm": 0.14826574334609666,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 4460
+    },
+    {
+      "epoch": 0.299225274172452,
+      "grad_norm": 0.15028397313407516,
+      "learning_rate": 2e-05,
+      "loss": 5.4854,
+      "step": 4461
+    },
+    {
+      "epoch": 0.2992923500016769,
+      "grad_norm": 0.141091426544405,
+      "learning_rate": 2e-05,
+      "loss": 5.5127,
+      "step": 4462
+    },
+    {
+      "epoch": 0.29935942583090186,
+      "grad_norm": 0.15684565249352467,
+      "learning_rate": 2e-05,
+      "loss": 5.3202,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2994265016601268,
+      "grad_norm": 0.14750241164647332,
+      "learning_rate": 2e-05,
+      "loss": 5.4648,
+      "step": 4464
+    },
+    {
+      "epoch": 0.29949357748935174,
+      "grad_norm": 0.15276471799981303,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2995606533185767,
+      "grad_norm": 0.15058568726440405,
+      "learning_rate": 2e-05,
+      "loss": 5.5332,
+      "step": 4466
+    },
+    {
+      "epoch": 0.2996277291478016,
+      "grad_norm": 0.1377913711056494,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 4467
+    },
+    {
+      "epoch": 0.29969480497702655,
+      "grad_norm": 0.14953139795101247,
+      "learning_rate": 2e-05,
+      "loss": 5.5087,
+      "step": 4468
+    },
+    {
+      "epoch": 0.2997618808062515,
+      "grad_norm": 0.14765105117349295,
+      "learning_rate": 2e-05,
+      "loss": 5.521,
+      "step": 4469
+    },
+    {
+      "epoch": 0.29982895663547643,
+      "grad_norm": 0.15276931099578667,
+      "learning_rate": 2e-05,
+      "loss": 5.5692,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29989603246470137,
+      "grad_norm": 0.14682924900176064,
+      "learning_rate": 2e-05,
+      "loss": 5.5785,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2999631082939263,
+      "grad_norm": 0.1486577590134233,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 4472
+    },
+    {
+      "epoch": 0.30003018412315124,
+      "grad_norm": 0.1511372544516596,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3000972599523762,
+      "grad_norm": 0.14873064689260387,
+      "learning_rate": 2e-05,
+      "loss": 5.4424,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3001643357816011,
+      "grad_norm": 0.15128949770112074,
+      "learning_rate": 2e-05,
+      "loss": 5.5394,
+      "step": 4475
+    },
+    {
+      "epoch": 0.30023141161082606,
+      "grad_norm": 0.15151526690915973,
+      "learning_rate": 2e-05,
+      "loss": 5.5082,
+      "step": 4476
+    },
+    {
+      "epoch": 0.300298487440051,
+      "grad_norm": 0.15006466407201843,
+      "learning_rate": 2e-05,
+      "loss": 5.3463,
+      "step": 4477
+    },
+    {
+      "epoch": 0.30036556326927594,
+      "grad_norm": 0.15122503133048854,
+      "learning_rate": 2e-05,
+      "loss": 5.4863,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3004326390985009,
+      "grad_norm": 0.1475357603410436,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3004997149277258,
+      "grad_norm": 0.1446218730006559,
+      "learning_rate": 2e-05,
+      "loss": 5.4818,
+      "step": 4480
+    },
+    {
+      "epoch": 0.30056679075695075,
+      "grad_norm": 0.14622607818349256,
+      "learning_rate": 2e-05,
+      "loss": 5.5227,
+      "step": 4481
+    },
+    {
+      "epoch": 0.3006338665861757,
+      "grad_norm": 0.1494675085489273,
+      "learning_rate": 2e-05,
+      "loss": 5.4127,
+      "step": 4482
+    },
+    {
+      "epoch": 0.30070094241540063,
+      "grad_norm": 0.15390793426753374,
+      "learning_rate": 2e-05,
+      "loss": 5.4985,
+      "step": 4483
+    },
+    {
+      "epoch": 0.30076801824462557,
+      "grad_norm": 0.14558580248121492,
+      "learning_rate": 2e-05,
+      "loss": 5.4427,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3008350940738505,
+      "grad_norm": 0.15178572945715418,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 4485
+    },
+    {
+      "epoch": 0.30090216990307544,
+      "grad_norm": 0.14795413195347523,
+      "learning_rate": 2e-05,
+      "loss": 5.4763,
+      "step": 4486
+    },
+    {
+      "epoch": 0.3009692457323004,
+      "grad_norm": 0.15333680684886275,
+      "learning_rate": 2e-05,
+      "loss": 5.577,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3010363215615253,
+      "grad_norm": 0.16100091253288232,
+      "learning_rate": 2e-05,
+      "loss": 5.4974,
+      "step": 4488
+    },
+    {
+      "epoch": 0.30110339739075026,
+      "grad_norm": 0.1463599293501929,
+      "learning_rate": 2e-05,
+      "loss": 5.4892,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3011704732199752,
+      "grad_norm": 0.15118176103732966,
+      "learning_rate": 2e-05,
+      "loss": 5.5019,
+      "step": 4490
+    },
+    {
+      "epoch": 0.30123754904920014,
+      "grad_norm": 0.14889063309329517,
+      "learning_rate": 2e-05,
+      "loss": 5.4498,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3013046248784251,
+      "grad_norm": 0.14311443594143256,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 4492
+    },
+    {
+      "epoch": 0.30137170070765,
+      "grad_norm": 0.14672927145288175,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 4493
+    },
+    {
+      "epoch": 0.30143877653687495,
+      "grad_norm": 0.14672386982399532,
+      "learning_rate": 2e-05,
+      "loss": 5.3343,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3015058523660999,
+      "grad_norm": 0.1464856276896854,
+      "learning_rate": 2e-05,
+      "loss": 5.4957,
+      "step": 4495
+    },
+    {
+      "epoch": 0.30157292819532483,
+      "grad_norm": 0.14516111755585517,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 4496
+    },
+    {
+      "epoch": 0.30164000402454977,
+      "grad_norm": 0.14823654075427836,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3017070798537747,
+      "grad_norm": 0.1532013366040041,
+      "learning_rate": 2e-05,
+      "loss": 5.3846,
+      "step": 4498
+    },
+    {
+      "epoch": 0.30177415568299965,
+      "grad_norm": 0.1414635707436847,
+      "learning_rate": 2e-05,
+      "loss": 5.4443,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3018412315122246,
+      "grad_norm": 0.14434118253351771,
+      "learning_rate": 2e-05,
+      "loss": 5.3085,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3019083073414495,
+      "grad_norm": 0.1412344956648059,
+      "learning_rate": 2e-05,
+      "loss": 5.4884,
+      "step": 4501
+    },
+    {
+      "epoch": 0.30197538317067446,
+      "grad_norm": 0.14170130000397244,
+      "learning_rate": 2e-05,
+      "loss": 5.5667,
+      "step": 4502
+    },
+    {
+      "epoch": 0.3020424589998994,
+      "grad_norm": 0.1508414484737552,
+      "learning_rate": 2e-05,
+      "loss": 5.4676,
+      "step": 4503
+    },
+    {
+      "epoch": 0.30210953482912434,
+      "grad_norm": 0.14430825440521997,
+      "learning_rate": 2e-05,
+      "loss": 5.4571,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3021766106583493,
+      "grad_norm": 0.14434438836071128,
+      "learning_rate": 2e-05,
+      "loss": 5.3389,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3022436864875742,
+      "grad_norm": 0.14636843486832962,
+      "learning_rate": 2e-05,
+      "loss": 5.457,
+      "step": 4506
+    },
+    {
+      "epoch": 0.30231076231679915,
+      "grad_norm": 0.14597823111871303,
+      "learning_rate": 2e-05,
+      "loss": 5.3266,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3023778381460241,
+      "grad_norm": 0.1444290340959737,
+      "learning_rate": 2e-05,
+      "loss": 5.3679,
+      "step": 4508
+    },
+    {
+      "epoch": 0.30244491397524903,
+      "grad_norm": 0.14435353593371408,
+      "learning_rate": 2e-05,
+      "loss": 5.4802,
+      "step": 4509
+    },
+    {
+      "epoch": 0.30251198980447397,
+      "grad_norm": 0.14427640301724307,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3025790656336989,
+      "grad_norm": 0.1493448607322142,
+      "learning_rate": 2e-05,
+      "loss": 5.3936,
+      "step": 4511
+    },
+    {
+      "epoch": 0.30264614146292385,
+      "grad_norm": 0.1511831575627833,
+      "learning_rate": 2e-05,
+      "loss": 5.5371,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3027132172921488,
+      "grad_norm": 0.14267054118883607,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3027802931213737,
+      "grad_norm": 0.14791303118277263,
+      "learning_rate": 2e-05,
+      "loss": 5.4798,
+      "step": 4514
+    },
+    {
+      "epoch": 0.30284736895059866,
+      "grad_norm": 0.1489424701503919,
+      "learning_rate": 2e-05,
+      "loss": 5.5886,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3029144447798236,
+      "grad_norm": 0.15856959404029955,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 4516
+    },
+    {
+      "epoch": 0.30298152060904854,
+      "grad_norm": 0.1468613996303402,
+      "learning_rate": 2e-05,
+      "loss": 5.3799,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3030485964382735,
+      "grad_norm": 0.14569601653461278,
+      "learning_rate": 2e-05,
+      "loss": 5.5053,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3031156722674984,
+      "grad_norm": 0.1437530933637486,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 4519
+    },
+    {
+      "epoch": 0.30318274809672335,
+      "grad_norm": 0.1429483075687242,
+      "learning_rate": 2e-05,
+      "loss": 5.3659,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3032498239259483,
+      "grad_norm": 0.14823019972115709,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 4521
+    },
+    {
+      "epoch": 0.30331689975517323,
+      "grad_norm": 0.14118602569998112,
+      "learning_rate": 2e-05,
+      "loss": 5.4246,
+      "step": 4522
+    },
+    {
+      "epoch": 0.30338397558439817,
+      "grad_norm": 0.1367046027499298,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3034510514136231,
+      "grad_norm": 0.14910539531423775,
+      "learning_rate": 2e-05,
+      "loss": 5.5252,
+      "step": 4524
+    },
+    {
+      "epoch": 0.30351812724284805,
+      "grad_norm": 0.14127945636164846,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 4525
+    },
+    {
+      "epoch": 0.303585203072073,
+      "grad_norm": 0.14298771721302891,
+      "learning_rate": 2e-05,
+      "loss": 5.4864,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3036522789012979,
+      "grad_norm": 0.14819389870927663,
+      "learning_rate": 2e-05,
+      "loss": 5.3822,
+      "step": 4527
+    },
+    {
+      "epoch": 0.30371935473052286,
+      "grad_norm": 0.15159666488472187,
+      "learning_rate": 2e-05,
+      "loss": 5.5236,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3037864305597478,
+      "grad_norm": 0.1519828903230427,
+      "learning_rate": 2e-05,
+      "loss": 5.5077,
+      "step": 4529
+    },
+    {
+      "epoch": 0.30385350638897274,
+      "grad_norm": 0.148337067669773,
+      "learning_rate": 2e-05,
+      "loss": 5.4608,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3039205822181977,
+      "grad_norm": 0.1539696762032026,
+      "learning_rate": 2e-05,
+      "loss": 5.5365,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3039876580474226,
+      "grad_norm": 0.1462703760903503,
+      "learning_rate": 2e-05,
+      "loss": 5.3581,
+      "step": 4532
+    },
+    {
+      "epoch": 0.30405473387664755,
+      "grad_norm": 0.14747877777618507,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 4533
+    },
+    {
+      "epoch": 0.3041218097058725,
+      "grad_norm": 0.14264505851686338,
+      "learning_rate": 2e-05,
+      "loss": 5.4042,
+      "step": 4534
+    },
+    {
+      "epoch": 0.30418888553509743,
+      "grad_norm": 0.14234372042259075,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 4535
+    },
+    {
+      "epoch": 0.30425596136432237,
+      "grad_norm": 0.14334791751630266,
+      "learning_rate": 2e-05,
+      "loss": 5.4058,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3043230371935473,
+      "grad_norm": 0.15336908681458847,
+      "learning_rate": 2e-05,
+      "loss": 5.3469,
+      "step": 4537
+    },
+    {
+      "epoch": 0.30439011302277225,
+      "grad_norm": 0.15145148553399412,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 4538
+    },
+    {
+      "epoch": 0.3044571888519972,
+      "grad_norm": 0.15233190838675115,
+      "learning_rate": 2e-05,
+      "loss": 5.3074,
+      "step": 4539
+    },
+    {
+      "epoch": 0.3045242646812221,
+      "grad_norm": 0.14294687996039007,
+      "learning_rate": 2e-05,
+      "loss": 5.3601,
+      "step": 4540
+    },
+    {
+      "epoch": 0.30459134051044706,
+      "grad_norm": 0.1481733919274403,
+      "learning_rate": 2e-05,
+      "loss": 5.354,
+      "step": 4541
+    },
+    {
+      "epoch": 0.304658416339672,
+      "grad_norm": 0.15027888735659364,
+      "learning_rate": 2e-05,
+      "loss": 5.5038,
+      "step": 4542
+    },
+    {
+      "epoch": 0.30472549216889694,
+      "grad_norm": 0.15709150889754572,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 4543
+    },
+    {
+      "epoch": 0.3047925679981219,
+      "grad_norm": 0.14912549952516282,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3048596438273468,
+      "grad_norm": 0.14223661685924016,
+      "learning_rate": 2e-05,
+      "loss": 5.4299,
+      "step": 4545
+    },
+    {
+      "epoch": 0.30492671965657175,
+      "grad_norm": 0.14841725745071396,
+      "learning_rate": 2e-05,
+      "loss": 5.3177,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3049937954857967,
+      "grad_norm": 0.15274045397357197,
+      "learning_rate": 2e-05,
+      "loss": 5.69,
+      "step": 4547
+    },
+    {
+      "epoch": 0.30506087131502163,
+      "grad_norm": 0.14604600946609206,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 4548
+    },
+    {
+      "epoch": 0.30512794714424657,
+      "grad_norm": 0.1457269834699606,
+      "learning_rate": 2e-05,
+      "loss": 5.461,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3051950229734715,
+      "grad_norm": 0.15264609031276527,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 4550
+    },
+    {
+      "epoch": 0.30526209880269645,
+      "grad_norm": 0.14868290403352563,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3053291746319214,
+      "grad_norm": 0.1457399706390801,
+      "learning_rate": 2e-05,
+      "loss": 5.3697,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3053962504611463,
+      "grad_norm": 0.1546530429916773,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 4553
+    },
+    {
+      "epoch": 0.30546332629037126,
+      "grad_norm": 0.15549605671735658,
+      "learning_rate": 2e-05,
+      "loss": 5.3758,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3055304021195962,
+      "grad_norm": 0.14831136184485863,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 4555
+    },
+    {
+      "epoch": 0.30559747794882114,
+      "grad_norm": 0.14717603234018364,
+      "learning_rate": 2e-05,
+      "loss": 5.4754,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3056645537780461,
+      "grad_norm": 0.15738969161474375,
+      "learning_rate": 2e-05,
+      "loss": 5.4012,
+      "step": 4557
+    },
+    {
+      "epoch": 0.305731629607271,
+      "grad_norm": 0.14801873205239519,
+      "learning_rate": 2e-05,
+      "loss": 5.4764,
+      "step": 4558
+    },
+    {
+      "epoch": 0.30579870543649595,
+      "grad_norm": 0.14780681581374525,
+      "learning_rate": 2e-05,
+      "loss": 5.5604,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3058657812657209,
+      "grad_norm": 0.14617369869722138,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 4560
+    },
+    {
+      "epoch": 0.30593285709494583,
+      "grad_norm": 0.15357989315622822,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 4561
+    },
+    {
+      "epoch": 0.30599993292417077,
+      "grad_norm": 0.15368770081449487,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3060670087533957,
+      "grad_norm": 0.1450983363320904,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 4563
+    },
+    {
+      "epoch": 0.30613408458262065,
+      "grad_norm": 0.15652310656206642,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3062011604118456,
+      "grad_norm": 0.15109419190851142,
+      "learning_rate": 2e-05,
+      "loss": 5.4071,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3062682362410705,
+      "grad_norm": 0.15211711780011758,
+      "learning_rate": 2e-05,
+      "loss": 5.3679,
+      "step": 4566
+    },
+    {
+      "epoch": 0.30633531207029546,
+      "grad_norm": 0.14703112798997636,
+      "learning_rate": 2e-05,
+      "loss": 5.4159,
+      "step": 4567
+    },
+    {
+      "epoch": 0.3064023878995204,
+      "grad_norm": 0.15022335343071944,
+      "learning_rate": 2e-05,
+      "loss": 5.3422,
+      "step": 4568
+    },
+    {
+      "epoch": 0.30646946372874534,
+      "grad_norm": 0.14104542680138143,
+      "learning_rate": 2e-05,
+      "loss": 5.3225,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3065365395579703,
+      "grad_norm": 0.1486877583072616,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3066036153871952,
+      "grad_norm": 0.14443600241853174,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 4571
+    },
+    {
+      "epoch": 0.30667069121642015,
+      "grad_norm": 0.14693169979282228,
+      "learning_rate": 2e-05,
+      "loss": 5.3677,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3067377670456451,
+      "grad_norm": 0.1413009762680042,
+      "learning_rate": 2e-05,
+      "loss": 5.3865,
+      "step": 4573
+    },
+    {
+      "epoch": 0.30680484287487003,
+      "grad_norm": 0.14801137075445153,
+      "learning_rate": 2e-05,
+      "loss": 5.4652,
+      "step": 4574
+    },
+    {
+      "epoch": 0.30687191870409497,
+      "grad_norm": 0.14248852367095532,
+      "learning_rate": 2e-05,
+      "loss": 5.3406,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3069389945333199,
+      "grad_norm": 0.13969273350202857,
+      "learning_rate": 2e-05,
+      "loss": 5.4499,
+      "step": 4576
+    },
+    {
+      "epoch": 0.30700607036254485,
+      "grad_norm": 0.15590339112791815,
+      "learning_rate": 2e-05,
+      "loss": 5.3756,
+      "step": 4577
+    },
+    {
+      "epoch": 0.3070731461917698,
+      "grad_norm": 0.15200230678681445,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3071402220209947,
+      "grad_norm": 0.14363550205383782,
+      "learning_rate": 2e-05,
+      "loss": 5.3754,
+      "step": 4579
+    },
+    {
+      "epoch": 0.30720729785021966,
+      "grad_norm": 0.1447628684834803,
+      "learning_rate": 2e-05,
+      "loss": 5.4548,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3072743736794446,
+      "grad_norm": 0.15753178554253797,
+      "learning_rate": 2e-05,
+      "loss": 5.3933,
+      "step": 4581
+    },
+    {
+      "epoch": 0.30734144950866954,
+      "grad_norm": 0.14679519097220706,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3074085253378945,
+      "grad_norm": 0.14363546232400354,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3074756011671194,
+      "grad_norm": 0.15183068539389968,
+      "learning_rate": 2e-05,
+      "loss": 5.5223,
+      "step": 4584
+    },
+    {
+      "epoch": 0.30754267699634436,
+      "grad_norm": 0.14038481092313318,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3076097528255693,
+      "grad_norm": 0.1418424237792973,
+      "learning_rate": 2e-05,
+      "loss": 5.3276,
+      "step": 4586
+    },
+    {
+      "epoch": 0.30767682865479423,
+      "grad_norm": 0.15383409663030792,
+      "learning_rate": 2e-05,
+      "loss": 5.444,
+      "step": 4587
+    },
+    {
+      "epoch": 0.30774390448401917,
+      "grad_norm": 0.15967788261473903,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 4588
+    },
+    {
+      "epoch": 0.3078109803132441,
+      "grad_norm": 0.14969094118129644,
+      "learning_rate": 2e-05,
+      "loss": 5.4152,
+      "step": 4589
+    },
+    {
+      "epoch": 0.30787805614246905,
+      "grad_norm": 0.14745971048770765,
+      "learning_rate": 2e-05,
+      "loss": 5.3486,
+      "step": 4590
+    },
+    {
+      "epoch": 0.307945131971694,
+      "grad_norm": 0.15745697954461246,
+      "learning_rate": 2e-05,
+      "loss": 5.6226,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3080122078009189,
+      "grad_norm": 0.15786906703880435,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 4592
+    },
+    {
+      "epoch": 0.30807928363014386,
+      "grad_norm": 0.14501931005816235,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 4593
+    },
+    {
+      "epoch": 0.3081463594593688,
+      "grad_norm": 0.14361598899551511,
+      "learning_rate": 2e-05,
+      "loss": 5.5272,
+      "step": 4594
+    },
+    {
+      "epoch": 0.30821343528859374,
+      "grad_norm": 0.14792513235163723,
+      "learning_rate": 2e-05,
+      "loss": 5.4956,
+      "step": 4595
+    },
+    {
+      "epoch": 0.3082805111178187,
+      "grad_norm": 0.15465535288458548,
+      "learning_rate": 2e-05,
+      "loss": 5.3506,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3083475869470436,
+      "grad_norm": 0.14495288651397753,
+      "learning_rate": 2e-05,
+      "loss": 5.4252,
+      "step": 4597
+    },
+    {
+      "epoch": 0.30841466277626856,
+      "grad_norm": 0.15300487205158467,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3084817386054935,
+      "grad_norm": 0.14876130399963258,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 4599
+    },
+    {
+      "epoch": 0.30854881443471843,
+      "grad_norm": 0.1423464324413054,
+      "learning_rate": 2e-05,
+      "loss": 5.3857,
+      "step": 4600
+    },
+    {
+      "epoch": 0.30861589026394337,
+      "grad_norm": 0.15731168224052566,
+      "learning_rate": 2e-05,
+      "loss": 5.5269,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3086829660931683,
+      "grad_norm": 0.1620010090212436,
+      "learning_rate": 2e-05,
+      "loss": 5.3271,
+      "step": 4602
+    },
+    {
+      "epoch": 0.30875004192239325,
+      "grad_norm": 0.14991006628716916,
+      "learning_rate": 2e-05,
+      "loss": 5.4247,
+      "step": 4603
+    },
+    {
+      "epoch": 0.3088171177516182,
+      "grad_norm": 0.14214095946765495,
+      "learning_rate": 2e-05,
+      "loss": 5.4,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3088841935808431,
+      "grad_norm": 0.16150230263866622,
+      "learning_rate": 2e-05,
+      "loss": 5.4253,
+      "step": 4605
+    },
+    {
+      "epoch": 0.30895126941006806,
+      "grad_norm": 0.16026261125907118,
+      "learning_rate": 2e-05,
+      "loss": 5.5439,
+      "step": 4606
+    },
+    {
+      "epoch": 0.309018345239293,
+      "grad_norm": 0.15536526973490877,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 4607
+    },
+    {
+      "epoch": 0.30908542106851794,
+      "grad_norm": 0.15216469849732597,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 4608
+    },
+    {
+      "epoch": 0.3091524968977429,
+      "grad_norm": 0.14824895901006058,
+      "learning_rate": 2e-05,
+      "loss": 5.407,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3092195727269678,
+      "grad_norm": 0.1522997881433914,
+      "learning_rate": 2e-05,
+      "loss": 5.3092,
+      "step": 4610
+    },
+    {
+      "epoch": 0.30928664855619276,
+      "grad_norm": 0.15110638672672808,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3093537243854177,
+      "grad_norm": 0.14537346530778378,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 4612
+    },
+    {
+      "epoch": 0.30942080021464263,
+      "grad_norm": 0.15775746568947666,
+      "learning_rate": 2e-05,
+      "loss": 5.4528,
+      "step": 4613
+    },
+    {
+      "epoch": 0.30948787604386757,
+      "grad_norm": 0.15533239438218802,
+      "learning_rate": 2e-05,
+      "loss": 5.3861,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3095549518730925,
+      "grad_norm": 0.14632138277532383,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 4615
+    },
+    {
+      "epoch": 0.30962202770231745,
+      "grad_norm": 0.15123899816137057,
+      "learning_rate": 2e-05,
+      "loss": 5.5185,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3096891035315424,
+      "grad_norm": 0.14271568522137856,
+      "learning_rate": 2e-05,
+      "loss": 5.347,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3097561793607673,
+      "grad_norm": 0.14510994714522116,
+      "learning_rate": 2e-05,
+      "loss": 5.4973,
+      "step": 4618
+    },
+    {
+      "epoch": 0.30982325518999226,
+      "grad_norm": 0.15017700552454685,
+      "learning_rate": 2e-05,
+      "loss": 5.5167,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3098903310192172,
+      "grad_norm": 0.15142894928214884,
+      "learning_rate": 2e-05,
+      "loss": 5.4095,
+      "step": 4620
+    },
+    {
+      "epoch": 0.30995740684844214,
+      "grad_norm": 0.14225233338234333,
+      "learning_rate": 2e-05,
+      "loss": 5.3923,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3100244826776671,
+      "grad_norm": 0.1473848677660478,
+      "learning_rate": 2e-05,
+      "loss": 5.4573,
+      "step": 4622
+    },
+    {
+      "epoch": 0.310091558506892,
+      "grad_norm": 0.14288511483486166,
+      "learning_rate": 2e-05,
+      "loss": 5.6104,
+      "step": 4623
+    },
+    {
+      "epoch": 0.31015863433611696,
+      "grad_norm": 0.15089326267976996,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3102257101653419,
+      "grad_norm": 0.14846862621841705,
+      "learning_rate": 2e-05,
+      "loss": 5.234,
+      "step": 4625
+    },
+    {
+      "epoch": 0.31029278599456683,
+      "grad_norm": 0.15060386874764256,
+      "learning_rate": 2e-05,
+      "loss": 5.5201,
+      "step": 4626
+    },
+    {
+      "epoch": 0.31035986182379177,
+      "grad_norm": 0.15090976330737346,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3104269376530167,
+      "grad_norm": 0.14073314404816134,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 4628
+    },
+    {
+      "epoch": 0.31049401348224165,
+      "grad_norm": 0.15323888058481783,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3105610893114666,
+      "grad_norm": 0.14716898982435206,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3106281651406915,
+      "grad_norm": 0.14605328877294618,
+      "learning_rate": 2e-05,
+      "loss": 5.5132,
+      "step": 4631
+    },
+    {
+      "epoch": 0.31069524096991646,
+      "grad_norm": 0.1425854716182849,
+      "learning_rate": 2e-05,
+      "loss": 5.585,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3107623167991414,
+      "grad_norm": 0.15604133264864842,
+      "learning_rate": 2e-05,
+      "loss": 5.4403,
+      "step": 4633
+    },
+    {
+      "epoch": 0.31082939262836634,
+      "grad_norm": 0.15036415422329483,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3108964684575913,
+      "grad_norm": 0.14424847348985592,
+      "learning_rate": 2e-05,
+      "loss": 5.4606,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3109635442868162,
+      "grad_norm": 0.1499394480389941,
+      "learning_rate": 2e-05,
+      "loss": 5.4863,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3110306201160412,
+      "grad_norm": 0.1407350789205912,
+      "learning_rate": 2e-05,
+      "loss": 5.3907,
+      "step": 4637
+    },
+    {
+      "epoch": 0.31109769594526615,
+      "grad_norm": 0.14619002858065103,
+      "learning_rate": 2e-05,
+      "loss": 5.3948,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3111647717744911,
+      "grad_norm": 0.1522452265177975,
+      "learning_rate": 2e-05,
+      "loss": 5.5317,
+      "step": 4639
+    },
+    {
+      "epoch": 0.311231847603716,
+      "grad_norm": 0.1431449930115255,
+      "learning_rate": 2e-05,
+      "loss": 5.4087,
+      "step": 4640
+    },
+    {
+      "epoch": 0.31129892343294097,
+      "grad_norm": 0.14315969648919447,
+      "learning_rate": 2e-05,
+      "loss": 5.3976,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3113659992621659,
+      "grad_norm": 0.13943149983887937,
+      "learning_rate": 2e-05,
+      "loss": 5.5114,
+      "step": 4642
+    },
+    {
+      "epoch": 0.31143307509139084,
+      "grad_norm": 0.14567244969960513,
+      "learning_rate": 2e-05,
+      "loss": 5.4642,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3115001509206158,
+      "grad_norm": 0.14120557048515864,
+      "learning_rate": 2e-05,
+      "loss": 5.5554,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3115672267498407,
+      "grad_norm": 0.15183074023545304,
+      "learning_rate": 2e-05,
+      "loss": 5.5392,
+      "step": 4645
+    },
+    {
+      "epoch": 0.31163430257906566,
+      "grad_norm": 0.14946198426061172,
+      "learning_rate": 2e-05,
+      "loss": 5.5169,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3117013784082906,
+      "grad_norm": 0.14417283524310517,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 4647
+    },
+    {
+      "epoch": 0.31176845423751554,
+      "grad_norm": 0.14732824013239007,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3118355300667405,
+      "grad_norm": 0.14412069077433934,
+      "learning_rate": 2e-05,
+      "loss": 5.3134,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3119026058959654,
+      "grad_norm": 0.14451732839962822,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 4650
+    },
+    {
+      "epoch": 0.31196968172519035,
+      "grad_norm": 0.14373478845230606,
+      "learning_rate": 2e-05,
+      "loss": 5.4058,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3120367575544153,
+      "grad_norm": 0.1436839560481511,
+      "learning_rate": 2e-05,
+      "loss": 5.4031,
+      "step": 4652
+    },
+    {
+      "epoch": 0.31210383338364023,
+      "grad_norm": 0.14591202190633557,
+      "learning_rate": 2e-05,
+      "loss": 5.3453,
+      "step": 4653
+    },
+    {
+      "epoch": 0.31217090921286517,
+      "grad_norm": 0.14228471315288413,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3122379850420901,
+      "grad_norm": 0.14237126606985576,
+      "learning_rate": 2e-05,
+      "loss": 5.2793,
+      "step": 4655
+    },
+    {
+      "epoch": 0.31230506087131504,
+      "grad_norm": 0.1410917291965898,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 4656
+    },
+    {
+      "epoch": 0.31237213670054,
+      "grad_norm": 0.14002631141178787,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3124392125297649,
+      "grad_norm": 0.1411799617673968,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 4658
+    },
+    {
+      "epoch": 0.31250628835898986,
+      "grad_norm": 0.1463924059688603,
+      "learning_rate": 2e-05,
+      "loss": 5.4871,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3125733641882148,
+      "grad_norm": 0.15064391107286537,
+      "learning_rate": 2e-05,
+      "loss": 5.5593,
+      "step": 4660
+    },
+    {
+      "epoch": 0.31264044001743974,
+      "grad_norm": 0.1461214882747298,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3127075158466647,
+      "grad_norm": 0.14840558265759152,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3127745916758896,
+      "grad_norm": 0.1536926969546078,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 4663
+    },
+    {
+      "epoch": 0.31284166750511455,
+      "grad_norm": 0.1445110462687453,
+      "learning_rate": 2e-05,
+      "loss": 5.4134,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3129087433343395,
+      "grad_norm": 0.14946865419914923,
+      "learning_rate": 2e-05,
+      "loss": 5.2978,
+      "step": 4665
+    },
+    {
+      "epoch": 0.31297581916356443,
+      "grad_norm": 0.15711096753468695,
+      "learning_rate": 2e-05,
+      "loss": 5.5948,
+      "step": 4666
+    },
+    {
+      "epoch": 0.31304289499278937,
+      "grad_norm": 0.14275900869180216,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3131099708220143,
+      "grad_norm": 0.1455105983809484,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 4668
+    },
+    {
+      "epoch": 0.31317704665123924,
+      "grad_norm": 0.15330006678307637,
+      "learning_rate": 2e-05,
+      "loss": 5.3295,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3132441224804642,
+      "grad_norm": 0.15319858311364123,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3133111983096891,
+      "grad_norm": 0.14349906790757497,
+      "learning_rate": 2e-05,
+      "loss": 5.5378,
+      "step": 4671
+    },
+    {
+      "epoch": 0.31337827413891406,
+      "grad_norm": 0.150311702395463,
+      "learning_rate": 2e-05,
+      "loss": 5.5919,
+      "step": 4672
+    },
+    {
+      "epoch": 0.313445349968139,
+      "grad_norm": 0.15396721973440974,
+      "learning_rate": 2e-05,
+      "loss": 5.4483,
+      "step": 4673
+    },
+    {
+      "epoch": 0.31351242579736394,
+      "grad_norm": 0.15316658396002145,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3135795016265889,
+      "grad_norm": 0.1478044495152701,
+      "learning_rate": 2e-05,
+      "loss": 5.6538,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3136465774558138,
+      "grad_norm": 0.14960402787176316,
+      "learning_rate": 2e-05,
+      "loss": 5.5027,
+      "step": 4676
+    },
+    {
+      "epoch": 0.31371365328503875,
+      "grad_norm": 0.15884806436096643,
+      "learning_rate": 2e-05,
+      "loss": 5.4894,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3137807291142637,
+      "grad_norm": 0.14946571708083475,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 4678
+    },
+    {
+      "epoch": 0.31384780494348863,
+      "grad_norm": 0.1503777717971093,
+      "learning_rate": 2e-05,
+      "loss": 5.5498,
+      "step": 4679
+    },
+    {
+      "epoch": 0.31391488077271357,
+      "grad_norm": 0.1440984086855228,
+      "learning_rate": 2e-05,
+      "loss": 5.3886,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3139819566019385,
+      "grad_norm": 0.14665558870279036,
+      "learning_rate": 2e-05,
+      "loss": 5.3958,
+      "step": 4681
+    },
+    {
+      "epoch": 0.31404903243116344,
+      "grad_norm": 0.14933084318356385,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3141161082603884,
+      "grad_norm": 0.14505875839129379,
+      "learning_rate": 2e-05,
+      "loss": 5.5749,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3141831840896133,
+      "grad_norm": 0.15484454505398815,
+      "learning_rate": 2e-05,
+      "loss": 5.3743,
+      "step": 4684
+    },
+    {
+      "epoch": 0.31425025991883826,
+      "grad_norm": 0.1420258687287775,
+      "learning_rate": 2e-05,
+      "loss": 5.5391,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3143173357480632,
+      "grad_norm": 0.14407515317823788,
+      "learning_rate": 2e-05,
+      "loss": 5.3228,
+      "step": 4686
+    },
+    {
+      "epoch": 0.31438441157728814,
+      "grad_norm": 0.15503544703983277,
+      "learning_rate": 2e-05,
+      "loss": 5.5453,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3144514874065131,
+      "grad_norm": 0.14749812950048496,
+      "learning_rate": 2e-05,
+      "loss": 5.3035,
+      "step": 4688
+    },
+    {
+      "epoch": 0.314518563235738,
+      "grad_norm": 0.1437354107601083,
+      "learning_rate": 2e-05,
+      "loss": 5.4418,
+      "step": 4689
+    },
+    {
+      "epoch": 0.31458563906496295,
+      "grad_norm": 0.14882058536314352,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3146527148941879,
+      "grad_norm": 0.14624550176637166,
+      "learning_rate": 2e-05,
+      "loss": 5.4727,
+      "step": 4691
+    },
+    {
+      "epoch": 0.31471979072341283,
+      "grad_norm": 0.1446039821704901,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 4692
+    },
+    {
+      "epoch": 0.31478686655263777,
+      "grad_norm": 0.14479325536998452,
+      "learning_rate": 2e-05,
+      "loss": 5.5231,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3148539423818627,
+      "grad_norm": 0.1579748244592186,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 4694
+    },
+    {
+      "epoch": 0.31492101821108764,
+      "grad_norm": 0.15290371296186447,
+      "learning_rate": 2e-05,
+      "loss": 5.3809,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3149880940403126,
+      "grad_norm": 0.14725952420414185,
+      "learning_rate": 2e-05,
+      "loss": 5.4038,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3150551698695375,
+      "grad_norm": 0.15125156856875202,
+      "learning_rate": 2e-05,
+      "loss": 5.4824,
+      "step": 4697
+    },
+    {
+      "epoch": 0.31512224569876246,
+      "grad_norm": 0.1531431360262526,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3151893215279874,
+      "grad_norm": 0.1517734706241143,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 4699
+    },
+    {
+      "epoch": 0.31525639735721234,
+      "grad_norm": 0.15012128768977626,
+      "learning_rate": 2e-05,
+      "loss": 5.5149,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3153234731864373,
+      "grad_norm": 0.1560666739708711,
+      "learning_rate": 2e-05,
+      "loss": 5.3988,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3153905490156622,
+      "grad_norm": 0.14615546409618727,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 4702
+    },
+    {
+      "epoch": 0.31545762484488715,
+      "grad_norm": 0.14962648930632258,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3155247006741121,
+      "grad_norm": 0.14319170306951273,
+      "learning_rate": 2e-05,
+      "loss": 5.3679,
+      "step": 4704
+    },
+    {
+      "epoch": 0.31559177650333703,
+      "grad_norm": 0.14483713386430172,
+      "learning_rate": 2e-05,
+      "loss": 5.5286,
+      "step": 4705
+    },
+    {
+      "epoch": 0.31565885233256197,
+      "grad_norm": 0.14592036567890643,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3157259281617869,
+      "grad_norm": 0.14722402600169546,
+      "learning_rate": 2e-05,
+      "loss": 5.3208,
+      "step": 4707
+    },
+    {
+      "epoch": 0.31579300399101184,
+      "grad_norm": 0.15117506411364792,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3158600798202368,
+      "grad_norm": 0.14185274858092073,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3159271556494617,
+      "grad_norm": 0.1534262341444003,
+      "learning_rate": 2e-05,
+      "loss": 5.3699,
+      "step": 4710
+    },
+    {
+      "epoch": 0.31599423147868666,
+      "grad_norm": 0.15169831531489686,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3160613073079116,
+      "grad_norm": 0.14446867376133682,
+      "learning_rate": 2e-05,
+      "loss": 5.3297,
+      "step": 4712
+    },
+    {
+      "epoch": 0.31612838313713654,
+      "grad_norm": 0.15366022082024802,
+      "learning_rate": 2e-05,
+      "loss": 5.3355,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3161954589663615,
+      "grad_norm": 0.14335065283408488,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3162625347955864,
+      "grad_norm": 0.15613503239847298,
+      "learning_rate": 2e-05,
+      "loss": 5.4607,
+      "step": 4715
+    },
+    {
+      "epoch": 0.31632961062481135,
+      "grad_norm": 0.14484645828050904,
+      "learning_rate": 2e-05,
+      "loss": 5.361,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3163966864540363,
+      "grad_norm": 0.1510328354254125,
+      "learning_rate": 2e-05,
+      "loss": 5.4835,
+      "step": 4717
+    },
+    {
+      "epoch": 0.31646376228326123,
+      "grad_norm": 0.14657485629072275,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 4718
+    },
+    {
+      "epoch": 0.31653083811248617,
+      "grad_norm": 0.1479704743625473,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3165979139417111,
+      "grad_norm": 0.1523568714420074,
+      "learning_rate": 2e-05,
+      "loss": 5.6355,
+      "step": 4720
+    },
+    {
+      "epoch": 0.31666498977093605,
+      "grad_norm": 0.14934225100829807,
+      "learning_rate": 2e-05,
+      "loss": 5.4797,
+      "step": 4721
+    },
+    {
+      "epoch": 0.316732065600161,
+      "grad_norm": 0.1470077641602299,
+      "learning_rate": 2e-05,
+      "loss": 5.3323,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3167991414293859,
+      "grad_norm": 0.1394436599797149,
+      "learning_rate": 2e-05,
+      "loss": 5.434,
+      "step": 4723
+    },
+    {
+      "epoch": 0.31686621725861086,
+      "grad_norm": 0.15439278899327236,
+      "learning_rate": 2e-05,
+      "loss": 5.5446,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3169332930878358,
+      "grad_norm": 0.14980120464455324,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 4725
+    },
+    {
+      "epoch": 0.31700036891706074,
+      "grad_norm": 0.1493621236521407,
+      "learning_rate": 2e-05,
+      "loss": 5.4467,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3170674447462857,
+      "grad_norm": 0.14429911955617658,
+      "learning_rate": 2e-05,
+      "loss": 5.4529,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3171345205755106,
+      "grad_norm": 0.1421161568822683,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 4728
+    },
+    {
+      "epoch": 0.31720159640473555,
+      "grad_norm": 0.14774180445775875,
+      "learning_rate": 2e-05,
+      "loss": 5.4473,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3172686722339605,
+      "grad_norm": 0.15995317087946379,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 4730
+    },
+    {
+      "epoch": 0.31733574806318543,
+      "grad_norm": 0.14866129658292343,
+      "learning_rate": 2e-05,
+      "loss": 5.5888,
+      "step": 4731
+    },
+    {
+      "epoch": 0.31740282389241037,
+      "grad_norm": 0.14985841669877367,
+      "learning_rate": 2e-05,
+      "loss": 5.6124,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3174698997216353,
+      "grad_norm": 0.16094435505092466,
+      "learning_rate": 2e-05,
+      "loss": 5.3304,
+      "step": 4733
+    },
+    {
+      "epoch": 0.31753697555086025,
+      "grad_norm": 0.15362408602840025,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3176040513800852,
+      "grad_norm": 0.143677071327506,
+      "learning_rate": 2e-05,
+      "loss": 5.3371,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3176711272093101,
+      "grad_norm": 0.14602987194825723,
+      "learning_rate": 2e-05,
+      "loss": 5.6483,
+      "step": 4736
+    },
+    {
+      "epoch": 0.31773820303853506,
+      "grad_norm": 0.15237480914998594,
+      "learning_rate": 2e-05,
+      "loss": 5.4591,
+      "step": 4737
+    },
+    {
+      "epoch": 0.31780527886776,
+      "grad_norm": 0.14824345319282412,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 4738
+    },
+    {
+      "epoch": 0.31787235469698494,
+      "grad_norm": 0.15883711706339843,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3179394305262099,
+      "grad_norm": 0.1454450550875837,
+      "learning_rate": 2e-05,
+      "loss": 5.6264,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3180065063554348,
+      "grad_norm": 0.14704749231694028,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 4741
+    },
+    {
+      "epoch": 0.31807358218465975,
+      "grad_norm": 0.14554647604095527,
+      "learning_rate": 2e-05,
+      "loss": 5.4988,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3181406580138847,
+      "grad_norm": 0.1511932394140102,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 4743
+    },
+    {
+      "epoch": 0.31820773384310963,
+      "grad_norm": 0.14707892299115866,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 4744
+    },
+    {
+      "epoch": 0.31827480967233457,
+      "grad_norm": 0.14467393200453876,
+      "learning_rate": 2e-05,
+      "loss": 5.5599,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3183418855015595,
+      "grad_norm": 0.15975680007363485,
+      "learning_rate": 2e-05,
+      "loss": 5.5861,
+      "step": 4746
+    },
+    {
+      "epoch": 0.31840896133078445,
+      "grad_norm": 0.15137620633119026,
+      "learning_rate": 2e-05,
+      "loss": 5.4424,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3184760371600094,
+      "grad_norm": 0.15268109377579053,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3185431129892343,
+      "grad_norm": 0.15046006544523788,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 4749
+    },
+    {
+      "epoch": 0.31861018881845926,
+      "grad_norm": 0.1658988526593255,
+      "learning_rate": 2e-05,
+      "loss": 5.5396,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3186772646476842,
+      "grad_norm": 0.14491326978578156,
+      "learning_rate": 2e-05,
+      "loss": 5.3779,
+      "step": 4751
+    },
+    {
+      "epoch": 0.31874434047690914,
+      "grad_norm": 0.15260638749243158,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3188114163061341,
+      "grad_norm": 0.14933540160204478,
+      "learning_rate": 2e-05,
+      "loss": 5.5724,
+      "step": 4753
+    },
+    {
+      "epoch": 0.318878492135359,
+      "grad_norm": 0.14962636880566121,
+      "learning_rate": 2e-05,
+      "loss": 5.4717,
+      "step": 4754
+    },
+    {
+      "epoch": 0.31894556796458395,
+      "grad_norm": 0.14644842386667636,
+      "learning_rate": 2e-05,
+      "loss": 5.3352,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3190126437938089,
+      "grad_norm": 0.15026614217265347,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 4756
+    },
+    {
+      "epoch": 0.31907971962303383,
+      "grad_norm": 0.14692957287857503,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 4757
+    },
+    {
+      "epoch": 0.31914679545225877,
+      "grad_norm": 0.14471908533492484,
+      "learning_rate": 2e-05,
+      "loss": 5.5519,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3192138712814837,
+      "grad_norm": 0.1491594873713398,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 4759
+    },
+    {
+      "epoch": 0.31928094711070865,
+      "grad_norm": 0.16606533319754455,
+      "learning_rate": 2e-05,
+      "loss": 5.4168,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3193480229399336,
+      "grad_norm": 0.14607620328453913,
+      "learning_rate": 2e-05,
+      "loss": 5.586,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3194150987691585,
+      "grad_norm": 0.14586120858903015,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 4762
+    },
+    {
+      "epoch": 0.31948217459838346,
+      "grad_norm": 0.15819008715948688,
+      "learning_rate": 2e-05,
+      "loss": 5.4789,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3195492504276084,
+      "grad_norm": 0.14949134406021342,
+      "learning_rate": 2e-05,
+      "loss": 5.5142,
+      "step": 4764
+    },
+    {
+      "epoch": 0.31961632625683334,
+      "grad_norm": 0.1466001625886673,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3196834020860583,
+      "grad_norm": 0.1477185343348984,
+      "learning_rate": 2e-05,
+      "loss": 5.4547,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3197504779152832,
+      "grad_norm": 0.14969009617074489,
+      "learning_rate": 2e-05,
+      "loss": 5.3961,
+      "step": 4767
+    },
+    {
+      "epoch": 0.31981755374450815,
+      "grad_norm": 0.1553312400981821,
+      "learning_rate": 2e-05,
+      "loss": 5.4109,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3198846295737331,
+      "grad_norm": 0.14465315410274657,
+      "learning_rate": 2e-05,
+      "loss": 5.4955,
+      "step": 4769
+    },
+    {
+      "epoch": 0.31995170540295803,
+      "grad_norm": 0.16184405939010724,
+      "learning_rate": 2e-05,
+      "loss": 5.4794,
+      "step": 4770
+    },
+    {
+      "epoch": 0.32001878123218297,
+      "grad_norm": 0.15251150252344112,
+      "learning_rate": 2e-05,
+      "loss": 5.2896,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3200858570614079,
+      "grad_norm": 0.15109039183776096,
+      "learning_rate": 2e-05,
+      "loss": 5.5683,
+      "step": 4772
+    },
+    {
+      "epoch": 0.32015293289063285,
+      "grad_norm": 0.1449767853838985,
+      "learning_rate": 2e-05,
+      "loss": 5.4546,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3202200087198578,
+      "grad_norm": 0.14694110909330158,
+      "learning_rate": 2e-05,
+      "loss": 5.3911,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3202870845490827,
+      "grad_norm": 0.15683261424371475,
+      "learning_rate": 2e-05,
+      "loss": 5.5406,
+      "step": 4775
+    },
+    {
+      "epoch": 0.32035416037830766,
+      "grad_norm": 0.14622637076989373,
+      "learning_rate": 2e-05,
+      "loss": 5.4904,
+      "step": 4776
+    },
+    {
+      "epoch": 0.3204212362075326,
+      "grad_norm": 0.14315566670875357,
+      "learning_rate": 2e-05,
+      "loss": 5.5556,
+      "step": 4777
+    },
+    {
+      "epoch": 0.32048831203675754,
+      "grad_norm": 0.1474608907668105,
+      "learning_rate": 2e-05,
+      "loss": 5.3365,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3205553878659825,
+      "grad_norm": 0.14853789818630225,
+      "learning_rate": 2e-05,
+      "loss": 5.5032,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3206224636952074,
+      "grad_norm": 0.15266675751804593,
+      "learning_rate": 2e-05,
+      "loss": 5.3744,
+      "step": 4780
+    },
+    {
+      "epoch": 0.32068953952443235,
+      "grad_norm": 0.14667154761382814,
+      "learning_rate": 2e-05,
+      "loss": 5.4037,
+      "step": 4781
+    },
+    {
+      "epoch": 0.3207566153536573,
+      "grad_norm": 0.14929532404162063,
+      "learning_rate": 2e-05,
+      "loss": 5.5752,
+      "step": 4782
+    },
+    {
+      "epoch": 0.32082369118288223,
+      "grad_norm": 0.1486821931201904,
+      "learning_rate": 2e-05,
+      "loss": 5.3801,
+      "step": 4783
+    },
+    {
+      "epoch": 0.32089076701210717,
+      "grad_norm": 0.14062531498184336,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3209578428413321,
+      "grad_norm": 0.14236755053261024,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 4785
+    },
+    {
+      "epoch": 0.32102491867055705,
+      "grad_norm": 0.14567562011579227,
+      "learning_rate": 2e-05,
+      "loss": 5.4372,
+      "step": 4786
+    },
+    {
+      "epoch": 0.321091994499782,
+      "grad_norm": 0.14268656111256953,
+      "learning_rate": 2e-05,
+      "loss": 5.5666,
+      "step": 4787
+    },
+    {
+      "epoch": 0.3211590703290069,
+      "grad_norm": 0.1482719639394982,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 4788
+    },
+    {
+      "epoch": 0.32122614615823186,
+      "grad_norm": 0.14691233222852676,
+      "learning_rate": 2e-05,
+      "loss": 5.4586,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3212932219874568,
+      "grad_norm": 0.14033597358985855,
+      "learning_rate": 2e-05,
+      "loss": 5.3343,
+      "step": 4790
+    },
+    {
+      "epoch": 0.32136029781668174,
+      "grad_norm": 0.14363998407166223,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 4791
+    },
+    {
+      "epoch": 0.3214273736459067,
+      "grad_norm": 0.1432881892323226,
+      "learning_rate": 2e-05,
+      "loss": 5.4977,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3214944494751316,
+      "grad_norm": 0.15048344149161572,
+      "learning_rate": 2e-05,
+      "loss": 5.6211,
+      "step": 4793
+    },
+    {
+      "epoch": 0.32156152530435655,
+      "grad_norm": 0.1421620678582059,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3216286011335815,
+      "grad_norm": 0.14581668119721064,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 4795
+    },
+    {
+      "epoch": 0.32169567696280643,
+      "grad_norm": 0.14136199042445885,
+      "learning_rate": 2e-05,
+      "loss": 5.5212,
+      "step": 4796
+    },
+    {
+      "epoch": 0.32176275279203137,
+      "grad_norm": 0.1454030708012873,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 4797
+    },
+    {
+      "epoch": 0.3218298286212563,
+      "grad_norm": 0.15106903827201995,
+      "learning_rate": 2e-05,
+      "loss": 5.4554,
+      "step": 4798
+    },
+    {
+      "epoch": 0.32189690445048125,
+      "grad_norm": 0.1455465547486163,
+      "learning_rate": 2e-05,
+      "loss": 5.4455,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3219639802797062,
+      "grad_norm": 0.14988136149476178,
+      "learning_rate": 2e-05,
+      "loss": 5.3967,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3220310561089311,
+      "grad_norm": 0.15126855183560264,
+      "learning_rate": 2e-05,
+      "loss": 5.4331,
+      "step": 4801
+    },
+    {
+      "epoch": 0.32209813193815606,
+      "grad_norm": 0.14620652027081413,
+      "learning_rate": 2e-05,
+      "loss": 5.4578,
+      "step": 4802
+    },
+    {
+      "epoch": 0.322165207767381,
+      "grad_norm": 0.14730289169786262,
+      "learning_rate": 2e-05,
+      "loss": 5.3637,
+      "step": 4803
+    },
+    {
+      "epoch": 0.32223228359660594,
+      "grad_norm": 0.15582217012477784,
+      "learning_rate": 2e-05,
+      "loss": 5.5836,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3222993594258309,
+      "grad_norm": 0.1487633828624673,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 4805
+    },
+    {
+      "epoch": 0.3223664352550558,
+      "grad_norm": 0.14510647203747284,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 4806
+    },
+    {
+      "epoch": 0.32243351108428076,
+      "grad_norm": 0.1496282338027695,
+      "learning_rate": 2e-05,
+      "loss": 5.3154,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3225005869135057,
+      "grad_norm": 0.14914789233926865,
+      "learning_rate": 2e-05,
+      "loss": 5.4238,
+      "step": 4808
+    },
+    {
+      "epoch": 0.32256766274273063,
+      "grad_norm": 0.14940360591559043,
+      "learning_rate": 2e-05,
+      "loss": 5.4358,
+      "step": 4809
+    },
+    {
+      "epoch": 0.32263473857195557,
+      "grad_norm": 0.15335938049795061,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3227018144011805,
+      "grad_norm": 0.14506045573241302,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 4811
+    },
+    {
+      "epoch": 0.32276889023040545,
+      "grad_norm": 0.14871601714412644,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3228359660596304,
+      "grad_norm": 0.14765537786976948,
+      "learning_rate": 2e-05,
+      "loss": 5.5435,
+      "step": 4813
+    },
+    {
+      "epoch": 0.3229030418888553,
+      "grad_norm": 0.14971927213614827,
+      "learning_rate": 2e-05,
+      "loss": 5.4723,
+      "step": 4814
+    },
+    {
+      "epoch": 0.32297011771808026,
+      "grad_norm": 0.1429795603115765,
+      "learning_rate": 2e-05,
+      "loss": 5.3743,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3230371935473052,
+      "grad_norm": 0.14007582103198876,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 4816
+    },
+    {
+      "epoch": 0.32310426937653014,
+      "grad_norm": 0.152811926415283,
+      "learning_rate": 2e-05,
+      "loss": 5.6733,
+      "step": 4817
+    },
+    {
+      "epoch": 0.32317134520575513,
+      "grad_norm": 0.15154827132119447,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 4818
+    },
+    {
+      "epoch": 0.3232384210349801,
+      "grad_norm": 0.16652475018066876,
+      "learning_rate": 2e-05,
+      "loss": 5.5609,
+      "step": 4819
+    },
+    {
+      "epoch": 0.323305496864205,
+      "grad_norm": 0.1625925691765137,
+      "learning_rate": 2e-05,
+      "loss": 5.5272,
+      "step": 4820
+    },
+    {
+      "epoch": 0.32337257269342995,
+      "grad_norm": 0.15142228053819187,
+      "learning_rate": 2e-05,
+      "loss": 5.3194,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3234396485226549,
+      "grad_norm": 0.16102270190331786,
+      "learning_rate": 2e-05,
+      "loss": 5.4048,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3235067243518798,
+      "grad_norm": 0.16367187994916846,
+      "learning_rate": 2e-05,
+      "loss": 5.3721,
+      "step": 4823
+    },
+    {
+      "epoch": 0.32357380018110476,
+      "grad_norm": 0.14454282646153518,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3236408760103297,
+      "grad_norm": 0.15310675745523536,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 4825
+    },
+    {
+      "epoch": 0.32370795183955464,
+      "grad_norm": 0.16102427002304626,
+      "learning_rate": 2e-05,
+      "loss": 5.4738,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3237750276687796,
+      "grad_norm": 0.15855364961893342,
+      "learning_rate": 2e-05,
+      "loss": 5.3506,
+      "step": 4827
+    },
+    {
+      "epoch": 0.3238421034980045,
+      "grad_norm": 0.15253864334315728,
+      "learning_rate": 2e-05,
+      "loss": 5.343,
+      "step": 4828
+    },
+    {
+      "epoch": 0.32390917932722946,
+      "grad_norm": 0.1553936645859567,
+      "learning_rate": 2e-05,
+      "loss": 5.4517,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3239762551564544,
+      "grad_norm": 0.15916161233210682,
+      "learning_rate": 2e-05,
+      "loss": 5.5743,
+      "step": 4830
+    },
+    {
+      "epoch": 0.32404333098567933,
+      "grad_norm": 0.15115005205816967,
+      "learning_rate": 2e-05,
+      "loss": 5.488,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3241104068149043,
+      "grad_norm": 0.1554158188105201,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3241774826441292,
+      "grad_norm": 0.15303457957679378,
+      "learning_rate": 2e-05,
+      "loss": 5.3729,
+      "step": 4833
+    },
+    {
+      "epoch": 0.32424455847335415,
+      "grad_norm": 0.14623170214985606,
+      "learning_rate": 2e-05,
+      "loss": 5.3064,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3243116343025791,
+      "grad_norm": 0.1466786378747016,
+      "learning_rate": 2e-05,
+      "loss": 5.3937,
+      "step": 4835
+    },
+    {
+      "epoch": 0.324378710131804,
+      "grad_norm": 0.1559167415693833,
+      "learning_rate": 2e-05,
+      "loss": 5.472,
+      "step": 4836
+    },
+    {
+      "epoch": 0.32444578596102897,
+      "grad_norm": 0.1464503423821888,
+      "learning_rate": 2e-05,
+      "loss": 5.3334,
+      "step": 4837
+    },
+    {
+      "epoch": 0.3245128617902539,
+      "grad_norm": 0.15580966582765746,
+      "learning_rate": 2e-05,
+      "loss": 5.6105,
+      "step": 4838
+    },
+    {
+      "epoch": 0.32457993761947884,
+      "grad_norm": 0.15107655181583143,
+      "learning_rate": 2e-05,
+      "loss": 5.5347,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3246470134487038,
+      "grad_norm": 0.14920950325154939,
+      "learning_rate": 2e-05,
+      "loss": 5.397,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3247140892779287,
+      "grad_norm": 0.1521811313005385,
+      "learning_rate": 2e-05,
+      "loss": 5.4813,
+      "step": 4841
+    },
+    {
+      "epoch": 0.32478116510715366,
+      "grad_norm": 0.15313905499000402,
+      "learning_rate": 2e-05,
+      "loss": 5.4505,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3248482409363786,
+      "grad_norm": 0.1465709951650294,
+      "learning_rate": 2e-05,
+      "loss": 5.5851,
+      "step": 4843
+    },
+    {
+      "epoch": 0.32491531676560353,
+      "grad_norm": 0.14911324354493985,
+      "learning_rate": 2e-05,
+      "loss": 5.2937,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3249823925948285,
+      "grad_norm": 0.14973730755077483,
+      "learning_rate": 2e-05,
+      "loss": 5.539,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3250494684240534,
+      "grad_norm": 0.14632297264620814,
+      "learning_rate": 2e-05,
+      "loss": 5.486,
+      "step": 4846
+    },
+    {
+      "epoch": 0.32511654425327835,
+      "grad_norm": 0.15194585941087044,
+      "learning_rate": 2e-05,
+      "loss": 5.4217,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3251836200825033,
+      "grad_norm": 0.15241222918971428,
+      "learning_rate": 2e-05,
+      "loss": 5.5148,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3252506959117282,
+      "grad_norm": 0.14737861749362985,
+      "learning_rate": 2e-05,
+      "loss": 5.4619,
+      "step": 4849
+    },
+    {
+      "epoch": 0.32531777174095317,
+      "grad_norm": 0.14699167338104402,
+      "learning_rate": 2e-05,
+      "loss": 5.3436,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3253848475701781,
+      "grad_norm": 0.14921530344209905,
+      "learning_rate": 2e-05,
+      "loss": 5.3577,
+      "step": 4851
+    },
+    {
+      "epoch": 0.32545192339940304,
+      "grad_norm": 0.14718435528823598,
+      "learning_rate": 2e-05,
+      "loss": 5.6717,
+      "step": 4852
+    },
+    {
+      "epoch": 0.325518999228628,
+      "grad_norm": 0.15499913816501207,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 4853
+    },
+    {
+      "epoch": 0.3255860750578529,
+      "grad_norm": 0.15253677496831808,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 4854
+    },
+    {
+      "epoch": 0.32565315088707786,
+      "grad_norm": 0.16627481275503478,
+      "learning_rate": 2e-05,
+      "loss": 5.3885,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3257202267163028,
+      "grad_norm": 0.1520583471337725,
+      "learning_rate": 2e-05,
+      "loss": 5.5676,
+      "step": 4856
+    },
+    {
+      "epoch": 0.32578730254552773,
+      "grad_norm": 0.14597748512585304,
+      "learning_rate": 2e-05,
+      "loss": 5.2946,
+      "step": 4857
+    },
+    {
+      "epoch": 0.3258543783747527,
+      "grad_norm": 0.15442009027535508,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 4858
+    },
+    {
+      "epoch": 0.3259214542039776,
+      "grad_norm": 0.1570651840626258,
+      "learning_rate": 2e-05,
+      "loss": 5.3472,
+      "step": 4859
+    },
+    {
+      "epoch": 0.32598853003320255,
+      "grad_norm": 0.14789565018701195,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3260556058624275,
+      "grad_norm": 0.15644992158727256,
+      "learning_rate": 2e-05,
+      "loss": 5.4177,
+      "step": 4861
+    },
+    {
+      "epoch": 0.3261226816916524,
+      "grad_norm": 0.1608760691965378,
+      "learning_rate": 2e-05,
+      "loss": 5.5442,
+      "step": 4862
+    },
+    {
+      "epoch": 0.32618975752087737,
+      "grad_norm": 0.15134519174996106,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3262568333501023,
+      "grad_norm": 0.1477824585246446,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 4864
+    },
+    {
+      "epoch": 0.32632390917932724,
+      "grad_norm": 0.16272921804413215,
+      "learning_rate": 2e-05,
+      "loss": 5.413,
+      "step": 4865
+    },
+    {
+      "epoch": 0.3263909850085522,
+      "grad_norm": 0.15041546324645658,
+      "learning_rate": 2e-05,
+      "loss": 5.3017,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3264580608377771,
+      "grad_norm": 0.1479048988365385,
+      "learning_rate": 2e-05,
+      "loss": 5.5009,
+      "step": 4867
+    },
+    {
+      "epoch": 0.32652513666700206,
+      "grad_norm": 0.16496502703298468,
+      "learning_rate": 2e-05,
+      "loss": 5.4904,
+      "step": 4868
+    },
+    {
+      "epoch": 0.326592212496227,
+      "grad_norm": 0.1623687571407885,
+      "learning_rate": 2e-05,
+      "loss": 5.4951,
+      "step": 4869
+    },
+    {
+      "epoch": 0.32665928832545194,
+      "grad_norm": 0.15266573526881952,
+      "learning_rate": 2e-05,
+      "loss": 5.3608,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3267263641546769,
+      "grad_norm": 0.15024551658641463,
+      "learning_rate": 2e-05,
+      "loss": 5.499,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3267934399839018,
+      "grad_norm": 0.15796720590133523,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 4872
+    },
+    {
+      "epoch": 0.32686051581312675,
+      "grad_norm": 0.15298261698983362,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 4873
+    },
+    {
+      "epoch": 0.3269275916423517,
+      "grad_norm": 0.14747903140178706,
+      "learning_rate": 2e-05,
+      "loss": 5.375,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3269946674715766,
+      "grad_norm": 0.1549224130383679,
+      "learning_rate": 2e-05,
+      "loss": 5.4119,
+      "step": 4875
+    },
+    {
+      "epoch": 0.32706174330080157,
+      "grad_norm": 0.15378800804499815,
+      "learning_rate": 2e-05,
+      "loss": 5.3748,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3271288191300265,
+      "grad_norm": 0.1464331346652091,
+      "learning_rate": 2e-05,
+      "loss": 5.534,
+      "step": 4877
+    },
+    {
+      "epoch": 0.32719589495925144,
+      "grad_norm": 0.1603634574953315,
+      "learning_rate": 2e-05,
+      "loss": 5.3508,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3272629707884764,
+      "grad_norm": 0.1601791384508407,
+      "learning_rate": 2e-05,
+      "loss": 5.3508,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3273300466177013,
+      "grad_norm": 0.15149161469502975,
+      "learning_rate": 2e-05,
+      "loss": 5.4547,
+      "step": 4880
+    },
+    {
+      "epoch": 0.32739712244692626,
+      "grad_norm": 0.15526259976898338,
+      "learning_rate": 2e-05,
+      "loss": 5.3564,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3274641982761512,
+      "grad_norm": 0.1437210514469127,
+      "learning_rate": 2e-05,
+      "loss": 5.5054,
+      "step": 4882
+    },
+    {
+      "epoch": 0.32753127410537614,
+      "grad_norm": 0.14867695342662923,
+      "learning_rate": 2e-05,
+      "loss": 5.5627,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3275983499346011,
+      "grad_norm": 0.15084129484228181,
+      "learning_rate": 2e-05,
+      "loss": 5.4474,
+      "step": 4884
+    },
+    {
+      "epoch": 0.327665425763826,
+      "grad_norm": 0.14889543516038425,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 4885
+    },
+    {
+      "epoch": 0.32773250159305095,
+      "grad_norm": 0.15551330624893006,
+      "learning_rate": 2e-05,
+      "loss": 5.4261,
+      "step": 4886
+    },
+    {
+      "epoch": 0.3277995774222759,
+      "grad_norm": 0.1447478114061947,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 4887
+    },
+    {
+      "epoch": 0.32786665325150083,
+      "grad_norm": 0.15084496870106934,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 4888
+    },
+    {
+      "epoch": 0.32793372908072577,
+      "grad_norm": 0.15374740906040166,
+      "learning_rate": 2e-05,
+      "loss": 5.3296,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3280008049099507,
+      "grad_norm": 0.1493727587292466,
+      "learning_rate": 2e-05,
+      "loss": 5.5519,
+      "step": 4890
+    },
+    {
+      "epoch": 0.32806788073917564,
+      "grad_norm": 0.14955456600521946,
+      "learning_rate": 2e-05,
+      "loss": 5.3655,
+      "step": 4891
+    },
+    {
+      "epoch": 0.3281349565684006,
+      "grad_norm": 0.15804807604157378,
+      "learning_rate": 2e-05,
+      "loss": 5.5491,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3282020323976255,
+      "grad_norm": 0.15204583204744546,
+      "learning_rate": 2e-05,
+      "loss": 5.3161,
+      "step": 4893
+    },
+    {
+      "epoch": 0.32826910822685046,
+      "grad_norm": 0.1603535604499304,
+      "learning_rate": 2e-05,
+      "loss": 5.45,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3283361840560754,
+      "grad_norm": 0.1537690131588415,
+      "learning_rate": 2e-05,
+      "loss": 5.3966,
+      "step": 4895
+    },
+    {
+      "epoch": 0.32840325988530034,
+      "grad_norm": 0.15015266451826612,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3284703357145253,
+      "grad_norm": 0.16490466382144775,
+      "learning_rate": 2e-05,
+      "loss": 5.4114,
+      "step": 4897
+    },
+    {
+      "epoch": 0.3285374115437502,
+      "grad_norm": 0.1487291331062549,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 4898
+    },
+    {
+      "epoch": 0.32860448737297515,
+      "grad_norm": 0.15158323681699107,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3286715632022001,
+      "grad_norm": 0.1527243872167189,
+      "learning_rate": 2e-05,
+      "loss": 5.3573,
+      "step": 4900
+    },
+    {
+      "epoch": 0.32873863903142503,
+      "grad_norm": 0.1586032242564745,
+      "learning_rate": 2e-05,
+      "loss": 5.3158,
+      "step": 4901
+    },
+    {
+      "epoch": 0.32880571486064997,
+      "grad_norm": 0.15546112987709979,
+      "learning_rate": 2e-05,
+      "loss": 5.4445,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3288727906898749,
+      "grad_norm": 0.15208789535982745,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 4903
+    },
+    {
+      "epoch": 0.32893986651909984,
+      "grad_norm": 0.15344555987248673,
+      "learning_rate": 2e-05,
+      "loss": 5.3794,
+      "step": 4904
+    },
+    {
+      "epoch": 0.3290069423483248,
+      "grad_norm": 0.1447745800852566,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3290740181775497,
+      "grad_norm": 0.16700366368168737,
+      "learning_rate": 2e-05,
+      "loss": 5.5159,
+      "step": 4906
+    },
+    {
+      "epoch": 0.32914109400677466,
+      "grad_norm": 0.13672254896418126,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3292081698359996,
+      "grad_norm": 0.14452504643997985,
+      "learning_rate": 2e-05,
+      "loss": 5.4799,
+      "step": 4908
+    },
+    {
+      "epoch": 0.32927524566522454,
+      "grad_norm": 0.1504504370855213,
+      "learning_rate": 2e-05,
+      "loss": 5.5415,
+      "step": 4909
+    },
+    {
+      "epoch": 0.3293423214944495,
+      "grad_norm": 0.14947634586000694,
+      "learning_rate": 2e-05,
+      "loss": 5.3722,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3294093973236744,
+      "grad_norm": 0.1438055020821984,
+      "learning_rate": 2e-05,
+      "loss": 5.4617,
+      "step": 4911
+    },
+    {
+      "epoch": 0.32947647315289935,
+      "grad_norm": 0.15095805781385754,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 4912
+    },
+    {
+      "epoch": 0.3295435489821243,
+      "grad_norm": 0.1524013112972752,
+      "learning_rate": 2e-05,
+      "loss": 5.515,
+      "step": 4913
+    },
+    {
+      "epoch": 0.32961062481134923,
+      "grad_norm": 0.15092124662112139,
+      "learning_rate": 2e-05,
+      "loss": 5.2374,
+      "step": 4914
+    },
+    {
+      "epoch": 0.32967770064057417,
+      "grad_norm": 0.15149304660865387,
+      "learning_rate": 2e-05,
+      "loss": 5.5233,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3297447764697991,
+      "grad_norm": 0.14506842477953535,
+      "learning_rate": 2e-05,
+      "loss": 5.313,
+      "step": 4916
+    },
+    {
+      "epoch": 0.32981185229902404,
+      "grad_norm": 0.14642296784713754,
+      "learning_rate": 2e-05,
+      "loss": 5.6227,
+      "step": 4917
+    },
+    {
+      "epoch": 0.329878928128249,
+      "grad_norm": 0.1581759764406734,
+      "learning_rate": 2e-05,
+      "loss": 5.3536,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3299460039574739,
+      "grad_norm": 0.1466015280249873,
+      "learning_rate": 2e-05,
+      "loss": 5.7307,
+      "step": 4919
+    },
+    {
+      "epoch": 0.33001307978669886,
+      "grad_norm": 0.14369858767757251,
+      "learning_rate": 2e-05,
+      "loss": 5.3318,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3300801556159238,
+      "grad_norm": 0.14950512662522442,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 4921
+    },
+    {
+      "epoch": 0.33014723144514874,
+      "grad_norm": 0.15335723440485768,
+      "learning_rate": 2e-05,
+      "loss": 5.5117,
+      "step": 4922
+    },
+    {
+      "epoch": 0.3302143072743737,
+      "grad_norm": 0.14728242672890066,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3302813831035986,
+      "grad_norm": 0.15779964111075048,
+      "learning_rate": 2e-05,
+      "loss": 5.368,
+      "step": 4924
+    },
+    {
+      "epoch": 0.33034845893282355,
+      "grad_norm": 0.15917094856818972,
+      "learning_rate": 2e-05,
+      "loss": 5.3708,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3304155347620485,
+      "grad_norm": 0.1570397447852906,
+      "learning_rate": 2e-05,
+      "loss": 5.5082,
+      "step": 4926
+    },
+    {
+      "epoch": 0.33048261059127343,
+      "grad_norm": 0.15361510983801951,
+      "learning_rate": 2e-05,
+      "loss": 5.2906,
+      "step": 4927
+    },
+    {
+      "epoch": 0.33054968642049837,
+      "grad_norm": 0.14771106042281923,
+      "learning_rate": 2e-05,
+      "loss": 5.4131,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3306167622497233,
+      "grad_norm": 0.14288778698684046,
+      "learning_rate": 2e-05,
+      "loss": 5.3912,
+      "step": 4929
+    },
+    {
+      "epoch": 0.33068383807894824,
+      "grad_norm": 0.16217206589541752,
+      "learning_rate": 2e-05,
+      "loss": 5.346,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3307509139081732,
+      "grad_norm": 0.1532795907629775,
+      "learning_rate": 2e-05,
+      "loss": 5.5483,
+      "step": 4931
+    },
+    {
+      "epoch": 0.3308179897373981,
+      "grad_norm": 0.14755566209017612,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 4932
+    },
+    {
+      "epoch": 0.33088506556662306,
+      "grad_norm": 0.15302190426729573,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 4933
+    },
+    {
+      "epoch": 0.330952141395848,
+      "grad_norm": 0.15829028014942445,
+      "learning_rate": 2e-05,
+      "loss": 5.2906,
+      "step": 4934
+    },
+    {
+      "epoch": 0.33101921722507294,
+      "grad_norm": 0.15157103143597975,
+      "learning_rate": 2e-05,
+      "loss": 5.5845,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3310862930542979,
+      "grad_norm": 0.15162863692490666,
+      "learning_rate": 2e-05,
+      "loss": 5.3532,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3311533688835228,
+      "grad_norm": 0.1532964949218231,
+      "learning_rate": 2e-05,
+      "loss": 5.5429,
+      "step": 4937
+    },
+    {
+      "epoch": 0.33122044471274775,
+      "grad_norm": 0.14313243243182644,
+      "learning_rate": 2e-05,
+      "loss": 5.3481,
+      "step": 4938
+    },
+    {
+      "epoch": 0.3312875205419727,
+      "grad_norm": 0.15005477530623978,
+      "learning_rate": 2e-05,
+      "loss": 5.4408,
+      "step": 4939
+    },
+    {
+      "epoch": 0.33135459637119763,
+      "grad_norm": 0.15320504116142544,
+      "learning_rate": 2e-05,
+      "loss": 5.3913,
+      "step": 4940
+    },
+    {
+      "epoch": 0.33142167220042257,
+      "grad_norm": 0.15112906135479812,
+      "learning_rate": 2e-05,
+      "loss": 5.314,
+      "step": 4941
+    },
+    {
+      "epoch": 0.3314887480296475,
+      "grad_norm": 0.15181364599303251,
+      "learning_rate": 2e-05,
+      "loss": 5.5866,
+      "step": 4942
+    },
+    {
+      "epoch": 0.33155582385887244,
+      "grad_norm": 0.16604673590594168,
+      "learning_rate": 2e-05,
+      "loss": 5.4114,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3316228996880974,
+      "grad_norm": 0.15145363424491115,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3316899755173223,
+      "grad_norm": 0.13881088136974254,
+      "learning_rate": 2e-05,
+      "loss": 5.3896,
+      "step": 4945
+    },
+    {
+      "epoch": 0.33175705134654726,
+      "grad_norm": 0.15263046131887917,
+      "learning_rate": 2e-05,
+      "loss": 5.4884,
+      "step": 4946
+    },
+    {
+      "epoch": 0.3318241271757722,
+      "grad_norm": 0.15458347920415602,
+      "learning_rate": 2e-05,
+      "loss": 5.335,
+      "step": 4947
+    },
+    {
+      "epoch": 0.33189120300499714,
+      "grad_norm": 0.15161263376761253,
+      "learning_rate": 2e-05,
+      "loss": 5.3876,
+      "step": 4948
+    },
+    {
+      "epoch": 0.3319582788342221,
+      "grad_norm": 0.16150234977754466,
+      "learning_rate": 2e-05,
+      "loss": 5.2274,
+      "step": 4949
+    },
+    {
+      "epoch": 0.332025354663447,
+      "grad_norm": 0.1576377745233706,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 4950
+    },
+    {
+      "epoch": 0.33209243049267195,
+      "grad_norm": 0.14322849080906128,
+      "learning_rate": 2e-05,
+      "loss": 5.3984,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3321595063218969,
+      "grad_norm": 0.1490021641208481,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 4952
+    },
+    {
+      "epoch": 0.33222658215112183,
+      "grad_norm": 0.15354156082442474,
+      "learning_rate": 2e-05,
+      "loss": 5.4984,
+      "step": 4953
+    },
+    {
+      "epoch": 0.33229365798034677,
+      "grad_norm": 0.1536383356974661,
+      "learning_rate": 2e-05,
+      "loss": 5.323,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3323607338095717,
+      "grad_norm": 0.14808021268218938,
+      "learning_rate": 2e-05,
+      "loss": 5.3162,
+      "step": 4955
+    },
+    {
+      "epoch": 0.33242780963879665,
+      "grad_norm": 0.14900601076435924,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3324948854680216,
+      "grad_norm": 0.15024403248768345,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 4957
+    },
+    {
+      "epoch": 0.3325619612972465,
+      "grad_norm": 0.1461424633271232,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 4958
+    },
+    {
+      "epoch": 0.33262903712647146,
+      "grad_norm": 0.15650437885931862,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3326961129556964,
+      "grad_norm": 0.15007645104289513,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 4960
+    },
+    {
+      "epoch": 0.33276318878492134,
+      "grad_norm": 0.14683110143038222,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 4961
+    },
+    {
+      "epoch": 0.3328302646141463,
+      "grad_norm": 0.15661969774289697,
+      "learning_rate": 2e-05,
+      "loss": 5.3209,
+      "step": 4962
+    },
+    {
+      "epoch": 0.3328973404433712,
+      "grad_norm": 0.1583038759251261,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 4963
+    },
+    {
+      "epoch": 0.33296441627259615,
+      "grad_norm": 0.14664745805358984,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 4964
+    },
+    {
+      "epoch": 0.3330314921018211,
+      "grad_norm": 0.1489718735058706,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 4965
+    },
+    {
+      "epoch": 0.33309856793104603,
+      "grad_norm": 0.14126125186740965,
+      "learning_rate": 2e-05,
+      "loss": 5.5438,
+      "step": 4966
+    },
+    {
+      "epoch": 0.33316564376027097,
+      "grad_norm": 0.15321073295001209,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 4967
+    },
+    {
+      "epoch": 0.3332327195894959,
+      "grad_norm": 0.14393108242375746,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 4968
+    },
+    {
+      "epoch": 0.33329979541872085,
+      "grad_norm": 0.14909762890709685,
+      "learning_rate": 2e-05,
+      "loss": 5.476,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3333668712479458,
+      "grad_norm": 0.14495631012481533,
+      "learning_rate": 2e-05,
+      "loss": 5.3794,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3334339470771707,
+      "grad_norm": 0.15022708991161957,
+      "learning_rate": 2e-05,
+      "loss": 5.5411,
+      "step": 4971
+    },
+    {
+      "epoch": 0.33350102290639566,
+      "grad_norm": 0.14458705809847766,
+      "learning_rate": 2e-05,
+      "loss": 5.3376,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3335680987356206,
+      "grad_norm": 0.14172222981178004,
+      "learning_rate": 2e-05,
+      "loss": 5.4787,
+      "step": 4973
+    },
+    {
+      "epoch": 0.33363517456484554,
+      "grad_norm": 0.14428962914230242,
+      "learning_rate": 2e-05,
+      "loss": 5.2374,
+      "step": 4974
+    },
+    {
+      "epoch": 0.3337022503940705,
+      "grad_norm": 0.14515079274135811,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3337693262232954,
+      "grad_norm": 0.14319913897522119,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 4976
+    },
+    {
+      "epoch": 0.33383640205252035,
+      "grad_norm": 0.14474456420284734,
+      "learning_rate": 2e-05,
+      "loss": 5.4133,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3339034778817453,
+      "grad_norm": 0.1491796631745555,
+      "learning_rate": 2e-05,
+      "loss": 5.3199,
+      "step": 4978
+    },
+    {
+      "epoch": 0.33397055371097023,
+      "grad_norm": 0.14527875385390626,
+      "learning_rate": 2e-05,
+      "loss": 5.3612,
+      "step": 4979
+    },
+    {
+      "epoch": 0.33403762954019517,
+      "grad_norm": 0.14267925447638172,
+      "learning_rate": 2e-05,
+      "loss": 5.3851,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3341047053694201,
+      "grad_norm": 0.14504510686271152,
+      "learning_rate": 2e-05,
+      "loss": 5.4632,
+      "step": 4981
+    },
+    {
+      "epoch": 0.33417178119864505,
+      "grad_norm": 0.14579766709189299,
+      "learning_rate": 2e-05,
+      "loss": 5.5059,
+      "step": 4982
+    },
+    {
+      "epoch": 0.33423885702787,
+      "grad_norm": 0.1418748241572961,
+      "learning_rate": 2e-05,
+      "loss": 5.5389,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3343059328570949,
+      "grad_norm": 0.14652261344353687,
+      "learning_rate": 2e-05,
+      "loss": 5.5008,
+      "step": 4984
+    },
+    {
+      "epoch": 0.33437300868631986,
+      "grad_norm": 0.15187027888096688,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3344400845155448,
+      "grad_norm": 0.148567254921888,
+      "learning_rate": 2e-05,
+      "loss": 5.5144,
+      "step": 4986
+    },
+    {
+      "epoch": 0.33450716034476974,
+      "grad_norm": 0.1387293717880951,
+      "learning_rate": 2e-05,
+      "loss": 5.4138,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3345742361739947,
+      "grad_norm": 0.13972123444970388,
+      "learning_rate": 2e-05,
+      "loss": 5.6487,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3346413120032196,
+      "grad_norm": 0.1549756640709451,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 4989
+    },
+    {
+      "epoch": 0.33470838783244455,
+      "grad_norm": 0.14460314065961194,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3347754636616695,
+      "grad_norm": 0.14568392348800133,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 4991
+    },
+    {
+      "epoch": 0.33484253949089443,
+      "grad_norm": 0.15775530389517214,
+      "learning_rate": 2e-05,
+      "loss": 5.5294,
+      "step": 4992
+    },
+    {
+      "epoch": 0.33490961532011937,
+      "grad_norm": 0.14623930618416095,
+      "learning_rate": 2e-05,
+      "loss": 5.563,
+      "step": 4993
+    },
+    {
+      "epoch": 0.3349766911493443,
+      "grad_norm": 0.14861483843638745,
+      "learning_rate": 2e-05,
+      "loss": 5.4286,
+      "step": 4994
+    },
+    {
+      "epoch": 0.33504376697856925,
+      "grad_norm": 0.14132549979065503,
+      "learning_rate": 2e-05,
+      "loss": 5.5107,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3351108428077942,
+      "grad_norm": 0.15705301663691884,
+      "learning_rate": 2e-05,
+      "loss": 5.5543,
+      "step": 4996
+    },
+    {
+      "epoch": 0.3351779186370191,
+      "grad_norm": 0.14995392048119274,
+      "learning_rate": 2e-05,
+      "loss": 5.39,
+      "step": 4997
+    },
+    {
+      "epoch": 0.33524499446624406,
+      "grad_norm": 0.14673390784312393,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 4998
+    },
+    {
+      "epoch": 0.335312070295469,
+      "grad_norm": 0.1521567390380464,
+      "learning_rate": 2e-05,
+      "loss": 5.6023,
+      "step": 4999
+    },
+    {
+      "epoch": 0.335379146124694,
+      "grad_norm": 0.14767372840167964,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 5000
+    },
+    {
+      "epoch": 0.33544622195391893,
+      "grad_norm": 0.14824962194192806,
+      "learning_rate": 2e-05,
+      "loss": 5.4797,
+      "step": 5001
+    },
+    {
+      "epoch": 0.33551329778314387,
+      "grad_norm": 0.14877878744279055,
+      "learning_rate": 2e-05,
+      "loss": 5.378,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3355803736123688,
+      "grad_norm": 0.15415716757625247,
+      "learning_rate": 2e-05,
+      "loss": 5.4589,
+      "step": 5003
+    },
+    {
+      "epoch": 0.33564744944159375,
+      "grad_norm": 0.14554973182504086,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 5004
+    },
+    {
+      "epoch": 0.3357145252708187,
+      "grad_norm": 0.1493642836268805,
+      "learning_rate": 2e-05,
+      "loss": 5.3014,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3357816011000436,
+      "grad_norm": 0.1485094523771149,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 5006
+    },
+    {
+      "epoch": 0.33584867692926856,
+      "grad_norm": 0.14556565883056555,
+      "learning_rate": 2e-05,
+      "loss": 5.3716,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3359157527584935,
+      "grad_norm": 0.1431052067349577,
+      "learning_rate": 2e-05,
+      "loss": 5.4625,
+      "step": 5008
+    },
+    {
+      "epoch": 0.33598282858771844,
+      "grad_norm": 0.1492047053114808,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 5009
+    },
+    {
+      "epoch": 0.3360499044169434,
+      "grad_norm": 0.14514087373012474,
+      "learning_rate": 2e-05,
+      "loss": 5.3163,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3361169802461683,
+      "grad_norm": 0.15200541244578164,
+      "learning_rate": 2e-05,
+      "loss": 5.4649,
+      "step": 5011
+    },
+    {
+      "epoch": 0.33618405607539326,
+      "grad_norm": 0.13969286625557828,
+      "learning_rate": 2e-05,
+      "loss": 5.4167,
+      "step": 5012
+    },
+    {
+      "epoch": 0.3362511319046182,
+      "grad_norm": 0.14555934913253313,
+      "learning_rate": 2e-05,
+      "loss": 5.3767,
+      "step": 5013
+    },
+    {
+      "epoch": 0.33631820773384313,
+      "grad_norm": 0.14367843444935469,
+      "learning_rate": 2e-05,
+      "loss": 5.4789,
+      "step": 5014
+    },
+    {
+      "epoch": 0.33638528356306807,
+      "grad_norm": 0.14357279506947693,
+      "learning_rate": 2e-05,
+      "loss": 5.3924,
+      "step": 5015
+    },
+    {
+      "epoch": 0.336452359392293,
+      "grad_norm": 0.1365281811335518,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 5016
+    },
+    {
+      "epoch": 0.33651943522151795,
+      "grad_norm": 0.14741406147724712,
+      "learning_rate": 2e-05,
+      "loss": 5.4845,
+      "step": 5017
+    },
+    {
+      "epoch": 0.3365865110507429,
+      "grad_norm": 0.14189689386419613,
+      "learning_rate": 2e-05,
+      "loss": 5.5267,
+      "step": 5018
+    },
+    {
+      "epoch": 0.3366535868799678,
+      "grad_norm": 0.14681574712704917,
+      "learning_rate": 2e-05,
+      "loss": 5.4156,
+      "step": 5019
+    },
+    {
+      "epoch": 0.33672066270919276,
+      "grad_norm": 0.1477096521882693,
+      "learning_rate": 2e-05,
+      "loss": 5.5242,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3367877385384177,
+      "grad_norm": 0.1505791829080233,
+      "learning_rate": 2e-05,
+      "loss": 5.4237,
+      "step": 5021
+    },
+    {
+      "epoch": 0.33685481436764264,
+      "grad_norm": 0.14989309802479092,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 5022
+    },
+    {
+      "epoch": 0.3369218901968676,
+      "grad_norm": 0.14229141917951757,
+      "learning_rate": 2e-05,
+      "loss": 5.4733,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3369889660260925,
+      "grad_norm": 0.14543049873946765,
+      "learning_rate": 2e-05,
+      "loss": 5.3328,
+      "step": 5024
+    },
+    {
+      "epoch": 0.33705604185531746,
+      "grad_norm": 0.15101729833307725,
+      "learning_rate": 2e-05,
+      "loss": 5.5538,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3371231176845424,
+      "grad_norm": 0.14748389433395945,
+      "learning_rate": 2e-05,
+      "loss": 5.3984,
+      "step": 5026
+    },
+    {
+      "epoch": 0.33719019351376733,
+      "grad_norm": 0.14136112746819238,
+      "learning_rate": 2e-05,
+      "loss": 5.3058,
+      "step": 5027
+    },
+    {
+      "epoch": 0.33725726934299227,
+      "grad_norm": 0.14540530396121973,
+      "learning_rate": 2e-05,
+      "loss": 5.4253,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3373243451722172,
+      "grad_norm": 0.15291602326456266,
+      "learning_rate": 2e-05,
+      "loss": 5.3431,
+      "step": 5029
+    },
+    {
+      "epoch": 0.33739142100144215,
+      "grad_norm": 0.14313508984520887,
+      "learning_rate": 2e-05,
+      "loss": 5.4755,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3374584968306671,
+      "grad_norm": 0.14927963314633153,
+      "learning_rate": 2e-05,
+      "loss": 5.185,
+      "step": 5031
+    },
+    {
+      "epoch": 0.337525572659892,
+      "grad_norm": 0.15513212878281346,
+      "learning_rate": 2e-05,
+      "loss": 5.3124,
+      "step": 5032
+    },
+    {
+      "epoch": 0.33759264848911696,
+      "grad_norm": 0.14636638004395433,
+      "learning_rate": 2e-05,
+      "loss": 5.3366,
+      "step": 5033
+    },
+    {
+      "epoch": 0.3376597243183419,
+      "grad_norm": 0.14909610356472933,
+      "learning_rate": 2e-05,
+      "loss": 5.5983,
+      "step": 5034
+    },
+    {
+      "epoch": 0.33772680014756684,
+      "grad_norm": 0.14992893648937955,
+      "learning_rate": 2e-05,
+      "loss": 5.3666,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3377938759767918,
+      "grad_norm": 0.14821226747375255,
+      "learning_rate": 2e-05,
+      "loss": 5.3006,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3378609518060167,
+      "grad_norm": 0.1483688164402614,
+      "learning_rate": 2e-05,
+      "loss": 5.4773,
+      "step": 5037
+    },
+    {
+      "epoch": 0.33792802763524166,
+      "grad_norm": 0.15691095802720081,
+      "learning_rate": 2e-05,
+      "loss": 5.4189,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3379951034644666,
+      "grad_norm": 0.14644764213583966,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 5039
+    },
+    {
+      "epoch": 0.33806217929369153,
+      "grad_norm": 0.14503174818408496,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 5040
+    },
+    {
+      "epoch": 0.33812925512291647,
+      "grad_norm": 0.15205580924210266,
+      "learning_rate": 2e-05,
+      "loss": 5.49,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3381963309521414,
+      "grad_norm": 0.15309680595084088,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 5042
+    },
+    {
+      "epoch": 0.33826340678136635,
+      "grad_norm": 0.15107317583840355,
+      "learning_rate": 2e-05,
+      "loss": 5.4355,
+      "step": 5043
+    },
+    {
+      "epoch": 0.3383304826105913,
+      "grad_norm": 0.15677018526578623,
+      "learning_rate": 2e-05,
+      "loss": 5.4304,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3383975584398162,
+      "grad_norm": 0.14535874965084616,
+      "learning_rate": 2e-05,
+      "loss": 5.449,
+      "step": 5045
+    },
+    {
+      "epoch": 0.33846463426904116,
+      "grad_norm": 0.15678572080709954,
+      "learning_rate": 2e-05,
+      "loss": 5.3081,
+      "step": 5046
+    },
+    {
+      "epoch": 0.3385317100982661,
+      "grad_norm": 0.14317364867309867,
+      "learning_rate": 2e-05,
+      "loss": 5.4476,
+      "step": 5047
+    },
+    {
+      "epoch": 0.33859878592749104,
+      "grad_norm": 0.15384037235000783,
+      "learning_rate": 2e-05,
+      "loss": 5.5361,
+      "step": 5048
+    },
+    {
+      "epoch": 0.338665861756716,
+      "grad_norm": 0.15514704007007674,
+      "learning_rate": 2e-05,
+      "loss": 5.4407,
+      "step": 5049
+    },
+    {
+      "epoch": 0.3387329375859409,
+      "grad_norm": 0.1537363737958236,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 5050
+    },
+    {
+      "epoch": 0.33880001341516586,
+      "grad_norm": 0.14866399911132988,
+      "learning_rate": 2e-05,
+      "loss": 5.4886,
+      "step": 5051
+    },
+    {
+      "epoch": 0.3388670892443908,
+      "grad_norm": 0.14764895178477416,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 5052
+    },
+    {
+      "epoch": 0.33893416507361573,
+      "grad_norm": 0.1528396259523453,
+      "learning_rate": 2e-05,
+      "loss": 5.5595,
+      "step": 5053
+    },
+    {
+      "epoch": 0.3390012409028407,
+      "grad_norm": 0.15076720498602397,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 5054
+    },
+    {
+      "epoch": 0.3390683167320656,
+      "grad_norm": 0.14771899224803817,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 5055
+    },
+    {
+      "epoch": 0.33913539256129055,
+      "grad_norm": 0.14164189609670205,
+      "learning_rate": 2e-05,
+      "loss": 5.4372,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3392024683905155,
+      "grad_norm": 0.1465670950394505,
+      "learning_rate": 2e-05,
+      "loss": 5.4566,
+      "step": 5057
+    },
+    {
+      "epoch": 0.3392695442197404,
+      "grad_norm": 0.16751985006456613,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 5058
+    },
+    {
+      "epoch": 0.33933662004896537,
+      "grad_norm": 0.15219993090641884,
+      "learning_rate": 2e-05,
+      "loss": 5.5595,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3394036958781903,
+      "grad_norm": 0.14815377771092478,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 5060
+    },
+    {
+      "epoch": 0.33947077170741524,
+      "grad_norm": 0.14071685835906506,
+      "learning_rate": 2e-05,
+      "loss": 5.333,
+      "step": 5061
+    },
+    {
+      "epoch": 0.3395378475366402,
+      "grad_norm": 0.15225623560302795,
+      "learning_rate": 2e-05,
+      "loss": 5.5012,
+      "step": 5062
+    },
+    {
+      "epoch": 0.3396049233658651,
+      "grad_norm": 0.14522381742278026,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5063
+    },
+    {
+      "epoch": 0.33967199919509006,
+      "grad_norm": 0.14571425962341106,
+      "learning_rate": 2e-05,
+      "loss": 5.3564,
+      "step": 5064
+    },
+    {
+      "epoch": 0.339739075024315,
+      "grad_norm": 0.14425833531598134,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 5065
+    },
+    {
+      "epoch": 0.33980615085353993,
+      "grad_norm": 0.14899106045045124,
+      "learning_rate": 2e-05,
+      "loss": 5.3626,
+      "step": 5066
+    },
+    {
+      "epoch": 0.3398732266827649,
+      "grad_norm": 0.1455809637009877,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 5067
+    },
+    {
+      "epoch": 0.3399403025119898,
+      "grad_norm": 0.14641716696758614,
+      "learning_rate": 2e-05,
+      "loss": 5.4307,
+      "step": 5068
+    },
+    {
+      "epoch": 0.34000737834121475,
+      "grad_norm": 0.160679892603039,
+      "learning_rate": 2e-05,
+      "loss": 5.3794,
+      "step": 5069
+    },
+    {
+      "epoch": 0.3400744541704397,
+      "grad_norm": 0.15284986917423876,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3401415299996646,
+      "grad_norm": 0.14810192522655632,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 5071
+    },
+    {
+      "epoch": 0.34020860582888957,
+      "grad_norm": 0.14716895700969587,
+      "learning_rate": 2e-05,
+      "loss": 5.4477,
+      "step": 5072
+    },
+    {
+      "epoch": 0.3402756816581145,
+      "grad_norm": 0.14654399690091047,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 5073
+    },
+    {
+      "epoch": 0.34034275748733944,
+      "grad_norm": 0.1482924774762828,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 5074
+    },
+    {
+      "epoch": 0.3404098333165644,
+      "grad_norm": 0.1442974232296339,
+      "learning_rate": 2e-05,
+      "loss": 5.4355,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3404769091457893,
+      "grad_norm": 0.15169316940290978,
+      "learning_rate": 2e-05,
+      "loss": 5.3812,
+      "step": 5076
+    },
+    {
+      "epoch": 0.34054398497501426,
+      "grad_norm": 0.14832829417408497,
+      "learning_rate": 2e-05,
+      "loss": 5.4387,
+      "step": 5077
+    },
+    {
+      "epoch": 0.3406110608042392,
+      "grad_norm": 0.14883942028416625,
+      "learning_rate": 2e-05,
+      "loss": 5.3984,
+      "step": 5078
+    },
+    {
+      "epoch": 0.34067813663346413,
+      "grad_norm": 0.1417248409493068,
+      "learning_rate": 2e-05,
+      "loss": 5.3005,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3407452124626891,
+      "grad_norm": 0.14230087649820689,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 5080
+    },
+    {
+      "epoch": 0.340812288291914,
+      "grad_norm": 0.15299695140911382,
+      "learning_rate": 2e-05,
+      "loss": 5.3335,
+      "step": 5081
+    },
+    {
+      "epoch": 0.34087936412113895,
+      "grad_norm": 0.14569532061635596,
+      "learning_rate": 2e-05,
+      "loss": 5.4136,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3409464399503639,
+      "grad_norm": 0.1511845084960759,
+      "learning_rate": 2e-05,
+      "loss": 5.53,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3410135157795888,
+      "grad_norm": 0.15180553823244108,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 5084
+    },
+    {
+      "epoch": 0.34108059160881377,
+      "grad_norm": 0.14377068642447766,
+      "learning_rate": 2e-05,
+      "loss": 5.5213,
+      "step": 5085
+    },
+    {
+      "epoch": 0.3411476674380387,
+      "grad_norm": 0.15102515874733421,
+      "learning_rate": 2e-05,
+      "loss": 5.346,
+      "step": 5086
+    },
+    {
+      "epoch": 0.34121474326726364,
+      "grad_norm": 0.1532780254807879,
+      "learning_rate": 2e-05,
+      "loss": 5.3399,
+      "step": 5087
+    },
+    {
+      "epoch": 0.3412818190964886,
+      "grad_norm": 0.14282291657302737,
+      "learning_rate": 2e-05,
+      "loss": 5.3364,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3413488949257135,
+      "grad_norm": 0.15601539436224252,
+      "learning_rate": 2e-05,
+      "loss": 5.476,
+      "step": 5089
+    },
+    {
+      "epoch": 0.34141597075493846,
+      "grad_norm": 0.1460241733026985,
+      "learning_rate": 2e-05,
+      "loss": 5.479,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3414830465841634,
+      "grad_norm": 0.13995626865829525,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 5091
+    },
+    {
+      "epoch": 0.34155012241338834,
+      "grad_norm": 0.15413332374449182,
+      "learning_rate": 2e-05,
+      "loss": 5.422,
+      "step": 5092
+    },
+    {
+      "epoch": 0.3416171982426133,
+      "grad_norm": 0.14663505786806075,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 5093
+    },
+    {
+      "epoch": 0.3416842740718382,
+      "grad_norm": 0.15039794221837058,
+      "learning_rate": 2e-05,
+      "loss": 5.3436,
+      "step": 5094
+    },
+    {
+      "epoch": 0.34175134990106315,
+      "grad_norm": 0.15232111935168902,
+      "learning_rate": 2e-05,
+      "loss": 5.6115,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3418184257302881,
+      "grad_norm": 0.15779900541038544,
+      "learning_rate": 2e-05,
+      "loss": 5.3535,
+      "step": 5096
+    },
+    {
+      "epoch": 0.341885501559513,
+      "grad_norm": 0.14906803140086322,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 5097
+    },
+    {
+      "epoch": 0.34195257738873797,
+      "grad_norm": 0.14552574898172513,
+      "learning_rate": 2e-05,
+      "loss": 5.4968,
+      "step": 5098
+    },
+    {
+      "epoch": 0.3420196532179629,
+      "grad_norm": 0.15545552641999075,
+      "learning_rate": 2e-05,
+      "loss": 5.3566,
+      "step": 5099
+    },
+    {
+      "epoch": 0.34208672904718784,
+      "grad_norm": 0.15216406205343933,
+      "learning_rate": 2e-05,
+      "loss": 5.5429,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3421538048764128,
+      "grad_norm": 0.14947469466902283,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 5101
+    },
+    {
+      "epoch": 0.3422208807056377,
+      "grad_norm": 0.14792125148436158,
+      "learning_rate": 2e-05,
+      "loss": 5.5185,
+      "step": 5102
+    },
+    {
+      "epoch": 0.34228795653486266,
+      "grad_norm": 0.14894503062645478,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 5103
+    },
+    {
+      "epoch": 0.3423550323640876,
+      "grad_norm": 0.1518954923295519,
+      "learning_rate": 2e-05,
+      "loss": 5.5866,
+      "step": 5104
+    },
+    {
+      "epoch": 0.34242210819331254,
+      "grad_norm": 0.16041759447284476,
+      "learning_rate": 2e-05,
+      "loss": 5.367,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3424891840225375,
+      "grad_norm": 0.14865956128812252,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 5106
+    },
+    {
+      "epoch": 0.3425562598517624,
+      "grad_norm": 0.1440020644400577,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 5107
+    },
+    {
+      "epoch": 0.34262333568098735,
+      "grad_norm": 0.14913045975732875,
+      "learning_rate": 2e-05,
+      "loss": 5.4837,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3426904115102123,
+      "grad_norm": 0.14733692277525778,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 5109
+    },
+    {
+      "epoch": 0.34275748733943723,
+      "grad_norm": 0.14721510119234313,
+      "learning_rate": 2e-05,
+      "loss": 5.3829,
+      "step": 5110
+    },
+    {
+      "epoch": 0.34282456316866217,
+      "grad_norm": 0.1476498774636057,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 5111
+    },
+    {
+      "epoch": 0.3428916389978871,
+      "grad_norm": 0.14555368426337806,
+      "learning_rate": 2e-05,
+      "loss": 5.3047,
+      "step": 5112
+    },
+    {
+      "epoch": 0.34295871482711204,
+      "grad_norm": 0.15102031664610527,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 5113
+    },
+    {
+      "epoch": 0.343025790656337,
+      "grad_norm": 0.14419373006420494,
+      "learning_rate": 2e-05,
+      "loss": 5.3179,
+      "step": 5114
+    },
+    {
+      "epoch": 0.3430928664855619,
+      "grad_norm": 0.1429482412966286,
+      "learning_rate": 2e-05,
+      "loss": 5.5208,
+      "step": 5115
+    },
+    {
+      "epoch": 0.34315994231478686,
+      "grad_norm": 0.15451069493650213,
+      "learning_rate": 2e-05,
+      "loss": 5.548,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3432270181440118,
+      "grad_norm": 0.14527823466699522,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 5117
+    },
+    {
+      "epoch": 0.34329409397323674,
+      "grad_norm": 0.15257992533036024,
+      "learning_rate": 2e-05,
+      "loss": 5.5369,
+      "step": 5118
+    },
+    {
+      "epoch": 0.3433611698024617,
+      "grad_norm": 0.14754850308439013,
+      "learning_rate": 2e-05,
+      "loss": 5.423,
+      "step": 5119
+    },
+    {
+      "epoch": 0.3434282456316866,
+      "grad_norm": 0.14535355419151572,
+      "learning_rate": 2e-05,
+      "loss": 5.52,
+      "step": 5120
+    },
+    {
+      "epoch": 0.34349532146091155,
+      "grad_norm": 0.16173474665370913,
+      "learning_rate": 2e-05,
+      "loss": 5.6329,
+      "step": 5121
+    },
+    {
+      "epoch": 0.3435623972901365,
+      "grad_norm": 0.15208462046423007,
+      "learning_rate": 2e-05,
+      "loss": 5.5284,
+      "step": 5122
+    },
+    {
+      "epoch": 0.34362947311936143,
+      "grad_norm": 0.1402745032975412,
+      "learning_rate": 2e-05,
+      "loss": 5.3445,
+      "step": 5123
+    },
+    {
+      "epoch": 0.34369654894858637,
+      "grad_norm": 0.15316425227719213,
+      "learning_rate": 2e-05,
+      "loss": 5.4263,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3437636247778113,
+      "grad_norm": 0.149886773350697,
+      "learning_rate": 2e-05,
+      "loss": 5.4182,
+      "step": 5125
+    },
+    {
+      "epoch": 0.34383070060703624,
+      "grad_norm": 0.15009277834640156,
+      "learning_rate": 2e-05,
+      "loss": 5.3994,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3438977764362612,
+      "grad_norm": 0.144365614773633,
+      "learning_rate": 2e-05,
+      "loss": 5.6116,
+      "step": 5127
+    },
+    {
+      "epoch": 0.3439648522654861,
+      "grad_norm": 0.16046795731004523,
+      "learning_rate": 2e-05,
+      "loss": 5.4369,
+      "step": 5128
+    },
+    {
+      "epoch": 0.34403192809471106,
+      "grad_norm": 0.1483597888309453,
+      "learning_rate": 2e-05,
+      "loss": 5.4992,
+      "step": 5129
+    },
+    {
+      "epoch": 0.344099003923936,
+      "grad_norm": 0.1503877305963891,
+      "learning_rate": 2e-05,
+      "loss": 5.4582,
+      "step": 5130
+    },
+    {
+      "epoch": 0.34416607975316094,
+      "grad_norm": 0.14869324559247787,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3442331555823859,
+      "grad_norm": 0.1526167963468802,
+      "learning_rate": 2e-05,
+      "loss": 5.4793,
+      "step": 5132
+    },
+    {
+      "epoch": 0.3443002314116108,
+      "grad_norm": 0.1452060920370571,
+      "learning_rate": 2e-05,
+      "loss": 5.4089,
+      "step": 5133
+    },
+    {
+      "epoch": 0.34436730724083575,
+      "grad_norm": 0.14308101700513118,
+      "learning_rate": 2e-05,
+      "loss": 5.5038,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3444343830700607,
+      "grad_norm": 0.16303349045104953,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 5135
+    },
+    {
+      "epoch": 0.34450145889928563,
+      "grad_norm": 0.15378575175382575,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5136
+    },
+    {
+      "epoch": 0.34456853472851057,
+      "grad_norm": 0.1466039791767733,
+      "learning_rate": 2e-05,
+      "loss": 5.4247,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3446356105577355,
+      "grad_norm": 0.1476463273896999,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 5138
+    },
+    {
+      "epoch": 0.34470268638696044,
+      "grad_norm": 0.14776687845951914,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 5139
+    },
+    {
+      "epoch": 0.3447697622161854,
+      "grad_norm": 0.14661487356212113,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3448368380454103,
+      "grad_norm": 0.15093893171586908,
+      "learning_rate": 2e-05,
+      "loss": 5.3159,
+      "step": 5141
+    },
+    {
+      "epoch": 0.34490391387463526,
+      "grad_norm": 0.15007865346381366,
+      "learning_rate": 2e-05,
+      "loss": 5.3043,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3449709897038602,
+      "grad_norm": 0.14445797477309952,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 5143
+    },
+    {
+      "epoch": 0.34503806553308514,
+      "grad_norm": 0.14729481011770787,
+      "learning_rate": 2e-05,
+      "loss": 5.3301,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3451051413623101,
+      "grad_norm": 0.15073863319676423,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 5145
+    },
+    {
+      "epoch": 0.345172217191535,
+      "grad_norm": 0.15076011306452702,
+      "learning_rate": 2e-05,
+      "loss": 5.3446,
+      "step": 5146
+    },
+    {
+      "epoch": 0.34523929302075995,
+      "grad_norm": 0.14581200851203543,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3453063688499849,
+      "grad_norm": 0.14669439325753705,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 5148
+    },
+    {
+      "epoch": 0.34537344467920983,
+      "grad_norm": 0.14904762507125538,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 5149
+    },
+    {
+      "epoch": 0.34544052050843477,
+      "grad_norm": 0.14697435863911823,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3455075963376597,
+      "grad_norm": 0.14037329701695794,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 5151
+    },
+    {
+      "epoch": 0.34557467216688464,
+      "grad_norm": 0.14301935901310492,
+      "learning_rate": 2e-05,
+      "loss": 5.384,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3456417479961096,
+      "grad_norm": 0.14697510037264214,
+      "learning_rate": 2e-05,
+      "loss": 5.6395,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3457088238253345,
+      "grad_norm": 0.14847370739450716,
+      "learning_rate": 2e-05,
+      "loss": 5.395,
+      "step": 5154
+    },
+    {
+      "epoch": 0.34577589965455946,
+      "grad_norm": 0.14069037252726643,
+      "learning_rate": 2e-05,
+      "loss": 5.3712,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3458429754837844,
+      "grad_norm": 0.14759606877900136,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 5156
+    },
+    {
+      "epoch": 0.34591005131300934,
+      "grad_norm": 0.14305307269265016,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3459771271422343,
+      "grad_norm": 0.15026908769664155,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3460442029714592,
+      "grad_norm": 0.1444880807410971,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 5159
+    },
+    {
+      "epoch": 0.34611127880068415,
+      "grad_norm": 0.1481152810532922,
+      "learning_rate": 2e-05,
+      "loss": 5.4217,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3461783546299091,
+      "grad_norm": 0.1458058337060952,
+      "learning_rate": 2e-05,
+      "loss": 5.333,
+      "step": 5161
+    },
+    {
+      "epoch": 0.34624543045913403,
+      "grad_norm": 0.14887701424791697,
+      "learning_rate": 2e-05,
+      "loss": 5.3154,
+      "step": 5162
+    },
+    {
+      "epoch": 0.34631250628835897,
+      "grad_norm": 0.15108812533369428,
+      "learning_rate": 2e-05,
+      "loss": 5.4646,
+      "step": 5163
+    },
+    {
+      "epoch": 0.3463795821175839,
+      "grad_norm": 0.14739915154505717,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 5164
+    },
+    {
+      "epoch": 0.34644665794680884,
+      "grad_norm": 0.15874554200389152,
+      "learning_rate": 2e-05,
+      "loss": 5.4369,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3465137337760338,
+      "grad_norm": 0.142606831432432,
+      "learning_rate": 2e-05,
+      "loss": 5.4613,
+      "step": 5166
+    },
+    {
+      "epoch": 0.3465808096052587,
+      "grad_norm": 0.15030030939323338,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 5167
+    },
+    {
+      "epoch": 0.34664788543448366,
+      "grad_norm": 0.14559824052826326,
+      "learning_rate": 2e-05,
+      "loss": 5.5129,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3467149612637086,
+      "grad_norm": 0.14322059123210432,
+      "learning_rate": 2e-05,
+      "loss": 5.2889,
+      "step": 5169
+    },
+    {
+      "epoch": 0.34678203709293354,
+      "grad_norm": 0.14349016301057502,
+      "learning_rate": 2e-05,
+      "loss": 5.3897,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3468491129221585,
+      "grad_norm": 0.15468662524943422,
+      "learning_rate": 2e-05,
+      "loss": 5.3195,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3469161887513834,
+      "grad_norm": 0.14714625895185038,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 5172
+    },
+    {
+      "epoch": 0.34698326458060835,
+      "grad_norm": 0.15362850127407718,
+      "learning_rate": 2e-05,
+      "loss": 5.592,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3470503404098333,
+      "grad_norm": 0.15295827249797345,
+      "learning_rate": 2e-05,
+      "loss": 5.502,
+      "step": 5174
+    },
+    {
+      "epoch": 0.34711741623905823,
+      "grad_norm": 0.15919253104373307,
+      "learning_rate": 2e-05,
+      "loss": 5.4479,
+      "step": 5175
+    },
+    {
+      "epoch": 0.34718449206828317,
+      "grad_norm": 0.14936731328691719,
+      "learning_rate": 2e-05,
+      "loss": 5.4146,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3472515678975081,
+      "grad_norm": 0.1508454914915626,
+      "learning_rate": 2e-05,
+      "loss": 5.4266,
+      "step": 5177
+    },
+    {
+      "epoch": 0.34731864372673305,
+      "grad_norm": 0.15419537860105784,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 5178
+    },
+    {
+      "epoch": 0.347385719555958,
+      "grad_norm": 0.15046529066226447,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3474527953851829,
+      "grad_norm": 0.15291162974212805,
+      "learning_rate": 2e-05,
+      "loss": 5.3986,
+      "step": 5180
+    },
+    {
+      "epoch": 0.34751987121440786,
+      "grad_norm": 0.155532902207235,
+      "learning_rate": 2e-05,
+      "loss": 5.3823,
+      "step": 5181
+    },
+    {
+      "epoch": 0.34758694704363285,
+      "grad_norm": 0.14503590324608562,
+      "learning_rate": 2e-05,
+      "loss": 5.4725,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3476540228728578,
+      "grad_norm": 0.14787679680568916,
+      "learning_rate": 2e-05,
+      "loss": 5.3663,
+      "step": 5183
+    },
+    {
+      "epoch": 0.34772109870208273,
+      "grad_norm": 0.1504502574308599,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 5184
+    },
+    {
+      "epoch": 0.34778817453130767,
+      "grad_norm": 0.15058335066354758,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3478552503605326,
+      "grad_norm": 0.15073542193278786,
+      "learning_rate": 2e-05,
+      "loss": 5.5058,
+      "step": 5186
+    },
+    {
+      "epoch": 0.34792232618975755,
+      "grad_norm": 0.15314935676932923,
+      "learning_rate": 2e-05,
+      "loss": 5.4658,
+      "step": 5187
+    },
+    {
+      "epoch": 0.3479894020189825,
+      "grad_norm": 0.14886798336462964,
+      "learning_rate": 2e-05,
+      "loss": 5.4823,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3480564778482074,
+      "grad_norm": 0.14869454585379094,
+      "learning_rate": 2e-05,
+      "loss": 5.5709,
+      "step": 5189
+    },
+    {
+      "epoch": 0.34812355367743236,
+      "grad_norm": 0.1410557695369779,
+      "learning_rate": 2e-05,
+      "loss": 5.3984,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3481906295066573,
+      "grad_norm": 0.14811737964250485,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 5191
+    },
+    {
+      "epoch": 0.34825770533588224,
+      "grad_norm": 0.1434071792216402,
+      "learning_rate": 2e-05,
+      "loss": 5.4756,
+      "step": 5192
+    },
+    {
+      "epoch": 0.3483247811651072,
+      "grad_norm": 0.1489233346729354,
+      "learning_rate": 2e-05,
+      "loss": 5.4931,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3483918569943321,
+      "grad_norm": 0.14805127229397902,
+      "learning_rate": 2e-05,
+      "loss": 5.5233,
+      "step": 5194
+    },
+    {
+      "epoch": 0.34845893282355705,
+      "grad_norm": 0.14244104141022254,
+      "learning_rate": 2e-05,
+      "loss": 5.4813,
+      "step": 5195
+    },
+    {
+      "epoch": 0.348526008652782,
+      "grad_norm": 0.14648226921756205,
+      "learning_rate": 2e-05,
+      "loss": 5.2961,
+      "step": 5196
+    },
+    {
+      "epoch": 0.34859308448200693,
+      "grad_norm": 0.15149624342804527,
+      "learning_rate": 2e-05,
+      "loss": 5.3099,
+      "step": 5197
+    },
+    {
+      "epoch": 0.34866016031123187,
+      "grad_norm": 0.14244434000572836,
+      "learning_rate": 2e-05,
+      "loss": 5.3768,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3487272361404568,
+      "grad_norm": 0.1499318000304649,
+      "learning_rate": 2e-05,
+      "loss": 5.3787,
+      "step": 5199
+    },
+    {
+      "epoch": 0.34879431196968175,
+      "grad_norm": 0.15587393974263983,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3488613877989067,
+      "grad_norm": 0.1460650093387309,
+      "learning_rate": 2e-05,
+      "loss": 5.3641,
+      "step": 5201
+    },
+    {
+      "epoch": 0.3489284636281316,
+      "grad_norm": 0.1524897863101171,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 5202
+    },
+    {
+      "epoch": 0.34899553945735656,
+      "grad_norm": 0.16377656855769968,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3490626152865815,
+      "grad_norm": 0.14848294017923336,
+      "learning_rate": 2e-05,
+      "loss": 5.3847,
+      "step": 5204
+    },
+    {
+      "epoch": 0.34912969111580644,
+      "grad_norm": 0.15062476247109152,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3491967669450314,
+      "grad_norm": 0.15473687546002152,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3492638427742563,
+      "grad_norm": 0.1423327004362045,
+      "learning_rate": 2e-05,
+      "loss": 5.3541,
+      "step": 5207
+    },
+    {
+      "epoch": 0.34933091860348126,
+      "grad_norm": 0.1438742541550418,
+      "learning_rate": 2e-05,
+      "loss": 5.3359,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3493979944327062,
+      "grad_norm": 0.146295309597274,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 5209
+    },
+    {
+      "epoch": 0.34946507026193113,
+      "grad_norm": 0.14936614206500587,
+      "learning_rate": 2e-05,
+      "loss": 5.3761,
+      "step": 5210
+    },
+    {
+      "epoch": 0.34953214609115607,
+      "grad_norm": 0.14602333132587048,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 5211
+    },
+    {
+      "epoch": 0.349599221920381,
+      "grad_norm": 0.14889872404205431,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 5212
+    },
+    {
+      "epoch": 0.34966629774960595,
+      "grad_norm": 0.15449491200237794,
+      "learning_rate": 2e-05,
+      "loss": 5.2685,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3497333735788309,
+      "grad_norm": 0.14976721474677326,
+      "learning_rate": 2e-05,
+      "loss": 5.5031,
+      "step": 5214
+    },
+    {
+      "epoch": 0.3498004494080558,
+      "grad_norm": 0.14063616649617897,
+      "learning_rate": 2e-05,
+      "loss": 5.318,
+      "step": 5215
+    },
+    {
+      "epoch": 0.34986752523728076,
+      "grad_norm": 0.1434515057512941,
+      "learning_rate": 2e-05,
+      "loss": 5.5416,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3499346010665057,
+      "grad_norm": 0.14927318510527998,
+      "learning_rate": 2e-05,
+      "loss": 5.3272,
+      "step": 5217
+    },
+    {
+      "epoch": 0.35000167689573064,
+      "grad_norm": 0.14895709984487637,
+      "learning_rate": 2e-05,
+      "loss": 5.3602,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3500687527249556,
+      "grad_norm": 0.1510104159538387,
+      "learning_rate": 2e-05,
+      "loss": 5.3817,
+      "step": 5219
+    },
+    {
+      "epoch": 0.3501358285541805,
+      "grad_norm": 0.15491433242463407,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 5220
+    },
+    {
+      "epoch": 0.35020290438340546,
+      "grad_norm": 0.14335719372174233,
+      "learning_rate": 2e-05,
+      "loss": 5.3551,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3502699802126304,
+      "grad_norm": 0.14766262764803334,
+      "learning_rate": 2e-05,
+      "loss": 5.5219,
+      "step": 5222
+    },
+    {
+      "epoch": 0.35033705604185533,
+      "grad_norm": 0.14373299802737988,
+      "learning_rate": 2e-05,
+      "loss": 5.6778,
+      "step": 5223
+    },
+    {
+      "epoch": 0.35040413187108027,
+      "grad_norm": 0.1486580266553973,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3504712077003052,
+      "grad_norm": 0.15251037334339454,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 5225
+    },
+    {
+      "epoch": 0.35053828352953015,
+      "grad_norm": 0.15153215903599207,
+      "learning_rate": 2e-05,
+      "loss": 5.4778,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3506053593587551,
+      "grad_norm": 0.14881816388196678,
+      "learning_rate": 2e-05,
+      "loss": 5.5016,
+      "step": 5227
+    },
+    {
+      "epoch": 0.35067243518798,
+      "grad_norm": 0.14068144698684956,
+      "learning_rate": 2e-05,
+      "loss": 5.3374,
+      "step": 5228
+    },
+    {
+      "epoch": 0.35073951101720496,
+      "grad_norm": 0.14902852359246077,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3508065868464299,
+      "grad_norm": 0.14621913159536207,
+      "learning_rate": 2e-05,
+      "loss": 5.4966,
+      "step": 5230
+    },
+    {
+      "epoch": 0.35087366267565484,
+      "grad_norm": 0.15361224348853364,
+      "learning_rate": 2e-05,
+      "loss": 5.5021,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3509407385048798,
+      "grad_norm": 0.14957295026917647,
+      "learning_rate": 2e-05,
+      "loss": 5.5107,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3510078143341047,
+      "grad_norm": 0.14907651719146167,
+      "learning_rate": 2e-05,
+      "loss": 5.5201,
+      "step": 5233
+    },
+    {
+      "epoch": 0.35107489016332966,
+      "grad_norm": 0.14736008225005953,
+      "learning_rate": 2e-05,
+      "loss": 5.3595,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3511419659925546,
+      "grad_norm": 0.14751305942518353,
+      "learning_rate": 2e-05,
+      "loss": 5.3125,
+      "step": 5235
+    },
+    {
+      "epoch": 0.35120904182177953,
+      "grad_norm": 0.1378198137665972,
+      "learning_rate": 2e-05,
+      "loss": 5.4597,
+      "step": 5236
+    },
+    {
+      "epoch": 0.35127611765100447,
+      "grad_norm": 0.14764281010984415,
+      "learning_rate": 2e-05,
+      "loss": 5.3325,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3513431934802294,
+      "grad_norm": 0.14602186175612378,
+      "learning_rate": 2e-05,
+      "loss": 5.3474,
+      "step": 5238
+    },
+    {
+      "epoch": 0.35141026930945435,
+      "grad_norm": 0.14590927229163556,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3514773451386793,
+      "grad_norm": 0.14789009444541937,
+      "learning_rate": 2e-05,
+      "loss": 5.3425,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3515444209679042,
+      "grad_norm": 0.15219684082113613,
+      "learning_rate": 2e-05,
+      "loss": 5.4379,
+      "step": 5241
+    },
+    {
+      "epoch": 0.35161149679712916,
+      "grad_norm": 0.15561737320847224,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 5242
+    },
+    {
+      "epoch": 0.3516785726263541,
+      "grad_norm": 0.15514816719554278,
+      "learning_rate": 2e-05,
+      "loss": 5.3938,
+      "step": 5243
+    },
+    {
+      "epoch": 0.35174564845557904,
+      "grad_norm": 0.1454694189921067,
+      "learning_rate": 2e-05,
+      "loss": 5.4823,
+      "step": 5244
+    },
+    {
+      "epoch": 0.351812724284804,
+      "grad_norm": 0.15181493966246498,
+      "learning_rate": 2e-05,
+      "loss": 5.6088,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3518798001140289,
+      "grad_norm": 0.1630078214754584,
+      "learning_rate": 2e-05,
+      "loss": 5.5013,
+      "step": 5246
+    },
+    {
+      "epoch": 0.35194687594325386,
+      "grad_norm": 0.14609829414048703,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3520139517724788,
+      "grad_norm": 0.1521682942782033,
+      "learning_rate": 2e-05,
+      "loss": 5.5515,
+      "step": 5248
+    },
+    {
+      "epoch": 0.35208102760170373,
+      "grad_norm": 0.1555422012177205,
+      "learning_rate": 2e-05,
+      "loss": 5.5736,
+      "step": 5249
+    },
+    {
+      "epoch": 0.35214810343092867,
+      "grad_norm": 0.14917210171013387,
+      "learning_rate": 2e-05,
+      "loss": 5.4361,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3522151792601536,
+      "grad_norm": 0.1505486066241177,
+      "learning_rate": 2e-05,
+      "loss": 5.336,
+      "step": 5251
+    },
+    {
+      "epoch": 0.35228225508937855,
+      "grad_norm": 0.14512335494369497,
+      "learning_rate": 2e-05,
+      "loss": 5.3559,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3523493309186035,
+      "grad_norm": 0.15670852378946906,
+      "learning_rate": 2e-05,
+      "loss": 5.2959,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3524164067478284,
+      "grad_norm": 0.14834890924122335,
+      "learning_rate": 2e-05,
+      "loss": 5.4883,
+      "step": 5254
+    },
+    {
+      "epoch": 0.35248348257705336,
+      "grad_norm": 0.15265941510568792,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3525505584062783,
+      "grad_norm": 0.14563361833170285,
+      "learning_rate": 2e-05,
+      "loss": 5.4364,
+      "step": 5256
+    },
+    {
+      "epoch": 0.35261763423550324,
+      "grad_norm": 0.14913907856542713,
+      "learning_rate": 2e-05,
+      "loss": 5.5352,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3526847100647282,
+      "grad_norm": 0.15518222921332991,
+      "learning_rate": 2e-05,
+      "loss": 5.3833,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3527517858939531,
+      "grad_norm": 0.15050071411139485,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 5259
+    },
+    {
+      "epoch": 0.35281886172317806,
+      "grad_norm": 0.14513971252592228,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 5260
+    },
+    {
+      "epoch": 0.352885937552403,
+      "grad_norm": 0.15031788218016473,
+      "learning_rate": 2e-05,
+      "loss": 5.371,
+      "step": 5261
+    },
+    {
+      "epoch": 0.35295301338162793,
+      "grad_norm": 0.15881602051917124,
+      "learning_rate": 2e-05,
+      "loss": 5.4628,
+      "step": 5262
+    },
+    {
+      "epoch": 0.35302008921085287,
+      "grad_norm": 0.15122978427203018,
+      "learning_rate": 2e-05,
+      "loss": 5.5502,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3530871650400778,
+      "grad_norm": 0.15265258557886197,
+      "learning_rate": 2e-05,
+      "loss": 5.5005,
+      "step": 5264
+    },
+    {
+      "epoch": 0.35315424086930275,
+      "grad_norm": 0.16096587586023706,
+      "learning_rate": 2e-05,
+      "loss": 5.3849,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3532213166985277,
+      "grad_norm": 0.14911735109177193,
+      "learning_rate": 2e-05,
+      "loss": 5.57,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3532883925277526,
+      "grad_norm": 0.16077662025712217,
+      "learning_rate": 2e-05,
+      "loss": 5.3453,
+      "step": 5267
+    },
+    {
+      "epoch": 0.35335546835697756,
+      "grad_norm": 0.14981137340515602,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3534225441862025,
+      "grad_norm": 0.1485026233643727,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 5269
+    },
+    {
+      "epoch": 0.35348962001542744,
+      "grad_norm": 0.15250737323010907,
+      "learning_rate": 2e-05,
+      "loss": 5.3803,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3535566958446524,
+      "grad_norm": 0.1439786701090003,
+      "learning_rate": 2e-05,
+      "loss": 5.4159,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3536237716738773,
+      "grad_norm": 0.14816280189115036,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 5272
+    },
+    {
+      "epoch": 0.35369084750310226,
+      "grad_norm": 0.15256926817735894,
+      "learning_rate": 2e-05,
+      "loss": 5.5311,
+      "step": 5273
+    },
+    {
+      "epoch": 0.3537579233323272,
+      "grad_norm": 0.15337196486874283,
+      "learning_rate": 2e-05,
+      "loss": 5.3014,
+      "step": 5274
+    },
+    {
+      "epoch": 0.35382499916155213,
+      "grad_norm": 0.1545734788996869,
+      "learning_rate": 2e-05,
+      "loss": 5.3558,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3538920749907771,
+      "grad_norm": 0.16148267909567288,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 5276
+    },
+    {
+      "epoch": 0.353959150820002,
+      "grad_norm": 0.14766623248719096,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 5277
+    },
+    {
+      "epoch": 0.35402622664922695,
+      "grad_norm": 0.14473777313014374,
+      "learning_rate": 2e-05,
+      "loss": 5.5388,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3540933024784519,
+      "grad_norm": 0.15306247346805718,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3541603783076768,
+      "grad_norm": 0.15725064282936788,
+      "learning_rate": 2e-05,
+      "loss": 5.3614,
+      "step": 5280
+    },
+    {
+      "epoch": 0.35422745413690176,
+      "grad_norm": 0.16076948380186573,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3542945299661267,
+      "grad_norm": 0.14621678588949044,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 5282
+    },
+    {
+      "epoch": 0.35436160579535164,
+      "grad_norm": 0.15770720871043084,
+      "learning_rate": 2e-05,
+      "loss": 5.2895,
+      "step": 5283
+    },
+    {
+      "epoch": 0.3544286816245766,
+      "grad_norm": 0.1456885552617241,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3544957574538015,
+      "grad_norm": 0.14696114398476084,
+      "learning_rate": 2e-05,
+      "loss": 5.4904,
+      "step": 5285
+    },
+    {
+      "epoch": 0.35456283328302646,
+      "grad_norm": 0.15091160431927883,
+      "learning_rate": 2e-05,
+      "loss": 5.5007,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3546299091122514,
+      "grad_norm": 0.15974368305000927,
+      "learning_rate": 2e-05,
+      "loss": 5.2927,
+      "step": 5287
+    },
+    {
+      "epoch": 0.35469698494147633,
+      "grad_norm": 0.1456612519839073,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3547640607707013,
+      "grad_norm": 0.14599105203998,
+      "learning_rate": 2e-05,
+      "loss": 5.4737,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3548311365999262,
+      "grad_norm": 0.14776374158847905,
+      "learning_rate": 2e-05,
+      "loss": 5.4719,
+      "step": 5290
+    },
+    {
+      "epoch": 0.35489821242915115,
+      "grad_norm": 0.15106502005936834,
+      "learning_rate": 2e-05,
+      "loss": 5.312,
+      "step": 5291
+    },
+    {
+      "epoch": 0.3549652882583761,
+      "grad_norm": 0.15885376766109824,
+      "learning_rate": 2e-05,
+      "loss": 5.316,
+      "step": 5292
+    },
+    {
+      "epoch": 0.355032364087601,
+      "grad_norm": 0.15661223744453795,
+      "learning_rate": 2e-05,
+      "loss": 5.3554,
+      "step": 5293
+    },
+    {
+      "epoch": 0.35509943991682597,
+      "grad_norm": 0.1540948577644022,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3551665157460509,
+      "grad_norm": 0.15323924790843024,
+      "learning_rate": 2e-05,
+      "loss": 5.31,
+      "step": 5295
+    },
+    {
+      "epoch": 0.35523359157527584,
+      "grad_norm": 0.15044656238938336,
+      "learning_rate": 2e-05,
+      "loss": 5.5043,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3553006674045008,
+      "grad_norm": 0.15970305640259166,
+      "learning_rate": 2e-05,
+      "loss": 5.323,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3553677432337257,
+      "grad_norm": 0.1465505615516917,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 5298
+    },
+    {
+      "epoch": 0.35543481906295066,
+      "grad_norm": 0.15189756429167792,
+      "learning_rate": 2e-05,
+      "loss": 5.5519,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3555018948921756,
+      "grad_norm": 0.14859591267904004,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 5300
+    },
+    {
+      "epoch": 0.35556897072140053,
+      "grad_norm": 0.15517661718147013,
+      "learning_rate": 2e-05,
+      "loss": 5.5167,
+      "step": 5301
+    },
+    {
+      "epoch": 0.3556360465506255,
+      "grad_norm": 0.14705188080798362,
+      "learning_rate": 2e-05,
+      "loss": 5.4043,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3557031223798504,
+      "grad_norm": 0.14556540002271984,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 5303
+    },
+    {
+      "epoch": 0.35577019820907535,
+      "grad_norm": 0.15165302983892465,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3558372740383003,
+      "grad_norm": 0.1440266313796821,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3559043498675252,
+      "grad_norm": 0.15664477010070635,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 5306
+    },
+    {
+      "epoch": 0.35597142569675017,
+      "grad_norm": 0.1440972793144366,
+      "learning_rate": 2e-05,
+      "loss": 5.3342,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3560385015259751,
+      "grad_norm": 0.14655709096450653,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 5308
+    },
+    {
+      "epoch": 0.35610557735520004,
+      "grad_norm": 0.1520733910085585,
+      "learning_rate": 2e-05,
+      "loss": 5.4547,
+      "step": 5309
+    },
+    {
+      "epoch": 0.356172653184425,
+      "grad_norm": 0.1480650319562172,
+      "learning_rate": 2e-05,
+      "loss": 5.5771,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3562397290136499,
+      "grad_norm": 0.1509816259325245,
+      "learning_rate": 2e-05,
+      "loss": 5.3344,
+      "step": 5311
+    },
+    {
+      "epoch": 0.35630680484287486,
+      "grad_norm": 0.15392105406832088,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3563738806720998,
+      "grad_norm": 0.14958227756520753,
+      "learning_rate": 2e-05,
+      "loss": 5.5027,
+      "step": 5313
+    },
+    {
+      "epoch": 0.35644095650132473,
+      "grad_norm": 0.1461636057737527,
+      "learning_rate": 2e-05,
+      "loss": 5.5272,
+      "step": 5314
+    },
+    {
+      "epoch": 0.3565080323305497,
+      "grad_norm": 0.15169033686389002,
+      "learning_rate": 2e-05,
+      "loss": 5.343,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3565751081597746,
+      "grad_norm": 0.14895563326329986,
+      "learning_rate": 2e-05,
+      "loss": 5.3513,
+      "step": 5316
+    },
+    {
+      "epoch": 0.35664218398899955,
+      "grad_norm": 0.14344043019087224,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3567092598182245,
+      "grad_norm": 0.14332848355280742,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 5318
+    },
+    {
+      "epoch": 0.3567763356474494,
+      "grad_norm": 0.14675192360663347,
+      "learning_rate": 2e-05,
+      "loss": 5.351,
+      "step": 5319
+    },
+    {
+      "epoch": 0.35684341147667437,
+      "grad_norm": 0.14824646454914178,
+      "learning_rate": 2e-05,
+      "loss": 5.5659,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3569104873058993,
+      "grad_norm": 0.15278185545125283,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 5321
+    },
+    {
+      "epoch": 0.35697756313512424,
+      "grad_norm": 0.15316388777417353,
+      "learning_rate": 2e-05,
+      "loss": 5.3931,
+      "step": 5322
+    },
+    {
+      "epoch": 0.3570446389643492,
+      "grad_norm": 0.1525510686770255,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3571117147935741,
+      "grad_norm": 0.14151410897521827,
+      "learning_rate": 2e-05,
+      "loss": 5.4694,
+      "step": 5324
+    },
+    {
+      "epoch": 0.35717879062279906,
+      "grad_norm": 0.15162587097774832,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 5325
+    },
+    {
+      "epoch": 0.357245866452024,
+      "grad_norm": 0.14809147603578782,
+      "learning_rate": 2e-05,
+      "loss": 5.4279,
+      "step": 5326
+    },
+    {
+      "epoch": 0.35731294228124894,
+      "grad_norm": 0.14911161831117317,
+      "learning_rate": 2e-05,
+      "loss": 5.5772,
+      "step": 5327
+    },
+    {
+      "epoch": 0.3573800181104739,
+      "grad_norm": 0.14839891753194492,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 5328
+    },
+    {
+      "epoch": 0.3574470939396988,
+      "grad_norm": 0.14365223923274434,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 5329
+    },
+    {
+      "epoch": 0.35751416976892375,
+      "grad_norm": 0.14560999660712617,
+      "learning_rate": 2e-05,
+      "loss": 5.3283,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3575812455981487,
+      "grad_norm": 0.15044343282976283,
+      "learning_rate": 2e-05,
+      "loss": 5.408,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3576483214273736,
+      "grad_norm": 0.1500135478602676,
+      "learning_rate": 2e-05,
+      "loss": 5.3144,
+      "step": 5332
+    },
+    {
+      "epoch": 0.35771539725659857,
+      "grad_norm": 0.15199188495596877,
+      "learning_rate": 2e-05,
+      "loss": 5.4011,
+      "step": 5333
+    },
+    {
+      "epoch": 0.3577824730858235,
+      "grad_norm": 0.1507397772219753,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 5334
+    },
+    {
+      "epoch": 0.35784954891504844,
+      "grad_norm": 0.14739607229388757,
+      "learning_rate": 2e-05,
+      "loss": 5.4397,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3579166247442734,
+      "grad_norm": 0.14968085721209112,
+      "learning_rate": 2e-05,
+      "loss": 5.4775,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3579837005734983,
+      "grad_norm": 0.14663965083012026,
+      "learning_rate": 2e-05,
+      "loss": 5.4913,
+      "step": 5337
+    },
+    {
+      "epoch": 0.35805077640272326,
+      "grad_norm": 0.1481863173808244,
+      "learning_rate": 2e-05,
+      "loss": 5.2868,
+      "step": 5338
+    },
+    {
+      "epoch": 0.3581178522319482,
+      "grad_norm": 0.1555077147956612,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 5339
+    },
+    {
+      "epoch": 0.35818492806117314,
+      "grad_norm": 0.1515380133496339,
+      "learning_rate": 2e-05,
+      "loss": 5.5339,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3582520038903981,
+      "grad_norm": 0.14918812740302537,
+      "learning_rate": 2e-05,
+      "loss": 5.6145,
+      "step": 5341
+    },
+    {
+      "epoch": 0.358319079719623,
+      "grad_norm": 0.1608628773390164,
+      "learning_rate": 2e-05,
+      "loss": 5.3109,
+      "step": 5342
+    },
+    {
+      "epoch": 0.35838615554884795,
+      "grad_norm": 0.15362708313248413,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5343
+    },
+    {
+      "epoch": 0.3584532313780729,
+      "grad_norm": 0.14345462382621876,
+      "learning_rate": 2e-05,
+      "loss": 5.4882,
+      "step": 5344
+    },
+    {
+      "epoch": 0.35852030720729783,
+      "grad_norm": 0.1470299667578743,
+      "learning_rate": 2e-05,
+      "loss": 5.4041,
+      "step": 5345
+    },
+    {
+      "epoch": 0.35858738303652277,
+      "grad_norm": 0.15783927682631868,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 5346
+    },
+    {
+      "epoch": 0.3586544588657477,
+      "grad_norm": 0.14550541028779748,
+      "learning_rate": 2e-05,
+      "loss": 5.3654,
+      "step": 5347
+    },
+    {
+      "epoch": 0.35872153469497264,
+      "grad_norm": 0.150669071069935,
+      "learning_rate": 2e-05,
+      "loss": 5.6414,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3587886105241976,
+      "grad_norm": 0.15218967312984985,
+      "learning_rate": 2e-05,
+      "loss": 5.316,
+      "step": 5349
+    },
+    {
+      "epoch": 0.3588556863534225,
+      "grad_norm": 0.1543267459541231,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 5350
+    },
+    {
+      "epoch": 0.35892276218264746,
+      "grad_norm": 0.14608692041068513,
+      "learning_rate": 2e-05,
+      "loss": 5.3546,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3589898380118724,
+      "grad_norm": 0.1595680367871526,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 5352
+    },
+    {
+      "epoch": 0.35905691384109734,
+      "grad_norm": 0.1464542929530747,
+      "learning_rate": 2e-05,
+      "loss": 5.3966,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3591239896703223,
+      "grad_norm": 0.14641171280390386,
+      "learning_rate": 2e-05,
+      "loss": 5.3183,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3591910654995472,
+      "grad_norm": 0.15953381045293194,
+      "learning_rate": 2e-05,
+      "loss": 5.4017,
+      "step": 5355
+    },
+    {
+      "epoch": 0.35925814132877215,
+      "grad_norm": 0.15792074369276402,
+      "learning_rate": 2e-05,
+      "loss": 5.5662,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3593252171579971,
+      "grad_norm": 0.14699273830162823,
+      "learning_rate": 2e-05,
+      "loss": 5.482,
+      "step": 5357
+    },
+    {
+      "epoch": 0.35939229298722203,
+      "grad_norm": 0.15850176461242924,
+      "learning_rate": 2e-05,
+      "loss": 5.4331,
+      "step": 5358
+    },
+    {
+      "epoch": 0.35945936881644697,
+      "grad_norm": 0.1512322786961055,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 5359
+    },
+    {
+      "epoch": 0.3595264446456719,
+      "grad_norm": 0.15394108059472913,
+      "learning_rate": 2e-05,
+      "loss": 5.3828,
+      "step": 5360
+    },
+    {
+      "epoch": 0.35959352047489684,
+      "grad_norm": 0.1493203874650817,
+      "learning_rate": 2e-05,
+      "loss": 5.3924,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3596605963041218,
+      "grad_norm": 0.14607735687094936,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 5362
+    },
+    {
+      "epoch": 0.3597276721333467,
+      "grad_norm": 0.15386593293453207,
+      "learning_rate": 2e-05,
+      "loss": 5.2804,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3597947479625717,
+      "grad_norm": 0.14414136763097873,
+      "learning_rate": 2e-05,
+      "loss": 5.3001,
+      "step": 5364
+    },
+    {
+      "epoch": 0.35986182379179665,
+      "grad_norm": 0.1532534020382846,
+      "learning_rate": 2e-05,
+      "loss": 5.5059,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3599288996210216,
+      "grad_norm": 0.14743252011124053,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 5366
+    },
+    {
+      "epoch": 0.35999597545024653,
+      "grad_norm": 0.16100725862688164,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 5367
+    },
+    {
+      "epoch": 0.36006305127947147,
+      "grad_norm": 0.150416245308718,
+      "learning_rate": 2e-05,
+      "loss": 5.525,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3601301271086964,
+      "grad_norm": 0.15840724586876206,
+      "learning_rate": 2e-05,
+      "loss": 5.3995,
+      "step": 5369
+    },
+    {
+      "epoch": 0.36019720293792135,
+      "grad_norm": 0.14870273044192034,
+      "learning_rate": 2e-05,
+      "loss": 5.4945,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3602642787671463,
+      "grad_norm": 0.16254098778521134,
+      "learning_rate": 2e-05,
+      "loss": 5.4029,
+      "step": 5371
+    },
+    {
+      "epoch": 0.3603313545963712,
+      "grad_norm": 0.1703353486986829,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 5372
+    },
+    {
+      "epoch": 0.36039843042559616,
+      "grad_norm": 0.15009792424180876,
+      "learning_rate": 2e-05,
+      "loss": 5.5125,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3604655062548211,
+      "grad_norm": 0.15857616372053626,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 5374
+    },
+    {
+      "epoch": 0.36053258208404604,
+      "grad_norm": 0.16660772089713893,
+      "learning_rate": 2e-05,
+      "loss": 5.4904,
+      "step": 5375
+    },
+    {
+      "epoch": 0.360599657913271,
+      "grad_norm": 0.14924306836895682,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3606667337424959,
+      "grad_norm": 0.15081122806904998,
+      "learning_rate": 2e-05,
+      "loss": 5.5376,
+      "step": 5377
+    },
+    {
+      "epoch": 0.36073380957172085,
+      "grad_norm": 0.15518068334042262,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 5378
+    },
+    {
+      "epoch": 0.3608008854009458,
+      "grad_norm": 0.1522546673898249,
+      "learning_rate": 2e-05,
+      "loss": 5.2947,
+      "step": 5379
+    },
+    {
+      "epoch": 0.36086796123017073,
+      "grad_norm": 0.1433819861182562,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 5380
+    },
+    {
+      "epoch": 0.36093503705939567,
+      "grad_norm": 0.15306982730261015,
+      "learning_rate": 2e-05,
+      "loss": 5.5207,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3610021128886206,
+      "grad_norm": 0.1458032464278277,
+      "learning_rate": 2e-05,
+      "loss": 5.5621,
+      "step": 5382
+    },
+    {
+      "epoch": 0.36106918871784555,
+      "grad_norm": 0.14792544767278443,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3611362645470705,
+      "grad_norm": 0.14310542516044272,
+      "learning_rate": 2e-05,
+      "loss": 5.3099,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3612033403762954,
+      "grad_norm": 0.14775757125420266,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 5385
+    },
+    {
+      "epoch": 0.36127041620552036,
+      "grad_norm": 0.14833978727553715,
+      "learning_rate": 2e-05,
+      "loss": 5.4484,
+      "step": 5386
+    },
+    {
+      "epoch": 0.3613374920347453,
+      "grad_norm": 0.1399412860996333,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 5387
+    },
+    {
+      "epoch": 0.36140456786397024,
+      "grad_norm": 0.145026817938437,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3614716436931952,
+      "grad_norm": 0.14282268215657698,
+      "learning_rate": 2e-05,
+      "loss": 5.3638,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3615387195224201,
+      "grad_norm": 0.14714781494314022,
+      "learning_rate": 2e-05,
+      "loss": 5.4545,
+      "step": 5390
+    },
+    {
+      "epoch": 0.36160579535164505,
+      "grad_norm": 0.1470453706252815,
+      "learning_rate": 2e-05,
+      "loss": 5.4694,
+      "step": 5391
+    },
+    {
+      "epoch": 0.36167287118087,
+      "grad_norm": 0.1489155368109233,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 5392
+    },
+    {
+      "epoch": 0.36173994701009493,
+      "grad_norm": 0.1444646203724267,
+      "learning_rate": 2e-05,
+      "loss": 5.3012,
+      "step": 5393
+    },
+    {
+      "epoch": 0.36180702283931987,
+      "grad_norm": 0.1564003168923802,
+      "learning_rate": 2e-05,
+      "loss": 5.4064,
+      "step": 5394
+    },
+    {
+      "epoch": 0.3618740986685448,
+      "grad_norm": 0.1398891599984906,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 5395
+    },
+    {
+      "epoch": 0.36194117449776975,
+      "grad_norm": 0.150794726486874,
+      "learning_rate": 2e-05,
+      "loss": 5.5206,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3620082503269947,
+      "grad_norm": 0.14903895088634517,
+      "learning_rate": 2e-05,
+      "loss": 5.321,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3620753261562196,
+      "grad_norm": 0.14358319098527464,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 5398
+    },
+    {
+      "epoch": 0.36214240198544456,
+      "grad_norm": 0.14777369226457143,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3622094778146695,
+      "grad_norm": 0.1542796190059665,
+      "learning_rate": 2e-05,
+      "loss": 5.3696,
+      "step": 5400
+    },
+    {
+      "epoch": 0.36227655364389444,
+      "grad_norm": 0.14224087005603966,
+      "learning_rate": 2e-05,
+      "loss": 5.4716,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3623436294731194,
+      "grad_norm": 0.1500075249906703,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3624107053023443,
+      "grad_norm": 0.15462757929106122,
+      "learning_rate": 2e-05,
+      "loss": 5.291,
+      "step": 5403
+    },
+    {
+      "epoch": 0.36247778113156925,
+      "grad_norm": 0.15360988145903695,
+      "learning_rate": 2e-05,
+      "loss": 5.4209,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3625448569607942,
+      "grad_norm": 0.15086114281030705,
+      "learning_rate": 2e-05,
+      "loss": 5.3927,
+      "step": 5405
+    },
+    {
+      "epoch": 0.36261193279001913,
+      "grad_norm": 0.1436206275871481,
+      "learning_rate": 2e-05,
+      "loss": 5.3536,
+      "step": 5406
+    },
+    {
+      "epoch": 0.36267900861924407,
+      "grad_norm": 0.14762145551076325,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 5407
+    },
+    {
+      "epoch": 0.362746084448469,
+      "grad_norm": 0.15879624935606546,
+      "learning_rate": 2e-05,
+      "loss": 5.4471,
+      "step": 5408
+    },
+    {
+      "epoch": 0.36281316027769395,
+      "grad_norm": 0.1508966526036397,
+      "learning_rate": 2e-05,
+      "loss": 5.3826,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3628802361069189,
+      "grad_norm": 0.15068304705657742,
+      "learning_rate": 2e-05,
+      "loss": 5.4649,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3629473119361438,
+      "grad_norm": 0.15227407450639333,
+      "learning_rate": 2e-05,
+      "loss": 5.4062,
+      "step": 5411
+    },
+    {
+      "epoch": 0.36301438776536876,
+      "grad_norm": 0.1590870686472759,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3630814635945937,
+      "grad_norm": 0.1509126420242703,
+      "learning_rate": 2e-05,
+      "loss": 5.466,
+      "step": 5413
+    },
+    {
+      "epoch": 0.36314853942381864,
+      "grad_norm": 0.1624335134070017,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 5414
+    },
+    {
+      "epoch": 0.3632156152530436,
+      "grad_norm": 0.14629461673886343,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3632826910822685,
+      "grad_norm": 0.14870930156941423,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 5416
+    },
+    {
+      "epoch": 0.36334976691149345,
+      "grad_norm": 0.1558993057785215,
+      "learning_rate": 2e-05,
+      "loss": 5.5399,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3634168427407184,
+      "grad_norm": 0.14767319969966913,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 5418
+    },
+    {
+      "epoch": 0.36348391856994333,
+      "grad_norm": 0.1605675770343836,
+      "learning_rate": 2e-05,
+      "loss": 5.5605,
+      "step": 5419
+    },
+    {
+      "epoch": 0.36355099439916827,
+      "grad_norm": 0.15380803120823688,
+      "learning_rate": 2e-05,
+      "loss": 5.4387,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3636180702283932,
+      "grad_norm": 0.16145630284462967,
+      "learning_rate": 2e-05,
+      "loss": 5.425,
+      "step": 5421
+    },
+    {
+      "epoch": 0.36368514605761815,
+      "grad_norm": 0.15643798809631754,
+      "learning_rate": 2e-05,
+      "loss": 5.3775,
+      "step": 5422
+    },
+    {
+      "epoch": 0.3637522218868431,
+      "grad_norm": 0.15595359961352345,
+      "learning_rate": 2e-05,
+      "loss": 5.5683,
+      "step": 5423
+    },
+    {
+      "epoch": 0.363819297716068,
+      "grad_norm": 0.1505847330340958,
+      "learning_rate": 2e-05,
+      "loss": 5.3621,
+      "step": 5424
+    },
+    {
+      "epoch": 0.36388637354529296,
+      "grad_norm": 0.15936828445546947,
+      "learning_rate": 2e-05,
+      "loss": 5.5236,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3639534493745179,
+      "grad_norm": 0.1478977494049453,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 5426
+    },
+    {
+      "epoch": 0.36402052520374284,
+      "grad_norm": 0.15198246059231868,
+      "learning_rate": 2e-05,
+      "loss": 5.4751,
+      "step": 5427
+    },
+    {
+      "epoch": 0.3640876010329678,
+      "grad_norm": 0.15321934417824123,
+      "learning_rate": 2e-05,
+      "loss": 5.5062,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3641546768621927,
+      "grad_norm": 0.14967420688506428,
+      "learning_rate": 2e-05,
+      "loss": 5.4504,
+      "step": 5429
+    },
+    {
+      "epoch": 0.36422175269141766,
+      "grad_norm": 0.15067380060033395,
+      "learning_rate": 2e-05,
+      "loss": 5.3571,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3642888285206426,
+      "grad_norm": 0.1505380840716092,
+      "learning_rate": 2e-05,
+      "loss": 5.3145,
+      "step": 5431
+    },
+    {
+      "epoch": 0.36435590434986753,
+      "grad_norm": 0.15551059745141374,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 5432
+    },
+    {
+      "epoch": 0.36442298017909247,
+      "grad_norm": 0.1466184925030557,
+      "learning_rate": 2e-05,
+      "loss": 5.5376,
+      "step": 5433
+    },
+    {
+      "epoch": 0.3644900560083174,
+      "grad_norm": 0.16632733730193522,
+      "learning_rate": 2e-05,
+      "loss": 5.5279,
+      "step": 5434
+    },
+    {
+      "epoch": 0.36455713183754235,
+      "grad_norm": 0.1497931854305357,
+      "learning_rate": 2e-05,
+      "loss": 5.4419,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3646242076667673,
+      "grad_norm": 0.15148278981268298,
+      "learning_rate": 2e-05,
+      "loss": 5.5459,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3646912834959922,
+      "grad_norm": 0.15290009328449256,
+      "learning_rate": 2e-05,
+      "loss": 5.5255,
+      "step": 5437
+    },
+    {
+      "epoch": 0.36475835932521716,
+      "grad_norm": 0.15196666914335688,
+      "learning_rate": 2e-05,
+      "loss": 5.4012,
+      "step": 5438
+    },
+    {
+      "epoch": 0.3648254351544421,
+      "grad_norm": 0.14977187589971255,
+      "learning_rate": 2e-05,
+      "loss": 5.6478,
+      "step": 5439
+    },
+    {
+      "epoch": 0.36489251098366704,
+      "grad_norm": 0.14462871132650695,
+      "learning_rate": 2e-05,
+      "loss": 5.4296,
+      "step": 5440
+    },
+    {
+      "epoch": 0.364959586812892,
+      "grad_norm": 0.1495789238067986,
+      "learning_rate": 2e-05,
+      "loss": 5.4062,
+      "step": 5441
+    },
+    {
+      "epoch": 0.3650266626421169,
+      "grad_norm": 0.1511770121431091,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 5442
+    },
+    {
+      "epoch": 0.36509373847134186,
+      "grad_norm": 0.14173752704836637,
+      "learning_rate": 2e-05,
+      "loss": 5.3286,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3651608143005668,
+      "grad_norm": 0.15472037603693517,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 5444
+    },
+    {
+      "epoch": 0.36522789012979173,
+      "grad_norm": 0.15621966570559556,
+      "learning_rate": 2e-05,
+      "loss": 5.4552,
+      "step": 5445
+    },
+    {
+      "epoch": 0.36529496595901667,
+      "grad_norm": 0.1419386137286694,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 5446
+    },
+    {
+      "epoch": 0.3653620417882416,
+      "grad_norm": 0.14884053371368822,
+      "learning_rate": 2e-05,
+      "loss": 5.4176,
+      "step": 5447
+    },
+    {
+      "epoch": 0.36542911761746655,
+      "grad_norm": 0.1416963671957745,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 5448
+    },
+    {
+      "epoch": 0.3654961934466915,
+      "grad_norm": 0.1440863624056914,
+      "learning_rate": 2e-05,
+      "loss": 5.3743,
+      "step": 5449
+    },
+    {
+      "epoch": 0.3655632692759164,
+      "grad_norm": 0.15269570296734628,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 5450
+    },
+    {
+      "epoch": 0.36563034510514136,
+      "grad_norm": 0.15025591423455026,
+      "learning_rate": 2e-05,
+      "loss": 5.3857,
+      "step": 5451
+    },
+    {
+      "epoch": 0.3656974209343663,
+      "grad_norm": 0.15290732430573467,
+      "learning_rate": 2e-05,
+      "loss": 5.5678,
+      "step": 5452
+    },
+    {
+      "epoch": 0.36576449676359124,
+      "grad_norm": 0.14807363916804586,
+      "learning_rate": 2e-05,
+      "loss": 5.499,
+      "step": 5453
+    },
+    {
+      "epoch": 0.3658315725928162,
+      "grad_norm": 0.14822621951617507,
+      "learning_rate": 2e-05,
+      "loss": 5.3129,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3658986484220411,
+      "grad_norm": 0.14739698964636905,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 5455
+    },
+    {
+      "epoch": 0.36596572425126606,
+      "grad_norm": 0.1443745378684563,
+      "learning_rate": 2e-05,
+      "loss": 5.6007,
+      "step": 5456
+    },
+    {
+      "epoch": 0.366032800080491,
+      "grad_norm": 0.1443358224374689,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 5457
+    },
+    {
+      "epoch": 0.36609987590971593,
+      "grad_norm": 0.15250601301458852,
+      "learning_rate": 2e-05,
+      "loss": 5.2992,
+      "step": 5458
+    },
+    {
+      "epoch": 0.36616695173894087,
+      "grad_norm": 0.14200029904806816,
+      "learning_rate": 2e-05,
+      "loss": 5.411,
+      "step": 5459
+    },
+    {
+      "epoch": 0.3662340275681658,
+      "grad_norm": 0.14385578709457708,
+      "learning_rate": 2e-05,
+      "loss": 5.4756,
+      "step": 5460
+    },
+    {
+      "epoch": 0.36630110339739075,
+      "grad_norm": 0.15190135863093124,
+      "learning_rate": 2e-05,
+      "loss": 5.3828,
+      "step": 5461
+    },
+    {
+      "epoch": 0.3663681792266157,
+      "grad_norm": 0.1539195873319875,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 5462
+    },
+    {
+      "epoch": 0.3664352550558406,
+      "grad_norm": 0.15307638184453828,
+      "learning_rate": 2e-05,
+      "loss": 5.4611,
+      "step": 5463
+    },
+    {
+      "epoch": 0.36650233088506556,
+      "grad_norm": 0.15440825635557992,
+      "learning_rate": 2e-05,
+      "loss": 5.4932,
+      "step": 5464
+    },
+    {
+      "epoch": 0.3665694067142905,
+      "grad_norm": 0.14173168835634256,
+      "learning_rate": 2e-05,
+      "loss": 5.3495,
+      "step": 5465
+    },
+    {
+      "epoch": 0.36663648254351544,
+      "grad_norm": 0.15348950288780439,
+      "learning_rate": 2e-05,
+      "loss": 5.537,
+      "step": 5466
+    },
+    {
+      "epoch": 0.3667035583727404,
+      "grad_norm": 0.15214400051550275,
+      "learning_rate": 2e-05,
+      "loss": 5.4101,
+      "step": 5467
+    },
+    {
+      "epoch": 0.3667706342019653,
+      "grad_norm": 0.14922806364617922,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 5468
+    },
+    {
+      "epoch": 0.36683771003119026,
+      "grad_norm": 0.15546918498452425,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 5469
+    },
+    {
+      "epoch": 0.3669047858604152,
+      "grad_norm": 0.15168429633956157,
+      "learning_rate": 2e-05,
+      "loss": 5.4877,
+      "step": 5470
+    },
+    {
+      "epoch": 0.36697186168964013,
+      "grad_norm": 0.14584836474609678,
+      "learning_rate": 2e-05,
+      "loss": 5.2908,
+      "step": 5471
+    },
+    {
+      "epoch": 0.36703893751886507,
+      "grad_norm": 0.14941324786022872,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 5472
+    },
+    {
+      "epoch": 0.36710601334809,
+      "grad_norm": 0.15120453299744185,
+      "learning_rate": 2e-05,
+      "loss": 5.3455,
+      "step": 5473
+    },
+    {
+      "epoch": 0.36717308917731495,
+      "grad_norm": 0.15076480378099286,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 5474
+    },
+    {
+      "epoch": 0.3672401650065399,
+      "grad_norm": 0.16239397512264894,
+      "learning_rate": 2e-05,
+      "loss": 5.5679,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3673072408357648,
+      "grad_norm": 0.14975349405691835,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 5476
+    },
+    {
+      "epoch": 0.36737431666498976,
+      "grad_norm": 0.15323943585424546,
+      "learning_rate": 2e-05,
+      "loss": 5.4453,
+      "step": 5477
+    },
+    {
+      "epoch": 0.3674413924942147,
+      "grad_norm": 0.15344945400186666,
+      "learning_rate": 2e-05,
+      "loss": 5.519,
+      "step": 5478
+    },
+    {
+      "epoch": 0.36750846832343964,
+      "grad_norm": 0.16044572615055286,
+      "learning_rate": 2e-05,
+      "loss": 5.4974,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3675755441526646,
+      "grad_norm": 0.15781973054004575,
+      "learning_rate": 2e-05,
+      "loss": 5.5384,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3676426199818895,
+      "grad_norm": 0.15150732087222665,
+      "learning_rate": 2e-05,
+      "loss": 5.467,
+      "step": 5481
+    },
+    {
+      "epoch": 0.36770969581111446,
+      "grad_norm": 0.1688274154571226,
+      "learning_rate": 2e-05,
+      "loss": 5.5013,
+      "step": 5482
+    },
+    {
+      "epoch": 0.3677767716403394,
+      "grad_norm": 0.15034195157754587,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 5483
+    },
+    {
+      "epoch": 0.36784384746956433,
+      "grad_norm": 0.14405254322538597,
+      "learning_rate": 2e-05,
+      "loss": 5.235,
+      "step": 5484
+    },
+    {
+      "epoch": 0.36791092329878927,
+      "grad_norm": 0.14182213398551044,
+      "learning_rate": 2e-05,
+      "loss": 5.4434,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3679779991280142,
+      "grad_norm": 0.14781041887249036,
+      "learning_rate": 2e-05,
+      "loss": 5.4737,
+      "step": 5486
+    },
+    {
+      "epoch": 0.36804507495723915,
+      "grad_norm": 0.16550221601479634,
+      "learning_rate": 2e-05,
+      "loss": 5.343,
+      "step": 5487
+    },
+    {
+      "epoch": 0.3681121507864641,
+      "grad_norm": 0.1437688995900029,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 5488
+    },
+    {
+      "epoch": 0.368179226615689,
+      "grad_norm": 0.14678792282682487,
+      "learning_rate": 2e-05,
+      "loss": 5.3134,
+      "step": 5489
+    },
+    {
+      "epoch": 0.36824630244491396,
+      "grad_norm": 0.14432726211253152,
+      "learning_rate": 2e-05,
+      "loss": 5.479,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3683133782741389,
+      "grad_norm": 0.1473166720103478,
+      "learning_rate": 2e-05,
+      "loss": 5.3963,
+      "step": 5491
+    },
+    {
+      "epoch": 0.36838045410336384,
+      "grad_norm": 0.15363349100641463,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 5492
+    },
+    {
+      "epoch": 0.3684475299325888,
+      "grad_norm": 0.14725497298077556,
+      "learning_rate": 2e-05,
+      "loss": 5.3065,
+      "step": 5493
+    },
+    {
+      "epoch": 0.3685146057618137,
+      "grad_norm": 0.1441947550795249,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 5494
+    },
+    {
+      "epoch": 0.36858168159103866,
+      "grad_norm": 0.14864274887202222,
+      "learning_rate": 2e-05,
+      "loss": 5.4553,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3686487574202636,
+      "grad_norm": 0.14916415961213877,
+      "learning_rate": 2e-05,
+      "loss": 5.3625,
+      "step": 5496
+    },
+    {
+      "epoch": 0.36871583324948853,
+      "grad_norm": 0.15043473156459355,
+      "learning_rate": 2e-05,
+      "loss": 5.5156,
+      "step": 5497
+    },
+    {
+      "epoch": 0.3687829090787135,
+      "grad_norm": 0.14777345471517014,
+      "learning_rate": 2e-05,
+      "loss": 5.5154,
+      "step": 5498
+    },
+    {
+      "epoch": 0.3688499849079384,
+      "grad_norm": 0.15852555066778976,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 5499
+    },
+    {
+      "epoch": 0.36891706073716335,
+      "grad_norm": 0.15841599653649124,
+      "learning_rate": 2e-05,
+      "loss": 5.3811,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3689841365663883,
+      "grad_norm": 0.14537778019808395,
+      "learning_rate": 2e-05,
+      "loss": 5.3768,
+      "step": 5501
+    },
+    {
+      "epoch": 0.3690512123956132,
+      "grad_norm": 0.14727291697818978,
+      "learning_rate": 2e-05,
+      "loss": 5.3023,
+      "step": 5502
+    },
+    {
+      "epoch": 0.36911828822483816,
+      "grad_norm": 0.16372196379078102,
+      "learning_rate": 2e-05,
+      "loss": 5.3893,
+      "step": 5503
+    },
+    {
+      "epoch": 0.3691853640540631,
+      "grad_norm": 0.15305358002299205,
+      "learning_rate": 2e-05,
+      "loss": 5.3671,
+      "step": 5504
+    },
+    {
+      "epoch": 0.36925243988328804,
+      "grad_norm": 0.1519634368839271,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 5505
+    },
+    {
+      "epoch": 0.369319515712513,
+      "grad_norm": 0.1534503443413111,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 5506
+    },
+    {
+      "epoch": 0.3693865915417379,
+      "grad_norm": 0.1530998253206066,
+      "learning_rate": 2e-05,
+      "loss": 5.4396,
+      "step": 5507
+    },
+    {
+      "epoch": 0.36945366737096286,
+      "grad_norm": 0.14408014426841553,
+      "learning_rate": 2e-05,
+      "loss": 5.2619,
+      "step": 5508
+    },
+    {
+      "epoch": 0.3695207432001878,
+      "grad_norm": 0.15259618736667727,
+      "learning_rate": 2e-05,
+      "loss": 5.4282,
+      "step": 5509
+    },
+    {
+      "epoch": 0.36958781902941273,
+      "grad_norm": 0.150064079456455,
+      "learning_rate": 2e-05,
+      "loss": 5.3441,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3696548948586377,
+      "grad_norm": 0.15020856718325531,
+      "learning_rate": 2e-05,
+      "loss": 5.3731,
+      "step": 5511
+    },
+    {
+      "epoch": 0.3697219706878626,
+      "grad_norm": 0.15269552182986748,
+      "learning_rate": 2e-05,
+      "loss": 5.2683,
+      "step": 5512
+    },
+    {
+      "epoch": 0.36978904651708755,
+      "grad_norm": 0.14814813382933187,
+      "learning_rate": 2e-05,
+      "loss": 5.5531,
+      "step": 5513
+    },
+    {
+      "epoch": 0.3698561223463125,
+      "grad_norm": 0.1452003217550247,
+      "learning_rate": 2e-05,
+      "loss": 5.4592,
+      "step": 5514
+    },
+    {
+      "epoch": 0.3699231981755374,
+      "grad_norm": 0.14937634031664626,
+      "learning_rate": 2e-05,
+      "loss": 5.3561,
+      "step": 5515
+    },
+    {
+      "epoch": 0.36999027400476237,
+      "grad_norm": 0.1479573666261167,
+      "learning_rate": 2e-05,
+      "loss": 5.3849,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3700573498339873,
+      "grad_norm": 0.14112520005164267,
+      "learning_rate": 2e-05,
+      "loss": 5.3403,
+      "step": 5517
+    },
+    {
+      "epoch": 0.37012442566321224,
+      "grad_norm": 0.1420855371835133,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 5518
+    },
+    {
+      "epoch": 0.3701915014924372,
+      "grad_norm": 0.1419970950961018,
+      "learning_rate": 2e-05,
+      "loss": 5.39,
+      "step": 5519
+    },
+    {
+      "epoch": 0.3702585773216621,
+      "grad_norm": 0.14304120751869323,
+      "learning_rate": 2e-05,
+      "loss": 5.5319,
+      "step": 5520
+    },
+    {
+      "epoch": 0.37032565315088706,
+      "grad_norm": 0.14352641900913535,
+      "learning_rate": 2e-05,
+      "loss": 5.4827,
+      "step": 5521
+    },
+    {
+      "epoch": 0.370392728980112,
+      "grad_norm": 0.14780231835756938,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 5522
+    },
+    {
+      "epoch": 0.37045980480933693,
+      "grad_norm": 0.13928018768756434,
+      "learning_rate": 2e-05,
+      "loss": 5.4516,
+      "step": 5523
+    },
+    {
+      "epoch": 0.3705268806385619,
+      "grad_norm": 0.15120962542831176,
+      "learning_rate": 2e-05,
+      "loss": 5.2851,
+      "step": 5524
+    },
+    {
+      "epoch": 0.3705939564677868,
+      "grad_norm": 0.14324955944748685,
+      "learning_rate": 2e-05,
+      "loss": 5.5173,
+      "step": 5525
+    },
+    {
+      "epoch": 0.37066103229701175,
+      "grad_norm": 0.1396918276681697,
+      "learning_rate": 2e-05,
+      "loss": 5.3201,
+      "step": 5526
+    },
+    {
+      "epoch": 0.3707281081262367,
+      "grad_norm": 0.15215704720028017,
+      "learning_rate": 2e-05,
+      "loss": 5.4143,
+      "step": 5527
+    },
+    {
+      "epoch": 0.3707951839554616,
+      "grad_norm": 0.1491225738479851,
+      "learning_rate": 2e-05,
+      "loss": 5.5254,
+      "step": 5528
+    },
+    {
+      "epoch": 0.37086225978468657,
+      "grad_norm": 0.14476099405627413,
+      "learning_rate": 2e-05,
+      "loss": 5.5344,
+      "step": 5529
+    },
+    {
+      "epoch": 0.3709293356139115,
+      "grad_norm": 0.16166656373773916,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 5530
+    },
+    {
+      "epoch": 0.37099641144313644,
+      "grad_norm": 0.14891293706509834,
+      "learning_rate": 2e-05,
+      "loss": 5.4113,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3710634872723614,
+      "grad_norm": 0.15803682294318247,
+      "learning_rate": 2e-05,
+      "loss": 5.5941,
+      "step": 5532
+    },
+    {
+      "epoch": 0.3711305631015863,
+      "grad_norm": 0.15554981965507686,
+      "learning_rate": 2e-05,
+      "loss": 5.4206,
+      "step": 5533
+    },
+    {
+      "epoch": 0.37119763893081126,
+      "grad_norm": 0.14352288001119287,
+      "learning_rate": 2e-05,
+      "loss": 5.4445,
+      "step": 5534
+    },
+    {
+      "epoch": 0.3712647147600362,
+      "grad_norm": 0.1499274501060584,
+      "learning_rate": 2e-05,
+      "loss": 5.5961,
+      "step": 5535
+    },
+    {
+      "epoch": 0.37133179058926113,
+      "grad_norm": 0.14810230421430595,
+      "learning_rate": 2e-05,
+      "loss": 5.3082,
+      "step": 5536
+    },
+    {
+      "epoch": 0.3713988664184861,
+      "grad_norm": 0.14721825291792134,
+      "learning_rate": 2e-05,
+      "loss": 5.4266,
+      "step": 5537
+    },
+    {
+      "epoch": 0.371465942247711,
+      "grad_norm": 0.15224724138266346,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 5538
+    },
+    {
+      "epoch": 0.37153301807693595,
+      "grad_norm": 0.14669044198908834,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 5539
+    },
+    {
+      "epoch": 0.3716000939061609,
+      "grad_norm": 0.1427712981154258,
+      "learning_rate": 2e-05,
+      "loss": 5.3259,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3716671697353858,
+      "grad_norm": 0.1540407373946393,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 5541
+    },
+    {
+      "epoch": 0.37173424556461077,
+      "grad_norm": 0.14843591557650118,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 5542
+    },
+    {
+      "epoch": 0.3718013213938357,
+      "grad_norm": 0.1474571034865597,
+      "learning_rate": 2e-05,
+      "loss": 5.3717,
+      "step": 5543
+    },
+    {
+      "epoch": 0.37186839722306064,
+      "grad_norm": 0.15227443277218533,
+      "learning_rate": 2e-05,
+      "loss": 5.4016,
+      "step": 5544
+    },
+    {
+      "epoch": 0.3719354730522856,
+      "grad_norm": 0.14123862346260413,
+      "learning_rate": 2e-05,
+      "loss": 5.3334,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3720025488815106,
+      "grad_norm": 0.14796602051134664,
+      "learning_rate": 2e-05,
+      "loss": 5.4211,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3720696247107355,
+      "grad_norm": 0.14072525105488323,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 5547
+    },
+    {
+      "epoch": 0.37213670053996045,
+      "grad_norm": 0.14607779296206377,
+      "learning_rate": 2e-05,
+      "loss": 5.5267,
+      "step": 5548
+    },
+    {
+      "epoch": 0.3722037763691854,
+      "grad_norm": 0.14266174746074073,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 5549
+    },
+    {
+      "epoch": 0.37227085219841033,
+      "grad_norm": 0.14416203884262846,
+      "learning_rate": 2e-05,
+      "loss": 5.633,
+      "step": 5550
+    },
+    {
+      "epoch": 0.37233792802763527,
+      "grad_norm": 0.14559050176652955,
+      "learning_rate": 2e-05,
+      "loss": 5.3644,
+      "step": 5551
+    },
+    {
+      "epoch": 0.3724050038568602,
+      "grad_norm": 0.15064941292615847,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 5552
+    },
+    {
+      "epoch": 0.37247207968608514,
+      "grad_norm": 0.15191189493947688,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 5553
+    },
+    {
+      "epoch": 0.3725391555153101,
+      "grad_norm": 0.14656529109820923,
+      "learning_rate": 2e-05,
+      "loss": 5.5203,
+      "step": 5554
+    },
+    {
+      "epoch": 0.372606231344535,
+      "grad_norm": 0.14839020812231873,
+      "learning_rate": 2e-05,
+      "loss": 5.3811,
+      "step": 5555
+    },
+    {
+      "epoch": 0.37267330717375996,
+      "grad_norm": 0.14054673116525768,
+      "learning_rate": 2e-05,
+      "loss": 5.6181,
+      "step": 5556
+    },
+    {
+      "epoch": 0.3727403830029849,
+      "grad_norm": 0.154988402156191,
+      "learning_rate": 2e-05,
+      "loss": 5.3881,
+      "step": 5557
+    },
+    {
+      "epoch": 0.37280745883220984,
+      "grad_norm": 0.1484114432266767,
+      "learning_rate": 2e-05,
+      "loss": 5.3575,
+      "step": 5558
+    },
+    {
+      "epoch": 0.3728745346614348,
+      "grad_norm": 0.1521287072490179,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 5559
+    },
+    {
+      "epoch": 0.3729416104906597,
+      "grad_norm": 0.15287835659323368,
+      "learning_rate": 2e-05,
+      "loss": 5.2856,
+      "step": 5560
+    },
+    {
+      "epoch": 0.37300868631988465,
+      "grad_norm": 0.15370088302341264,
+      "learning_rate": 2e-05,
+      "loss": 5.5718,
+      "step": 5561
+    },
+    {
+      "epoch": 0.3730757621491096,
+      "grad_norm": 0.1668478097540375,
+      "learning_rate": 2e-05,
+      "loss": 5.4343,
+      "step": 5562
+    },
+    {
+      "epoch": 0.37314283797833453,
+      "grad_norm": 0.1574447959381691,
+      "learning_rate": 2e-05,
+      "loss": 5.3504,
+      "step": 5563
+    },
+    {
+      "epoch": 0.37320991380755947,
+      "grad_norm": 0.14594117683507551,
+      "learning_rate": 2e-05,
+      "loss": 5.444,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3732769896367844,
+      "grad_norm": 0.1510638839218263,
+      "learning_rate": 2e-05,
+      "loss": 5.3343,
+      "step": 5565
+    },
+    {
+      "epoch": 0.37334406546600934,
+      "grad_norm": 0.15916342192380262,
+      "learning_rate": 2e-05,
+      "loss": 5.499,
+      "step": 5566
+    },
+    {
+      "epoch": 0.3734111412952343,
+      "grad_norm": 0.16071424671918755,
+      "learning_rate": 2e-05,
+      "loss": 5.4411,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3734782171244592,
+      "grad_norm": 0.14609988837184734,
+      "learning_rate": 2e-05,
+      "loss": 5.2535,
+      "step": 5568
+    },
+    {
+      "epoch": 0.37354529295368416,
+      "grad_norm": 0.156520907572076,
+      "learning_rate": 2e-05,
+      "loss": 5.2823,
+      "step": 5569
+    },
+    {
+      "epoch": 0.3736123687829091,
+      "grad_norm": 0.15498779068862176,
+      "learning_rate": 2e-05,
+      "loss": 5.5263,
+      "step": 5570
+    },
+    {
+      "epoch": 0.37367944461213404,
+      "grad_norm": 0.14710194112710948,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 5571
+    },
+    {
+      "epoch": 0.373746520441359,
+      "grad_norm": 0.15671377966416863,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 5572
+    },
+    {
+      "epoch": 0.3738135962705839,
+      "grad_norm": 0.15139907147728582,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 5573
+    },
+    {
+      "epoch": 0.37388067209980885,
+      "grad_norm": 0.14450880234482671,
+      "learning_rate": 2e-05,
+      "loss": 5.3829,
+      "step": 5574
+    },
+    {
+      "epoch": 0.3739477479290338,
+      "grad_norm": 0.14864949500504998,
+      "learning_rate": 2e-05,
+      "loss": 5.4772,
+      "step": 5575
+    },
+    {
+      "epoch": 0.37401482375825873,
+      "grad_norm": 0.13850890009342715,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 5576
+    },
+    {
+      "epoch": 0.37408189958748367,
+      "grad_norm": 0.14917010126674998,
+      "learning_rate": 2e-05,
+      "loss": 5.5175,
+      "step": 5577
+    },
+    {
+      "epoch": 0.3741489754167086,
+      "grad_norm": 0.15160585090757833,
+      "learning_rate": 2e-05,
+      "loss": 5.354,
+      "step": 5578
+    },
+    {
+      "epoch": 0.37421605124593355,
+      "grad_norm": 0.15124008302812078,
+      "learning_rate": 2e-05,
+      "loss": 5.3575,
+      "step": 5579
+    },
+    {
+      "epoch": 0.3742831270751585,
+      "grad_norm": 0.14741318550399948,
+      "learning_rate": 2e-05,
+      "loss": 5.4706,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3743502029043834,
+      "grad_norm": 0.14668569054791308,
+      "learning_rate": 2e-05,
+      "loss": 5.3874,
+      "step": 5581
+    },
+    {
+      "epoch": 0.37441727873360836,
+      "grad_norm": 0.15186573618789523,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 5582
+    },
+    {
+      "epoch": 0.3744843545628333,
+      "grad_norm": 0.15094078552355342,
+      "learning_rate": 2e-05,
+      "loss": 5.4831,
+      "step": 5583
+    },
+    {
+      "epoch": 0.37455143039205824,
+      "grad_norm": 0.14354134814834418,
+      "learning_rate": 2e-05,
+      "loss": 5.3419,
+      "step": 5584
+    },
+    {
+      "epoch": 0.3746185062212832,
+      "grad_norm": 0.14454810641850804,
+      "learning_rate": 2e-05,
+      "loss": 5.4407,
+      "step": 5585
+    },
+    {
+      "epoch": 0.3746855820505081,
+      "grad_norm": 0.15097254138078867,
+      "learning_rate": 2e-05,
+      "loss": 5.3702,
+      "step": 5586
+    },
+    {
+      "epoch": 0.37475265787973305,
+      "grad_norm": 0.1477994122020231,
+      "learning_rate": 2e-05,
+      "loss": 5.5111,
+      "step": 5587
+    },
+    {
+      "epoch": 0.374819733708958,
+      "grad_norm": 0.15447008820843638,
+      "learning_rate": 2e-05,
+      "loss": 5.5062,
+      "step": 5588
+    },
+    {
+      "epoch": 0.37488680953818293,
+      "grad_norm": 0.15077334713700072,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 5589
+    },
+    {
+      "epoch": 0.37495388536740787,
+      "grad_norm": 0.15414863951257665,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3750209611966328,
+      "grad_norm": 0.15460815325086488,
+      "learning_rate": 2e-05,
+      "loss": 5.6081,
+      "step": 5591
+    },
+    {
+      "epoch": 0.37508803702585775,
+      "grad_norm": 0.15461685778274153,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 5592
+    },
+    {
+      "epoch": 0.3751551128550827,
+      "grad_norm": 0.14842581322187065,
+      "learning_rate": 2e-05,
+      "loss": 5.3438,
+      "step": 5593
+    },
+    {
+      "epoch": 0.3752221886843076,
+      "grad_norm": 0.151392575111644,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 5594
+    },
+    {
+      "epoch": 0.37528926451353256,
+      "grad_norm": 0.1434304955963111,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3753563403427575,
+      "grad_norm": 0.1501258694477132,
+      "learning_rate": 2e-05,
+      "loss": 5.4388,
+      "step": 5596
+    },
+    {
+      "epoch": 0.37542341617198244,
+      "grad_norm": 0.15456145768057547,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 5597
+    },
+    {
+      "epoch": 0.3754904920012074,
+      "grad_norm": 0.14233773086637688,
+      "learning_rate": 2e-05,
+      "loss": 5.4888,
+      "step": 5598
+    },
+    {
+      "epoch": 0.3755575678304323,
+      "grad_norm": 0.14721919463821326,
+      "learning_rate": 2e-05,
+      "loss": 5.3096,
+      "step": 5599
+    },
+    {
+      "epoch": 0.37562464365965725,
+      "grad_norm": 0.1536187722150757,
+      "learning_rate": 2e-05,
+      "loss": 5.559,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3756917194888822,
+      "grad_norm": 0.1489342866086286,
+      "learning_rate": 2e-05,
+      "loss": 5.2914,
+      "step": 5601
+    },
+    {
+      "epoch": 0.37575879531810713,
+      "grad_norm": 0.1433655519035909,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 5602
+    },
+    {
+      "epoch": 0.37582587114733207,
+      "grad_norm": 0.1466145011373058,
+      "learning_rate": 2e-05,
+      "loss": 5.3028,
+      "step": 5603
+    },
+    {
+      "epoch": 0.375892946976557,
+      "grad_norm": 0.14402930264128616,
+      "learning_rate": 2e-05,
+      "loss": 5.4552,
+      "step": 5604
+    },
+    {
+      "epoch": 0.37596002280578195,
+      "grad_norm": 0.1542693291420134,
+      "learning_rate": 2e-05,
+      "loss": 5.5299,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3760270986350069,
+      "grad_norm": 0.14201318792822237,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 5606
+    },
+    {
+      "epoch": 0.3760941744642318,
+      "grad_norm": 0.1535874154397052,
+      "learning_rate": 2e-05,
+      "loss": 5.5668,
+      "step": 5607
+    },
+    {
+      "epoch": 0.37616125029345676,
+      "grad_norm": 0.15188263496620896,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3762283261226817,
+      "grad_norm": 0.1468698805824067,
+      "learning_rate": 2e-05,
+      "loss": 5.5188,
+      "step": 5609
+    },
+    {
+      "epoch": 0.37629540195190664,
+      "grad_norm": 0.14428681036956412,
+      "learning_rate": 2e-05,
+      "loss": 5.5626,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3763624777811316,
+      "grad_norm": 0.1478433486706171,
+      "learning_rate": 2e-05,
+      "loss": 5.3745,
+      "step": 5611
+    },
+    {
+      "epoch": 0.3764295536103565,
+      "grad_norm": 0.14071286527197796,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 5612
+    },
+    {
+      "epoch": 0.37649662943958145,
+      "grad_norm": 0.1474376572648173,
+      "learning_rate": 2e-05,
+      "loss": 5.5427,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3765637052688064,
+      "grad_norm": 0.1552786033378589,
+      "learning_rate": 2e-05,
+      "loss": 5.5162,
+      "step": 5614
+    },
+    {
+      "epoch": 0.37663078109803133,
+      "grad_norm": 0.1457005390957894,
+      "learning_rate": 2e-05,
+      "loss": 5.3894,
+      "step": 5615
+    },
+    {
+      "epoch": 0.37669785692725627,
+      "grad_norm": 0.14954680571033063,
+      "learning_rate": 2e-05,
+      "loss": 5.4366,
+      "step": 5616
+    },
+    {
+      "epoch": 0.3767649327564812,
+      "grad_norm": 0.15264090241256678,
+      "learning_rate": 2e-05,
+      "loss": 5.3803,
+      "step": 5617
+    },
+    {
+      "epoch": 0.37683200858570615,
+      "grad_norm": 0.148066593143168,
+      "learning_rate": 2e-05,
+      "loss": 5.356,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3768990844149311,
+      "grad_norm": 0.15190660855045435,
+      "learning_rate": 2e-05,
+      "loss": 5.3864,
+      "step": 5619
+    },
+    {
+      "epoch": 0.376966160244156,
+      "grad_norm": 0.1464637384228485,
+      "learning_rate": 2e-05,
+      "loss": 5.6101,
+      "step": 5620
+    },
+    {
+      "epoch": 0.37703323607338096,
+      "grad_norm": 0.1521534079648647,
+      "learning_rate": 2e-05,
+      "loss": 5.44,
+      "step": 5621
+    },
+    {
+      "epoch": 0.3771003119026059,
+      "grad_norm": 0.15571326185839346,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 5622
+    },
+    {
+      "epoch": 0.37716738773183084,
+      "grad_norm": 0.14753796488181944,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 5623
+    },
+    {
+      "epoch": 0.3772344635610558,
+      "grad_norm": 0.14379622495281016,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 5624
+    },
+    {
+      "epoch": 0.3773015393902807,
+      "grad_norm": 0.15049692803299444,
+      "learning_rate": 2e-05,
+      "loss": 5.3839,
+      "step": 5625
+    },
+    {
+      "epoch": 0.37736861521950565,
+      "grad_norm": 0.14867128586019063,
+      "learning_rate": 2e-05,
+      "loss": 5.3128,
+      "step": 5626
+    },
+    {
+      "epoch": 0.3774356910487306,
+      "grad_norm": 0.14085668815504895,
+      "learning_rate": 2e-05,
+      "loss": 5.4422,
+      "step": 5627
+    },
+    {
+      "epoch": 0.37750276687795553,
+      "grad_norm": 0.15811497814131273,
+      "learning_rate": 2e-05,
+      "loss": 5.3583,
+      "step": 5628
+    },
+    {
+      "epoch": 0.37756984270718047,
+      "grad_norm": 0.15050840673456664,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 5629
+    },
+    {
+      "epoch": 0.3776369185364054,
+      "grad_norm": 0.14464895717444495,
+      "learning_rate": 2e-05,
+      "loss": 5.485,
+      "step": 5630
+    },
+    {
+      "epoch": 0.37770399436563035,
+      "grad_norm": 0.14537872128695847,
+      "learning_rate": 2e-05,
+      "loss": 5.4772,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3777710701948553,
+      "grad_norm": 0.14929018457211307,
+      "learning_rate": 2e-05,
+      "loss": 5.4757,
+      "step": 5632
+    },
+    {
+      "epoch": 0.3778381460240802,
+      "grad_norm": 0.143446398956801,
+      "learning_rate": 2e-05,
+      "loss": 5.4993,
+      "step": 5633
+    },
+    {
+      "epoch": 0.37790522185330516,
+      "grad_norm": 0.14230341964755744,
+      "learning_rate": 2e-05,
+      "loss": 5.4279,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3779722976825301,
+      "grad_norm": 0.14735019389697465,
+      "learning_rate": 2e-05,
+      "loss": 5.3356,
+      "step": 5635
+    },
+    {
+      "epoch": 0.37803937351175504,
+      "grad_norm": 0.14242151603472458,
+      "learning_rate": 2e-05,
+      "loss": 5.3347,
+      "step": 5636
+    },
+    {
+      "epoch": 0.37810644934098,
+      "grad_norm": 0.14187749570962258,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 5637
+    },
+    {
+      "epoch": 0.3781735251702049,
+      "grad_norm": 0.14714986016908446,
+      "learning_rate": 2e-05,
+      "loss": 5.5736,
+      "step": 5638
+    },
+    {
+      "epoch": 0.37824060099942985,
+      "grad_norm": 0.15659338169965129,
+      "learning_rate": 2e-05,
+      "loss": 5.388,
+      "step": 5639
+    },
+    {
+      "epoch": 0.3783076768286548,
+      "grad_norm": 0.15024832631959611,
+      "learning_rate": 2e-05,
+      "loss": 5.3646,
+      "step": 5640
+    },
+    {
+      "epoch": 0.37837475265787973,
+      "grad_norm": 0.14398377667448664,
+      "learning_rate": 2e-05,
+      "loss": 5.5238,
+      "step": 5641
+    },
+    {
+      "epoch": 0.37844182848710467,
+      "grad_norm": 0.1549214381118798,
+      "learning_rate": 2e-05,
+      "loss": 5.5697,
+      "step": 5642
+    },
+    {
+      "epoch": 0.3785089043163296,
+      "grad_norm": 0.1455794045705954,
+      "learning_rate": 2e-05,
+      "loss": 5.5641,
+      "step": 5643
+    },
+    {
+      "epoch": 0.37857598014555455,
+      "grad_norm": 0.1450838024664113,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 5644
+    },
+    {
+      "epoch": 0.3786430559747795,
+      "grad_norm": 0.1435940082297661,
+      "learning_rate": 2e-05,
+      "loss": 5.3042,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3787101318040044,
+      "grad_norm": 0.14534365034970545,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 5646
+    },
+    {
+      "epoch": 0.37877720763322936,
+      "grad_norm": 0.15246764940722343,
+      "learning_rate": 2e-05,
+      "loss": 5.4133,
+      "step": 5647
+    },
+    {
+      "epoch": 0.3788442834624543,
+      "grad_norm": 0.15168053306963955,
+      "learning_rate": 2e-05,
+      "loss": 5.5226,
+      "step": 5648
+    },
+    {
+      "epoch": 0.37891135929167924,
+      "grad_norm": 0.14991256578405002,
+      "learning_rate": 2e-05,
+      "loss": 5.5371,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3789784351209042,
+      "grad_norm": 0.14811622140820938,
+      "learning_rate": 2e-05,
+      "loss": 5.4283,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3790455109501291,
+      "grad_norm": 0.14884936514944025,
+      "learning_rate": 2e-05,
+      "loss": 5.3068,
+      "step": 5651
+    },
+    {
+      "epoch": 0.37911258677935405,
+      "grad_norm": 0.14754440824666046,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 5652
+    },
+    {
+      "epoch": 0.379179662608579,
+      "grad_norm": 0.14091289456339712,
+      "learning_rate": 2e-05,
+      "loss": 5.4343,
+      "step": 5653
+    },
+    {
+      "epoch": 0.37924673843780393,
+      "grad_norm": 0.14257657239024757,
+      "learning_rate": 2e-05,
+      "loss": 5.4038,
+      "step": 5654
+    },
+    {
+      "epoch": 0.37931381426702887,
+      "grad_norm": 0.14558516288391837,
+      "learning_rate": 2e-05,
+      "loss": 5.3652,
+      "step": 5655
+    },
+    {
+      "epoch": 0.3793808900962538,
+      "grad_norm": 0.14703277495103326,
+      "learning_rate": 2e-05,
+      "loss": 5.3626,
+      "step": 5656
+    },
+    {
+      "epoch": 0.37944796592547875,
+      "grad_norm": 0.14658116734706209,
+      "learning_rate": 2e-05,
+      "loss": 5.2893,
+      "step": 5657
+    },
+    {
+      "epoch": 0.3795150417547037,
+      "grad_norm": 0.14189107681409643,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 5658
+    },
+    {
+      "epoch": 0.3795821175839286,
+      "grad_norm": 0.1438752452153265,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 5659
+    },
+    {
+      "epoch": 0.37964919341315356,
+      "grad_norm": 0.14903073496653785,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3797162692423785,
+      "grad_norm": 0.14403824052363287,
+      "learning_rate": 2e-05,
+      "loss": 5.5573,
+      "step": 5661
+    },
+    {
+      "epoch": 0.37978334507160344,
+      "grad_norm": 0.14064920027906896,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3798504209008284,
+      "grad_norm": 0.1439063417224582,
+      "learning_rate": 2e-05,
+      "loss": 5.3619,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3799174967300533,
+      "grad_norm": 0.14853065497675222,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 5664
+    },
+    {
+      "epoch": 0.37998457255927826,
+      "grad_norm": 0.1449137590231032,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3800516483885032,
+      "grad_norm": 0.15270187462220255,
+      "learning_rate": 2e-05,
+      "loss": 5.3982,
+      "step": 5666
+    },
+    {
+      "epoch": 0.38011872421772813,
+      "grad_norm": 0.1519795213197557,
+      "learning_rate": 2e-05,
+      "loss": 5.3699,
+      "step": 5667
+    },
+    {
+      "epoch": 0.38018580004695307,
+      "grad_norm": 0.14917106869611083,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 5668
+    },
+    {
+      "epoch": 0.380252875876178,
+      "grad_norm": 0.1532996875820362,
+      "learning_rate": 2e-05,
+      "loss": 5.3641,
+      "step": 5669
+    },
+    {
+      "epoch": 0.38031995170540295,
+      "grad_norm": 0.1574437664426534,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3803870275346279,
+      "grad_norm": 0.15080333735405138,
+      "learning_rate": 2e-05,
+      "loss": 5.3798,
+      "step": 5671
+    },
+    {
+      "epoch": 0.3804541033638528,
+      "grad_norm": 0.1484597728671174,
+      "learning_rate": 2e-05,
+      "loss": 5.4924,
+      "step": 5672
+    },
+    {
+      "epoch": 0.38052117919307776,
+      "grad_norm": 0.16006431831640297,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 5673
+    },
+    {
+      "epoch": 0.3805882550223027,
+      "grad_norm": 0.14630625317965001,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 5674
+    },
+    {
+      "epoch": 0.38065533085152764,
+      "grad_norm": 0.1521513921257831,
+      "learning_rate": 2e-05,
+      "loss": 5.4455,
+      "step": 5675
+    },
+    {
+      "epoch": 0.3807224066807526,
+      "grad_norm": 0.1564945510547526,
+      "learning_rate": 2e-05,
+      "loss": 5.5749,
+      "step": 5676
+    },
+    {
+      "epoch": 0.3807894825099775,
+      "grad_norm": 0.16097006498441713,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 5677
+    },
+    {
+      "epoch": 0.38085655833920246,
+      "grad_norm": 0.14333959791860254,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 5678
+    },
+    {
+      "epoch": 0.3809236341684274,
+      "grad_norm": 0.14800044537478288,
+      "learning_rate": 2e-05,
+      "loss": 5.4103,
+      "step": 5679
+    },
+    {
+      "epoch": 0.38099070999765233,
+      "grad_norm": 0.15605846115922317,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 5680
+    },
+    {
+      "epoch": 0.38105778582687727,
+      "grad_norm": 0.14679593402373675,
+      "learning_rate": 2e-05,
+      "loss": 5.3302,
+      "step": 5681
+    },
+    {
+      "epoch": 0.3811248616561022,
+      "grad_norm": 0.1448909781964578,
+      "learning_rate": 2e-05,
+      "loss": 5.5888,
+      "step": 5682
+    },
+    {
+      "epoch": 0.38119193748532715,
+      "grad_norm": 0.14623760113700685,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 5683
+    },
+    {
+      "epoch": 0.3812590133145521,
+      "grad_norm": 0.15143568485298878,
+      "learning_rate": 2e-05,
+      "loss": 5.4996,
+      "step": 5684
+    },
+    {
+      "epoch": 0.381326089143777,
+      "grad_norm": 0.14353407874719562,
+      "learning_rate": 2e-05,
+      "loss": 5.4701,
+      "step": 5685
+    },
+    {
+      "epoch": 0.38139316497300196,
+      "grad_norm": 0.15022663469564407,
+      "learning_rate": 2e-05,
+      "loss": 5.3713,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3814602408022269,
+      "grad_norm": 0.15614605463027664,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 5687
+    },
+    {
+      "epoch": 0.38152731663145184,
+      "grad_norm": 0.16275272368386842,
+      "learning_rate": 2e-05,
+      "loss": 5.4579,
+      "step": 5688
+    },
+    {
+      "epoch": 0.3815943924606768,
+      "grad_norm": 0.14376849861535926,
+      "learning_rate": 2e-05,
+      "loss": 5.3594,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3816614682899017,
+      "grad_norm": 0.1570172844027234,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 5690
+    },
+    {
+      "epoch": 0.38172854411912666,
+      "grad_norm": 0.15652575741503774,
+      "learning_rate": 2e-05,
+      "loss": 5.41,
+      "step": 5691
+    },
+    {
+      "epoch": 0.3817956199483516,
+      "grad_norm": 0.15601793407937498,
+      "learning_rate": 2e-05,
+      "loss": 5.4781,
+      "step": 5692
+    },
+    {
+      "epoch": 0.38186269577757653,
+      "grad_norm": 0.14965045503480381,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 5693
+    },
+    {
+      "epoch": 0.38192977160680147,
+      "grad_norm": 0.16159814389069738,
+      "learning_rate": 2e-05,
+      "loss": 5.4623,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3819968474360264,
+      "grad_norm": 0.16346474939585,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 5695
+    },
+    {
+      "epoch": 0.38206392326525135,
+      "grad_norm": 0.1456984278186836,
+      "learning_rate": 2e-05,
+      "loss": 5.3662,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3821309990944763,
+      "grad_norm": 0.16095060096330588,
+      "learning_rate": 2e-05,
+      "loss": 5.6356,
+      "step": 5697
+    },
+    {
+      "epoch": 0.3821980749237012,
+      "grad_norm": 0.15258029562712908,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 5698
+    },
+    {
+      "epoch": 0.38226515075292616,
+      "grad_norm": 0.14498546407750915,
+      "learning_rate": 2e-05,
+      "loss": 5.4709,
+      "step": 5699
+    },
+    {
+      "epoch": 0.3823322265821511,
+      "grad_norm": 0.142385901424097,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 5700
+    },
+    {
+      "epoch": 0.38239930241137604,
+      "grad_norm": 0.1517974455756827,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 5701
+    },
+    {
+      "epoch": 0.382466378240601,
+      "grad_norm": 0.1509436184821043,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 5702
+    },
+    {
+      "epoch": 0.3825334540698259,
+      "grad_norm": 0.15461911059893751,
+      "learning_rate": 2e-05,
+      "loss": 5.4528,
+      "step": 5703
+    },
+    {
+      "epoch": 0.38260052989905086,
+      "grad_norm": 0.1571875701235248,
+      "learning_rate": 2e-05,
+      "loss": 5.3931,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3826676057282758,
+      "grad_norm": 0.1586439368063672,
+      "learning_rate": 2e-05,
+      "loss": 5.336,
+      "step": 5705
+    },
+    {
+      "epoch": 0.38273468155750073,
+      "grad_norm": 0.1542788368779236,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5706
+    },
+    {
+      "epoch": 0.38280175738672567,
+      "grad_norm": 0.1711005333485567,
+      "learning_rate": 2e-05,
+      "loss": 5.4486,
+      "step": 5707
+    },
+    {
+      "epoch": 0.3828688332159506,
+      "grad_norm": 0.15487664183397837,
+      "learning_rate": 2e-05,
+      "loss": 5.4729,
+      "step": 5708
+    },
+    {
+      "epoch": 0.38293590904517555,
+      "grad_norm": 0.15636687783978193,
+      "learning_rate": 2e-05,
+      "loss": 5.5162,
+      "step": 5709
+    },
+    {
+      "epoch": 0.3830029848744005,
+      "grad_norm": 0.1564644942005319,
+      "learning_rate": 2e-05,
+      "loss": 5.2824,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3830700607036254,
+      "grad_norm": 0.1551127600198856,
+      "learning_rate": 2e-05,
+      "loss": 5.4194,
+      "step": 5711
+    },
+    {
+      "epoch": 0.38313713653285036,
+      "grad_norm": 0.15365072501782398,
+      "learning_rate": 2e-05,
+      "loss": 5.3954,
+      "step": 5712
+    },
+    {
+      "epoch": 0.3832042123620753,
+      "grad_norm": 0.15454466869657202,
+      "learning_rate": 2e-05,
+      "loss": 5.5393,
+      "step": 5713
+    },
+    {
+      "epoch": 0.38327128819130024,
+      "grad_norm": 0.16584922791907644,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 5714
+    },
+    {
+      "epoch": 0.3833383640205252,
+      "grad_norm": 0.1510082370082647,
+      "learning_rate": 2e-05,
+      "loss": 5.4851,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3834054398497501,
+      "grad_norm": 0.14796427203702556,
+      "learning_rate": 2e-05,
+      "loss": 5.3059,
+      "step": 5716
+    },
+    {
+      "epoch": 0.38347251567897506,
+      "grad_norm": 0.16349384370009598,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 5717
+    },
+    {
+      "epoch": 0.3835395915082,
+      "grad_norm": 0.14752524127202854,
+      "learning_rate": 2e-05,
+      "loss": 5.543,
+      "step": 5718
+    },
+    {
+      "epoch": 0.38360666733742493,
+      "grad_norm": 0.15256596198128272,
+      "learning_rate": 2e-05,
+      "loss": 5.4807,
+      "step": 5719
+    },
+    {
+      "epoch": 0.38367374316664987,
+      "grad_norm": 0.15816991975309536,
+      "learning_rate": 2e-05,
+      "loss": 5.3645,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3837408189958748,
+      "grad_norm": 0.1411437855310054,
+      "learning_rate": 2e-05,
+      "loss": 5.4019,
+      "step": 5721
+    },
+    {
+      "epoch": 0.38380789482509975,
+      "grad_norm": 0.16252922503005524,
+      "learning_rate": 2e-05,
+      "loss": 5.4664,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3838749706543247,
+      "grad_norm": 0.15784878041507622,
+      "learning_rate": 2e-05,
+      "loss": 5.3616,
+      "step": 5723
+    },
+    {
+      "epoch": 0.3839420464835496,
+      "grad_norm": 0.1531011345695869,
+      "learning_rate": 2e-05,
+      "loss": 5.3384,
+      "step": 5724
+    },
+    {
+      "epoch": 0.38400912231277456,
+      "grad_norm": 0.14909718419093362,
+      "learning_rate": 2e-05,
+      "loss": 5.5092,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3840761981419995,
+      "grad_norm": 0.15896620395054176,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 5726
+    },
+    {
+      "epoch": 0.38414327397122444,
+      "grad_norm": 0.1563974475413855,
+      "learning_rate": 2e-05,
+      "loss": 5.5092,
+      "step": 5727
+    },
+    {
+      "epoch": 0.38421034980044944,
+      "grad_norm": 0.15720141549015046,
+      "learning_rate": 2e-05,
+      "loss": 5.5211,
+      "step": 5728
+    },
+    {
+      "epoch": 0.3842774256296744,
+      "grad_norm": 0.14847943955483478,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 5729
+    },
+    {
+      "epoch": 0.3843445014588993,
+      "grad_norm": 0.15235343021192485,
+      "learning_rate": 2e-05,
+      "loss": 5.2857,
+      "step": 5730
+    },
+    {
+      "epoch": 0.38441157728812425,
+      "grad_norm": 0.15042910767708836,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 5731
+    },
+    {
+      "epoch": 0.3844786531173492,
+      "grad_norm": 0.148202929430328,
+      "learning_rate": 2e-05,
+      "loss": 5.5483,
+      "step": 5732
+    },
+    {
+      "epoch": 0.38454572894657413,
+      "grad_norm": 0.14470823721825973,
+      "learning_rate": 2e-05,
+      "loss": 5.4887,
+      "step": 5733
+    },
+    {
+      "epoch": 0.38461280477579907,
+      "grad_norm": 0.15656849397332648,
+      "learning_rate": 2e-05,
+      "loss": 5.5195,
+      "step": 5734
+    },
+    {
+      "epoch": 0.384679880605024,
+      "grad_norm": 0.1467692187052694,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 5735
+    },
+    {
+      "epoch": 0.38474695643424894,
+      "grad_norm": 0.1473241433039828,
+      "learning_rate": 2e-05,
+      "loss": 5.374,
+      "step": 5736
+    },
+    {
+      "epoch": 0.3848140322634739,
+      "grad_norm": 0.1470127234729622,
+      "learning_rate": 2e-05,
+      "loss": 5.4792,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3848811080926988,
+      "grad_norm": 0.1550364845840921,
+      "learning_rate": 2e-05,
+      "loss": 5.569,
+      "step": 5738
+    },
+    {
+      "epoch": 0.38494818392192376,
+      "grad_norm": 0.14720413973134358,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3850152597511487,
+      "grad_norm": 0.14855027141189991,
+      "learning_rate": 2e-05,
+      "loss": 5.4781,
+      "step": 5740
+    },
+    {
+      "epoch": 0.38508233558037364,
+      "grad_norm": 0.15840216420667838,
+      "learning_rate": 2e-05,
+      "loss": 5.6191,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3851494114095986,
+      "grad_norm": 0.15391712522459483,
+      "learning_rate": 2e-05,
+      "loss": 5.4976,
+      "step": 5742
+    },
+    {
+      "epoch": 0.3852164872388235,
+      "grad_norm": 0.1420166719611395,
+      "learning_rate": 2e-05,
+      "loss": 5.1992,
+      "step": 5743
+    },
+    {
+      "epoch": 0.38528356306804845,
+      "grad_norm": 0.1496049195570291,
+      "learning_rate": 2e-05,
+      "loss": 5.4821,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3853506388972734,
+      "grad_norm": 0.1534824380901906,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 5745
+    },
+    {
+      "epoch": 0.38541771472649833,
+      "grad_norm": 0.1494685385901173,
+      "learning_rate": 2e-05,
+      "loss": 5.4107,
+      "step": 5746
+    },
+    {
+      "epoch": 0.38548479055572327,
+      "grad_norm": 0.15050772958602043,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 5747
+    },
+    {
+      "epoch": 0.3855518663849482,
+      "grad_norm": 0.1438705252563661,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 5748
+    },
+    {
+      "epoch": 0.38561894221417314,
+      "grad_norm": 0.1525319704529148,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3856860180433981,
+      "grad_norm": 0.15788115560852764,
+      "learning_rate": 2e-05,
+      "loss": 5.4025,
+      "step": 5750
+    },
+    {
+      "epoch": 0.385753093872623,
+      "grad_norm": 0.14117548226093063,
+      "learning_rate": 2e-05,
+      "loss": 5.3824,
+      "step": 5751
+    },
+    {
+      "epoch": 0.38582016970184796,
+      "grad_norm": 0.14697515097674868,
+      "learning_rate": 2e-05,
+      "loss": 5.4763,
+      "step": 5752
+    },
+    {
+      "epoch": 0.3858872455310729,
+      "grad_norm": 0.1438792041875919,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 5753
+    },
+    {
+      "epoch": 0.38595432136029784,
+      "grad_norm": 0.1487794927224431,
+      "learning_rate": 2e-05,
+      "loss": 5.4429,
+      "step": 5754
+    },
+    {
+      "epoch": 0.3860213971895228,
+      "grad_norm": 0.1448073346110761,
+      "learning_rate": 2e-05,
+      "loss": 5.5125,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3860884730187477,
+      "grad_norm": 0.1408077581019159,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 5756
+    },
+    {
+      "epoch": 0.38615554884797265,
+      "grad_norm": 0.14679152330572764,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 5757
+    },
+    {
+      "epoch": 0.3862226246771976,
+      "grad_norm": 0.15291326219680593,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 5758
+    },
+    {
+      "epoch": 0.38628970050642253,
+      "grad_norm": 0.1458944869191814,
+      "learning_rate": 2e-05,
+      "loss": 5.5791,
+      "step": 5759
+    },
+    {
+      "epoch": 0.38635677633564747,
+      "grad_norm": 0.15233042403233657,
+      "learning_rate": 2e-05,
+      "loss": 5.4575,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3864238521648724,
+      "grad_norm": 0.15606303618977047,
+      "learning_rate": 2e-05,
+      "loss": 5.2747,
+      "step": 5761
+    },
+    {
+      "epoch": 0.38649092799409734,
+      "grad_norm": 0.14921354907674902,
+      "learning_rate": 2e-05,
+      "loss": 5.5192,
+      "step": 5762
+    },
+    {
+      "epoch": 0.3865580038233223,
+      "grad_norm": 0.15098907198049402,
+      "learning_rate": 2e-05,
+      "loss": 5.3491,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3866250796525472,
+      "grad_norm": 0.17200953241576947,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 5764
+    },
+    {
+      "epoch": 0.38669215548177216,
+      "grad_norm": 0.14997870724062318,
+      "learning_rate": 2e-05,
+      "loss": 5.3868,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3867592313109971,
+      "grad_norm": 0.1627316920003786,
+      "learning_rate": 2e-05,
+      "loss": 5.3421,
+      "step": 5766
+    },
+    {
+      "epoch": 0.38682630714022204,
+      "grad_norm": 0.15829288915484274,
+      "learning_rate": 2e-05,
+      "loss": 5.3309,
+      "step": 5767
+    },
+    {
+      "epoch": 0.386893382969447,
+      "grad_norm": 0.15506305779774382,
+      "learning_rate": 2e-05,
+      "loss": 5.3094,
+      "step": 5768
+    },
+    {
+      "epoch": 0.3869604587986719,
+      "grad_norm": 0.15238213721438468,
+      "learning_rate": 2e-05,
+      "loss": 5.4083,
+      "step": 5769
+    },
+    {
+      "epoch": 0.38702753462789685,
+      "grad_norm": 0.1472984929465323,
+      "learning_rate": 2e-05,
+      "loss": 5.4418,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3870946104571218,
+      "grad_norm": 0.1645741855048162,
+      "learning_rate": 2e-05,
+      "loss": 5.3464,
+      "step": 5771
+    },
+    {
+      "epoch": 0.38716168628634673,
+      "grad_norm": 0.16197113427904658,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 5772
+    },
+    {
+      "epoch": 0.38722876211557167,
+      "grad_norm": 0.16307425069815754,
+      "learning_rate": 2e-05,
+      "loss": 5.3956,
+      "step": 5773
+    },
+    {
+      "epoch": 0.3872958379447966,
+      "grad_norm": 0.14666860758345238,
+      "learning_rate": 2e-05,
+      "loss": 5.4078,
+      "step": 5774
+    },
+    {
+      "epoch": 0.38736291377402154,
+      "grad_norm": 0.1559956699809593,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3874299896032465,
+      "grad_norm": 0.1505131884595001,
+      "learning_rate": 2e-05,
+      "loss": 5.5274,
+      "step": 5776
+    },
+    {
+      "epoch": 0.3874970654324714,
+      "grad_norm": 0.1608953465500763,
+      "learning_rate": 2e-05,
+      "loss": 5.4966,
+      "step": 5777
+    },
+    {
+      "epoch": 0.38756414126169636,
+      "grad_norm": 0.16026293246168155,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 5778
+    },
+    {
+      "epoch": 0.3876312170909213,
+      "grad_norm": 0.1514043498248274,
+      "learning_rate": 2e-05,
+      "loss": 5.4426,
+      "step": 5779
+    },
+    {
+      "epoch": 0.38769829292014624,
+      "grad_norm": 0.15251188741441846,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3877653687493712,
+      "grad_norm": 0.15189685578113293,
+      "learning_rate": 2e-05,
+      "loss": 5.4664,
+      "step": 5781
+    },
+    {
+      "epoch": 0.3878324445785961,
+      "grad_norm": 0.14565848766337064,
+      "learning_rate": 2e-05,
+      "loss": 5.2978,
+      "step": 5782
+    },
+    {
+      "epoch": 0.38789952040782105,
+      "grad_norm": 0.1541200326792232,
+      "learning_rate": 2e-05,
+      "loss": 5.426,
+      "step": 5783
+    },
+    {
+      "epoch": 0.387966596237046,
+      "grad_norm": 0.1470771222969767,
+      "learning_rate": 2e-05,
+      "loss": 5.414,
+      "step": 5784
+    },
+    {
+      "epoch": 0.38803367206627093,
+      "grad_norm": 0.14215110031266326,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 5785
+    },
+    {
+      "epoch": 0.38810074789549587,
+      "grad_norm": 0.1467245333970438,
+      "learning_rate": 2e-05,
+      "loss": 5.3082,
+      "step": 5786
+    },
+    {
+      "epoch": 0.3881678237247208,
+      "grad_norm": 0.14995550201238386,
+      "learning_rate": 2e-05,
+      "loss": 5.4666,
+      "step": 5787
+    },
+    {
+      "epoch": 0.38823489955394574,
+      "grad_norm": 0.14821301929964303,
+      "learning_rate": 2e-05,
+      "loss": 5.4591,
+      "step": 5788
+    },
+    {
+      "epoch": 0.3883019753831707,
+      "grad_norm": 0.14506341417876575,
+      "learning_rate": 2e-05,
+      "loss": 5.4522,
+      "step": 5789
+    },
+    {
+      "epoch": 0.3883690512123956,
+      "grad_norm": 0.15474739733267345,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 5790
+    },
+    {
+      "epoch": 0.38843612704162056,
+      "grad_norm": 0.14848442985121812,
+      "learning_rate": 2e-05,
+      "loss": 5.5167,
+      "step": 5791
+    },
+    {
+      "epoch": 0.3885032028708455,
+      "grad_norm": 0.152070366352867,
+      "learning_rate": 2e-05,
+      "loss": 5.4312,
+      "step": 5792
+    },
+    {
+      "epoch": 0.38857027870007044,
+      "grad_norm": 0.14417811567412508,
+      "learning_rate": 2e-05,
+      "loss": 5.438,
+      "step": 5793
+    },
+    {
+      "epoch": 0.3886373545292954,
+      "grad_norm": 0.1511343063681683,
+      "learning_rate": 2e-05,
+      "loss": 5.4834,
+      "step": 5794
+    },
+    {
+      "epoch": 0.3887044303585203,
+      "grad_norm": 0.15159324006010275,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 5795
+    },
+    {
+      "epoch": 0.38877150618774525,
+      "grad_norm": 0.14816401371673357,
+      "learning_rate": 2e-05,
+      "loss": 5.3282,
+      "step": 5796
+    },
+    {
+      "epoch": 0.3888385820169702,
+      "grad_norm": 0.14531865875849878,
+      "learning_rate": 2e-05,
+      "loss": 5.4478,
+      "step": 5797
+    },
+    {
+      "epoch": 0.38890565784619513,
+      "grad_norm": 0.15292925069954352,
+      "learning_rate": 2e-05,
+      "loss": 5.3936,
+      "step": 5798
+    },
+    {
+      "epoch": 0.38897273367542007,
+      "grad_norm": 0.14952620997963448,
+      "learning_rate": 2e-05,
+      "loss": 5.2906,
+      "step": 5799
+    },
+    {
+      "epoch": 0.389039809504645,
+      "grad_norm": 0.14727831874981087,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 5800
+    },
+    {
+      "epoch": 0.38910688533386995,
+      "grad_norm": 0.15525239130754134,
+      "learning_rate": 2e-05,
+      "loss": 5.4701,
+      "step": 5801
+    },
+    {
+      "epoch": 0.3891739611630949,
+      "grad_norm": 0.14341059362473732,
+      "learning_rate": 2e-05,
+      "loss": 5.6545,
+      "step": 5802
+    },
+    {
+      "epoch": 0.3892410369923198,
+      "grad_norm": 0.15048413618361994,
+      "learning_rate": 2e-05,
+      "loss": 5.4475,
+      "step": 5803
+    },
+    {
+      "epoch": 0.38930811282154476,
+      "grad_norm": 0.14394208697221159,
+      "learning_rate": 2e-05,
+      "loss": 5.4728,
+      "step": 5804
+    },
+    {
+      "epoch": 0.3893751886507697,
+      "grad_norm": 0.15132828534446185,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 5805
+    },
+    {
+      "epoch": 0.38944226447999464,
+      "grad_norm": 0.14587403127016033,
+      "learning_rate": 2e-05,
+      "loss": 5.486,
+      "step": 5806
+    },
+    {
+      "epoch": 0.3895093403092196,
+      "grad_norm": 0.14911332405043554,
+      "learning_rate": 2e-05,
+      "loss": 5.4372,
+      "step": 5807
+    },
+    {
+      "epoch": 0.3895764161384445,
+      "grad_norm": 0.14405820374580666,
+      "learning_rate": 2e-05,
+      "loss": 5.3667,
+      "step": 5808
+    },
+    {
+      "epoch": 0.38964349196766945,
+      "grad_norm": 0.14570594554687002,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 5809
+    },
+    {
+      "epoch": 0.3897105677968944,
+      "grad_norm": 0.14082092912148184,
+      "learning_rate": 2e-05,
+      "loss": 5.6008,
+      "step": 5810
+    },
+    {
+      "epoch": 0.38977764362611933,
+      "grad_norm": 0.1536444396204144,
+      "learning_rate": 2e-05,
+      "loss": 5.514,
+      "step": 5811
+    },
+    {
+      "epoch": 0.38984471945534427,
+      "grad_norm": 0.14320647192058983,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 5812
+    },
+    {
+      "epoch": 0.3899117952845692,
+      "grad_norm": 0.14633644126544296,
+      "learning_rate": 2e-05,
+      "loss": 5.4566,
+      "step": 5813
+    },
+    {
+      "epoch": 0.38997887111379415,
+      "grad_norm": 0.14607862508416297,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 5814
+    },
+    {
+      "epoch": 0.3900459469430191,
+      "grad_norm": 0.1543896778461844,
+      "learning_rate": 2e-05,
+      "loss": 5.385,
+      "step": 5815
+    },
+    {
+      "epoch": 0.390113022772244,
+      "grad_norm": 0.14770766013001516,
+      "learning_rate": 2e-05,
+      "loss": 5.2307,
+      "step": 5816
+    },
+    {
+      "epoch": 0.39018009860146896,
+      "grad_norm": 0.1465830124752459,
+      "learning_rate": 2e-05,
+      "loss": 5.3346,
+      "step": 5817
+    },
+    {
+      "epoch": 0.3902471744306939,
+      "grad_norm": 0.1467139878006663,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 5818
+    },
+    {
+      "epoch": 0.39031425025991884,
+      "grad_norm": 0.15251057024865108,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 5819
+    },
+    {
+      "epoch": 0.3903813260891438,
+      "grad_norm": 0.14980022891124575,
+      "learning_rate": 2e-05,
+      "loss": 5.504,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3904484019183687,
+      "grad_norm": 0.14578889858537006,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 5821
+    },
+    {
+      "epoch": 0.39051547774759365,
+      "grad_norm": 0.15746604341929926,
+      "learning_rate": 2e-05,
+      "loss": 5.4043,
+      "step": 5822
+    },
+    {
+      "epoch": 0.3905825535768186,
+      "grad_norm": 0.14763029429451663,
+      "learning_rate": 2e-05,
+      "loss": 5.4789,
+      "step": 5823
+    },
+    {
+      "epoch": 0.39064962940604353,
+      "grad_norm": 0.14453762727861416,
+      "learning_rate": 2e-05,
+      "loss": 5.4434,
+      "step": 5824
+    },
+    {
+      "epoch": 0.39071670523526847,
+      "grad_norm": 0.1482714117806355,
+      "learning_rate": 2e-05,
+      "loss": 5.3103,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3907837810644934,
+      "grad_norm": 0.14277396511915819,
+      "learning_rate": 2e-05,
+      "loss": 5.5002,
+      "step": 5826
+    },
+    {
+      "epoch": 0.39085085689371835,
+      "grad_norm": 0.14031335417959032,
+      "learning_rate": 2e-05,
+      "loss": 5.2838,
+      "step": 5827
+    },
+    {
+      "epoch": 0.3909179327229433,
+      "grad_norm": 0.15033184572344638,
+      "learning_rate": 2e-05,
+      "loss": 5.3506,
+      "step": 5828
+    },
+    {
+      "epoch": 0.3909850085521682,
+      "grad_norm": 0.14949829543893822,
+      "learning_rate": 2e-05,
+      "loss": 5.4066,
+      "step": 5829
+    },
+    {
+      "epoch": 0.39105208438139316,
+      "grad_norm": 0.14406053932317042,
+      "learning_rate": 2e-05,
+      "loss": 5.3856,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3911191602106181,
+      "grad_norm": 0.14901425167444243,
+      "learning_rate": 2e-05,
+      "loss": 5.5368,
+      "step": 5831
+    },
+    {
+      "epoch": 0.39118623603984304,
+      "grad_norm": 0.147227249338175,
+      "learning_rate": 2e-05,
+      "loss": 5.4054,
+      "step": 5832
+    },
+    {
+      "epoch": 0.391253311869068,
+      "grad_norm": 0.14394813686051822,
+      "learning_rate": 2e-05,
+      "loss": 5.2778,
+      "step": 5833
+    },
+    {
+      "epoch": 0.3913203876982929,
+      "grad_norm": 0.14337469739320138,
+      "learning_rate": 2e-05,
+      "loss": 5.5388,
+      "step": 5834
+    },
+    {
+      "epoch": 0.39138746352751785,
+      "grad_norm": 0.14108897499139236,
+      "learning_rate": 2e-05,
+      "loss": 5.525,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3914545393567428,
+      "grad_norm": 0.1505485702488886,
+      "learning_rate": 2e-05,
+      "loss": 5.5474,
+      "step": 5836
+    },
+    {
+      "epoch": 0.39152161518596773,
+      "grad_norm": 0.14798527418835022,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 5837
+    },
+    {
+      "epoch": 0.39158869101519267,
+      "grad_norm": 0.1473902958249018,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 5838
+    },
+    {
+      "epoch": 0.3916557668444176,
+      "grad_norm": 0.1466486655459091,
+      "learning_rate": 2e-05,
+      "loss": 5.5092,
+      "step": 5839
+    },
+    {
+      "epoch": 0.39172284267364255,
+      "grad_norm": 0.14960641071607422,
+      "learning_rate": 2e-05,
+      "loss": 5.4205,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3917899185028675,
+      "grad_norm": 0.14857934054427546,
+      "learning_rate": 2e-05,
+      "loss": 5.2852,
+      "step": 5841
+    },
+    {
+      "epoch": 0.3918569943320924,
+      "grad_norm": 0.15083213291660186,
+      "learning_rate": 2e-05,
+      "loss": 5.3686,
+      "step": 5842
+    },
+    {
+      "epoch": 0.39192407016131736,
+      "grad_norm": 0.14442706168103567,
+      "learning_rate": 2e-05,
+      "loss": 5.4237,
+      "step": 5843
+    },
+    {
+      "epoch": 0.3919911459905423,
+      "grad_norm": 0.1420124008956615,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 5844
+    },
+    {
+      "epoch": 0.39205822181976724,
+      "grad_norm": 0.15042761987372208,
+      "learning_rate": 2e-05,
+      "loss": 5.338,
+      "step": 5845
+    },
+    {
+      "epoch": 0.3921252976489922,
+      "grad_norm": 0.14532655998388777,
+      "learning_rate": 2e-05,
+      "loss": 5.5238,
+      "step": 5846
+    },
+    {
+      "epoch": 0.3921923734782171,
+      "grad_norm": 0.1506725701683815,
+      "learning_rate": 2e-05,
+      "loss": 5.5372,
+      "step": 5847
+    },
+    {
+      "epoch": 0.39225944930744205,
+      "grad_norm": 0.14950196532591545,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 5848
+    },
+    {
+      "epoch": 0.392326525136667,
+      "grad_norm": 0.1544108596795565,
+      "learning_rate": 2e-05,
+      "loss": 5.38,
+      "step": 5849
+    },
+    {
+      "epoch": 0.39239360096589193,
+      "grad_norm": 0.14869836247070767,
+      "learning_rate": 2e-05,
+      "loss": 5.5782,
+      "step": 5850
+    },
+    {
+      "epoch": 0.39246067679511687,
+      "grad_norm": 0.1529561095614913,
+      "learning_rate": 2e-05,
+      "loss": 5.468,
+      "step": 5851
+    },
+    {
+      "epoch": 0.3925277526243418,
+      "grad_norm": 0.15041447037843858,
+      "learning_rate": 2e-05,
+      "loss": 5.3211,
+      "step": 5852
+    },
+    {
+      "epoch": 0.39259482845356675,
+      "grad_norm": 0.15204153751880556,
+      "learning_rate": 2e-05,
+      "loss": 5.4607,
+      "step": 5853
+    },
+    {
+      "epoch": 0.3926619042827917,
+      "grad_norm": 0.15440588993360396,
+      "learning_rate": 2e-05,
+      "loss": 5.4773,
+      "step": 5854
+    },
+    {
+      "epoch": 0.3927289801120166,
+      "grad_norm": 0.15605505745468676,
+      "learning_rate": 2e-05,
+      "loss": 5.4669,
+      "step": 5855
+    },
+    {
+      "epoch": 0.39279605594124156,
+      "grad_norm": 0.14216839607768803,
+      "learning_rate": 2e-05,
+      "loss": 5.4061,
+      "step": 5856
+    },
+    {
+      "epoch": 0.3928631317704665,
+      "grad_norm": 0.14545635650789174,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 5857
+    },
+    {
+      "epoch": 0.39293020759969144,
+      "grad_norm": 0.16329896297836766,
+      "learning_rate": 2e-05,
+      "loss": 5.6048,
+      "step": 5858
+    },
+    {
+      "epoch": 0.3929972834289164,
+      "grad_norm": 0.15041615828376692,
+      "learning_rate": 2e-05,
+      "loss": 5.4713,
+      "step": 5859
+    },
+    {
+      "epoch": 0.3930643592581413,
+      "grad_norm": 0.14195131104651246,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 5860
+    },
+    {
+      "epoch": 0.39313143508736625,
+      "grad_norm": 0.15789515463820297,
+      "learning_rate": 2e-05,
+      "loss": 5.3867,
+      "step": 5861
+    },
+    {
+      "epoch": 0.3931985109165912,
+      "grad_norm": 0.14681535617021604,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 5862
+    },
+    {
+      "epoch": 0.39326558674581613,
+      "grad_norm": 0.14581865040491213,
+      "learning_rate": 2e-05,
+      "loss": 5.5053,
+      "step": 5863
+    },
+    {
+      "epoch": 0.39333266257504107,
+      "grad_norm": 0.15761384304150605,
+      "learning_rate": 2e-05,
+      "loss": 5.2644,
+      "step": 5864
+    },
+    {
+      "epoch": 0.393399738404266,
+      "grad_norm": 0.1459885046920075,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 5865
+    },
+    {
+      "epoch": 0.39346681423349095,
+      "grad_norm": 0.15497000026868515,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 5866
+    },
+    {
+      "epoch": 0.3935338900627159,
+      "grad_norm": 0.15118018149901188,
+      "learning_rate": 2e-05,
+      "loss": 5.3361,
+      "step": 5867
+    },
+    {
+      "epoch": 0.3936009658919408,
+      "grad_norm": 0.151844572609273,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 5868
+    },
+    {
+      "epoch": 0.39366804172116576,
+      "grad_norm": 0.1572812820599554,
+      "learning_rate": 2e-05,
+      "loss": 5.4731,
+      "step": 5869
+    },
+    {
+      "epoch": 0.3937351175503907,
+      "grad_norm": 0.1470762124958615,
+      "learning_rate": 2e-05,
+      "loss": 5.3743,
+      "step": 5870
+    },
+    {
+      "epoch": 0.39380219337961564,
+      "grad_norm": 0.15127325905130617,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 5871
+    },
+    {
+      "epoch": 0.3938692692088406,
+      "grad_norm": 0.15123197393215188,
+      "learning_rate": 2e-05,
+      "loss": 5.524,
+      "step": 5872
+    },
+    {
+      "epoch": 0.3939363450380655,
+      "grad_norm": 0.1539080030887222,
+      "learning_rate": 2e-05,
+      "loss": 5.3228,
+      "step": 5873
+    },
+    {
+      "epoch": 0.39400342086729045,
+      "grad_norm": 0.15331816121825945,
+      "learning_rate": 2e-05,
+      "loss": 5.5709,
+      "step": 5874
+    },
+    {
+      "epoch": 0.3940704966965154,
+      "grad_norm": 0.15446404531313307,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 5875
+    },
+    {
+      "epoch": 0.39413757252574033,
+      "grad_norm": 0.14911869554800308,
+      "learning_rate": 2e-05,
+      "loss": 5.3263,
+      "step": 5876
+    },
+    {
+      "epoch": 0.39420464835496527,
+      "grad_norm": 0.14649057994646156,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 5877
+    },
+    {
+      "epoch": 0.3942717241841902,
+      "grad_norm": 0.15676194385574097,
+      "learning_rate": 2e-05,
+      "loss": 5.4417,
+      "step": 5878
+    },
+    {
+      "epoch": 0.39433880001341515,
+      "grad_norm": 0.15387481551009605,
+      "learning_rate": 2e-05,
+      "loss": 5.2368,
+      "step": 5879
+    },
+    {
+      "epoch": 0.3944058758426401,
+      "grad_norm": 0.14892873339514628,
+      "learning_rate": 2e-05,
+      "loss": 5.3044,
+      "step": 5880
+    },
+    {
+      "epoch": 0.394472951671865,
+      "grad_norm": 0.1469748621388426,
+      "learning_rate": 2e-05,
+      "loss": 5.3817,
+      "step": 5881
+    },
+    {
+      "epoch": 0.39454002750108996,
+      "grad_norm": 0.1584177163221102,
+      "learning_rate": 2e-05,
+      "loss": 5.4134,
+      "step": 5882
+    },
+    {
+      "epoch": 0.3946071033303149,
+      "grad_norm": 0.15940586393223796,
+      "learning_rate": 2e-05,
+      "loss": 5.3838,
+      "step": 5883
+    },
+    {
+      "epoch": 0.39467417915953984,
+      "grad_norm": 0.14695627370702188,
+      "learning_rate": 2e-05,
+      "loss": 5.5238,
+      "step": 5884
+    },
+    {
+      "epoch": 0.3947412549887648,
+      "grad_norm": 0.15578020267473816,
+      "learning_rate": 2e-05,
+      "loss": 5.4212,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3948083308179897,
+      "grad_norm": 0.15271411276089863,
+      "learning_rate": 2e-05,
+      "loss": 5.454,
+      "step": 5886
+    },
+    {
+      "epoch": 0.39487540664721466,
+      "grad_norm": 0.14702760504436627,
+      "learning_rate": 2e-05,
+      "loss": 5.4571,
+      "step": 5887
+    },
+    {
+      "epoch": 0.3949424824764396,
+      "grad_norm": 0.15330898358937073,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 5888
+    },
+    {
+      "epoch": 0.39500955830566453,
+      "grad_norm": 0.16088087387838515,
+      "learning_rate": 2e-05,
+      "loss": 5.4522,
+      "step": 5889
+    },
+    {
+      "epoch": 0.39507663413488947,
+      "grad_norm": 0.1472815247733883,
+      "learning_rate": 2e-05,
+      "loss": 5.5046,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3951437099641144,
+      "grad_norm": 0.14426069406292338,
+      "learning_rate": 2e-05,
+      "loss": 5.2891,
+      "step": 5891
+    },
+    {
+      "epoch": 0.39521078579333935,
+      "grad_norm": 0.16311130732212054,
+      "learning_rate": 2e-05,
+      "loss": 5.489,
+      "step": 5892
+    },
+    {
+      "epoch": 0.3952778616225643,
+      "grad_norm": 0.154712861052737,
+      "learning_rate": 2e-05,
+      "loss": 5.4121,
+      "step": 5893
+    },
+    {
+      "epoch": 0.3953449374517892,
+      "grad_norm": 0.15025477551754537,
+      "learning_rate": 2e-05,
+      "loss": 5.4913,
+      "step": 5894
+    },
+    {
+      "epoch": 0.39541201328101416,
+      "grad_norm": 0.14803051416651397,
+      "learning_rate": 2e-05,
+      "loss": 5.2498,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3954790891102391,
+      "grad_norm": 0.16080372268882145,
+      "learning_rate": 2e-05,
+      "loss": 5.4435,
+      "step": 5896
+    },
+    {
+      "epoch": 0.39554616493946404,
+      "grad_norm": 0.15706826761860182,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 5897
+    },
+    {
+      "epoch": 0.395613240768689,
+      "grad_norm": 0.16270738593117287,
+      "learning_rate": 2e-05,
+      "loss": 5.5277,
+      "step": 5898
+    },
+    {
+      "epoch": 0.3956803165979139,
+      "grad_norm": 0.14919863538352876,
+      "learning_rate": 2e-05,
+      "loss": 5.3161,
+      "step": 5899
+    },
+    {
+      "epoch": 0.39574739242713886,
+      "grad_norm": 0.15746024032806694,
+      "learning_rate": 2e-05,
+      "loss": 5.3774,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3958144682563638,
+      "grad_norm": 0.15285489980845465,
+      "learning_rate": 2e-05,
+      "loss": 5.4144,
+      "step": 5901
+    },
+    {
+      "epoch": 0.39588154408558873,
+      "grad_norm": 0.15249777193890982,
+      "learning_rate": 2e-05,
+      "loss": 5.3758,
+      "step": 5902
+    },
+    {
+      "epoch": 0.39594861991481367,
+      "grad_norm": 0.15013348964677142,
+      "learning_rate": 2e-05,
+      "loss": 5.4919,
+      "step": 5903
+    },
+    {
+      "epoch": 0.3960156957440386,
+      "grad_norm": 0.14916448559579706,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 5904
+    },
+    {
+      "epoch": 0.39608277157326355,
+      "grad_norm": 0.14584303489138212,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 5905
+    },
+    {
+      "epoch": 0.3961498474024885,
+      "grad_norm": 0.14647074535356697,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 5906
+    },
+    {
+      "epoch": 0.3962169232317134,
+      "grad_norm": 0.15266346090206342,
+      "learning_rate": 2e-05,
+      "loss": 5.5534,
+      "step": 5907
+    },
+    {
+      "epoch": 0.39628399906093836,
+      "grad_norm": 0.1488024725750587,
+      "learning_rate": 2e-05,
+      "loss": 5.5299,
+      "step": 5908
+    },
+    {
+      "epoch": 0.39635107489016336,
+      "grad_norm": 0.14912862642532618,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3964181507193883,
+      "grad_norm": 0.15443940992326785,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 5910
+    },
+    {
+      "epoch": 0.39648522654861323,
+      "grad_norm": 0.14633777271339263,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 5911
+    },
+    {
+      "epoch": 0.3965523023778382,
+      "grad_norm": 0.15088324330558475,
+      "learning_rate": 2e-05,
+      "loss": 5.4363,
+      "step": 5912
+    },
+    {
+      "epoch": 0.3966193782070631,
+      "grad_norm": 0.15398689821084152,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 5913
+    },
+    {
+      "epoch": 0.39668645403628805,
+      "grad_norm": 0.15037335404171037,
+      "learning_rate": 2e-05,
+      "loss": 5.4467,
+      "step": 5914
+    },
+    {
+      "epoch": 0.396753529865513,
+      "grad_norm": 0.15470222329142982,
+      "learning_rate": 2e-05,
+      "loss": 5.4111,
+      "step": 5915
+    },
+    {
+      "epoch": 0.3968206056947379,
+      "grad_norm": 0.1534803105244773,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 5916
+    },
+    {
+      "epoch": 0.39688768152396287,
+      "grad_norm": 0.15381044530912338,
+      "learning_rate": 2e-05,
+      "loss": 5.3416,
+      "step": 5917
+    },
+    {
+      "epoch": 0.3969547573531878,
+      "grad_norm": 0.1614030520270631,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 5918
+    },
+    {
+      "epoch": 0.39702183318241274,
+      "grad_norm": 0.1611834406810516,
+      "learning_rate": 2e-05,
+      "loss": 5.5282,
+      "step": 5919
+    },
+    {
+      "epoch": 0.3970889090116377,
+      "grad_norm": 0.1675946964905524,
+      "learning_rate": 2e-05,
+      "loss": 5.5445,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3971559848408626,
+      "grad_norm": 0.1527152525581282,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 5921
+    },
+    {
+      "epoch": 0.39722306067008756,
+      "grad_norm": 0.1524762159041056,
+      "learning_rate": 2e-05,
+      "loss": 5.519,
+      "step": 5922
+    },
+    {
+      "epoch": 0.3972901364993125,
+      "grad_norm": 0.1631318010120498,
+      "learning_rate": 2e-05,
+      "loss": 5.4225,
+      "step": 5923
+    },
+    {
+      "epoch": 0.39735721232853743,
+      "grad_norm": 0.1467361516176354,
+      "learning_rate": 2e-05,
+      "loss": 5.5318,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3974242881577624,
+      "grad_norm": 0.15252256919287552,
+      "learning_rate": 2e-05,
+      "loss": 5.5398,
+      "step": 5925
+    },
+    {
+      "epoch": 0.3974913639869873,
+      "grad_norm": 0.14448913482194106,
+      "learning_rate": 2e-05,
+      "loss": 5.3063,
+      "step": 5926
+    },
+    {
+      "epoch": 0.39755843981621225,
+      "grad_norm": 0.14529512100080058,
+      "learning_rate": 2e-05,
+      "loss": 5.581,
+      "step": 5927
+    },
+    {
+      "epoch": 0.3976255156454372,
+      "grad_norm": 0.15256322503581018,
+      "learning_rate": 2e-05,
+      "loss": 5.5163,
+      "step": 5928
+    },
+    {
+      "epoch": 0.3976925914746621,
+      "grad_norm": 0.14893159733161832,
+      "learning_rate": 2e-05,
+      "loss": 5.3897,
+      "step": 5929
+    },
+    {
+      "epoch": 0.39775966730388707,
+      "grad_norm": 0.15580480659438692,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 5930
+    },
+    {
+      "epoch": 0.397826743133112,
+      "grad_norm": 0.14780760967655887,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 5931
+    },
+    {
+      "epoch": 0.39789381896233694,
+      "grad_norm": 0.14510909445715506,
+      "learning_rate": 2e-05,
+      "loss": 5.347,
+      "step": 5932
+    },
+    {
+      "epoch": 0.3979608947915619,
+      "grad_norm": 0.14370781476981787,
+      "learning_rate": 2e-05,
+      "loss": 5.2568,
+      "step": 5933
+    },
+    {
+      "epoch": 0.3980279706207868,
+      "grad_norm": 0.14292931180412027,
+      "learning_rate": 2e-05,
+      "loss": 5.5042,
+      "step": 5934
+    },
+    {
+      "epoch": 0.39809504645001176,
+      "grad_norm": 0.15030992316036293,
+      "learning_rate": 2e-05,
+      "loss": 5.5282,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3981621222792367,
+      "grad_norm": 0.14887999047050238,
+      "learning_rate": 2e-05,
+      "loss": 5.3641,
+      "step": 5936
+    },
+    {
+      "epoch": 0.39822919810846164,
+      "grad_norm": 0.14644776344215313,
+      "learning_rate": 2e-05,
+      "loss": 5.5784,
+      "step": 5937
+    },
+    {
+      "epoch": 0.3982962739376866,
+      "grad_norm": 0.14610512744194495,
+      "learning_rate": 2e-05,
+      "loss": 5.4894,
+      "step": 5938
+    },
+    {
+      "epoch": 0.3983633497669115,
+      "grad_norm": 0.14738273914162306,
+      "learning_rate": 2e-05,
+      "loss": 5.3288,
+      "step": 5939
+    },
+    {
+      "epoch": 0.39843042559613645,
+      "grad_norm": 0.14322379972702287,
+      "learning_rate": 2e-05,
+      "loss": 5.4571,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3984975014253614,
+      "grad_norm": 0.14731376169643412,
+      "learning_rate": 2e-05,
+      "loss": 5.3314,
+      "step": 5941
+    },
+    {
+      "epoch": 0.3985645772545863,
+      "grad_norm": 0.15307151395920085,
+      "learning_rate": 2e-05,
+      "loss": 5.5605,
+      "step": 5942
+    },
+    {
+      "epoch": 0.39863165308381127,
+      "grad_norm": 0.1470009280172678,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 5943
+    },
+    {
+      "epoch": 0.3986987289130362,
+      "grad_norm": 0.14758255697515557,
+      "learning_rate": 2e-05,
+      "loss": 5.3098,
+      "step": 5944
+    },
+    {
+      "epoch": 0.39876580474226114,
+      "grad_norm": 0.15022116546064915,
+      "learning_rate": 2e-05,
+      "loss": 5.5179,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3988328805714861,
+      "grad_norm": 0.1457235433148265,
+      "learning_rate": 2e-05,
+      "loss": 5.2951,
+      "step": 5946
+    },
+    {
+      "epoch": 0.398899956400711,
+      "grad_norm": 0.14833353760049367,
+      "learning_rate": 2e-05,
+      "loss": 5.4332,
+      "step": 5947
+    },
+    {
+      "epoch": 0.39896703222993596,
+      "grad_norm": 0.149166510993386,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 5948
+    },
+    {
+      "epoch": 0.3990341080591609,
+      "grad_norm": 0.15680672811473656,
+      "learning_rate": 2e-05,
+      "loss": 5.5172,
+      "step": 5949
+    },
+    {
+      "epoch": 0.39910118388838584,
+      "grad_norm": 0.1450015466693006,
+      "learning_rate": 2e-05,
+      "loss": 5.5007,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3991682597176108,
+      "grad_norm": 0.14599289201203672,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 5951
+    },
+    {
+      "epoch": 0.3992353355468357,
+      "grad_norm": 0.15817724896534816,
+      "learning_rate": 2e-05,
+      "loss": 5.5397,
+      "step": 5952
+    },
+    {
+      "epoch": 0.39930241137606065,
+      "grad_norm": 0.14423517782586429,
+      "learning_rate": 2e-05,
+      "loss": 5.3355,
+      "step": 5953
+    },
+    {
+      "epoch": 0.3993694872052856,
+      "grad_norm": 0.15015891395980277,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 5954
+    },
+    {
+      "epoch": 0.39943656303451053,
+      "grad_norm": 0.15910903161894122,
+      "learning_rate": 2e-05,
+      "loss": 5.4983,
+      "step": 5955
+    },
+    {
+      "epoch": 0.39950363886373547,
+      "grad_norm": 0.1446941819929459,
+      "learning_rate": 2e-05,
+      "loss": 5.3558,
+      "step": 5956
+    },
+    {
+      "epoch": 0.3995707146929604,
+      "grad_norm": 0.14999820687559004,
+      "learning_rate": 2e-05,
+      "loss": 5.544,
+      "step": 5957
+    },
+    {
+      "epoch": 0.39963779052218534,
+      "grad_norm": 0.1549669433363154,
+      "learning_rate": 2e-05,
+      "loss": 5.3761,
+      "step": 5958
+    },
+    {
+      "epoch": 0.3997048663514103,
+      "grad_norm": 0.14458153614935193,
+      "learning_rate": 2e-05,
+      "loss": 5.4589,
+      "step": 5959
+    },
+    {
+      "epoch": 0.3997719421806352,
+      "grad_norm": 0.14452673981854028,
+      "learning_rate": 2e-05,
+      "loss": 5.355,
+      "step": 5960
+    },
+    {
+      "epoch": 0.39983901800986016,
+      "grad_norm": 0.15211697981582967,
+      "learning_rate": 2e-05,
+      "loss": 5.3976,
+      "step": 5961
+    },
+    {
+      "epoch": 0.3999060938390851,
+      "grad_norm": 0.15274857465846034,
+      "learning_rate": 2e-05,
+      "loss": 5.3967,
+      "step": 5962
+    },
+    {
+      "epoch": 0.39997316966831004,
+      "grad_norm": 0.15072527679708161,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 5963
+    },
+    {
+      "epoch": 0.400040245497535,
+      "grad_norm": 0.15427341397887093,
+      "learning_rate": 2e-05,
+      "loss": 5.5552,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4001073213267599,
+      "grad_norm": 0.15524144569325887,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 5965
+    },
+    {
+      "epoch": 0.40017439715598485,
+      "grad_norm": 0.15974259740772193,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4002414729852098,
+      "grad_norm": 0.15414255761401013,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 5967
+    },
+    {
+      "epoch": 0.40030854881443473,
+      "grad_norm": 0.16015912465443163,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 5968
+    },
+    {
+      "epoch": 0.40037562464365967,
+      "grad_norm": 0.1464117026057262,
+      "learning_rate": 2e-05,
+      "loss": 5.5139,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4004427004728846,
+      "grad_norm": 0.1457975488459741,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 5970
+    },
+    {
+      "epoch": 0.40050977630210954,
+      "grad_norm": 0.14849575474811771,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4005768521313345,
+      "grad_norm": 0.15138786328699524,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4006439279605594,
+      "grad_norm": 0.15515794784077275,
+      "learning_rate": 2e-05,
+      "loss": 5.3223,
+      "step": 5973
+    },
+    {
+      "epoch": 0.40071100378978436,
+      "grad_norm": 0.14931956508386474,
+      "learning_rate": 2e-05,
+      "loss": 5.319,
+      "step": 5974
+    },
+    {
+      "epoch": 0.4007780796190093,
+      "grad_norm": 0.1523360570959072,
+      "learning_rate": 2e-05,
+      "loss": 5.3647,
+      "step": 5975
+    },
+    {
+      "epoch": 0.40084515544823424,
+      "grad_norm": 0.15030047581861997,
+      "learning_rate": 2e-05,
+      "loss": 5.3425,
+      "step": 5976
+    },
+    {
+      "epoch": 0.4009122312774592,
+      "grad_norm": 0.1566461279408005,
+      "learning_rate": 2e-05,
+      "loss": 5.6564,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4009793071066841,
+      "grad_norm": 0.16793460818724498,
+      "learning_rate": 2e-05,
+      "loss": 5.5158,
+      "step": 5978
+    },
+    {
+      "epoch": 0.40104638293590905,
+      "grad_norm": 0.1480784151958493,
+      "learning_rate": 2e-05,
+      "loss": 5.3516,
+      "step": 5979
+    },
+    {
+      "epoch": 0.401113458765134,
+      "grad_norm": 0.15780079254317383,
+      "learning_rate": 2e-05,
+      "loss": 5.6116,
+      "step": 5980
+    },
+    {
+      "epoch": 0.40118053459435893,
+      "grad_norm": 0.15974484794590924,
+      "learning_rate": 2e-05,
+      "loss": 5.5555,
+      "step": 5981
+    },
+    {
+      "epoch": 0.40124761042358387,
+      "grad_norm": 0.15533366773400992,
+      "learning_rate": 2e-05,
+      "loss": 5.5404,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4013146862528088,
+      "grad_norm": 0.14960791855137717,
+      "learning_rate": 2e-05,
+      "loss": 5.3608,
+      "step": 5983
+    },
+    {
+      "epoch": 0.40138176208203374,
+      "grad_norm": 0.15606271060187718,
+      "learning_rate": 2e-05,
+      "loss": 5.4299,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4014488379112587,
+      "grad_norm": 0.15213952142674927,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4015159137404836,
+      "grad_norm": 0.15492798596919966,
+      "learning_rate": 2e-05,
+      "loss": 5.4458,
+      "step": 5986
+    },
+    {
+      "epoch": 0.40158298956970856,
+      "grad_norm": 0.1502859466266917,
+      "learning_rate": 2e-05,
+      "loss": 5.305,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4016500653989335,
+      "grad_norm": 0.15053162582835142,
+      "learning_rate": 2e-05,
+      "loss": 5.472,
+      "step": 5988
+    },
+    {
+      "epoch": 0.40171714122815844,
+      "grad_norm": 0.1560287783529022,
+      "learning_rate": 2e-05,
+      "loss": 5.4753,
+      "step": 5989
+    },
+    {
+      "epoch": 0.4017842170573834,
+      "grad_norm": 0.14237230547652055,
+      "learning_rate": 2e-05,
+      "loss": 5.2587,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4018512928866083,
+      "grad_norm": 0.15172545088934214,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 5991
+    },
+    {
+      "epoch": 0.40191836871583325,
+      "grad_norm": 0.16324104770353742,
+      "learning_rate": 2e-05,
+      "loss": 5.644,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4019854445450582,
+      "grad_norm": 0.15443232929486764,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 5993
+    },
+    {
+      "epoch": 0.40205252037428313,
+      "grad_norm": 0.14998236761273226,
+      "learning_rate": 2e-05,
+      "loss": 5.3467,
+      "step": 5994
+    },
+    {
+      "epoch": 0.40211959620350807,
+      "grad_norm": 0.15949096548765676,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 5995
+    },
+    {
+      "epoch": 0.402186672032733,
+      "grad_norm": 0.15393847386347603,
+      "learning_rate": 2e-05,
+      "loss": 5.4736,
+      "step": 5996
+    },
+    {
+      "epoch": 0.40225374786195794,
+      "grad_norm": 0.14873881419982177,
+      "learning_rate": 2e-05,
+      "loss": 5.536,
+      "step": 5997
+    },
+    {
+      "epoch": 0.4023208236911829,
+      "grad_norm": 0.15832075068230408,
+      "learning_rate": 2e-05,
+      "loss": 5.5771,
+      "step": 5998
+    },
+    {
+      "epoch": 0.4023878995204078,
+      "grad_norm": 0.1651991586554004,
+      "learning_rate": 2e-05,
+      "loss": 5.2404,
+      "step": 5999
+    },
+    {
+      "epoch": 0.40245497534963276,
+      "grad_norm": 0.15260305276965905,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4025220511788577,
+      "grad_norm": 0.15845256948084602,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 6001
+    },
+    {
+      "epoch": 0.40258912700808264,
+      "grad_norm": 0.1533038381959413,
+      "learning_rate": 2e-05,
+      "loss": 5.5904,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4026562028373076,
+      "grad_norm": 0.1476470063009068,
+      "learning_rate": 2e-05,
+      "loss": 5.3976,
+      "step": 6003
+    },
+    {
+      "epoch": 0.4027232786665325,
+      "grad_norm": 0.1524396060890464,
+      "learning_rate": 2e-05,
+      "loss": 5.3979,
+      "step": 6004
+    },
+    {
+      "epoch": 0.40279035449575745,
+      "grad_norm": 0.1698079006375316,
+      "learning_rate": 2e-05,
+      "loss": 5.5146,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4028574303249824,
+      "grad_norm": 0.15222545116242248,
+      "learning_rate": 2e-05,
+      "loss": 5.3234,
+      "step": 6006
+    },
+    {
+      "epoch": 0.40292450615420733,
+      "grad_norm": 0.15095382845030694,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 6007
+    },
+    {
+      "epoch": 0.40299158198343227,
+      "grad_norm": 0.16410280022962706,
+      "learning_rate": 2e-05,
+      "loss": 5.5087,
+      "step": 6008
+    },
+    {
+      "epoch": 0.4030586578126572,
+      "grad_norm": 0.15946479861421245,
+      "learning_rate": 2e-05,
+      "loss": 5.4589,
+      "step": 6009
+    },
+    {
+      "epoch": 0.40312573364188214,
+      "grad_norm": 0.1601739879462838,
+      "learning_rate": 2e-05,
+      "loss": 5.4183,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4031928094711071,
+      "grad_norm": 0.15039068286143933,
+      "learning_rate": 2e-05,
+      "loss": 5.5256,
+      "step": 6011
+    },
+    {
+      "epoch": 0.403259885300332,
+      "grad_norm": 0.15624060887507452,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 6012
+    },
+    {
+      "epoch": 0.40332696112955696,
+      "grad_norm": 0.1539690397156179,
+      "learning_rate": 2e-05,
+      "loss": 5.4258,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4033940369587819,
+      "grad_norm": 0.1515976927373858,
+      "learning_rate": 2e-05,
+      "loss": 5.4026,
+      "step": 6014
+    },
+    {
+      "epoch": 0.40346111278800684,
+      "grad_norm": 0.15704954616036437,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4035281886172318,
+      "grad_norm": 0.16024113471010473,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4035952644464567,
+      "grad_norm": 0.15420382377101463,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 6017
+    },
+    {
+      "epoch": 0.40366234027568165,
+      "grad_norm": 0.15664716558324795,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 6018
+    },
+    {
+      "epoch": 0.4037294161049066,
+      "grad_norm": 0.15180064671749896,
+      "learning_rate": 2e-05,
+      "loss": 5.4129,
+      "step": 6019
+    },
+    {
+      "epoch": 0.40379649193413153,
+      "grad_norm": 0.1547786024366051,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 6020
+    },
+    {
+      "epoch": 0.40386356776335647,
+      "grad_norm": 0.1587669350642931,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 6021
+    },
+    {
+      "epoch": 0.4039306435925814,
+      "grad_norm": 0.15693967334666328,
+      "learning_rate": 2e-05,
+      "loss": 5.6413,
+      "step": 6022
+    },
+    {
+      "epoch": 0.40399771942180634,
+      "grad_norm": 0.16285352493060998,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 6023
+    },
+    {
+      "epoch": 0.4040647952510313,
+      "grad_norm": 0.15292817468719166,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4041318710802562,
+      "grad_norm": 0.14462371822268005,
+      "learning_rate": 2e-05,
+      "loss": 5.4783,
+      "step": 6025
+    },
+    {
+      "epoch": 0.40419894690948116,
+      "grad_norm": 0.14739639699681178,
+      "learning_rate": 2e-05,
+      "loss": 5.4944,
+      "step": 6026
+    },
+    {
+      "epoch": 0.4042660227387061,
+      "grad_norm": 0.16108022105434244,
+      "learning_rate": 2e-05,
+      "loss": 5.3953,
+      "step": 6027
+    },
+    {
+      "epoch": 0.40433309856793104,
+      "grad_norm": 0.1474509768870367,
+      "learning_rate": 2e-05,
+      "loss": 5.5286,
+      "step": 6028
+    },
+    {
+      "epoch": 0.404400174397156,
+      "grad_norm": 0.15079750264331543,
+      "learning_rate": 2e-05,
+      "loss": 5.4186,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4044672502263809,
+      "grad_norm": 0.1458719980847305,
+      "learning_rate": 2e-05,
+      "loss": 5.3563,
+      "step": 6030
+    },
+    {
+      "epoch": 0.40453432605560585,
+      "grad_norm": 0.15304975621717112,
+      "learning_rate": 2e-05,
+      "loss": 5.5925,
+      "step": 6031
+    },
+    {
+      "epoch": 0.4046014018848308,
+      "grad_norm": 0.1520346560029922,
+      "learning_rate": 2e-05,
+      "loss": 5.4501,
+      "step": 6032
+    },
+    {
+      "epoch": 0.40466847771405573,
+      "grad_norm": 0.15064836907560086,
+      "learning_rate": 2e-05,
+      "loss": 5.4288,
+      "step": 6033
+    },
+    {
+      "epoch": 0.40473555354328067,
+      "grad_norm": 0.14603597324586737,
+      "learning_rate": 2e-05,
+      "loss": 5.4439,
+      "step": 6034
+    },
+    {
+      "epoch": 0.4048026293725056,
+      "grad_norm": 0.15385398407709727,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 6035
+    },
+    {
+      "epoch": 0.40486970520173055,
+      "grad_norm": 0.14137539672449997,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 6036
+    },
+    {
+      "epoch": 0.4049367810309555,
+      "grad_norm": 0.14316446522327167,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4050038568601804,
+      "grad_norm": 0.14543957958051615,
+      "learning_rate": 2e-05,
+      "loss": 5.3809,
+      "step": 6038
+    },
+    {
+      "epoch": 0.40507093268940536,
+      "grad_norm": 0.14384966970502341,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 6039
+    },
+    {
+      "epoch": 0.4051380085186303,
+      "grad_norm": 0.14261973756452165,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 6040
+    },
+    {
+      "epoch": 0.40520508434785524,
+      "grad_norm": 0.14594225185179877,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 6041
+    },
+    {
+      "epoch": 0.4052721601770802,
+      "grad_norm": 0.14238136018724745,
+      "learning_rate": 2e-05,
+      "loss": 5.4951,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4053392360063051,
+      "grad_norm": 0.14718572165026994,
+      "learning_rate": 2e-05,
+      "loss": 5.4058,
+      "step": 6043
+    },
+    {
+      "epoch": 0.40540631183553005,
+      "grad_norm": 0.14363938442517427,
+      "learning_rate": 2e-05,
+      "loss": 5.3782,
+      "step": 6044
+    },
+    {
+      "epoch": 0.405473387664755,
+      "grad_norm": 0.1479353910642996,
+      "learning_rate": 2e-05,
+      "loss": 5.5865,
+      "step": 6045
+    },
+    {
+      "epoch": 0.40554046349397993,
+      "grad_norm": 0.14623340023390166,
+      "learning_rate": 2e-05,
+      "loss": 5.4653,
+      "step": 6046
+    },
+    {
+      "epoch": 0.40560753932320487,
+      "grad_norm": 0.14461855324385894,
+      "learning_rate": 2e-05,
+      "loss": 5.4816,
+      "step": 6047
+    },
+    {
+      "epoch": 0.4056746151524298,
+      "grad_norm": 0.1527639109903678,
+      "learning_rate": 2e-05,
+      "loss": 5.3776,
+      "step": 6048
+    },
+    {
+      "epoch": 0.40574169098165475,
+      "grad_norm": 0.15501073851554598,
+      "learning_rate": 2e-05,
+      "loss": 5.4025,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4058087668108797,
+      "grad_norm": 0.14969192504580944,
+      "learning_rate": 2e-05,
+      "loss": 5.3329,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4058758426401046,
+      "grad_norm": 0.14699245386539642,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 6051
+    },
+    {
+      "epoch": 0.40594291846932956,
+      "grad_norm": 0.15232413374637474,
+      "learning_rate": 2e-05,
+      "loss": 5.5125,
+      "step": 6052
+    },
+    {
+      "epoch": 0.4060099942985545,
+      "grad_norm": 0.15442690656464464,
+      "learning_rate": 2e-05,
+      "loss": 5.3806,
+      "step": 6053
+    },
+    {
+      "epoch": 0.40607707012777944,
+      "grad_norm": 0.16098324783117463,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4061441459570044,
+      "grad_norm": 0.1513771087140424,
+      "learning_rate": 2e-05,
+      "loss": 5.5473,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4062112217862293,
+      "grad_norm": 0.16258286685946519,
+      "learning_rate": 2e-05,
+      "loss": 5.4633,
+      "step": 6056
+    },
+    {
+      "epoch": 0.40627829761545425,
+      "grad_norm": 0.1457785852182703,
+      "learning_rate": 2e-05,
+      "loss": 5.3305,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4063453734446792,
+      "grad_norm": 0.1446502311723748,
+      "learning_rate": 2e-05,
+      "loss": 5.3584,
+      "step": 6058
+    },
+    {
+      "epoch": 0.40641244927390413,
+      "grad_norm": 0.15360605323060894,
+      "learning_rate": 2e-05,
+      "loss": 5.4209,
+      "step": 6059
+    },
+    {
+      "epoch": 0.40647952510312907,
+      "grad_norm": 0.15331275633886143,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 6060
+    },
+    {
+      "epoch": 0.406546600932354,
+      "grad_norm": 0.153156307319625,
+      "learning_rate": 2e-05,
+      "loss": 5.3788,
+      "step": 6061
+    },
+    {
+      "epoch": 0.40661367676157895,
+      "grad_norm": 0.15708553491669636,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4066807525908039,
+      "grad_norm": 0.1479300552202045,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 6063
+    },
+    {
+      "epoch": 0.4067478284200288,
+      "grad_norm": 0.15279132418782454,
+      "learning_rate": 2e-05,
+      "loss": 5.4119,
+      "step": 6064
+    },
+    {
+      "epoch": 0.40681490424925376,
+      "grad_norm": 0.14962557499541565,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4068819800784787,
+      "grad_norm": 0.1540267964870355,
+      "learning_rate": 2e-05,
+      "loss": 5.2877,
+      "step": 6066
+    },
+    {
+      "epoch": 0.40694905590770364,
+      "grad_norm": 0.15368758941890037,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 6067
+    },
+    {
+      "epoch": 0.4070161317369286,
+      "grad_norm": 0.15500344012336434,
+      "learning_rate": 2e-05,
+      "loss": 5.4851,
+      "step": 6068
+    },
+    {
+      "epoch": 0.4070832075661535,
+      "grad_norm": 0.14754607150670423,
+      "learning_rate": 2e-05,
+      "loss": 5.5639,
+      "step": 6069
+    },
+    {
+      "epoch": 0.40715028339537845,
+      "grad_norm": 0.15337883625537224,
+      "learning_rate": 2e-05,
+      "loss": 5.3378,
+      "step": 6070
+    },
+    {
+      "epoch": 0.4072173592246034,
+      "grad_norm": 0.15227212821048294,
+      "learning_rate": 2e-05,
+      "loss": 5.3013,
+      "step": 6071
+    },
+    {
+      "epoch": 0.40728443505382833,
+      "grad_norm": 0.14900960291244386,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 6072
+    },
+    {
+      "epoch": 0.40735151088305327,
+      "grad_norm": 0.14201703042706224,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4074185867122782,
+      "grad_norm": 0.1464483912732817,
+      "learning_rate": 2e-05,
+      "loss": 5.5129,
+      "step": 6074
+    },
+    {
+      "epoch": 0.40748566254150315,
+      "grad_norm": 0.15308299310567888,
+      "learning_rate": 2e-05,
+      "loss": 5.3051,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4075527383707281,
+      "grad_norm": 0.1462770854150524,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 6076
+    },
+    {
+      "epoch": 0.407619814199953,
+      "grad_norm": 0.15044370225557685,
+      "learning_rate": 2e-05,
+      "loss": 5.5122,
+      "step": 6077
+    },
+    {
+      "epoch": 0.40768689002917796,
+      "grad_norm": 0.1676513113476112,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 6078
+    },
+    {
+      "epoch": 0.4077539658584029,
+      "grad_norm": 0.14575690015912224,
+      "learning_rate": 2e-05,
+      "loss": 5.4962,
+      "step": 6079
+    },
+    {
+      "epoch": 0.40782104168762784,
+      "grad_norm": 0.15726576245724802,
+      "learning_rate": 2e-05,
+      "loss": 5.3551,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4078881175168528,
+      "grad_norm": 0.14541743146383224,
+      "learning_rate": 2e-05,
+      "loss": 5.4716,
+      "step": 6081
+    },
+    {
+      "epoch": 0.4079551933460777,
+      "grad_norm": 0.16319908595976848,
+      "learning_rate": 2e-05,
+      "loss": 5.5158,
+      "step": 6082
+    },
+    {
+      "epoch": 0.40802226917530265,
+      "grad_norm": 0.15575388296546167,
+      "learning_rate": 2e-05,
+      "loss": 5.5094,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4080893450045276,
+      "grad_norm": 0.14871726357723505,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 6084
+    },
+    {
+      "epoch": 0.40815642083375253,
+      "grad_norm": 0.1518768114233469,
+      "learning_rate": 2e-05,
+      "loss": 5.3644,
+      "step": 6085
+    },
+    {
+      "epoch": 0.40822349666297747,
+      "grad_norm": 0.15589791501222594,
+      "learning_rate": 2e-05,
+      "loss": 5.4819,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4082905724922024,
+      "grad_norm": 0.14531584804929493,
+      "learning_rate": 2e-05,
+      "loss": 5.2746,
+      "step": 6087
+    },
+    {
+      "epoch": 0.40835764832142735,
+      "grad_norm": 0.14744504757742968,
+      "learning_rate": 2e-05,
+      "loss": 5.4029,
+      "step": 6088
+    },
+    {
+      "epoch": 0.4084247241506523,
+      "grad_norm": 0.1485517928386306,
+      "learning_rate": 2e-05,
+      "loss": 5.4761,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4084917999798772,
+      "grad_norm": 0.14919780234425425,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4085588758091022,
+      "grad_norm": 0.14446332847796833,
+      "learning_rate": 2e-05,
+      "loss": 5.4434,
+      "step": 6091
+    },
+    {
+      "epoch": 0.40862595163832716,
+      "grad_norm": 0.15123167708792787,
+      "learning_rate": 2e-05,
+      "loss": 5.3837,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4086930274675521,
+      "grad_norm": 0.14717028567044393,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 6093
+    },
+    {
+      "epoch": 0.40876010329677703,
+      "grad_norm": 0.15042712068782174,
+      "learning_rate": 2e-05,
+      "loss": 5.4027,
+      "step": 6094
+    },
+    {
+      "epoch": 0.40882717912600197,
+      "grad_norm": 0.14375038162597945,
+      "learning_rate": 2e-05,
+      "loss": 5.3387,
+      "step": 6095
+    },
+    {
+      "epoch": 0.4088942549552269,
+      "grad_norm": 0.15291742270012196,
+      "learning_rate": 2e-05,
+      "loss": 5.5601,
+      "step": 6096
+    },
+    {
+      "epoch": 0.40896133078445185,
+      "grad_norm": 0.14806975628068123,
+      "learning_rate": 2e-05,
+      "loss": 5.5391,
+      "step": 6097
+    },
+    {
+      "epoch": 0.4090284066136768,
+      "grad_norm": 0.14601110429422573,
+      "learning_rate": 2e-05,
+      "loss": 5.6242,
+      "step": 6098
+    },
+    {
+      "epoch": 0.4090954824429017,
+      "grad_norm": 0.14263892148998944,
+      "learning_rate": 2e-05,
+      "loss": 5.4961,
+      "step": 6099
+    },
+    {
+      "epoch": 0.40916255827212666,
+      "grad_norm": 0.15065958473053842,
+      "learning_rate": 2e-05,
+      "loss": 5.3485,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4092296341013516,
+      "grad_norm": 0.14694193541917297,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 6101
+    },
+    {
+      "epoch": 0.40929670993057654,
+      "grad_norm": 0.15278120479286747,
+      "learning_rate": 2e-05,
+      "loss": 5.5714,
+      "step": 6102
+    },
+    {
+      "epoch": 0.4093637857598015,
+      "grad_norm": 0.16054130302940997,
+      "learning_rate": 2e-05,
+      "loss": 5.208,
+      "step": 6103
+    },
+    {
+      "epoch": 0.4094308615890264,
+      "grad_norm": 0.1446162996874019,
+      "learning_rate": 2e-05,
+      "loss": 5.4623,
+      "step": 6104
+    },
+    {
+      "epoch": 0.40949793741825136,
+      "grad_norm": 0.14882105241487117,
+      "learning_rate": 2e-05,
+      "loss": 5.3525,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4095650132474763,
+      "grad_norm": 0.15989120155100828,
+      "learning_rate": 2e-05,
+      "loss": 5.378,
+      "step": 6106
+    },
+    {
+      "epoch": 0.40963208907670123,
+      "grad_norm": 0.14458361666581818,
+      "learning_rate": 2e-05,
+      "loss": 5.4872,
+      "step": 6107
+    },
+    {
+      "epoch": 0.40969916490592617,
+      "grad_norm": 0.14737736523841655,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 6108
+    },
+    {
+      "epoch": 0.4097662407351511,
+      "grad_norm": 0.16387172711224748,
+      "learning_rate": 2e-05,
+      "loss": 5.5087,
+      "step": 6109
+    },
+    {
+      "epoch": 0.40983331656437605,
+      "grad_norm": 0.1480723348815157,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 6110
+    },
+    {
+      "epoch": 0.409900392393601,
+      "grad_norm": 0.15257662510780653,
+      "learning_rate": 2e-05,
+      "loss": 5.5466,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4099674682228259,
+      "grad_norm": 0.15904679556609308,
+      "learning_rate": 2e-05,
+      "loss": 5.3305,
+      "step": 6112
+    },
+    {
+      "epoch": 0.41003454405205086,
+      "grad_norm": 0.15154523568578862,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 6113
+    },
+    {
+      "epoch": 0.4101016198812758,
+      "grad_norm": 0.15098305445598653,
+      "learning_rate": 2e-05,
+      "loss": 5.5554,
+      "step": 6114
+    },
+    {
+      "epoch": 0.41016869571050074,
+      "grad_norm": 0.1589589698493392,
+      "learning_rate": 2e-05,
+      "loss": 5.3591,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4102357715397257,
+      "grad_norm": 0.15656133212438542,
+      "learning_rate": 2e-05,
+      "loss": 5.5106,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4103028473689506,
+      "grad_norm": 0.15140365208840909,
+      "learning_rate": 2e-05,
+      "loss": 5.3766,
+      "step": 6117
+    },
+    {
+      "epoch": 0.41036992319817556,
+      "grad_norm": 0.1434327408736165,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 6118
+    },
+    {
+      "epoch": 0.4104369990274005,
+      "grad_norm": 0.15115410891386852,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 6119
+    },
+    {
+      "epoch": 0.41050407485662543,
+      "grad_norm": 0.1497631913037576,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4105711506858504,
+      "grad_norm": 0.14568586051385393,
+      "learning_rate": 2e-05,
+      "loss": 5.4176,
+      "step": 6121
+    },
+    {
+      "epoch": 0.4106382265150753,
+      "grad_norm": 0.15127203960472932,
+      "learning_rate": 2e-05,
+      "loss": 5.3339,
+      "step": 6122
+    },
+    {
+      "epoch": 0.41070530234430025,
+      "grad_norm": 0.15076954556867875,
+      "learning_rate": 2e-05,
+      "loss": 5.4308,
+      "step": 6123
+    },
+    {
+      "epoch": 0.4107723781735252,
+      "grad_norm": 0.14902824739542336,
+      "learning_rate": 2e-05,
+      "loss": 5.3892,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4108394540027501,
+      "grad_norm": 0.15369660370591312,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 6125
+    },
+    {
+      "epoch": 0.41090652983197506,
+      "grad_norm": 0.1517700878127306,
+      "learning_rate": 2e-05,
+      "loss": 5.54,
+      "step": 6126
+    },
+    {
+      "epoch": 0.4109736056612,
+      "grad_norm": 0.1478424008628996,
+      "learning_rate": 2e-05,
+      "loss": 5.4344,
+      "step": 6127
+    },
+    {
+      "epoch": 0.41104068149042494,
+      "grad_norm": 0.15971026450019005,
+      "learning_rate": 2e-05,
+      "loss": 5.3187,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4111077573196499,
+      "grad_norm": 0.14988302029310308,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4111748331488748,
+      "grad_norm": 0.149698365106905,
+      "learning_rate": 2e-05,
+      "loss": 5.5568,
+      "step": 6130
+    },
+    {
+      "epoch": 0.41124190897809976,
+      "grad_norm": 0.15212818817153279,
+      "learning_rate": 2e-05,
+      "loss": 5.4725,
+      "step": 6131
+    },
+    {
+      "epoch": 0.4113089848073247,
+      "grad_norm": 0.15848372340059896,
+      "learning_rate": 2e-05,
+      "loss": 5.5815,
+      "step": 6132
+    },
+    {
+      "epoch": 0.41137606063654963,
+      "grad_norm": 0.15423699786438205,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 6133
+    },
+    {
+      "epoch": 0.4114431364657746,
+      "grad_norm": 0.14893908040288306,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 6134
+    },
+    {
+      "epoch": 0.4115102122949995,
+      "grad_norm": 0.15242625319854541,
+      "learning_rate": 2e-05,
+      "loss": 5.338,
+      "step": 6135
+    },
+    {
+      "epoch": 0.41157728812422445,
+      "grad_norm": 0.15771477847133705,
+      "learning_rate": 2e-05,
+      "loss": 5.4753,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4116443639534494,
+      "grad_norm": 0.15334869569459353,
+      "learning_rate": 2e-05,
+      "loss": 5.393,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4117114397826743,
+      "grad_norm": 0.14610029309748182,
+      "learning_rate": 2e-05,
+      "loss": 5.3249,
+      "step": 6138
+    },
+    {
+      "epoch": 0.41177851561189927,
+      "grad_norm": 0.15157673023211954,
+      "learning_rate": 2e-05,
+      "loss": 5.4579,
+      "step": 6139
+    },
+    {
+      "epoch": 0.4118455914411242,
+      "grad_norm": 0.14346479193707015,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 6140
+    },
+    {
+      "epoch": 0.41191266727034914,
+      "grad_norm": 0.14838402000327952,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 6141
+    },
+    {
+      "epoch": 0.4119797430995741,
+      "grad_norm": 0.1430616001646572,
+      "learning_rate": 2e-05,
+      "loss": 5.3267,
+      "step": 6142
+    },
+    {
+      "epoch": 0.412046818928799,
+      "grad_norm": 0.15980469581124987,
+      "learning_rate": 2e-05,
+      "loss": 5.3938,
+      "step": 6143
+    },
+    {
+      "epoch": 0.41211389475802396,
+      "grad_norm": 0.14205378270823804,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4121809705872489,
+      "grad_norm": 0.14997561769203577,
+      "learning_rate": 2e-05,
+      "loss": 5.3378,
+      "step": 6145
+    },
+    {
+      "epoch": 0.41224804641647383,
+      "grad_norm": 0.15366498258968467,
+      "learning_rate": 2e-05,
+      "loss": 5.5612,
+      "step": 6146
+    },
+    {
+      "epoch": 0.4123151222456988,
+      "grad_norm": 0.15688349912975239,
+      "learning_rate": 2e-05,
+      "loss": 5.4789,
+      "step": 6147
+    },
+    {
+      "epoch": 0.4123821980749237,
+      "grad_norm": 0.14456526939338343,
+      "learning_rate": 2e-05,
+      "loss": 5.3515,
+      "step": 6148
+    },
+    {
+      "epoch": 0.41244927390414865,
+      "grad_norm": 0.15172809385312647,
+      "learning_rate": 2e-05,
+      "loss": 5.5482,
+      "step": 6149
+    },
+    {
+      "epoch": 0.4125163497333736,
+      "grad_norm": 0.1496753254493908,
+      "learning_rate": 2e-05,
+      "loss": 5.3616,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4125834255625985,
+      "grad_norm": 0.14683100707445412,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 6151
+    },
+    {
+      "epoch": 0.41265050139182347,
+      "grad_norm": 0.15170101686392787,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 6152
+    },
+    {
+      "epoch": 0.4127175772210484,
+      "grad_norm": 0.15250098212703309,
+      "learning_rate": 2e-05,
+      "loss": 5.5805,
+      "step": 6153
+    },
+    {
+      "epoch": 0.41278465305027334,
+      "grad_norm": 0.15117191472511607,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 6154
+    },
+    {
+      "epoch": 0.4128517288794983,
+      "grad_norm": 0.14958652625255425,
+      "learning_rate": 2e-05,
+      "loss": 5.5265,
+      "step": 6155
+    },
+    {
+      "epoch": 0.4129188047087232,
+      "grad_norm": 0.1549552328992079,
+      "learning_rate": 2e-05,
+      "loss": 5.4175,
+      "step": 6156
+    },
+    {
+      "epoch": 0.41298588053794816,
+      "grad_norm": 0.14840966341952094,
+      "learning_rate": 2e-05,
+      "loss": 5.5041,
+      "step": 6157
+    },
+    {
+      "epoch": 0.4130529563671731,
+      "grad_norm": 0.14837594762594064,
+      "learning_rate": 2e-05,
+      "loss": 5.4823,
+      "step": 6158
+    },
+    {
+      "epoch": 0.41312003219639803,
+      "grad_norm": 0.14633692723179006,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 6159
+    },
+    {
+      "epoch": 0.413187108025623,
+      "grad_norm": 0.15219694866476308,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4132541838548479,
+      "grad_norm": 0.146770936016127,
+      "learning_rate": 2e-05,
+      "loss": 5.426,
+      "step": 6161
+    },
+    {
+      "epoch": 0.41332125968407285,
+      "grad_norm": 0.14531835629164536,
+      "learning_rate": 2e-05,
+      "loss": 5.2618,
+      "step": 6162
+    },
+    {
+      "epoch": 0.4133883355132978,
+      "grad_norm": 0.1486287444434929,
+      "learning_rate": 2e-05,
+      "loss": 5.4312,
+      "step": 6163
+    },
+    {
+      "epoch": 0.4134554113425227,
+      "grad_norm": 0.14921852784556688,
+      "learning_rate": 2e-05,
+      "loss": 5.4557,
+      "step": 6164
+    },
+    {
+      "epoch": 0.41352248717174767,
+      "grad_norm": 0.14637180480236725,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4135895630009726,
+      "grad_norm": 0.1724012154011378,
+      "learning_rate": 2e-05,
+      "loss": 5.3923,
+      "step": 6166
+    },
+    {
+      "epoch": 0.41365663883019754,
+      "grad_norm": 0.15004133266552003,
+      "learning_rate": 2e-05,
+      "loss": 5.3885,
+      "step": 6167
+    },
+    {
+      "epoch": 0.4137237146594225,
+      "grad_norm": 0.15316988103669055,
+      "learning_rate": 2e-05,
+      "loss": 5.4615,
+      "step": 6168
+    },
+    {
+      "epoch": 0.4137907904886474,
+      "grad_norm": 0.14623395463643032,
+      "learning_rate": 2e-05,
+      "loss": 5.5389,
+      "step": 6169
+    },
+    {
+      "epoch": 0.41385786631787236,
+      "grad_norm": 0.14808101918085212,
+      "learning_rate": 2e-05,
+      "loss": 5.4192,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4139249421470973,
+      "grad_norm": 0.149877247337902,
+      "learning_rate": 2e-05,
+      "loss": 5.3277,
+      "step": 6171
+    },
+    {
+      "epoch": 0.41399201797632224,
+      "grad_norm": 0.14189501524026232,
+      "learning_rate": 2e-05,
+      "loss": 5.4958,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4140590938055472,
+      "grad_norm": 0.14489508681495938,
+      "learning_rate": 2e-05,
+      "loss": 5.2224,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4141261696347721,
+      "grad_norm": 0.14271955104483994,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 6174
+    },
+    {
+      "epoch": 0.41419324546399705,
+      "grad_norm": 0.1432449732709902,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 6175
+    },
+    {
+      "epoch": 0.414260321293222,
+      "grad_norm": 0.1429603245203277,
+      "learning_rate": 2e-05,
+      "loss": 5.3307,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4143273971224469,
+      "grad_norm": 0.1434477456316279,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 6177
+    },
+    {
+      "epoch": 0.41439447295167187,
+      "grad_norm": 0.15518142713920297,
+      "learning_rate": 2e-05,
+      "loss": 5.352,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4144615487808968,
+      "grad_norm": 0.14401172798104242,
+      "learning_rate": 2e-05,
+      "loss": 5.4644,
+      "step": 6179
+    },
+    {
+      "epoch": 0.41452862461012174,
+      "grad_norm": 0.15729530516313117,
+      "learning_rate": 2e-05,
+      "loss": 5.4151,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4145957004393467,
+      "grad_norm": 0.14732355971864336,
+      "learning_rate": 2e-05,
+      "loss": 5.3683,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4146627762685716,
+      "grad_norm": 0.14565955506072412,
+      "learning_rate": 2e-05,
+      "loss": 5.5106,
+      "step": 6182
+    },
+    {
+      "epoch": 0.41472985209779656,
+      "grad_norm": 0.15234533745162324,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4147969279270215,
+      "grad_norm": 0.15354433454569177,
+      "learning_rate": 2e-05,
+      "loss": 5.3758,
+      "step": 6184
+    },
+    {
+      "epoch": 0.41486400375624644,
+      "grad_norm": 0.15406927654932262,
+      "learning_rate": 2e-05,
+      "loss": 5.4858,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4149310795854714,
+      "grad_norm": 0.14512594791563746,
+      "learning_rate": 2e-05,
+      "loss": 5.4833,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4149981554146963,
+      "grad_norm": 0.14833918876614743,
+      "learning_rate": 2e-05,
+      "loss": 5.3592,
+      "step": 6187
+    },
+    {
+      "epoch": 0.41506523124392125,
+      "grad_norm": 0.162823890699379,
+      "learning_rate": 2e-05,
+      "loss": 5.6159,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4151323070731462,
+      "grad_norm": 0.15248920966437957,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 6189
+    },
+    {
+      "epoch": 0.41519938290237113,
+      "grad_norm": 0.14752179454132905,
+      "learning_rate": 2e-05,
+      "loss": 5.5288,
+      "step": 6190
+    },
+    {
+      "epoch": 0.41526645873159607,
+      "grad_norm": 0.15382748152958348,
+      "learning_rate": 2e-05,
+      "loss": 5.3071,
+      "step": 6191
+    },
+    {
+      "epoch": 0.415333534560821,
+      "grad_norm": 0.15206752100770493,
+      "learning_rate": 2e-05,
+      "loss": 5.2151,
+      "step": 6192
+    },
+    {
+      "epoch": 0.41540061039004594,
+      "grad_norm": 0.15166292563124945,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4154676862192709,
+      "grad_norm": 0.14353771090088854,
+      "learning_rate": 2e-05,
+      "loss": 5.3598,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4155347620484958,
+      "grad_norm": 0.15607206891195802,
+      "learning_rate": 2e-05,
+      "loss": 5.4813,
+      "step": 6195
+    },
+    {
+      "epoch": 0.41560183787772076,
+      "grad_norm": 0.15240955471116122,
+      "learning_rate": 2e-05,
+      "loss": 5.3344,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4156689137069457,
+      "grad_norm": 0.14220299492229682,
+      "learning_rate": 2e-05,
+      "loss": 5.4989,
+      "step": 6197
+    },
+    {
+      "epoch": 0.41573598953617064,
+      "grad_norm": 0.14448273926202573,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4158030653653956,
+      "grad_norm": 0.14752277696179278,
+      "learning_rate": 2e-05,
+      "loss": 5.3598,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4158701411946205,
+      "grad_norm": 0.1487673169243866,
+      "learning_rate": 2e-05,
+      "loss": 5.4494,
+      "step": 6200
+    },
+    {
+      "epoch": 0.41593721702384545,
+      "grad_norm": 0.15423482820114895,
+      "learning_rate": 2e-05,
+      "loss": 5.369,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4160042928530704,
+      "grad_norm": 0.15224545886837434,
+      "learning_rate": 2e-05,
+      "loss": 5.3478,
+      "step": 6202
+    },
+    {
+      "epoch": 0.41607136868229533,
+      "grad_norm": 0.15111854897418345,
+      "learning_rate": 2e-05,
+      "loss": 5.4217,
+      "step": 6203
+    },
+    {
+      "epoch": 0.41613844451152027,
+      "grad_norm": 0.14513002352064175,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4162055203407452,
+      "grad_norm": 0.15278045129940115,
+      "learning_rate": 2e-05,
+      "loss": 5.4462,
+      "step": 6205
+    },
+    {
+      "epoch": 0.41627259616997014,
+      "grad_norm": 0.15353011611301318,
+      "learning_rate": 2e-05,
+      "loss": 5.4453,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4163396719991951,
+      "grad_norm": 0.14586770030273707,
+      "learning_rate": 2e-05,
+      "loss": 5.4109,
+      "step": 6207
+    },
+    {
+      "epoch": 0.41640674782842,
+      "grad_norm": 0.1423842447594983,
+      "learning_rate": 2e-05,
+      "loss": 5.3681,
+      "step": 6208
+    },
+    {
+      "epoch": 0.41647382365764496,
+      "grad_norm": 0.14882019391754242,
+      "learning_rate": 2e-05,
+      "loss": 5.539,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4165408994868699,
+      "grad_norm": 0.14348557963834951,
+      "learning_rate": 2e-05,
+      "loss": 5.4452,
+      "step": 6210
+    },
+    {
+      "epoch": 0.41660797531609484,
+      "grad_norm": 0.15039619927816222,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4166750511453198,
+      "grad_norm": 0.13939458139050415,
+      "learning_rate": 2e-05,
+      "loss": 5.4613,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4167421269745447,
+      "grad_norm": 0.14689257277495957,
+      "learning_rate": 2e-05,
+      "loss": 5.223,
+      "step": 6213
+    },
+    {
+      "epoch": 0.41680920280376965,
+      "grad_norm": 0.150487088584561,
+      "learning_rate": 2e-05,
+      "loss": 5.5122,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4168762786329946,
+      "grad_norm": 0.15026701198479003,
+      "learning_rate": 2e-05,
+      "loss": 5.4937,
+      "step": 6215
+    },
+    {
+      "epoch": 0.41694335446221953,
+      "grad_norm": 0.14851473736825993,
+      "learning_rate": 2e-05,
+      "loss": 5.4574,
+      "step": 6216
+    },
+    {
+      "epoch": 0.41701043029144447,
+      "grad_norm": 0.1438750385177889,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4170775061206694,
+      "grad_norm": 0.1482205165061603,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 6218
+    },
+    {
+      "epoch": 0.41714458194989434,
+      "grad_norm": 0.14799252323494966,
+      "learning_rate": 2e-05,
+      "loss": 5.3691,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4172116577791193,
+      "grad_norm": 0.14614882601745888,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4172787336083442,
+      "grad_norm": 0.1457305357527663,
+      "learning_rate": 2e-05,
+      "loss": 5.3378,
+      "step": 6221
+    },
+    {
+      "epoch": 0.41734580943756916,
+      "grad_norm": 0.15300223575820496,
+      "learning_rate": 2e-05,
+      "loss": 5.5145,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4174128852667941,
+      "grad_norm": 0.1478922301907365,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 6223
+    },
+    {
+      "epoch": 0.41747996109601904,
+      "grad_norm": 0.14879945433933275,
+      "learning_rate": 2e-05,
+      "loss": 5.3496,
+      "step": 6224
+    },
+    {
+      "epoch": 0.417547036925244,
+      "grad_norm": 0.1507505739326256,
+      "learning_rate": 2e-05,
+      "loss": 5.3937,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4176141127544689,
+      "grad_norm": 0.1507091153310175,
+      "learning_rate": 2e-05,
+      "loss": 5.2893,
+      "step": 6226
+    },
+    {
+      "epoch": 0.41768118858369385,
+      "grad_norm": 0.15565932321239265,
+      "learning_rate": 2e-05,
+      "loss": 5.5619,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4177482644129188,
+      "grad_norm": 0.153089756977703,
+      "learning_rate": 2e-05,
+      "loss": 5.5483,
+      "step": 6228
+    },
+    {
+      "epoch": 0.41781534024214373,
+      "grad_norm": 0.14653417287645107,
+      "learning_rate": 2e-05,
+      "loss": 5.3846,
+      "step": 6229
+    },
+    {
+      "epoch": 0.41788241607136867,
+      "grad_norm": 0.16677507039010597,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4179494919005936,
+      "grad_norm": 0.15151076827398888,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 6231
+    },
+    {
+      "epoch": 0.41801656772981854,
+      "grad_norm": 0.151191684336144,
+      "learning_rate": 2e-05,
+      "loss": 5.3537,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4180836435590435,
+      "grad_norm": 0.14520346000136614,
+      "learning_rate": 2e-05,
+      "loss": 5.3477,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4181507193882684,
+      "grad_norm": 0.15630497739757923,
+      "learning_rate": 2e-05,
+      "loss": 5.3903,
+      "step": 6234
+    },
+    {
+      "epoch": 0.41821779521749336,
+      "grad_norm": 0.14745242313948076,
+      "learning_rate": 2e-05,
+      "loss": 5.3193,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4182848710467183,
+      "grad_norm": 0.15475620807369375,
+      "learning_rate": 2e-05,
+      "loss": 5.4467,
+      "step": 6236
+    },
+    {
+      "epoch": 0.41835194687594324,
+      "grad_norm": 0.15597631423108876,
+      "learning_rate": 2e-05,
+      "loss": 5.4333,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4184190227051682,
+      "grad_norm": 0.1454715923665536,
+      "learning_rate": 2e-05,
+      "loss": 5.3532,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4184860985343931,
+      "grad_norm": 0.15405110337621403,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 6239
+    },
+    {
+      "epoch": 0.41855317436361805,
+      "grad_norm": 0.14510767484656545,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 6240
+    },
+    {
+      "epoch": 0.418620250192843,
+      "grad_norm": 0.14327506672282547,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 6241
+    },
+    {
+      "epoch": 0.41868732602206793,
+      "grad_norm": 0.1436085215241924,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 6242
+    },
+    {
+      "epoch": 0.41875440185129287,
+      "grad_norm": 0.16043325284090895,
+      "learning_rate": 2e-05,
+      "loss": 5.5273,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4188214776805178,
+      "grad_norm": 0.1471908179961017,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 6244
+    },
+    {
+      "epoch": 0.41888855350974274,
+      "grad_norm": 0.15056358505816222,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4189556293389677,
+      "grad_norm": 0.14354892536811004,
+      "learning_rate": 2e-05,
+      "loss": 5.2549,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4190227051681926,
+      "grad_norm": 0.15385620857295468,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 6247
+    },
+    {
+      "epoch": 0.41908978099741756,
+      "grad_norm": 0.1424732070591553,
+      "learning_rate": 2e-05,
+      "loss": 5.4353,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4191568568266425,
+      "grad_norm": 0.15001966269652894,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 6249
+    },
+    {
+      "epoch": 0.41922393265586744,
+      "grad_norm": 0.14770740112158803,
+      "learning_rate": 2e-05,
+      "loss": 5.5145,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4192910084850924,
+      "grad_norm": 0.15044719622659666,
+      "learning_rate": 2e-05,
+      "loss": 5.3205,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4193580843143173,
+      "grad_norm": 0.1527564091363946,
+      "learning_rate": 2e-05,
+      "loss": 5.5283,
+      "step": 6252
+    },
+    {
+      "epoch": 0.41942516014354225,
+      "grad_norm": 0.15188573785871098,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4194922359727672,
+      "grad_norm": 0.15294602291747922,
+      "learning_rate": 2e-05,
+      "loss": 5.522,
+      "step": 6254
+    },
+    {
+      "epoch": 0.41955931180199213,
+      "grad_norm": 0.15843716597399227,
+      "learning_rate": 2e-05,
+      "loss": 5.3152,
+      "step": 6255
+    },
+    {
+      "epoch": 0.41962638763121707,
+      "grad_norm": 0.15179939584651417,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 6256
+    },
+    {
+      "epoch": 0.419693463460442,
+      "grad_norm": 0.14471588666040858,
+      "learning_rate": 2e-05,
+      "loss": 5.5224,
+      "step": 6257
+    },
+    {
+      "epoch": 0.41976053928966695,
+      "grad_norm": 0.15220375224074542,
+      "learning_rate": 2e-05,
+      "loss": 5.3866,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4198276151188919,
+      "grad_norm": 0.15891180059895374,
+      "learning_rate": 2e-05,
+      "loss": 5.4016,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4198946909481168,
+      "grad_norm": 0.15281374239137188,
+      "learning_rate": 2e-05,
+      "loss": 5.4259,
+      "step": 6260
+    },
+    {
+      "epoch": 0.41996176677734176,
+      "grad_norm": 0.1447042177596159,
+      "learning_rate": 2e-05,
+      "loss": 5.4588,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4200288426065667,
+      "grad_norm": 0.14936128720877714,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 6262
+    },
+    {
+      "epoch": 0.42009591843579164,
+      "grad_norm": 0.1480513829830632,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 6263
+    },
+    {
+      "epoch": 0.4201629942650166,
+      "grad_norm": 0.14375693722623836,
+      "learning_rate": 2e-05,
+      "loss": 5.3675,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4202300700942415,
+      "grad_norm": 0.14299636149433792,
+      "learning_rate": 2e-05,
+      "loss": 5.5178,
+      "step": 6265
+    },
+    {
+      "epoch": 0.42029714592346645,
+      "grad_norm": 0.15514933069794587,
+      "learning_rate": 2e-05,
+      "loss": 5.5361,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4203642217526914,
+      "grad_norm": 0.14467812185900947,
+      "learning_rate": 2e-05,
+      "loss": 5.5949,
+      "step": 6267
+    },
+    {
+      "epoch": 0.42043129758191633,
+      "grad_norm": 0.1521273815934375,
+      "learning_rate": 2e-05,
+      "loss": 5.3192,
+      "step": 6268
+    },
+    {
+      "epoch": 0.42049837341114127,
+      "grad_norm": 0.1464862777116028,
+      "learning_rate": 2e-05,
+      "loss": 5.4545,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4205654492403662,
+      "grad_norm": 0.16111925823418224,
+      "learning_rate": 2e-05,
+      "loss": 5.6047,
+      "step": 6270
+    },
+    {
+      "epoch": 0.42063252506959115,
+      "grad_norm": 0.14796975252678068,
+      "learning_rate": 2e-05,
+      "loss": 5.3861,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4206996008988161,
+      "grad_norm": 0.14952287798702274,
+      "learning_rate": 2e-05,
+      "loss": 5.3915,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4207666767280411,
+      "grad_norm": 0.15282892046203028,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 6273
+    },
+    {
+      "epoch": 0.420833752557266,
+      "grad_norm": 0.1589997171380678,
+      "learning_rate": 2e-05,
+      "loss": 5.4139,
+      "step": 6274
+    },
+    {
+      "epoch": 0.42090082838649095,
+      "grad_norm": 0.14862137958017682,
+      "learning_rate": 2e-05,
+      "loss": 5.3871,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4209679042157159,
+      "grad_norm": 0.14549293419163686,
+      "learning_rate": 2e-05,
+      "loss": 5.4168,
+      "step": 6276
+    },
+    {
+      "epoch": 0.42103498004494083,
+      "grad_norm": 0.15642185184478385,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 6277
+    },
+    {
+      "epoch": 0.42110205587416577,
+      "grad_norm": 0.14301914804485114,
+      "learning_rate": 2e-05,
+      "loss": 5.2758,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4211691317033907,
+      "grad_norm": 0.14699058275200158,
+      "learning_rate": 2e-05,
+      "loss": 5.4027,
+      "step": 6279
+    },
+    {
+      "epoch": 0.42123620753261565,
+      "grad_norm": 0.14616535676223494,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4213032833618406,
+      "grad_norm": 0.15220548597395242,
+      "learning_rate": 2e-05,
+      "loss": 5.5928,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4213703591910655,
+      "grad_norm": 0.15010775257668782,
+      "learning_rate": 2e-05,
+      "loss": 5.3152,
+      "step": 6282
+    },
+    {
+      "epoch": 0.42143743502029046,
+      "grad_norm": 0.1518526357259836,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4215045108495154,
+      "grad_norm": 0.14904138457613592,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 6284
+    },
+    {
+      "epoch": 0.42157158667874034,
+      "grad_norm": 0.15080750347040855,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4216386625079653,
+      "grad_norm": 0.1440761690030597,
+      "learning_rate": 2e-05,
+      "loss": 5.5277,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4217057383371902,
+      "grad_norm": 0.14424318522665128,
+      "learning_rate": 2e-05,
+      "loss": 5.5545,
+      "step": 6287
+    },
+    {
+      "epoch": 0.42177281416641516,
+      "grad_norm": 0.1472559350447186,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4218398899956401,
+      "grad_norm": 0.14790661005189126,
+      "learning_rate": 2e-05,
+      "loss": 5.3892,
+      "step": 6289
+    },
+    {
+      "epoch": 0.42190696582486503,
+      "grad_norm": 0.15631683783403477,
+      "learning_rate": 2e-05,
+      "loss": 5.4418,
+      "step": 6290
+    },
+    {
+      "epoch": 0.42197404165408997,
+      "grad_norm": 0.1538317415095551,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4220411174833149,
+      "grad_norm": 0.1465802683346149,
+      "learning_rate": 2e-05,
+      "loss": 5.3616,
+      "step": 6292
+    },
+    {
+      "epoch": 0.42210819331253985,
+      "grad_norm": 0.14910748102731003,
+      "learning_rate": 2e-05,
+      "loss": 5.4102,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4221752691417648,
+      "grad_norm": 0.16212135060554408,
+      "learning_rate": 2e-05,
+      "loss": 5.5169,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4222423449709897,
+      "grad_norm": 0.1582649965838994,
+      "learning_rate": 2e-05,
+      "loss": 5.3175,
+      "step": 6295
+    },
+    {
+      "epoch": 0.42230942080021466,
+      "grad_norm": 0.1468493767731153,
+      "learning_rate": 2e-05,
+      "loss": 5.4917,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4223764966294396,
+      "grad_norm": 0.15839304377719646,
+      "learning_rate": 2e-05,
+      "loss": 5.502,
+      "step": 6297
+    },
+    {
+      "epoch": 0.42244357245866454,
+      "grad_norm": 0.16364533413533558,
+      "learning_rate": 2e-05,
+      "loss": 5.5592,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4225106482878895,
+      "grad_norm": 0.15234920820946346,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4225777241171144,
+      "grad_norm": 0.1615345449521551,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 6300
+    },
+    {
+      "epoch": 0.42264479994633936,
+      "grad_norm": 0.15130417365066476,
+      "learning_rate": 2e-05,
+      "loss": 5.3677,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4227118757755643,
+      "grad_norm": 0.15068306036917925,
+      "learning_rate": 2e-05,
+      "loss": 5.4435,
+      "step": 6302
+    },
+    {
+      "epoch": 0.42277895160478923,
+      "grad_norm": 0.147663147686365,
+      "learning_rate": 2e-05,
+      "loss": 5.339,
+      "step": 6303
+    },
+    {
+      "epoch": 0.42284602743401417,
+      "grad_norm": 0.13883749752196478,
+      "learning_rate": 2e-05,
+      "loss": 5.4882,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4229131032632391,
+      "grad_norm": 0.14375279316401002,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 6305
+    },
+    {
+      "epoch": 0.42298017909246405,
+      "grad_norm": 0.14555707459421724,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 6306
+    },
+    {
+      "epoch": 0.423047254921689,
+      "grad_norm": 0.15391036158667443,
+      "learning_rate": 2e-05,
+      "loss": 5.3035,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4231143307509139,
+      "grad_norm": 0.1457817489900408,
+      "learning_rate": 2e-05,
+      "loss": 5.5036,
+      "step": 6308
+    },
+    {
+      "epoch": 0.42318140658013886,
+      "grad_norm": 0.14328876678178493,
+      "learning_rate": 2e-05,
+      "loss": 5.3407,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4232484824093638,
+      "grad_norm": 0.1477254926824025,
+      "learning_rate": 2e-05,
+      "loss": 5.453,
+      "step": 6310
+    },
+    {
+      "epoch": 0.42331555823858874,
+      "grad_norm": 0.1481214922280087,
+      "learning_rate": 2e-05,
+      "loss": 5.5024,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4233826340678137,
+      "grad_norm": 0.15426413393993568,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4234497098970386,
+      "grad_norm": 0.1525589847575938,
+      "learning_rate": 2e-05,
+      "loss": 5.4712,
+      "step": 6313
+    },
+    {
+      "epoch": 0.42351678572626356,
+      "grad_norm": 0.15218552624249482,
+      "learning_rate": 2e-05,
+      "loss": 5.4856,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4235838615554885,
+      "grad_norm": 0.14912497423905594,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 6315
+    },
+    {
+      "epoch": 0.42365093738471343,
+      "grad_norm": 0.1499585820884491,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 6316
+    },
+    {
+      "epoch": 0.42371801321393837,
+      "grad_norm": 0.1502630969709438,
+      "learning_rate": 2e-05,
+      "loss": 5.297,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4237850890431633,
+      "grad_norm": 0.1525721676751632,
+      "learning_rate": 2e-05,
+      "loss": 5.473,
+      "step": 6318
+    },
+    {
+      "epoch": 0.42385216487238825,
+      "grad_norm": 0.15229626044829908,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4239192407016132,
+      "grad_norm": 0.15255199292726954,
+      "learning_rate": 2e-05,
+      "loss": 5.3864,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4239863165308381,
+      "grad_norm": 0.14961846356886088,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 6321
+    },
+    {
+      "epoch": 0.42405339236006306,
+      "grad_norm": 0.14489946545823082,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 6322
+    },
+    {
+      "epoch": 0.424120468189288,
+      "grad_norm": 0.15234288818410083,
+      "learning_rate": 2e-05,
+      "loss": 5.2633,
+      "step": 6323
+    },
+    {
+      "epoch": 0.42418754401851294,
+      "grad_norm": 0.14812149276654887,
+      "learning_rate": 2e-05,
+      "loss": 5.4852,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4242546198477379,
+      "grad_norm": 0.14841254149381464,
+      "learning_rate": 2e-05,
+      "loss": 5.5512,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4243216956769628,
+      "grad_norm": 0.15082652186283776,
+      "learning_rate": 2e-05,
+      "loss": 5.5245,
+      "step": 6326
+    },
+    {
+      "epoch": 0.42438877150618776,
+      "grad_norm": 0.14353582523216818,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 6327
+    },
+    {
+      "epoch": 0.4244558473354127,
+      "grad_norm": 0.14934034716833278,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 6328
+    },
+    {
+      "epoch": 0.42452292316463763,
+      "grad_norm": 0.15915198119848054,
+      "learning_rate": 2e-05,
+      "loss": 5.2098,
+      "step": 6329
+    },
+    {
+      "epoch": 0.42458999899386257,
+      "grad_norm": 0.14782800402427604,
+      "learning_rate": 2e-05,
+      "loss": 5.5379,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4246570748230875,
+      "grad_norm": 0.14846977375180034,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 6331
+    },
+    {
+      "epoch": 0.42472415065231245,
+      "grad_norm": 0.1500486209479468,
+      "learning_rate": 2e-05,
+      "loss": 5.5086,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4247912264815374,
+      "grad_norm": 0.15060898056889266,
+      "learning_rate": 2e-05,
+      "loss": 5.309,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4248583023107623,
+      "grad_norm": 0.1508606600946992,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 6334
+    },
+    {
+      "epoch": 0.42492537813998726,
+      "grad_norm": 0.14921555929636063,
+      "learning_rate": 2e-05,
+      "loss": 5.5702,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4249924539692122,
+      "grad_norm": 0.15041149736880607,
+      "learning_rate": 2e-05,
+      "loss": 5.5617,
+      "step": 6336
+    },
+    {
+      "epoch": 0.42505952979843714,
+      "grad_norm": 0.1451681716019452,
+      "learning_rate": 2e-05,
+      "loss": 5.3624,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4251266056276621,
+      "grad_norm": 0.14880848314998352,
+      "learning_rate": 2e-05,
+      "loss": 5.3402,
+      "step": 6338
+    },
+    {
+      "epoch": 0.425193681456887,
+      "grad_norm": 0.14524561265530886,
+      "learning_rate": 2e-05,
+      "loss": 5.3628,
+      "step": 6339
+    },
+    {
+      "epoch": 0.42526075728611196,
+      "grad_norm": 0.14800925243553942,
+      "learning_rate": 2e-05,
+      "loss": 5.422,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4253278331153369,
+      "grad_norm": 0.14181305057171026,
+      "learning_rate": 2e-05,
+      "loss": 5.27,
+      "step": 6341
+    },
+    {
+      "epoch": 0.42539490894456183,
+      "grad_norm": 0.14544460065508572,
+      "learning_rate": 2e-05,
+      "loss": 5.4706,
+      "step": 6342
+    },
+    {
+      "epoch": 0.42546198477378677,
+      "grad_norm": 0.14885509482560755,
+      "learning_rate": 2e-05,
+      "loss": 5.4048,
+      "step": 6343
+    },
+    {
+      "epoch": 0.4255290606030117,
+      "grad_norm": 0.15540881001470022,
+      "learning_rate": 2e-05,
+      "loss": 5.3827,
+      "step": 6344
+    },
+    {
+      "epoch": 0.42559613643223665,
+      "grad_norm": 0.14196113215334985,
+      "learning_rate": 2e-05,
+      "loss": 5.558,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4256632122614616,
+      "grad_norm": 0.15090964564556247,
+      "learning_rate": 2e-05,
+      "loss": 5.5721,
+      "step": 6346
+    },
+    {
+      "epoch": 0.4257302880906865,
+      "grad_norm": 0.15310971759885308,
+      "learning_rate": 2e-05,
+      "loss": 5.2969,
+      "step": 6347
+    },
+    {
+      "epoch": 0.42579736391991146,
+      "grad_norm": 0.14690489069101612,
+      "learning_rate": 2e-05,
+      "loss": 5.3421,
+      "step": 6348
+    },
+    {
+      "epoch": 0.4258644397491364,
+      "grad_norm": 0.14299142792452332,
+      "learning_rate": 2e-05,
+      "loss": 5.5065,
+      "step": 6349
+    },
+    {
+      "epoch": 0.42593151557836134,
+      "grad_norm": 0.14402663086109949,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4259985914075863,
+      "grad_norm": 0.14384231402536837,
+      "learning_rate": 2e-05,
+      "loss": 5.2908,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4260656672368112,
+      "grad_norm": 0.14189841708225695,
+      "learning_rate": 2e-05,
+      "loss": 5.4331,
+      "step": 6352
+    },
+    {
+      "epoch": 0.42613274306603616,
+      "grad_norm": 0.14909073763349645,
+      "learning_rate": 2e-05,
+      "loss": 5.5054,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4261998188952611,
+      "grad_norm": 0.14853813067716531,
+      "learning_rate": 2e-05,
+      "loss": 5.5366,
+      "step": 6354
+    },
+    {
+      "epoch": 0.42626689472448603,
+      "grad_norm": 0.14873100129395245,
+      "learning_rate": 2e-05,
+      "loss": 5.5537,
+      "step": 6355
+    },
+    {
+      "epoch": 0.426333970553711,
+      "grad_norm": 0.15875127778007253,
+      "learning_rate": 2e-05,
+      "loss": 5.4858,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4264010463829359,
+      "grad_norm": 0.1501812051895804,
+      "learning_rate": 2e-05,
+      "loss": 5.362,
+      "step": 6357
+    },
+    {
+      "epoch": 0.42646812221216085,
+      "grad_norm": 0.14884330491326195,
+      "learning_rate": 2e-05,
+      "loss": 5.3117,
+      "step": 6358
+    },
+    {
+      "epoch": 0.4265351980413858,
+      "grad_norm": 0.16006967063380614,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 6359
+    },
+    {
+      "epoch": 0.4266022738706107,
+      "grad_norm": 0.14776442218168975,
+      "learning_rate": 2e-05,
+      "loss": 5.3439,
+      "step": 6360
+    },
+    {
+      "epoch": 0.42666934969983566,
+      "grad_norm": 0.15011998717111025,
+      "learning_rate": 2e-05,
+      "loss": 5.5161,
+      "step": 6361
+    },
+    {
+      "epoch": 0.4267364255290606,
+      "grad_norm": 0.15000862265307194,
+      "learning_rate": 2e-05,
+      "loss": 5.4363,
+      "step": 6362
+    },
+    {
+      "epoch": 0.42680350135828554,
+      "grad_norm": 0.14806097712621338,
+      "learning_rate": 2e-05,
+      "loss": 5.3629,
+      "step": 6363
+    },
+    {
+      "epoch": 0.4268705771875105,
+      "grad_norm": 0.15530771936686952,
+      "learning_rate": 2e-05,
+      "loss": 5.3554,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4269376530167354,
+      "grad_norm": 0.15988492115346004,
+      "learning_rate": 2e-05,
+      "loss": 5.3779,
+      "step": 6365
+    },
+    {
+      "epoch": 0.42700472884596036,
+      "grad_norm": 0.15049775642801386,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4270718046751853,
+      "grad_norm": 0.14977308658474148,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 6367
+    },
+    {
+      "epoch": 0.42713888050441023,
+      "grad_norm": 0.15419666328961787,
+      "learning_rate": 2e-05,
+      "loss": 5.3701,
+      "step": 6368
+    },
+    {
+      "epoch": 0.4272059563336352,
+      "grad_norm": 0.14984688694792614,
+      "learning_rate": 2e-05,
+      "loss": 5.3434,
+      "step": 6369
+    },
+    {
+      "epoch": 0.4272730321628601,
+      "grad_norm": 0.1489388949148364,
+      "learning_rate": 2e-05,
+      "loss": 5.3495,
+      "step": 6370
+    },
+    {
+      "epoch": 0.42734010799208505,
+      "grad_norm": 0.1531576232434007,
+      "learning_rate": 2e-05,
+      "loss": 5.4124,
+      "step": 6371
+    },
+    {
+      "epoch": 0.42740718382131,
+      "grad_norm": 0.14872504331147854,
+      "learning_rate": 2e-05,
+      "loss": 5.4013,
+      "step": 6372
+    },
+    {
+      "epoch": 0.4274742596505349,
+      "grad_norm": 0.14672099809144326,
+      "learning_rate": 2e-05,
+      "loss": 5.5647,
+      "step": 6373
+    },
+    {
+      "epoch": 0.42754133547975987,
+      "grad_norm": 0.1454822892726034,
+      "learning_rate": 2e-05,
+      "loss": 5.3571,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4276084113089848,
+      "grad_norm": 0.15603740321447734,
+      "learning_rate": 2e-05,
+      "loss": 5.3858,
+      "step": 6375
+    },
+    {
+      "epoch": 0.42767548713820974,
+      "grad_norm": 0.15098213421693266,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 6376
+    },
+    {
+      "epoch": 0.4277425629674347,
+      "grad_norm": 0.154305484759387,
+      "learning_rate": 2e-05,
+      "loss": 5.2736,
+      "step": 6377
+    },
+    {
+      "epoch": 0.4278096387966596,
+      "grad_norm": 0.15745747840784957,
+      "learning_rate": 2e-05,
+      "loss": 5.355,
+      "step": 6378
+    },
+    {
+      "epoch": 0.42787671462588456,
+      "grad_norm": 0.14742988821197034,
+      "learning_rate": 2e-05,
+      "loss": 5.3308,
+      "step": 6379
+    },
+    {
+      "epoch": 0.4279437904551095,
+      "grad_norm": 0.152115945607217,
+      "learning_rate": 2e-05,
+      "loss": 5.3687,
+      "step": 6380
+    },
+    {
+      "epoch": 0.42801086628433443,
+      "grad_norm": 0.1479827086120508,
+      "learning_rate": 2e-05,
+      "loss": 5.4318,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4280779421135594,
+      "grad_norm": 0.1553773220878037,
+      "learning_rate": 2e-05,
+      "loss": 5.427,
+      "step": 6382
+    },
+    {
+      "epoch": 0.4281450179427843,
+      "grad_norm": 0.145888337533544,
+      "learning_rate": 2e-05,
+      "loss": 5.2282,
+      "step": 6383
+    },
+    {
+      "epoch": 0.42821209377200925,
+      "grad_norm": 0.1545810029462627,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 6384
+    },
+    {
+      "epoch": 0.4282791696012342,
+      "grad_norm": 0.14737752906559962,
+      "learning_rate": 2e-05,
+      "loss": 5.5683,
+      "step": 6385
+    },
+    {
+      "epoch": 0.4283462454304591,
+      "grad_norm": 0.15292685924566551,
+      "learning_rate": 2e-05,
+      "loss": 5.4204,
+      "step": 6386
+    },
+    {
+      "epoch": 0.42841332125968407,
+      "grad_norm": 0.1433130993420583,
+      "learning_rate": 2e-05,
+      "loss": 5.501,
+      "step": 6387
+    },
+    {
+      "epoch": 0.428480397088909,
+      "grad_norm": 0.14312607300289157,
+      "learning_rate": 2e-05,
+      "loss": 5.3602,
+      "step": 6388
+    },
+    {
+      "epoch": 0.42854747291813394,
+      "grad_norm": 0.14680712931886405,
+      "learning_rate": 2e-05,
+      "loss": 5.2793,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4286145487473589,
+      "grad_norm": 0.14981783086726577,
+      "learning_rate": 2e-05,
+      "loss": 5.4758,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4286816245765838,
+      "grad_norm": 0.14593466764486915,
+      "learning_rate": 2e-05,
+      "loss": 5.5443,
+      "step": 6391
+    },
+    {
+      "epoch": 0.42874870040580876,
+      "grad_norm": 0.1493379664286977,
+      "learning_rate": 2e-05,
+      "loss": 5.3108,
+      "step": 6392
+    },
+    {
+      "epoch": 0.4288157762350337,
+      "grad_norm": 0.15134819385202816,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 6393
+    },
+    {
+      "epoch": 0.42888285206425864,
+      "grad_norm": 0.15297423734525215,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 6394
+    },
+    {
+      "epoch": 0.4289499278934836,
+      "grad_norm": 0.1425429299323518,
+      "learning_rate": 2e-05,
+      "loss": 5.5074,
+      "step": 6395
+    },
+    {
+      "epoch": 0.4290170037227085,
+      "grad_norm": 0.15897539228962473,
+      "learning_rate": 2e-05,
+      "loss": 5.5658,
+      "step": 6396
+    },
+    {
+      "epoch": 0.42908407955193345,
+      "grad_norm": 0.1490652384228625,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 6397
+    },
+    {
+      "epoch": 0.4291511553811584,
+      "grad_norm": 0.1547315814636295,
+      "learning_rate": 2e-05,
+      "loss": 5.4819,
+      "step": 6398
+    },
+    {
+      "epoch": 0.4292182312103833,
+      "grad_norm": 0.15763984885845436,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 6399
+    },
+    {
+      "epoch": 0.42928530703960827,
+      "grad_norm": 0.14781203099497478,
+      "learning_rate": 2e-05,
+      "loss": 5.5324,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4293523828688332,
+      "grad_norm": 0.14572788008268758,
+      "learning_rate": 2e-05,
+      "loss": 5.3654,
+      "step": 6401
+    },
+    {
+      "epoch": 0.42941945869805814,
+      "grad_norm": 0.1458755520601082,
+      "learning_rate": 2e-05,
+      "loss": 5.2775,
+      "step": 6402
+    },
+    {
+      "epoch": 0.4294865345272831,
+      "grad_norm": 0.1529328654873612,
+      "learning_rate": 2e-05,
+      "loss": 5.4064,
+      "step": 6403
+    },
+    {
+      "epoch": 0.429553610356508,
+      "grad_norm": 0.15261991308240883,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 6404
+    },
+    {
+      "epoch": 0.42962068618573296,
+      "grad_norm": 0.15905577783875505,
+      "learning_rate": 2e-05,
+      "loss": 5.5258,
+      "step": 6405
+    },
+    {
+      "epoch": 0.4296877620149579,
+      "grad_norm": 0.15064357757502875,
+      "learning_rate": 2e-05,
+      "loss": 5.3624,
+      "step": 6406
+    },
+    {
+      "epoch": 0.42975483784418284,
+      "grad_norm": 0.14840854283968974,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 6407
+    },
+    {
+      "epoch": 0.4298219136734078,
+      "grad_norm": 0.14504654214928808,
+      "learning_rate": 2e-05,
+      "loss": 5.2371,
+      "step": 6408
+    },
+    {
+      "epoch": 0.4298889895026327,
+      "grad_norm": 0.14911189443394368,
+      "learning_rate": 2e-05,
+      "loss": 5.4908,
+      "step": 6409
+    },
+    {
+      "epoch": 0.42995606533185765,
+      "grad_norm": 0.14773236405709883,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4300231411610826,
+      "grad_norm": 0.14157823409930834,
+      "learning_rate": 2e-05,
+      "loss": 5.2504,
+      "step": 6411
+    },
+    {
+      "epoch": 0.43009021699030753,
+      "grad_norm": 0.14853204921675014,
+      "learning_rate": 2e-05,
+      "loss": 5.4086,
+      "step": 6412
+    },
+    {
+      "epoch": 0.43015729281953247,
+      "grad_norm": 0.15265822388382758,
+      "learning_rate": 2e-05,
+      "loss": 5.5443,
+      "step": 6413
+    },
+    {
+      "epoch": 0.4302243686487574,
+      "grad_norm": 0.1480659069604244,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 6414
+    },
+    {
+      "epoch": 0.43029144447798234,
+      "grad_norm": 0.15424386731286038,
+      "learning_rate": 2e-05,
+      "loss": 5.5073,
+      "step": 6415
+    },
+    {
+      "epoch": 0.4303585203072073,
+      "grad_norm": 0.1510670254529116,
+      "learning_rate": 2e-05,
+      "loss": 5.3653,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4304255961364322,
+      "grad_norm": 0.14643493882806216,
+      "learning_rate": 2e-05,
+      "loss": 5.3166,
+      "step": 6417
+    },
+    {
+      "epoch": 0.43049267196565716,
+      "grad_norm": 0.14985285621780203,
+      "learning_rate": 2e-05,
+      "loss": 5.3932,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4305597477948821,
+      "grad_norm": 0.1467991038780593,
+      "learning_rate": 2e-05,
+      "loss": 5.4269,
+      "step": 6419
+    },
+    {
+      "epoch": 0.43062682362410704,
+      "grad_norm": 0.15460907464710014,
+      "learning_rate": 2e-05,
+      "loss": 5.4842,
+      "step": 6420
+    },
+    {
+      "epoch": 0.430693899453332,
+      "grad_norm": 0.15710265971778103,
+      "learning_rate": 2e-05,
+      "loss": 5.339,
+      "step": 6421
+    },
+    {
+      "epoch": 0.4307609752825569,
+      "grad_norm": 0.15701538159094788,
+      "learning_rate": 2e-05,
+      "loss": 5.367,
+      "step": 6422
+    },
+    {
+      "epoch": 0.43082805111178185,
+      "grad_norm": 0.15376326424912343,
+      "learning_rate": 2e-05,
+      "loss": 5.4265,
+      "step": 6423
+    },
+    {
+      "epoch": 0.4308951269410068,
+      "grad_norm": 0.15355941175582652,
+      "learning_rate": 2e-05,
+      "loss": 5.3635,
+      "step": 6424
+    },
+    {
+      "epoch": 0.43096220277023173,
+      "grad_norm": 0.15126633900874148,
+      "learning_rate": 2e-05,
+      "loss": 5.3284,
+      "step": 6425
+    },
+    {
+      "epoch": 0.43102927859945667,
+      "grad_norm": 0.15409141857616335,
+      "learning_rate": 2e-05,
+      "loss": 5.3698,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4310963544286816,
+      "grad_norm": 0.14770617826253574,
+      "learning_rate": 2e-05,
+      "loss": 5.4317,
+      "step": 6427
+    },
+    {
+      "epoch": 0.43116343025790654,
+      "grad_norm": 0.1622167039033906,
+      "learning_rate": 2e-05,
+      "loss": 5.4944,
+      "step": 6428
+    },
+    {
+      "epoch": 0.4312305060871315,
+      "grad_norm": 0.14485107421766014,
+      "learning_rate": 2e-05,
+      "loss": 5.5541,
+      "step": 6429
+    },
+    {
+      "epoch": 0.4312975819163564,
+      "grad_norm": 0.14596453018935882,
+      "learning_rate": 2e-05,
+      "loss": 5.4006,
+      "step": 6430
+    },
+    {
+      "epoch": 0.43136465774558136,
+      "grad_norm": 0.15839891096282108,
+      "learning_rate": 2e-05,
+      "loss": 5.416,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4314317335748063,
+      "grad_norm": 0.1527165643426115,
+      "learning_rate": 2e-05,
+      "loss": 5.2916,
+      "step": 6432
+    },
+    {
+      "epoch": 0.43149880940403124,
+      "grad_norm": 0.15332935619037338,
+      "learning_rate": 2e-05,
+      "loss": 5.348,
+      "step": 6433
+    },
+    {
+      "epoch": 0.4315658852332562,
+      "grad_norm": 0.14979485726620326,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 6434
+    },
+    {
+      "epoch": 0.4316329610624811,
+      "grad_norm": 0.14712689426092052,
+      "learning_rate": 2e-05,
+      "loss": 5.4517,
+      "step": 6435
+    },
+    {
+      "epoch": 0.43170003689170605,
+      "grad_norm": 0.15328383600397746,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 6436
+    },
+    {
+      "epoch": 0.431767112720931,
+      "grad_norm": 0.1489991525094696,
+      "learning_rate": 2e-05,
+      "loss": 5.5447,
+      "step": 6437
+    },
+    {
+      "epoch": 0.43183418855015593,
+      "grad_norm": 0.14363200920344524,
+      "learning_rate": 2e-05,
+      "loss": 5.366,
+      "step": 6438
+    },
+    {
+      "epoch": 0.43190126437938087,
+      "grad_norm": 0.14443407148890078,
+      "learning_rate": 2e-05,
+      "loss": 5.3599,
+      "step": 6439
+    },
+    {
+      "epoch": 0.4319683402086058,
+      "grad_norm": 0.144891659748065,
+      "learning_rate": 2e-05,
+      "loss": 5.5854,
+      "step": 6440
+    },
+    {
+      "epoch": 0.43203541603783074,
+      "grad_norm": 0.1485065639946896,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 6441
+    },
+    {
+      "epoch": 0.4321024918670557,
+      "grad_norm": 0.14810432321507336,
+      "learning_rate": 2e-05,
+      "loss": 5.415,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4321695676962806,
+      "grad_norm": 0.1466022937421589,
+      "learning_rate": 2e-05,
+      "loss": 5.602,
+      "step": 6443
+    },
+    {
+      "epoch": 0.43223664352550556,
+      "grad_norm": 0.14297251774357253,
+      "learning_rate": 2e-05,
+      "loss": 5.6131,
+      "step": 6444
+    },
+    {
+      "epoch": 0.4323037193547305,
+      "grad_norm": 0.1518150357426711,
+      "learning_rate": 2e-05,
+      "loss": 5.4819,
+      "step": 6445
+    },
+    {
+      "epoch": 0.43237079518395544,
+      "grad_norm": 0.14803697992867357,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4324378710131804,
+      "grad_norm": 0.14926622405614468,
+      "learning_rate": 2e-05,
+      "loss": 5.3633,
+      "step": 6447
+    },
+    {
+      "epoch": 0.4325049468424053,
+      "grad_norm": 0.14268681145376752,
+      "learning_rate": 2e-05,
+      "loss": 5.3373,
+      "step": 6448
+    },
+    {
+      "epoch": 0.43257202267163025,
+      "grad_norm": 0.14327260077481244,
+      "learning_rate": 2e-05,
+      "loss": 5.4831,
+      "step": 6449
+    },
+    {
+      "epoch": 0.4326390985008552,
+      "grad_norm": 0.1471321941779846,
+      "learning_rate": 2e-05,
+      "loss": 5.3186,
+      "step": 6450
+    },
+    {
+      "epoch": 0.43270617433008013,
+      "grad_norm": 0.15236364389787693,
+      "learning_rate": 2e-05,
+      "loss": 5.3837,
+      "step": 6451
+    },
+    {
+      "epoch": 0.43277325015930507,
+      "grad_norm": 0.1484765755628171,
+      "learning_rate": 2e-05,
+      "loss": 5.5478,
+      "step": 6452
+    },
+    {
+      "epoch": 0.43284032598853,
+      "grad_norm": 0.15094969861886232,
+      "learning_rate": 2e-05,
+      "loss": 5.4776,
+      "step": 6453
+    },
+    {
+      "epoch": 0.43290740181775494,
+      "grad_norm": 0.15007158354212266,
+      "learning_rate": 2e-05,
+      "loss": 5.3285,
+      "step": 6454
+    },
+    {
+      "epoch": 0.43297447764697994,
+      "grad_norm": 0.1482849425774349,
+      "learning_rate": 2e-05,
+      "loss": 5.3695,
+      "step": 6455
+    },
+    {
+      "epoch": 0.4330415534762049,
+      "grad_norm": 0.14078213246126908,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 6456
+    },
+    {
+      "epoch": 0.4331086293054298,
+      "grad_norm": 0.1492265619889194,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 6457
+    },
+    {
+      "epoch": 0.43317570513465475,
+      "grad_norm": 0.15498551491931645,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 6458
+    },
+    {
+      "epoch": 0.4332427809638797,
+      "grad_norm": 0.15429732876846353,
+      "learning_rate": 2e-05,
+      "loss": 5.4166,
+      "step": 6459
+    },
+    {
+      "epoch": 0.43330985679310463,
+      "grad_norm": 0.1541213168498063,
+      "learning_rate": 2e-05,
+      "loss": 5.483,
+      "step": 6460
+    },
+    {
+      "epoch": 0.43337693262232957,
+      "grad_norm": 0.14344646607348122,
+      "learning_rate": 2e-05,
+      "loss": 5.2714,
+      "step": 6461
+    },
+    {
+      "epoch": 0.4334440084515545,
+      "grad_norm": 0.14830674169560729,
+      "learning_rate": 2e-05,
+      "loss": 5.6067,
+      "step": 6462
+    },
+    {
+      "epoch": 0.43351108428077945,
+      "grad_norm": 0.1523386618868509,
+      "learning_rate": 2e-05,
+      "loss": 5.4715,
+      "step": 6463
+    },
+    {
+      "epoch": 0.4335781601100044,
+      "grad_norm": 0.17719785006734695,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 6464
+    },
+    {
+      "epoch": 0.4336452359392293,
+      "grad_norm": 0.15559283243082037,
+      "learning_rate": 2e-05,
+      "loss": 5.4168,
+      "step": 6465
+    },
+    {
+      "epoch": 0.43371231176845426,
+      "grad_norm": 0.14910494517060077,
+      "learning_rate": 2e-05,
+      "loss": 5.3749,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4337793875976792,
+      "grad_norm": 0.14835677456319674,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 6467
+    },
+    {
+      "epoch": 0.43384646342690414,
+      "grad_norm": 0.1504196016445764,
+      "learning_rate": 2e-05,
+      "loss": 5.4627,
+      "step": 6468
+    },
+    {
+      "epoch": 0.4339135392561291,
+      "grad_norm": 0.1473813709397406,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 6469
+    },
+    {
+      "epoch": 0.433980615085354,
+      "grad_norm": 0.1511866008539804,
+      "learning_rate": 2e-05,
+      "loss": 5.3579,
+      "step": 6470
+    },
+    {
+      "epoch": 0.43404769091457895,
+      "grad_norm": 0.1495947706842585,
+      "learning_rate": 2e-05,
+      "loss": 5.3498,
+      "step": 6471
+    },
+    {
+      "epoch": 0.4341147667438039,
+      "grad_norm": 0.14203377167250106,
+      "learning_rate": 2e-05,
+      "loss": 5.2866,
+      "step": 6472
+    },
+    {
+      "epoch": 0.43418184257302883,
+      "grad_norm": 0.14415821735177195,
+      "learning_rate": 2e-05,
+      "loss": 5.4387,
+      "step": 6473
+    },
+    {
+      "epoch": 0.43424891840225377,
+      "grad_norm": 0.1431905738346033,
+      "learning_rate": 2e-05,
+      "loss": 5.4473,
+      "step": 6474
+    },
+    {
+      "epoch": 0.4343159942314787,
+      "grad_norm": 0.15292637277025203,
+      "learning_rate": 2e-05,
+      "loss": 5.4572,
+      "step": 6475
+    },
+    {
+      "epoch": 0.43438307006070365,
+      "grad_norm": 0.15599992981061672,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 6476
+    },
+    {
+      "epoch": 0.4344501458899286,
+      "grad_norm": 0.15095865002659434,
+      "learning_rate": 2e-05,
+      "loss": 5.4943,
+      "step": 6477
+    },
+    {
+      "epoch": 0.4345172217191535,
+      "grad_norm": 0.1486092603320759,
+      "learning_rate": 2e-05,
+      "loss": 5.5389,
+      "step": 6478
+    },
+    {
+      "epoch": 0.43458429754837846,
+      "grad_norm": 0.15130782165134798,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 6479
+    },
+    {
+      "epoch": 0.4346513733776034,
+      "grad_norm": 0.15282293421995943,
+      "learning_rate": 2e-05,
+      "loss": 5.4994,
+      "step": 6480
+    },
+    {
+      "epoch": 0.43471844920682834,
+      "grad_norm": 0.14666742082905437,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 6481
+    },
+    {
+      "epoch": 0.4347855250360533,
+      "grad_norm": 0.1550133982274994,
+      "learning_rate": 2e-05,
+      "loss": 5.4567,
+      "step": 6482
+    },
+    {
+      "epoch": 0.4348526008652782,
+      "grad_norm": 0.14962044434609417,
+      "learning_rate": 2e-05,
+      "loss": 5.4674,
+      "step": 6483
+    },
+    {
+      "epoch": 0.43491967669450315,
+      "grad_norm": 0.14995713809178896,
+      "learning_rate": 2e-05,
+      "loss": 5.3438,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4349867525237281,
+      "grad_norm": 0.1462576779334566,
+      "learning_rate": 2e-05,
+      "loss": 5.3963,
+      "step": 6485
+    },
+    {
+      "epoch": 0.43505382835295303,
+      "grad_norm": 0.14641371939113967,
+      "learning_rate": 2e-05,
+      "loss": 5.636,
+      "step": 6486
+    },
+    {
+      "epoch": 0.43512090418217797,
+      "grad_norm": 0.14976000061783576,
+      "learning_rate": 2e-05,
+      "loss": 5.4595,
+      "step": 6487
+    },
+    {
+      "epoch": 0.4351879800114029,
+      "grad_norm": 0.1465188810384411,
+      "learning_rate": 2e-05,
+      "loss": 5.3589,
+      "step": 6488
+    },
+    {
+      "epoch": 0.43525505584062785,
+      "grad_norm": 0.1467777269924352,
+      "learning_rate": 2e-05,
+      "loss": 5.563,
+      "step": 6489
+    },
+    {
+      "epoch": 0.4353221316698528,
+      "grad_norm": 0.14719069998299456,
+      "learning_rate": 2e-05,
+      "loss": 5.4876,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4353892074990777,
+      "grad_norm": 0.14905333933902756,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 6491
+    },
+    {
+      "epoch": 0.43545628332830266,
+      "grad_norm": 0.16115701880204109,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 6492
+    },
+    {
+      "epoch": 0.4355233591575276,
+      "grad_norm": 0.15439904773209534,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 6493
+    },
+    {
+      "epoch": 0.43559043498675254,
+      "grad_norm": 0.15388884718054893,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 6494
+    },
+    {
+      "epoch": 0.4356575108159775,
+      "grad_norm": 0.14936544203231644,
+      "learning_rate": 2e-05,
+      "loss": 5.5318,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4357245866452024,
+      "grad_norm": 0.14723547816319466,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 6496
+    },
+    {
+      "epoch": 0.43579166247442735,
+      "grad_norm": 0.15809724177393877,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 6497
+    },
+    {
+      "epoch": 0.4358587383036523,
+      "grad_norm": 0.1452806178436295,
+      "learning_rate": 2e-05,
+      "loss": 5.3799,
+      "step": 6498
+    },
+    {
+      "epoch": 0.43592581413287723,
+      "grad_norm": 0.1409622576039092,
+      "learning_rate": 2e-05,
+      "loss": 5.306,
+      "step": 6499
+    },
+    {
+      "epoch": 0.43599288996210217,
+      "grad_norm": 0.1463196957721756,
+      "learning_rate": 2e-05,
+      "loss": 5.4365,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4360599657913271,
+      "grad_norm": 0.14430325126467639,
+      "learning_rate": 2e-05,
+      "loss": 5.313,
+      "step": 6501
+    },
+    {
+      "epoch": 0.43612704162055205,
+      "grad_norm": 0.15007517797316197,
+      "learning_rate": 2e-05,
+      "loss": 5.3587,
+      "step": 6502
+    },
+    {
+      "epoch": 0.436194117449777,
+      "grad_norm": 0.14310658126585693,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 6503
+    },
+    {
+      "epoch": 0.4362611932790019,
+      "grad_norm": 0.15042199815199803,
+      "learning_rate": 2e-05,
+      "loss": 5.3219,
+      "step": 6504
+    },
+    {
+      "epoch": 0.43632826910822686,
+      "grad_norm": 0.15125545114388883,
+      "learning_rate": 2e-05,
+      "loss": 5.5031,
+      "step": 6505
+    },
+    {
+      "epoch": 0.4363953449374518,
+      "grad_norm": 0.14299969768165352,
+      "learning_rate": 2e-05,
+      "loss": 5.3501,
+      "step": 6506
+    },
+    {
+      "epoch": 0.43646242076667674,
+      "grad_norm": 0.1452451071863212,
+      "learning_rate": 2e-05,
+      "loss": 5.525,
+      "step": 6507
+    },
+    {
+      "epoch": 0.4365294965959017,
+      "grad_norm": 0.14930959899317486,
+      "learning_rate": 2e-05,
+      "loss": 5.4911,
+      "step": 6508
+    },
+    {
+      "epoch": 0.4365965724251266,
+      "grad_norm": 0.14047390647584607,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 6509
+    },
+    {
+      "epoch": 0.43666364825435156,
+      "grad_norm": 0.14901258047951998,
+      "learning_rate": 2e-05,
+      "loss": 5.464,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4367307240835765,
+      "grad_norm": 0.15298619110651454,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 6511
+    },
+    {
+      "epoch": 0.43679779991280143,
+      "grad_norm": 0.14570068922637158,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 6512
+    },
+    {
+      "epoch": 0.43686487574202637,
+      "grad_norm": 0.1523595927264857,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 6513
+    },
+    {
+      "epoch": 0.4369319515712513,
+      "grad_norm": 0.15238609487331148,
+      "learning_rate": 2e-05,
+      "loss": 5.3042,
+      "step": 6514
+    },
+    {
+      "epoch": 0.43699902740047625,
+      "grad_norm": 0.1488463714195997,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 6515
+    },
+    {
+      "epoch": 0.4370661032297012,
+      "grad_norm": 0.14770437396759872,
+      "learning_rate": 2e-05,
+      "loss": 5.423,
+      "step": 6516
+    },
+    {
+      "epoch": 0.4371331790589261,
+      "grad_norm": 0.14454179724666164,
+      "learning_rate": 2e-05,
+      "loss": 5.3732,
+      "step": 6517
+    },
+    {
+      "epoch": 0.43720025488815106,
+      "grad_norm": 0.14851305582091523,
+      "learning_rate": 2e-05,
+      "loss": 5.3884,
+      "step": 6518
+    },
+    {
+      "epoch": 0.437267330717376,
+      "grad_norm": 0.14038262755243677,
+      "learning_rate": 2e-05,
+      "loss": 5.5154,
+      "step": 6519
+    },
+    {
+      "epoch": 0.43733440654660094,
+      "grad_norm": 0.14600285379994998,
+      "learning_rate": 2e-05,
+      "loss": 5.5234,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4374014823758259,
+      "grad_norm": 0.1662547744434856,
+      "learning_rate": 2e-05,
+      "loss": 5.3796,
+      "step": 6521
+    },
+    {
+      "epoch": 0.4374685582050508,
+      "grad_norm": 0.15084473245905944,
+      "learning_rate": 2e-05,
+      "loss": 5.3938,
+      "step": 6522
+    },
+    {
+      "epoch": 0.43753563403427576,
+      "grad_norm": 0.1470581665265498,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 6523
+    },
+    {
+      "epoch": 0.4376027098635007,
+      "grad_norm": 0.15348952143952527,
+      "learning_rate": 2e-05,
+      "loss": 5.4175,
+      "step": 6524
+    },
+    {
+      "epoch": 0.43766978569272563,
+      "grad_norm": 0.1499339267554661,
+      "learning_rate": 2e-05,
+      "loss": 5.2976,
+      "step": 6525
+    },
+    {
+      "epoch": 0.43773686152195057,
+      "grad_norm": 0.14902037168569038,
+      "learning_rate": 2e-05,
+      "loss": 5.5762,
+      "step": 6526
+    },
+    {
+      "epoch": 0.4378039373511755,
+      "grad_norm": 0.15127749371146954,
+      "learning_rate": 2e-05,
+      "loss": 5.2917,
+      "step": 6527
+    },
+    {
+      "epoch": 0.43787101318040045,
+      "grad_norm": 0.15370515151762248,
+      "learning_rate": 2e-05,
+      "loss": 5.2111,
+      "step": 6528
+    },
+    {
+      "epoch": 0.4379380890096254,
+      "grad_norm": 0.1427975743863519,
+      "learning_rate": 2e-05,
+      "loss": 5.3715,
+      "step": 6529
+    },
+    {
+      "epoch": 0.4380051648388503,
+      "grad_norm": 0.1468642762682309,
+      "learning_rate": 2e-05,
+      "loss": 5.4783,
+      "step": 6530
+    },
+    {
+      "epoch": 0.43807224066807526,
+      "grad_norm": 0.14758104624157203,
+      "learning_rate": 2e-05,
+      "loss": 5.3457,
+      "step": 6531
+    },
+    {
+      "epoch": 0.4381393164973002,
+      "grad_norm": 0.16034040238928493,
+      "learning_rate": 2e-05,
+      "loss": 5.4666,
+      "step": 6532
+    },
+    {
+      "epoch": 0.43820639232652514,
+      "grad_norm": 0.1613649878780952,
+      "learning_rate": 2e-05,
+      "loss": 5.5423,
+      "step": 6533
+    },
+    {
+      "epoch": 0.4382734681557501,
+      "grad_norm": 0.15086251310094687,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 6534
+    },
+    {
+      "epoch": 0.438340543984975,
+      "grad_norm": 0.14440208073266078,
+      "learning_rate": 2e-05,
+      "loss": 5.3114,
+      "step": 6535
+    },
+    {
+      "epoch": 0.43840761981419996,
+      "grad_norm": 0.1477984132465928,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 6536
+    },
+    {
+      "epoch": 0.4384746956434249,
+      "grad_norm": 0.1574625039631118,
+      "learning_rate": 2e-05,
+      "loss": 5.5576,
+      "step": 6537
+    },
+    {
+      "epoch": 0.43854177147264983,
+      "grad_norm": 0.14927344219827374,
+      "learning_rate": 2e-05,
+      "loss": 5.4314,
+      "step": 6538
+    },
+    {
+      "epoch": 0.43860884730187477,
+      "grad_norm": 0.15343294613825711,
+      "learning_rate": 2e-05,
+      "loss": 5.4566,
+      "step": 6539
+    },
+    {
+      "epoch": 0.4386759231310997,
+      "grad_norm": 0.15106177547644953,
+      "learning_rate": 2e-05,
+      "loss": 5.268,
+      "step": 6540
+    },
+    {
+      "epoch": 0.43874299896032465,
+      "grad_norm": 0.14839898937449458,
+      "learning_rate": 2e-05,
+      "loss": 5.3612,
+      "step": 6541
+    },
+    {
+      "epoch": 0.4388100747895496,
+      "grad_norm": 0.1452975566917359,
+      "learning_rate": 2e-05,
+      "loss": 5.378,
+      "step": 6542
+    },
+    {
+      "epoch": 0.4388771506187745,
+      "grad_norm": 0.1528872497564754,
+      "learning_rate": 2e-05,
+      "loss": 5.5507,
+      "step": 6543
+    },
+    {
+      "epoch": 0.43894422644799946,
+      "grad_norm": 0.14814977167338475,
+      "learning_rate": 2e-05,
+      "loss": 5.4942,
+      "step": 6544
+    },
+    {
+      "epoch": 0.4390113022772244,
+      "grad_norm": 0.15630813117447975,
+      "learning_rate": 2e-05,
+      "loss": 5.4068,
+      "step": 6545
+    },
+    {
+      "epoch": 0.43907837810644934,
+      "grad_norm": 0.14504560674121356,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 6546
+    },
+    {
+      "epoch": 0.4391454539356743,
+      "grad_norm": 0.1537792325928432,
+      "learning_rate": 2e-05,
+      "loss": 5.4435,
+      "step": 6547
+    },
+    {
+      "epoch": 0.4392125297648992,
+      "grad_norm": 0.14606777762464698,
+      "learning_rate": 2e-05,
+      "loss": 5.39,
+      "step": 6548
+    },
+    {
+      "epoch": 0.43927960559412416,
+      "grad_norm": 0.15548643503207374,
+      "learning_rate": 2e-05,
+      "loss": 5.439,
+      "step": 6549
+    },
+    {
+      "epoch": 0.4393466814233491,
+      "grad_norm": 0.15103014263965145,
+      "learning_rate": 2e-05,
+      "loss": 5.4321,
+      "step": 6550
+    },
+    {
+      "epoch": 0.43941375725257403,
+      "grad_norm": 0.15093954075195562,
+      "learning_rate": 2e-05,
+      "loss": 5.3862,
+      "step": 6551
+    },
+    {
+      "epoch": 0.43948083308179897,
+      "grad_norm": 0.15037246276553526,
+      "learning_rate": 2e-05,
+      "loss": 5.3707,
+      "step": 6552
+    },
+    {
+      "epoch": 0.4395479089110239,
+      "grad_norm": 0.15247950946008182,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 6553
+    },
+    {
+      "epoch": 0.43961498474024885,
+      "grad_norm": 0.15038146258967283,
+      "learning_rate": 2e-05,
+      "loss": 5.3753,
+      "step": 6554
+    },
+    {
+      "epoch": 0.4396820605694738,
+      "grad_norm": 0.14980913944491045,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4397491363986987,
+      "grad_norm": 0.1491539124173332,
+      "learning_rate": 2e-05,
+      "loss": 5.4644,
+      "step": 6556
+    },
+    {
+      "epoch": 0.43981621222792366,
+      "grad_norm": 0.14781749003329037,
+      "learning_rate": 2e-05,
+      "loss": 5.3387,
+      "step": 6557
+    },
+    {
+      "epoch": 0.4398832880571486,
+      "grad_norm": 0.1580412696537096,
+      "learning_rate": 2e-05,
+      "loss": 5.5622,
+      "step": 6558
+    },
+    {
+      "epoch": 0.43995036388637354,
+      "grad_norm": 0.15187060333007285,
+      "learning_rate": 2e-05,
+      "loss": 5.3755,
+      "step": 6559
+    },
+    {
+      "epoch": 0.4400174397155985,
+      "grad_norm": 0.1432836723540337,
+      "learning_rate": 2e-05,
+      "loss": 5.5426,
+      "step": 6560
+    },
+    {
+      "epoch": 0.4400845155448234,
+      "grad_norm": 0.15986158637794615,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 6561
+    },
+    {
+      "epoch": 0.44015159137404836,
+      "grad_norm": 0.15896925265072182,
+      "learning_rate": 2e-05,
+      "loss": 5.4009,
+      "step": 6562
+    },
+    {
+      "epoch": 0.4402186672032733,
+      "grad_norm": 0.15357300265078128,
+      "learning_rate": 2e-05,
+      "loss": 5.4508,
+      "step": 6563
+    },
+    {
+      "epoch": 0.44028574303249823,
+      "grad_norm": 0.152656001905032,
+      "learning_rate": 2e-05,
+      "loss": 5.2975,
+      "step": 6564
+    },
+    {
+      "epoch": 0.44035281886172317,
+      "grad_norm": 0.1608679636651188,
+      "learning_rate": 2e-05,
+      "loss": 5.4545,
+      "step": 6565
+    },
+    {
+      "epoch": 0.4404198946909481,
+      "grad_norm": 0.152795132149424,
+      "learning_rate": 2e-05,
+      "loss": 5.4073,
+      "step": 6566
+    },
+    {
+      "epoch": 0.44048697052017305,
+      "grad_norm": 0.1653384127840587,
+      "learning_rate": 2e-05,
+      "loss": 5.5434,
+      "step": 6567
+    },
+    {
+      "epoch": 0.440554046349398,
+      "grad_norm": 0.1539568031670306,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 6568
+    },
+    {
+      "epoch": 0.4406211221786229,
+      "grad_norm": 0.15488489013475615,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 6569
+    },
+    {
+      "epoch": 0.44068819800784786,
+      "grad_norm": 0.14798659625317692,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4407552738370728,
+      "grad_norm": 0.14602038084829724,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 6571
+    },
+    {
+      "epoch": 0.44082234966629774,
+      "grad_norm": 0.15183956923687641,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 6572
+    },
+    {
+      "epoch": 0.4408894254955227,
+      "grad_norm": 0.15177138316041358,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 6573
+    },
+    {
+      "epoch": 0.4409565013247476,
+      "grad_norm": 0.14919998202441334,
+      "learning_rate": 2e-05,
+      "loss": 5.4998,
+      "step": 6574
+    },
+    {
+      "epoch": 0.44102357715397256,
+      "grad_norm": 0.15182148395386746,
+      "learning_rate": 2e-05,
+      "loss": 5.4422,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4410906529831975,
+      "grad_norm": 0.15089528183703504,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 6576
+    },
+    {
+      "epoch": 0.44115772881242243,
+      "grad_norm": 0.1517086275742126,
+      "learning_rate": 2e-05,
+      "loss": 5.4344,
+      "step": 6577
+    },
+    {
+      "epoch": 0.4412248046416474,
+      "grad_norm": 0.16172123307532343,
+      "learning_rate": 2e-05,
+      "loss": 5.4671,
+      "step": 6578
+    },
+    {
+      "epoch": 0.4412918804708723,
+      "grad_norm": 0.15396792642323331,
+      "learning_rate": 2e-05,
+      "loss": 5.4733,
+      "step": 6579
+    },
+    {
+      "epoch": 0.44135895630009725,
+      "grad_norm": 0.15277016028578194,
+      "learning_rate": 2e-05,
+      "loss": 5.357,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4414260321293222,
+      "grad_norm": 0.1463870273372572,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4414931079585471,
+      "grad_norm": 0.15159104671233314,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 6582
+    },
+    {
+      "epoch": 0.44156018378777206,
+      "grad_norm": 0.14567407904338778,
+      "learning_rate": 2e-05,
+      "loss": 5.4093,
+      "step": 6583
+    },
+    {
+      "epoch": 0.441627259616997,
+      "grad_norm": 0.15287392129191243,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 6584
+    },
+    {
+      "epoch": 0.44169433544622194,
+      "grad_norm": 0.14788633372542784,
+      "learning_rate": 2e-05,
+      "loss": 5.4876,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4417614112754469,
+      "grad_norm": 0.148144864133421,
+      "learning_rate": 2e-05,
+      "loss": 5.6115,
+      "step": 6586
+    },
+    {
+      "epoch": 0.4418284871046718,
+      "grad_norm": 0.15444238012652672,
+      "learning_rate": 2e-05,
+      "loss": 5.5204,
+      "step": 6587
+    },
+    {
+      "epoch": 0.44189556293389676,
+      "grad_norm": 0.1571012716824601,
+      "learning_rate": 2e-05,
+      "loss": 5.5957,
+      "step": 6588
+    },
+    {
+      "epoch": 0.4419626387631217,
+      "grad_norm": 0.14195006591498083,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 6589
+    },
+    {
+      "epoch": 0.44202971459234663,
+      "grad_norm": 0.15175842101533307,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4420967904215716,
+      "grad_norm": 0.1493847203135792,
+      "learning_rate": 2e-05,
+      "loss": 5.468,
+      "step": 6591
+    },
+    {
+      "epoch": 0.4421638662507965,
+      "grad_norm": 0.14766716104716443,
+      "learning_rate": 2e-05,
+      "loss": 5.2206,
+      "step": 6592
+    },
+    {
+      "epoch": 0.44223094208002145,
+      "grad_norm": 0.14628343335823263,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 6593
+    },
+    {
+      "epoch": 0.4422980179092464,
+      "grad_norm": 0.148206000410834,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4423650937384713,
+      "grad_norm": 0.1533235544995044,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 6595
+    },
+    {
+      "epoch": 0.44243216956769627,
+      "grad_norm": 0.14423944046304926,
+      "learning_rate": 2e-05,
+      "loss": 5.4586,
+      "step": 6596
+    },
+    {
+      "epoch": 0.4424992453969212,
+      "grad_norm": 0.15122814754896363,
+      "learning_rate": 2e-05,
+      "loss": 5.4319,
+      "step": 6597
+    },
+    {
+      "epoch": 0.44256632122614614,
+      "grad_norm": 0.15347044874275947,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4426333970553711,
+      "grad_norm": 0.1552555881488444,
+      "learning_rate": 2e-05,
+      "loss": 5.2656,
+      "step": 6599
+    },
+    {
+      "epoch": 0.442700472884596,
+      "grad_norm": 0.14903007994141543,
+      "learning_rate": 2e-05,
+      "loss": 5.3637,
+      "step": 6600
+    },
+    {
+      "epoch": 0.44276754871382096,
+      "grad_norm": 0.15301671759935032,
+      "learning_rate": 2e-05,
+      "loss": 5.5157,
+      "step": 6601
+    },
+    {
+      "epoch": 0.4428346245430459,
+      "grad_norm": 0.16217529028592492,
+      "learning_rate": 2e-05,
+      "loss": 5.3721,
+      "step": 6602
+    },
+    {
+      "epoch": 0.44290170037227083,
+      "grad_norm": 0.1492554611312957,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 6603
+    },
+    {
+      "epoch": 0.4429687762014958,
+      "grad_norm": 0.14724820169561914,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4430358520307207,
+      "grad_norm": 0.15722310116303093,
+      "learning_rate": 2e-05,
+      "loss": 5.4769,
+      "step": 6605
+    },
+    {
+      "epoch": 0.44310292785994565,
+      "grad_norm": 0.15177971493888537,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4431700036891706,
+      "grad_norm": 0.1486480076807384,
+      "learning_rate": 2e-05,
+      "loss": 5.4301,
+      "step": 6607
+    },
+    {
+      "epoch": 0.4432370795183955,
+      "grad_norm": 0.1544667587987344,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 6608
+    },
+    {
+      "epoch": 0.44330415534762047,
+      "grad_norm": 0.14577529758454774,
+      "learning_rate": 2e-05,
+      "loss": 5.4974,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4433712311768454,
+      "grad_norm": 0.15151826044760752,
+      "learning_rate": 2e-05,
+      "loss": 5.5422,
+      "step": 6610
+    },
+    {
+      "epoch": 0.44343830700607034,
+      "grad_norm": 0.14732449594468477,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 6611
+    },
+    {
+      "epoch": 0.4435053828352953,
+      "grad_norm": 0.1451523064773278,
+      "learning_rate": 2e-05,
+      "loss": 5.5987,
+      "step": 6612
+    },
+    {
+      "epoch": 0.4435724586645202,
+      "grad_norm": 0.15571006021049275,
+      "learning_rate": 2e-05,
+      "loss": 5.4156,
+      "step": 6613
+    },
+    {
+      "epoch": 0.44363953449374516,
+      "grad_norm": 0.14665291146215828,
+      "learning_rate": 2e-05,
+      "loss": 5.196,
+      "step": 6614
+    },
+    {
+      "epoch": 0.4437066103229701,
+      "grad_norm": 0.14406411710919398,
+      "learning_rate": 2e-05,
+      "loss": 5.5187,
+      "step": 6615
+    },
+    {
+      "epoch": 0.44377368615219503,
+      "grad_norm": 0.15753104141748633,
+      "learning_rate": 2e-05,
+      "loss": 5.53,
+      "step": 6616
+    },
+    {
+      "epoch": 0.44384076198142,
+      "grad_norm": 0.15194025773441955,
+      "learning_rate": 2e-05,
+      "loss": 5.4815,
+      "step": 6617
+    },
+    {
+      "epoch": 0.4439078378106449,
+      "grad_norm": 0.1562071577079159,
+      "learning_rate": 2e-05,
+      "loss": 5.3143,
+      "step": 6618
+    },
+    {
+      "epoch": 0.44397491363986985,
+      "grad_norm": 0.15128631934668688,
+      "learning_rate": 2e-05,
+      "loss": 5.5328,
+      "step": 6619
+    },
+    {
+      "epoch": 0.4440419894690948,
+      "grad_norm": 0.15632108888487933,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4441090652983197,
+      "grad_norm": 0.15336313212275132,
+      "learning_rate": 2e-05,
+      "loss": 5.3664,
+      "step": 6621
+    },
+    {
+      "epoch": 0.44417614112754467,
+      "grad_norm": 0.1561996519035676,
+      "learning_rate": 2e-05,
+      "loss": 5.2853,
+      "step": 6622
+    },
+    {
+      "epoch": 0.4442432169567696,
+      "grad_norm": 0.14862237098363446,
+      "learning_rate": 2e-05,
+      "loss": 5.3588,
+      "step": 6623
+    },
+    {
+      "epoch": 0.44431029278599454,
+      "grad_norm": 0.15443687795325156,
+      "learning_rate": 2e-05,
+      "loss": 5.5403,
+      "step": 6624
+    },
+    {
+      "epoch": 0.4443773686152195,
+      "grad_norm": 0.15622178675636322,
+      "learning_rate": 2e-05,
+      "loss": 5.5337,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.14773188331480097,
+      "learning_rate": 2e-05,
+      "loss": 5.4149,
+      "step": 6626
+    },
+    {
+      "epoch": 0.44451152027366936,
+      "grad_norm": 0.14742614398500345,
+      "learning_rate": 2e-05,
+      "loss": 5.5661,
+      "step": 6627
+    },
+    {
+      "epoch": 0.4445785961028943,
+      "grad_norm": 0.14851724833893878,
+      "learning_rate": 2e-05,
+      "loss": 5.3201,
+      "step": 6628
+    },
+    {
+      "epoch": 0.44464567193211924,
+      "grad_norm": 0.15060657430155902,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 6629
+    },
+    {
+      "epoch": 0.4447127477613442,
+      "grad_norm": 0.14709491409047068,
+      "learning_rate": 2e-05,
+      "loss": 5.3822,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4447798235905691,
+      "grad_norm": 0.1459893070829904,
+      "learning_rate": 2e-05,
+      "loss": 5.412,
+      "step": 6631
+    },
+    {
+      "epoch": 0.44484689941979405,
+      "grad_norm": 0.14381043355326667,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 6632
+    },
+    {
+      "epoch": 0.444913975249019,
+      "grad_norm": 0.14698452418498303,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 6633
+    },
+    {
+      "epoch": 0.4449810510782439,
+      "grad_norm": 0.14356264084865447,
+      "learning_rate": 2e-05,
+      "loss": 5.2334,
+      "step": 6634
+    },
+    {
+      "epoch": 0.44504812690746887,
+      "grad_norm": 0.154447934251591,
+      "learning_rate": 2e-05,
+      "loss": 5.3944,
+      "step": 6635
+    },
+    {
+      "epoch": 0.4451152027366938,
+      "grad_norm": 0.15239359236927882,
+      "learning_rate": 2e-05,
+      "loss": 5.327,
+      "step": 6636
+    },
+    {
+      "epoch": 0.4451822785659188,
+      "grad_norm": 0.14594201532369264,
+      "learning_rate": 2e-05,
+      "loss": 5.4264,
+      "step": 6637
+    },
+    {
+      "epoch": 0.44524935439514374,
+      "grad_norm": 0.1506752930741795,
+      "learning_rate": 2e-05,
+      "loss": 5.3287,
+      "step": 6638
+    },
+    {
+      "epoch": 0.4453164302243687,
+      "grad_norm": 0.14816198753858636,
+      "learning_rate": 2e-05,
+      "loss": 5.4084,
+      "step": 6639
+    },
+    {
+      "epoch": 0.4453835060535936,
+      "grad_norm": 0.14410547625150685,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 6640
+    },
+    {
+      "epoch": 0.44545058188281855,
+      "grad_norm": 0.14613228580684476,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 6641
+    },
+    {
+      "epoch": 0.4455176577120435,
+      "grad_norm": 0.1481184147095058,
+      "learning_rate": 2e-05,
+      "loss": 5.3997,
+      "step": 6642
+    },
+    {
+      "epoch": 0.44558473354126843,
+      "grad_norm": 0.15588938931119636,
+      "learning_rate": 2e-05,
+      "loss": 5.3531,
+      "step": 6643
+    },
+    {
+      "epoch": 0.44565180937049337,
+      "grad_norm": 0.149203954360798,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 6644
+    },
+    {
+      "epoch": 0.4457188851997183,
+      "grad_norm": 0.1516249028898083,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 6645
+    },
+    {
+      "epoch": 0.44578596102894325,
+      "grad_norm": 0.14662033084046575,
+      "learning_rate": 2e-05,
+      "loss": 5.4291,
+      "step": 6646
+    },
+    {
+      "epoch": 0.4458530368581682,
+      "grad_norm": 0.144564878101825,
+      "learning_rate": 2e-05,
+      "loss": 5.4444,
+      "step": 6647
+    },
+    {
+      "epoch": 0.4459201126873931,
+      "grad_norm": 0.14591729091901104,
+      "learning_rate": 2e-05,
+      "loss": 5.5002,
+      "step": 6648
+    },
+    {
+      "epoch": 0.44598718851661806,
+      "grad_norm": 0.1566092777577826,
+      "learning_rate": 2e-05,
+      "loss": 5.3468,
+      "step": 6649
+    },
+    {
+      "epoch": 0.446054264345843,
+      "grad_norm": 0.1490968430319641,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 6650
+    },
+    {
+      "epoch": 0.44612134017506794,
+      "grad_norm": 0.1490571968351877,
+      "learning_rate": 2e-05,
+      "loss": 5.3728,
+      "step": 6651
+    },
+    {
+      "epoch": 0.4461884160042929,
+      "grad_norm": 0.15226288731932097,
+      "learning_rate": 2e-05,
+      "loss": 5.4961,
+      "step": 6652
+    },
+    {
+      "epoch": 0.4462554918335178,
+      "grad_norm": 0.15069540839649254,
+      "learning_rate": 2e-05,
+      "loss": 5.3484,
+      "step": 6653
+    },
+    {
+      "epoch": 0.44632256766274275,
+      "grad_norm": 0.15312319666116453,
+      "learning_rate": 2e-05,
+      "loss": 5.5559,
+      "step": 6654
+    },
+    {
+      "epoch": 0.4463896434919677,
+      "grad_norm": 0.1482282393806249,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 6655
+    },
+    {
+      "epoch": 0.44645671932119263,
+      "grad_norm": 0.14707636960585865,
+      "learning_rate": 2e-05,
+      "loss": 5.4523,
+      "step": 6656
+    },
+    {
+      "epoch": 0.44652379515041757,
+      "grad_norm": 0.16539437872407012,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 6657
+    },
+    {
+      "epoch": 0.4465908709796425,
+      "grad_norm": 0.15399100925060719,
+      "learning_rate": 2e-05,
+      "loss": 5.3626,
+      "step": 6658
+    },
+    {
+      "epoch": 0.44665794680886745,
+      "grad_norm": 0.14583619880442414,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 6659
+    },
+    {
+      "epoch": 0.4467250226380924,
+      "grad_norm": 0.15921102018043792,
+      "learning_rate": 2e-05,
+      "loss": 5.4614,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4467920984673173,
+      "grad_norm": 0.14617146686480678,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 6661
+    },
+    {
+      "epoch": 0.44685917429654226,
+      "grad_norm": 0.14846827221132186,
+      "learning_rate": 2e-05,
+      "loss": 5.4321,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4469262501257672,
+      "grad_norm": 0.14977002561281297,
+      "learning_rate": 2e-05,
+      "loss": 5.2613,
+      "step": 6663
+    },
+    {
+      "epoch": 0.44699332595499214,
+      "grad_norm": 0.1533965718609225,
+      "learning_rate": 2e-05,
+      "loss": 5.4142,
+      "step": 6664
+    },
+    {
+      "epoch": 0.4470604017842171,
+      "grad_norm": 0.1543869201892017,
+      "learning_rate": 2e-05,
+      "loss": 5.6953,
+      "step": 6665
+    },
+    {
+      "epoch": 0.447127477613442,
+      "grad_norm": 0.15190651132875177,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 6666
+    },
+    {
+      "epoch": 0.44719455344266695,
+      "grad_norm": 0.15244172044299725,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 6667
+    },
+    {
+      "epoch": 0.4472616292718919,
+      "grad_norm": 0.14882740896208693,
+      "learning_rate": 2e-05,
+      "loss": 5.6013,
+      "step": 6668
+    },
+    {
+      "epoch": 0.44732870510111683,
+      "grad_norm": 0.149972826343418,
+      "learning_rate": 2e-05,
+      "loss": 5.4807,
+      "step": 6669
+    },
+    {
+      "epoch": 0.44739578093034177,
+      "grad_norm": 0.15139272906864057,
+      "learning_rate": 2e-05,
+      "loss": 5.5085,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4474628567595667,
+      "grad_norm": 0.14986785548922218,
+      "learning_rate": 2e-05,
+      "loss": 5.4266,
+      "step": 6671
+    },
+    {
+      "epoch": 0.44752993258879165,
+      "grad_norm": 0.14820078598781897,
+      "learning_rate": 2e-05,
+      "loss": 5.4291,
+      "step": 6672
+    },
+    {
+      "epoch": 0.4475970084180166,
+      "grad_norm": 0.15623121689510372,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 6673
+    },
+    {
+      "epoch": 0.4476640842472415,
+      "grad_norm": 0.14960149526230757,
+      "learning_rate": 2e-05,
+      "loss": 5.5171,
+      "step": 6674
+    },
+    {
+      "epoch": 0.44773116007646646,
+      "grad_norm": 0.157126873968632,
+      "learning_rate": 2e-05,
+      "loss": 5.3985,
+      "step": 6675
+    },
+    {
+      "epoch": 0.4477982359056914,
+      "grad_norm": 0.15025003725263988,
+      "learning_rate": 2e-05,
+      "loss": 5.3939,
+      "step": 6676
+    },
+    {
+      "epoch": 0.44786531173491634,
+      "grad_norm": 0.15522741177620403,
+      "learning_rate": 2e-05,
+      "loss": 5.6191,
+      "step": 6677
+    },
+    {
+      "epoch": 0.4479323875641413,
+      "grad_norm": 0.1604623257892669,
+      "learning_rate": 2e-05,
+      "loss": 5.3341,
+      "step": 6678
+    },
+    {
+      "epoch": 0.4479994633933662,
+      "grad_norm": 0.15533784618387583,
+      "learning_rate": 2e-05,
+      "loss": 5.3595,
+      "step": 6679
+    },
+    {
+      "epoch": 0.44806653922259115,
+      "grad_norm": 0.1587687647845349,
+      "learning_rate": 2e-05,
+      "loss": 5.3895,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4481336150518161,
+      "grad_norm": 0.15805917726926447,
+      "learning_rate": 2e-05,
+      "loss": 5.3696,
+      "step": 6681
+    },
+    {
+      "epoch": 0.44820069088104103,
+      "grad_norm": 0.16132127236066854,
+      "learning_rate": 2e-05,
+      "loss": 5.4513,
+      "step": 6682
+    },
+    {
+      "epoch": 0.44826776671026597,
+      "grad_norm": 0.16199038202555024,
+      "learning_rate": 2e-05,
+      "loss": 5.3482,
+      "step": 6683
+    },
+    {
+      "epoch": 0.4483348425394909,
+      "grad_norm": 0.15773989836509403,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 6684
+    },
+    {
+      "epoch": 0.44840191836871585,
+      "grad_norm": 0.16537203932462535,
+      "learning_rate": 2e-05,
+      "loss": 5.5128,
+      "step": 6685
+    },
+    {
+      "epoch": 0.4484689941979408,
+      "grad_norm": 0.15473641644344394,
+      "learning_rate": 2e-05,
+      "loss": 5.4736,
+      "step": 6686
+    },
+    {
+      "epoch": 0.4485360700271657,
+      "grad_norm": 0.15995678927274165,
+      "learning_rate": 2e-05,
+      "loss": 5.3176,
+      "step": 6687
+    },
+    {
+      "epoch": 0.44860314585639066,
+      "grad_norm": 0.14992014956088429,
+      "learning_rate": 2e-05,
+      "loss": 5.3749,
+      "step": 6688
+    },
+    {
+      "epoch": 0.4486702216856156,
+      "grad_norm": 0.15060858864674606,
+      "learning_rate": 2e-05,
+      "loss": 5.2935,
+      "step": 6689
+    },
+    {
+      "epoch": 0.44873729751484054,
+      "grad_norm": 0.14958871032634782,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4488043733440655,
+      "grad_norm": 0.14305401649297447,
+      "learning_rate": 2e-05,
+      "loss": 5.3295,
+      "step": 6691
+    },
+    {
+      "epoch": 0.4488714491732904,
+      "grad_norm": 0.1477895959811945,
+      "learning_rate": 2e-05,
+      "loss": 5.4475,
+      "step": 6692
+    },
+    {
+      "epoch": 0.44893852500251535,
+      "grad_norm": 0.1524956871448201,
+      "learning_rate": 2e-05,
+      "loss": 5.4578,
+      "step": 6693
+    },
+    {
+      "epoch": 0.4490056008317403,
+      "grad_norm": 0.14719777161321448,
+      "learning_rate": 2e-05,
+      "loss": 5.482,
+      "step": 6694
+    },
+    {
+      "epoch": 0.44907267666096523,
+      "grad_norm": 0.15556864685203514,
+      "learning_rate": 2e-05,
+      "loss": 5.5769,
+      "step": 6695
+    },
+    {
+      "epoch": 0.44913975249019017,
+      "grad_norm": 0.1645124007062323,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 6696
+    },
+    {
+      "epoch": 0.4492068283194151,
+      "grad_norm": 0.1518412650160944,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 6697
+    },
+    {
+      "epoch": 0.44927390414864005,
+      "grad_norm": 0.15197246649475754,
+      "learning_rate": 2e-05,
+      "loss": 5.4547,
+      "step": 6698
+    },
+    {
+      "epoch": 0.449340979977865,
+      "grad_norm": 0.1595081676304039,
+      "learning_rate": 2e-05,
+      "loss": 5.3386,
+      "step": 6699
+    },
+    {
+      "epoch": 0.4494080558070899,
+      "grad_norm": 0.1526204590875003,
+      "learning_rate": 2e-05,
+      "loss": 5.5002,
+      "step": 6700
+    },
+    {
+      "epoch": 0.44947513163631486,
+      "grad_norm": 0.1551256988403826,
+      "learning_rate": 2e-05,
+      "loss": 5.4193,
+      "step": 6701
+    },
+    {
+      "epoch": 0.4495422074655398,
+      "grad_norm": 0.14460995143112626,
+      "learning_rate": 2e-05,
+      "loss": 5.4147,
+      "step": 6702
+    },
+    {
+      "epoch": 0.44960928329476474,
+      "grad_norm": 0.1458574211779147,
+      "learning_rate": 2e-05,
+      "loss": 5.4886,
+      "step": 6703
+    },
+    {
+      "epoch": 0.4496763591239897,
+      "grad_norm": 0.1483865999453952,
+      "learning_rate": 2e-05,
+      "loss": 5.3956,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4497434349532146,
+      "grad_norm": 0.15139650193523052,
+      "learning_rate": 2e-05,
+      "loss": 5.4573,
+      "step": 6705
+    },
+    {
+      "epoch": 0.44981051078243955,
+      "grad_norm": 0.15623143453998375,
+      "learning_rate": 2e-05,
+      "loss": 5.5624,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4498775866116645,
+      "grad_norm": 0.15032086072233322,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 6707
+    },
+    {
+      "epoch": 0.44994466244088943,
+      "grad_norm": 0.15340291619810498,
+      "learning_rate": 2e-05,
+      "loss": 5.4293,
+      "step": 6708
+    },
+    {
+      "epoch": 0.45001173827011437,
+      "grad_norm": 0.15540585247498404,
+      "learning_rate": 2e-05,
+      "loss": 5.472,
+      "step": 6709
+    },
+    {
+      "epoch": 0.4500788140993393,
+      "grad_norm": 0.15093565922251084,
+      "learning_rate": 2e-05,
+      "loss": 5.4898,
+      "step": 6710
+    },
+    {
+      "epoch": 0.45014588992856425,
+      "grad_norm": 0.15303216862799726,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 6711
+    },
+    {
+      "epoch": 0.4502129657577892,
+      "grad_norm": 0.14651228911259123,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 6712
+    },
+    {
+      "epoch": 0.4502800415870141,
+      "grad_norm": 0.14262593862720402,
+      "learning_rate": 2e-05,
+      "loss": 5.4632,
+      "step": 6713
+    },
+    {
+      "epoch": 0.45034711741623906,
+      "grad_norm": 0.14605982034033693,
+      "learning_rate": 2e-05,
+      "loss": 5.4776,
+      "step": 6714
+    },
+    {
+      "epoch": 0.450414193245464,
+      "grad_norm": 0.14674471624171198,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 6715
+    },
+    {
+      "epoch": 0.45048126907468894,
+      "grad_norm": 0.14475982916311003,
+      "learning_rate": 2e-05,
+      "loss": 5.3348,
+      "step": 6716
+    },
+    {
+      "epoch": 0.4505483449039139,
+      "grad_norm": 0.14258355146235574,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 6717
+    },
+    {
+      "epoch": 0.4506154207331388,
+      "grad_norm": 0.15603985907555126,
+      "learning_rate": 2e-05,
+      "loss": 5.498,
+      "step": 6718
+    },
+    {
+      "epoch": 0.45068249656236375,
+      "grad_norm": 0.15225058433383828,
+      "learning_rate": 2e-05,
+      "loss": 5.3479,
+      "step": 6719
+    },
+    {
+      "epoch": 0.4507495723915887,
+      "grad_norm": 0.15314594114711044,
+      "learning_rate": 2e-05,
+      "loss": 5.3656,
+      "step": 6720
+    },
+    {
+      "epoch": 0.45081664822081363,
+      "grad_norm": 0.14940584863351314,
+      "learning_rate": 2e-05,
+      "loss": 5.3748,
+      "step": 6721
+    },
+    {
+      "epoch": 0.45088372405003857,
+      "grad_norm": 0.1489712657962272,
+      "learning_rate": 2e-05,
+      "loss": 5.3171,
+      "step": 6722
+    },
+    {
+      "epoch": 0.4509507998792635,
+      "grad_norm": 0.14869446175442633,
+      "learning_rate": 2e-05,
+      "loss": 5.3705,
+      "step": 6723
+    },
+    {
+      "epoch": 0.45101787570848845,
+      "grad_norm": 0.1665371875529824,
+      "learning_rate": 2e-05,
+      "loss": 5.5524,
+      "step": 6724
+    },
+    {
+      "epoch": 0.4510849515377134,
+      "grad_norm": 0.15795641471037727,
+      "learning_rate": 2e-05,
+      "loss": 5.42,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4511520273669383,
+      "grad_norm": 0.1484251275736474,
+      "learning_rate": 2e-05,
+      "loss": 5.4447,
+      "step": 6726
+    },
+    {
+      "epoch": 0.45121910319616326,
+      "grad_norm": 0.15490746345234377,
+      "learning_rate": 2e-05,
+      "loss": 5.4625,
+      "step": 6727
+    },
+    {
+      "epoch": 0.4512861790253882,
+      "grad_norm": 0.15040446963836424,
+      "learning_rate": 2e-05,
+      "loss": 5.4584,
+      "step": 6728
+    },
+    {
+      "epoch": 0.45135325485461314,
+      "grad_norm": 0.14891769234703464,
+      "learning_rate": 2e-05,
+      "loss": 5.4338,
+      "step": 6729
+    },
+    {
+      "epoch": 0.4514203306838381,
+      "grad_norm": 0.15191405531174987,
+      "learning_rate": 2e-05,
+      "loss": 5.4062,
+      "step": 6730
+    },
+    {
+      "epoch": 0.451487406513063,
+      "grad_norm": 0.15469071893892103,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 6731
+    },
+    {
+      "epoch": 0.45155448234228796,
+      "grad_norm": 0.1506459573573634,
+      "learning_rate": 2e-05,
+      "loss": 5.4145,
+      "step": 6732
+    },
+    {
+      "epoch": 0.4516215581715129,
+      "grad_norm": 0.15841825134238746,
+      "learning_rate": 2e-05,
+      "loss": 5.4135,
+      "step": 6733
+    },
+    {
+      "epoch": 0.45168863400073783,
+      "grad_norm": 0.15244260472909607,
+      "learning_rate": 2e-05,
+      "loss": 5.427,
+      "step": 6734
+    },
+    {
+      "epoch": 0.45175570982996277,
+      "grad_norm": 0.15578113130632093,
+      "learning_rate": 2e-05,
+      "loss": 5.5414,
+      "step": 6735
+    },
+    {
+      "epoch": 0.4518227856591877,
+      "grad_norm": 0.15184408795210916,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 6736
+    },
+    {
+      "epoch": 0.45188986148841265,
+      "grad_norm": 0.15974802711725725,
+      "learning_rate": 2e-05,
+      "loss": 5.3543,
+      "step": 6737
+    },
+    {
+      "epoch": 0.4519569373176376,
+      "grad_norm": 0.1518013073962905,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 6738
+    },
+    {
+      "epoch": 0.4520240131468625,
+      "grad_norm": 0.1506694285647593,
+      "learning_rate": 2e-05,
+      "loss": 5.3523,
+      "step": 6739
+    },
+    {
+      "epoch": 0.45209108897608746,
+      "grad_norm": 0.15241078006350828,
+      "learning_rate": 2e-05,
+      "loss": 5.5658,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4521581648053124,
+      "grad_norm": 0.15886447918685906,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 6741
+    },
+    {
+      "epoch": 0.45222524063453734,
+      "grad_norm": 0.14747066942844014,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 6742
+    },
+    {
+      "epoch": 0.4522923164637623,
+      "grad_norm": 0.1441903088604782,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 6743
+    },
+    {
+      "epoch": 0.4523593922929872,
+      "grad_norm": 0.15416502428577003,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 6744
+    },
+    {
+      "epoch": 0.45242646812221216,
+      "grad_norm": 0.15596868958194618,
+      "learning_rate": 2e-05,
+      "loss": 5.3459,
+      "step": 6745
+    },
+    {
+      "epoch": 0.4524935439514371,
+      "grad_norm": 0.14449522692098185,
+      "learning_rate": 2e-05,
+      "loss": 5.2653,
+      "step": 6746
+    },
+    {
+      "epoch": 0.45256061978066203,
+      "grad_norm": 0.1436775697201396,
+      "learning_rate": 2e-05,
+      "loss": 5.497,
+      "step": 6747
+    },
+    {
+      "epoch": 0.45262769560988697,
+      "grad_norm": 0.15880022317882891,
+      "learning_rate": 2e-05,
+      "loss": 5.5393,
+      "step": 6748
+    },
+    {
+      "epoch": 0.4526947714391119,
+      "grad_norm": 0.14361066667199573,
+      "learning_rate": 2e-05,
+      "loss": 5.4315,
+      "step": 6749
+    },
+    {
+      "epoch": 0.45276184726833685,
+      "grad_norm": 0.14894503374532814,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4528289230975618,
+      "grad_norm": 0.14499630710687358,
+      "learning_rate": 2e-05,
+      "loss": 5.5352,
+      "step": 6751
+    },
+    {
+      "epoch": 0.4528959989267867,
+      "grad_norm": 0.1493419188654647,
+      "learning_rate": 2e-05,
+      "loss": 5.4632,
+      "step": 6752
+    },
+    {
+      "epoch": 0.45296307475601166,
+      "grad_norm": 0.14700421925374885,
+      "learning_rate": 2e-05,
+      "loss": 5.4791,
+      "step": 6753
+    },
+    {
+      "epoch": 0.4530301505852366,
+      "grad_norm": 0.15355700882945647,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 6754
+    },
+    {
+      "epoch": 0.45309722641446154,
+      "grad_norm": 0.15006343098444855,
+      "learning_rate": 2e-05,
+      "loss": 5.2803,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4531643022436865,
+      "grad_norm": 0.15568878795923285,
+      "learning_rate": 2e-05,
+      "loss": 5.6067,
+      "step": 6756
+    },
+    {
+      "epoch": 0.4532313780729114,
+      "grad_norm": 0.15151165694055962,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 6757
+    },
+    {
+      "epoch": 0.45329845390213636,
+      "grad_norm": 0.1490430211160049,
+      "learning_rate": 2e-05,
+      "loss": 5.3991,
+      "step": 6758
+    },
+    {
+      "epoch": 0.4533655297313613,
+      "grad_norm": 0.1473519034352782,
+      "learning_rate": 2e-05,
+      "loss": 5.3893,
+      "step": 6759
+    },
+    {
+      "epoch": 0.45343260556058623,
+      "grad_norm": 0.15555970069085348,
+      "learning_rate": 2e-05,
+      "loss": 5.3927,
+      "step": 6760
+    },
+    {
+      "epoch": 0.45349968138981117,
+      "grad_norm": 0.14868522304892845,
+      "learning_rate": 2e-05,
+      "loss": 5.5196,
+      "step": 6761
+    },
+    {
+      "epoch": 0.4535667572190361,
+      "grad_norm": 0.1453499895823372,
+      "learning_rate": 2e-05,
+      "loss": 5.2662,
+      "step": 6762
+    },
+    {
+      "epoch": 0.45363383304826105,
+      "grad_norm": 0.15195613885814688,
+      "learning_rate": 2e-05,
+      "loss": 5.5273,
+      "step": 6763
+    },
+    {
+      "epoch": 0.453700908877486,
+      "grad_norm": 0.1572684460117502,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 6764
+    },
+    {
+      "epoch": 0.4537679847067109,
+      "grad_norm": 0.14908801833054158,
+      "learning_rate": 2e-05,
+      "loss": 5.6312,
+      "step": 6765
+    },
+    {
+      "epoch": 0.45383506053593586,
+      "grad_norm": 0.15087374879702167,
+      "learning_rate": 2e-05,
+      "loss": 5.4087,
+      "step": 6766
+    },
+    {
+      "epoch": 0.4539021363651608,
+      "grad_norm": 0.14241983601471628,
+      "learning_rate": 2e-05,
+      "loss": 5.5734,
+      "step": 6767
+    },
+    {
+      "epoch": 0.45396921219438574,
+      "grad_norm": 0.1495772741834931,
+      "learning_rate": 2e-05,
+      "loss": 5.3459,
+      "step": 6768
+    },
+    {
+      "epoch": 0.4540362880236107,
+      "grad_norm": 0.14309825889399053,
+      "learning_rate": 2e-05,
+      "loss": 5.4752,
+      "step": 6769
+    },
+    {
+      "epoch": 0.4541033638528356,
+      "grad_norm": 0.1471018627262288,
+      "learning_rate": 2e-05,
+      "loss": 5.252,
+      "step": 6770
+    },
+    {
+      "epoch": 0.45417043968206056,
+      "grad_norm": 0.14697599348999096,
+      "learning_rate": 2e-05,
+      "loss": 5.4274,
+      "step": 6771
+    },
+    {
+      "epoch": 0.4542375155112855,
+      "grad_norm": 0.1468864888267532,
+      "learning_rate": 2e-05,
+      "loss": 5.3886,
+      "step": 6772
+    },
+    {
+      "epoch": 0.45430459134051043,
+      "grad_norm": 0.14896204709292452,
+      "learning_rate": 2e-05,
+      "loss": 5.5465,
+      "step": 6773
+    },
+    {
+      "epoch": 0.45437166716973537,
+      "grad_norm": 0.1416643042434383,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 6774
+    },
+    {
+      "epoch": 0.4544387429989603,
+      "grad_norm": 0.14482135846457422,
+      "learning_rate": 2e-05,
+      "loss": 5.4253,
+      "step": 6775
+    },
+    {
+      "epoch": 0.45450581882818525,
+      "grad_norm": 0.14889897420253156,
+      "learning_rate": 2e-05,
+      "loss": 5.3665,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4545728946574102,
+      "grad_norm": 0.14840040106952188,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 6777
+    },
+    {
+      "epoch": 0.4546399704866351,
+      "grad_norm": 0.15272561991420458,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 6778
+    },
+    {
+      "epoch": 0.45470704631586006,
+      "grad_norm": 0.16162495644330718,
+      "learning_rate": 2e-05,
+      "loss": 5.4434,
+      "step": 6779
+    },
+    {
+      "epoch": 0.454774122145085,
+      "grad_norm": 0.14529100969601305,
+      "learning_rate": 2e-05,
+      "loss": 5.3613,
+      "step": 6780
+    },
+    {
+      "epoch": 0.45484119797430994,
+      "grad_norm": 0.14945675857808685,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 6781
+    },
+    {
+      "epoch": 0.4549082738035349,
+      "grad_norm": 0.15159926655540495,
+      "learning_rate": 2e-05,
+      "loss": 5.4658,
+      "step": 6782
+    },
+    {
+      "epoch": 0.4549753496327598,
+      "grad_norm": 0.153422925138813,
+      "learning_rate": 2e-05,
+      "loss": 5.3999,
+      "step": 6783
+    },
+    {
+      "epoch": 0.45504242546198476,
+      "grad_norm": 0.15508182220241562,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4551095012912097,
+      "grad_norm": 0.15153719394643492,
+      "learning_rate": 2e-05,
+      "loss": 5.3045,
+      "step": 6785
+    },
+    {
+      "epoch": 0.45517657712043463,
+      "grad_norm": 0.15580567264626596,
+      "learning_rate": 2e-05,
+      "loss": 5.4171,
+      "step": 6786
+    },
+    {
+      "epoch": 0.45524365294965957,
+      "grad_norm": 0.1473647873169069,
+      "learning_rate": 2e-05,
+      "loss": 5.319,
+      "step": 6787
+    },
+    {
+      "epoch": 0.4553107287788845,
+      "grad_norm": 0.15272026070819308,
+      "learning_rate": 2e-05,
+      "loss": 5.4639,
+      "step": 6788
+    },
+    {
+      "epoch": 0.45537780460810945,
+      "grad_norm": 0.153414290462464,
+      "learning_rate": 2e-05,
+      "loss": 5.3883,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4554448804373344,
+      "grad_norm": 0.15090595890991929,
+      "learning_rate": 2e-05,
+      "loss": 5.3126,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4555119562665593,
+      "grad_norm": 0.15281037121437713,
+      "learning_rate": 2e-05,
+      "loss": 5.3654,
+      "step": 6791
+    },
+    {
+      "epoch": 0.45557903209578426,
+      "grad_norm": 0.1486893429212238,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 6792
+    },
+    {
+      "epoch": 0.4556461079250092,
+      "grad_norm": 0.15197232490173232,
+      "learning_rate": 2e-05,
+      "loss": 5.3909,
+      "step": 6793
+    },
+    {
+      "epoch": 0.45571318375423414,
+      "grad_norm": 0.14574549099874146,
+      "learning_rate": 2e-05,
+      "loss": 5.2675,
+      "step": 6794
+    },
+    {
+      "epoch": 0.4557802595834591,
+      "grad_norm": 0.140932002100928,
+      "learning_rate": 2e-05,
+      "loss": 5.4156,
+      "step": 6795
+    },
+    {
+      "epoch": 0.455847335412684,
+      "grad_norm": 0.1497684036686364,
+      "learning_rate": 2e-05,
+      "loss": 5.5309,
+      "step": 6796
+    },
+    {
+      "epoch": 0.45591441124190896,
+      "grad_norm": 0.1504370141736865,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 6797
+    },
+    {
+      "epoch": 0.4559814870711339,
+      "grad_norm": 0.1481266171180037,
+      "learning_rate": 2e-05,
+      "loss": 5.5354,
+      "step": 6798
+    },
+    {
+      "epoch": 0.45604856290035883,
+      "grad_norm": 0.15065593803871136,
+      "learning_rate": 2e-05,
+      "loss": 5.5571,
+      "step": 6799
+    },
+    {
+      "epoch": 0.45611563872958377,
+      "grad_norm": 0.1427567008983785,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4561827145588087,
+      "grad_norm": 0.14413870741633147,
+      "learning_rate": 2e-05,
+      "loss": 5.298,
+      "step": 6801
+    },
+    {
+      "epoch": 0.45624979038803365,
+      "grad_norm": 0.14613927866435575,
+      "learning_rate": 2e-05,
+      "loss": 5.5427,
+      "step": 6802
+    },
+    {
+      "epoch": 0.4563168662172586,
+      "grad_norm": 0.1572618437385827,
+      "learning_rate": 2e-05,
+      "loss": 5.4561,
+      "step": 6803
+    },
+    {
+      "epoch": 0.4563839420464835,
+      "grad_norm": 0.15392001653236845,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 6804
+    },
+    {
+      "epoch": 0.45645101787570846,
+      "grad_norm": 0.15513710660861638,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 6805
+    },
+    {
+      "epoch": 0.4565180937049334,
+      "grad_norm": 0.1531837537968394,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 6806
+    },
+    {
+      "epoch": 0.45658516953415834,
+      "grad_norm": 0.14549312799348676,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 6807
+    },
+    {
+      "epoch": 0.4566522453633833,
+      "grad_norm": 0.1515604646861352,
+      "learning_rate": 2e-05,
+      "loss": 5.5358,
+      "step": 6808
+    },
+    {
+      "epoch": 0.4567193211926082,
+      "grad_norm": 0.15554387475075712,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 6809
+    },
+    {
+      "epoch": 0.45678639702183316,
+      "grad_norm": 0.14248033371355506,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4568534728510581,
+      "grad_norm": 0.14552718053370287,
+      "learning_rate": 2e-05,
+      "loss": 5.4726,
+      "step": 6811
+    },
+    {
+      "epoch": 0.45692054868028303,
+      "grad_norm": 0.1584713522793728,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 6812
+    },
+    {
+      "epoch": 0.456987624509508,
+      "grad_norm": 0.15189221423506102,
+      "learning_rate": 2e-05,
+      "loss": 5.3784,
+      "step": 6813
+    },
+    {
+      "epoch": 0.4570547003387329,
+      "grad_norm": 0.15098210163678075,
+      "learning_rate": 2e-05,
+      "loss": 5.5663,
+      "step": 6814
+    },
+    {
+      "epoch": 0.45712177616795785,
+      "grad_norm": 0.14998442816389243,
+      "learning_rate": 2e-05,
+      "loss": 5.412,
+      "step": 6815
+    },
+    {
+      "epoch": 0.4571888519971828,
+      "grad_norm": 0.15148875580852159,
+      "learning_rate": 2e-05,
+      "loss": 5.4968,
+      "step": 6816
+    },
+    {
+      "epoch": 0.4572559278264077,
+      "grad_norm": 0.1488532936339803,
+      "learning_rate": 2e-05,
+      "loss": 5.3634,
+      "step": 6817
+    },
+    {
+      "epoch": 0.4573230036556327,
+      "grad_norm": 0.14864038668991458,
+      "learning_rate": 2e-05,
+      "loss": 5.443,
+      "step": 6818
+    },
+    {
+      "epoch": 0.45739007948485766,
+      "grad_norm": 0.14905748875340388,
+      "learning_rate": 2e-05,
+      "loss": 5.3715,
+      "step": 6819
+    },
+    {
+      "epoch": 0.4574571553140826,
+      "grad_norm": 0.1542489156970249,
+      "learning_rate": 2e-05,
+      "loss": 5.2481,
+      "step": 6820
+    },
+    {
+      "epoch": 0.45752423114330754,
+      "grad_norm": 0.14465655347565348,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 6821
+    },
+    {
+      "epoch": 0.4575913069725325,
+      "grad_norm": 0.14383478993836624,
+      "learning_rate": 2e-05,
+      "loss": 5.3609,
+      "step": 6822
+    },
+    {
+      "epoch": 0.4576583828017574,
+      "grad_norm": 0.14920575818076662,
+      "learning_rate": 2e-05,
+      "loss": 5.5277,
+      "step": 6823
+    },
+    {
+      "epoch": 0.45772545863098235,
+      "grad_norm": 0.15398943892602626,
+      "learning_rate": 2e-05,
+      "loss": 5.443,
+      "step": 6824
+    },
+    {
+      "epoch": 0.4577925344602073,
+      "grad_norm": 0.15903410715419328,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 6825
+    },
+    {
+      "epoch": 0.45785961028943223,
+      "grad_norm": 0.14944428611257876,
+      "learning_rate": 2e-05,
+      "loss": 5.4037,
+      "step": 6826
+    },
+    {
+      "epoch": 0.45792668611865717,
+      "grad_norm": 0.14609668423029729,
+      "learning_rate": 2e-05,
+      "loss": 5.4424,
+      "step": 6827
+    },
+    {
+      "epoch": 0.4579937619478821,
+      "grad_norm": 0.1490239788627957,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 6828
+    },
+    {
+      "epoch": 0.45806083777710704,
+      "grad_norm": 0.1560167274124761,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 6829
+    },
+    {
+      "epoch": 0.458127913606332,
+      "grad_norm": 0.1484792875021223,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 6830
+    },
+    {
+      "epoch": 0.4581949894355569,
+      "grad_norm": 0.14644493131324363,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 6831
+    },
+    {
+      "epoch": 0.45826206526478186,
+      "grad_norm": 0.14963182525052476,
+      "learning_rate": 2e-05,
+      "loss": 5.5202,
+      "step": 6832
+    },
+    {
+      "epoch": 0.4583291410940068,
+      "grad_norm": 0.1492016315959773,
+      "learning_rate": 2e-05,
+      "loss": 5.4887,
+      "step": 6833
+    },
+    {
+      "epoch": 0.45839621692323174,
+      "grad_norm": 0.1606279012582248,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 6834
+    },
+    {
+      "epoch": 0.4584632927524567,
+      "grad_norm": 0.15341594066844494,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4585303685816816,
+      "grad_norm": 0.1486740356555224,
+      "learning_rate": 2e-05,
+      "loss": 5.5475,
+      "step": 6836
+    },
+    {
+      "epoch": 0.45859744441090655,
+      "grad_norm": 0.14941081289396368,
+      "learning_rate": 2e-05,
+      "loss": 5.4845,
+      "step": 6837
+    },
+    {
+      "epoch": 0.4586645202401315,
+      "grad_norm": 0.1437072389361915,
+      "learning_rate": 2e-05,
+      "loss": 5.3068,
+      "step": 6838
+    },
+    {
+      "epoch": 0.45873159606935643,
+      "grad_norm": 0.16230741934824738,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 6839
+    },
+    {
+      "epoch": 0.45879867189858137,
+      "grad_norm": 0.1509260401817456,
+      "learning_rate": 2e-05,
+      "loss": 5.2249,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4588657477278063,
+      "grad_norm": 0.1480138557745186,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 6841
+    },
+    {
+      "epoch": 0.45893282355703124,
+      "grad_norm": 0.1573583391060196,
+      "learning_rate": 2e-05,
+      "loss": 5.5945,
+      "step": 6842
+    },
+    {
+      "epoch": 0.4589998993862562,
+      "grad_norm": 0.1563945714113123,
+      "learning_rate": 2e-05,
+      "loss": 5.2916,
+      "step": 6843
+    },
+    {
+      "epoch": 0.4590669752154811,
+      "grad_norm": 0.15197606726096863,
+      "learning_rate": 2e-05,
+      "loss": 5.3326,
+      "step": 6844
+    },
+    {
+      "epoch": 0.45913405104470606,
+      "grad_norm": 0.1475515685179669,
+      "learning_rate": 2e-05,
+      "loss": 5.5122,
+      "step": 6845
+    },
+    {
+      "epoch": 0.459201126873931,
+      "grad_norm": 0.1586890013726817,
+      "learning_rate": 2e-05,
+      "loss": 5.3445,
+      "step": 6846
+    },
+    {
+      "epoch": 0.45926820270315594,
+      "grad_norm": 0.16522729500786631,
+      "learning_rate": 2e-05,
+      "loss": 5.556,
+      "step": 6847
+    },
+    {
+      "epoch": 0.4593352785323809,
+      "grad_norm": 0.15332365514919233,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 6848
+    },
+    {
+      "epoch": 0.4594023543616058,
+      "grad_norm": 0.14468835222295723,
+      "learning_rate": 2e-05,
+      "loss": 5.1447,
+      "step": 6849
+    },
+    {
+      "epoch": 0.45946943019083075,
+      "grad_norm": 0.1633879437538502,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4595365060200557,
+      "grad_norm": 0.16367654628024456,
+      "learning_rate": 2e-05,
+      "loss": 5.5638,
+      "step": 6851
+    },
+    {
+      "epoch": 0.45960358184928063,
+      "grad_norm": 0.14792256133534445,
+      "learning_rate": 2e-05,
+      "loss": 5.5333,
+      "step": 6852
+    },
+    {
+      "epoch": 0.45967065767850557,
+      "grad_norm": 0.155761577735702,
+      "learning_rate": 2e-05,
+      "loss": 5.4167,
+      "step": 6853
+    },
+    {
+      "epoch": 0.4597377335077305,
+      "grad_norm": 0.17104952415285546,
+      "learning_rate": 2e-05,
+      "loss": 5.2985,
+      "step": 6854
+    },
+    {
+      "epoch": 0.45980480933695544,
+      "grad_norm": 0.16979817934350178,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 6855
+    },
+    {
+      "epoch": 0.4598718851661804,
+      "grad_norm": 0.15138516126144502,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 6856
+    },
+    {
+      "epoch": 0.4599389609954053,
+      "grad_norm": 0.17166294857898953,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 6857
+    },
+    {
+      "epoch": 0.46000603682463026,
+      "grad_norm": 0.16126027210637386,
+      "learning_rate": 2e-05,
+      "loss": 5.4556,
+      "step": 6858
+    },
+    {
+      "epoch": 0.4600731126538552,
+      "grad_norm": 0.16306187911632683,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 6859
+    },
+    {
+      "epoch": 0.46014018848308014,
+      "grad_norm": 0.1540449653355481,
+      "learning_rate": 2e-05,
+      "loss": 5.4879,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4602072643123051,
+      "grad_norm": 0.15980252006265108,
+      "learning_rate": 2e-05,
+      "loss": 5.3491,
+      "step": 6861
+    },
+    {
+      "epoch": 0.46027434014153,
+      "grad_norm": 0.15988193402297754,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 6862
+    },
+    {
+      "epoch": 0.46034141597075495,
+      "grad_norm": 0.16255255547307942,
+      "learning_rate": 2e-05,
+      "loss": 5.4825,
+      "step": 6863
+    },
+    {
+      "epoch": 0.4604084917999799,
+      "grad_norm": 0.15395051994633366,
+      "learning_rate": 2e-05,
+      "loss": 5.4932,
+      "step": 6864
+    },
+    {
+      "epoch": 0.46047556762920483,
+      "grad_norm": 0.15954079484248598,
+      "learning_rate": 2e-05,
+      "loss": 5.4023,
+      "step": 6865
+    },
+    {
+      "epoch": 0.46054264345842977,
+      "grad_norm": 0.15368879107035383,
+      "learning_rate": 2e-05,
+      "loss": 5.4606,
+      "step": 6866
+    },
+    {
+      "epoch": 0.4606097192876547,
+      "grad_norm": 0.1596556019262668,
+      "learning_rate": 2e-05,
+      "loss": 5.5666,
+      "step": 6867
+    },
+    {
+      "epoch": 0.46067679511687964,
+      "grad_norm": 0.15638060299038178,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 6868
+    },
+    {
+      "epoch": 0.4607438709461046,
+      "grad_norm": 0.1543171123634257,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 6869
+    },
+    {
+      "epoch": 0.4608109467753295,
+      "grad_norm": 0.1620174065511477,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 6870
+    },
+    {
+      "epoch": 0.46087802260455446,
+      "grad_norm": 0.15923071950255707,
+      "learning_rate": 2e-05,
+      "loss": 5.383,
+      "step": 6871
+    },
+    {
+      "epoch": 0.4609450984337794,
+      "grad_norm": 0.1560591044547252,
+      "learning_rate": 2e-05,
+      "loss": 5.3896,
+      "step": 6872
+    },
+    {
+      "epoch": 0.46101217426300434,
+      "grad_norm": 0.16056990164153642,
+      "learning_rate": 2e-05,
+      "loss": 5.2891,
+      "step": 6873
+    },
+    {
+      "epoch": 0.4610792500922293,
+      "grad_norm": 0.15345297141644507,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 6874
+    },
+    {
+      "epoch": 0.4611463259214542,
+      "grad_norm": 0.15042588598379872,
+      "learning_rate": 2e-05,
+      "loss": 5.5809,
+      "step": 6875
+    },
+    {
+      "epoch": 0.46121340175067915,
+      "grad_norm": 0.15313454691390052,
+      "learning_rate": 2e-05,
+      "loss": 5.3473,
+      "step": 6876
+    },
+    {
+      "epoch": 0.4612804775799041,
+      "grad_norm": 0.15215817610877422,
+      "learning_rate": 2e-05,
+      "loss": 5.447,
+      "step": 6877
+    },
+    {
+      "epoch": 0.46134755340912903,
+      "grad_norm": 0.14576484478761279,
+      "learning_rate": 2e-05,
+      "loss": 5.4818,
+      "step": 6878
+    },
+    {
+      "epoch": 0.46141462923835397,
+      "grad_norm": 0.15770568272415206,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 6879
+    },
+    {
+      "epoch": 0.4614817050675789,
+      "grad_norm": 0.15302481145260463,
+      "learning_rate": 2e-05,
+      "loss": 5.4615,
+      "step": 6880
+    },
+    {
+      "epoch": 0.46154878089680385,
+      "grad_norm": 0.16030093390564054,
+      "learning_rate": 2e-05,
+      "loss": 5.4299,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4616158567260288,
+      "grad_norm": 0.16396075011348726,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 6882
+    },
+    {
+      "epoch": 0.4616829325552537,
+      "grad_norm": 0.14585112902347835,
+      "learning_rate": 2e-05,
+      "loss": 5.3386,
+      "step": 6883
+    },
+    {
+      "epoch": 0.46175000838447866,
+      "grad_norm": 0.14606060603619322,
+      "learning_rate": 2e-05,
+      "loss": 5.4251,
+      "step": 6884
+    },
+    {
+      "epoch": 0.4618170842137036,
+      "grad_norm": 0.151404840348132,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 6885
+    },
+    {
+      "epoch": 0.46188416004292854,
+      "grad_norm": 0.15324013379395765,
+      "learning_rate": 2e-05,
+      "loss": 5.3451,
+      "step": 6886
+    },
+    {
+      "epoch": 0.4619512358721535,
+      "grad_norm": 0.15543860580247973,
+      "learning_rate": 2e-05,
+      "loss": 5.4014,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4620183117013784,
+      "grad_norm": 0.15281911805965062,
+      "learning_rate": 2e-05,
+      "loss": 5.5368,
+      "step": 6888
+    },
+    {
+      "epoch": 0.46208538753060335,
+      "grad_norm": 0.14666811658672133,
+      "learning_rate": 2e-05,
+      "loss": 5.5135,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4621524633598283,
+      "grad_norm": 0.16017857389825435,
+      "learning_rate": 2e-05,
+      "loss": 5.3202,
+      "step": 6890
+    },
+    {
+      "epoch": 0.46221953918905323,
+      "grad_norm": 0.14742991055372673,
+      "learning_rate": 2e-05,
+      "loss": 5.3528,
+      "step": 6891
+    },
+    {
+      "epoch": 0.46228661501827817,
+      "grad_norm": 0.1516833967947273,
+      "learning_rate": 2e-05,
+      "loss": 5.3472,
+      "step": 6892
+    },
+    {
+      "epoch": 0.4623536908475031,
+      "grad_norm": 0.15554706278451133,
+      "learning_rate": 2e-05,
+      "loss": 5.418,
+      "step": 6893
+    },
+    {
+      "epoch": 0.46242076667672805,
+      "grad_norm": 0.15292364999403638,
+      "learning_rate": 2e-05,
+      "loss": 5.5843,
+      "step": 6894
+    },
+    {
+      "epoch": 0.462487842505953,
+      "grad_norm": 0.1456045806223181,
+      "learning_rate": 2e-05,
+      "loss": 5.482,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4625549183351779,
+      "grad_norm": 0.14744795452139045,
+      "learning_rate": 2e-05,
+      "loss": 5.3165,
+      "step": 6896
+    },
+    {
+      "epoch": 0.46262199416440286,
+      "grad_norm": 0.15987486523007366,
+      "learning_rate": 2e-05,
+      "loss": 5.5192,
+      "step": 6897
+    },
+    {
+      "epoch": 0.4626890699936278,
+      "grad_norm": 0.14634283140843818,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 6898
+    },
+    {
+      "epoch": 0.46275614582285274,
+      "grad_norm": 0.15461747961660474,
+      "learning_rate": 2e-05,
+      "loss": 5.3416,
+      "step": 6899
+    },
+    {
+      "epoch": 0.4628232216520777,
+      "grad_norm": 0.14885018183181015,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4628902974813026,
+      "grad_norm": 0.15677463588083337,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 6901
+    },
+    {
+      "epoch": 0.46295737331052755,
+      "grad_norm": 0.15884081096055716,
+      "learning_rate": 2e-05,
+      "loss": 5.4438,
+      "step": 6902
+    },
+    {
+      "epoch": 0.4630244491397525,
+      "grad_norm": 0.15696451354859234,
+      "learning_rate": 2e-05,
+      "loss": 5.5286,
+      "step": 6903
+    },
+    {
+      "epoch": 0.46309152496897743,
+      "grad_norm": 0.15193108496692742,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 6904
+    },
+    {
+      "epoch": 0.46315860079820237,
+      "grad_norm": 0.15623535823596021,
+      "learning_rate": 2e-05,
+      "loss": 5.2706,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4632256766274273,
+      "grad_norm": 0.15077800595437188,
+      "learning_rate": 2e-05,
+      "loss": 5.377,
+      "step": 6906
+    },
+    {
+      "epoch": 0.46329275245665225,
+      "grad_norm": 0.14441539537996173,
+      "learning_rate": 2e-05,
+      "loss": 5.3369,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4633598282858772,
+      "grad_norm": 0.15836694628089862,
+      "learning_rate": 2e-05,
+      "loss": 5.5044,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4634269041151021,
+      "grad_norm": 0.15390534497418698,
+      "learning_rate": 2e-05,
+      "loss": 5.383,
+      "step": 6909
+    },
+    {
+      "epoch": 0.46349397994432706,
+      "grad_norm": 0.1487593422830048,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 6910
+    },
+    {
+      "epoch": 0.463561055773552,
+      "grad_norm": 0.14991572565228262,
+      "learning_rate": 2e-05,
+      "loss": 5.515,
+      "step": 6911
+    },
+    {
+      "epoch": 0.46362813160277694,
+      "grad_norm": 0.16462358333259483,
+      "learning_rate": 2e-05,
+      "loss": 5.3525,
+      "step": 6912
+    },
+    {
+      "epoch": 0.4636952074320019,
+      "grad_norm": 0.15021014157892695,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 6913
+    },
+    {
+      "epoch": 0.4637622832612268,
+      "grad_norm": 0.1503320038624078,
+      "learning_rate": 2e-05,
+      "loss": 5.3151,
+      "step": 6914
+    },
+    {
+      "epoch": 0.46382935909045175,
+      "grad_norm": 0.15281805741785356,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4638964349196767,
+      "grad_norm": 0.14528153041482786,
+      "learning_rate": 2e-05,
+      "loss": 5.4037,
+      "step": 6916
+    },
+    {
+      "epoch": 0.46396351074890163,
+      "grad_norm": 0.14586281185449368,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 6917
+    },
+    {
+      "epoch": 0.46403058657812657,
+      "grad_norm": 0.15134114694695414,
+      "learning_rate": 2e-05,
+      "loss": 5.5017,
+      "step": 6918
+    },
+    {
+      "epoch": 0.4640976624073515,
+      "grad_norm": 0.14987264986425503,
+      "learning_rate": 2e-05,
+      "loss": 5.36,
+      "step": 6919
+    },
+    {
+      "epoch": 0.46416473823657645,
+      "grad_norm": 0.150838808903471,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4642318140658014,
+      "grad_norm": 0.16607860069060745,
+      "learning_rate": 2e-05,
+      "loss": 5.3559,
+      "step": 6921
+    },
+    {
+      "epoch": 0.4642988898950263,
+      "grad_norm": 0.15087429925567544,
+      "learning_rate": 2e-05,
+      "loss": 5.3145,
+      "step": 6922
+    },
+    {
+      "epoch": 0.46436596572425126,
+      "grad_norm": 0.1498953318541709,
+      "learning_rate": 2e-05,
+      "loss": 5.5187,
+      "step": 6923
+    },
+    {
+      "epoch": 0.4644330415534762,
+      "grad_norm": 0.15463250418527355,
+      "learning_rate": 2e-05,
+      "loss": 5.2846,
+      "step": 6924
+    },
+    {
+      "epoch": 0.46450011738270114,
+      "grad_norm": 0.15255499592922106,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4645671932119261,
+      "grad_norm": 0.1493142687296615,
+      "learning_rate": 2e-05,
+      "loss": 5.4225,
+      "step": 6926
+    },
+    {
+      "epoch": 0.464634269041151,
+      "grad_norm": 0.15243735812981327,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 6927
+    },
+    {
+      "epoch": 0.46470134487037595,
+      "grad_norm": 0.15202042563306756,
+      "learning_rate": 2e-05,
+      "loss": 5.5682,
+      "step": 6928
+    },
+    {
+      "epoch": 0.4647684206996009,
+      "grad_norm": 0.15083950388730794,
+      "learning_rate": 2e-05,
+      "loss": 5.4143,
+      "step": 6929
+    },
+    {
+      "epoch": 0.46483549652882583,
+      "grad_norm": 0.15756573244834723,
+      "learning_rate": 2e-05,
+      "loss": 5.3373,
+      "step": 6930
+    },
+    {
+      "epoch": 0.46490257235805077,
+      "grad_norm": 0.14565159462507912,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 6931
+    },
+    {
+      "epoch": 0.4649696481872757,
+      "grad_norm": 0.15538131716712208,
+      "learning_rate": 2e-05,
+      "loss": 5.4973,
+      "step": 6932
+    },
+    {
+      "epoch": 0.46503672401650065,
+      "grad_norm": 0.14658396990056732,
+      "learning_rate": 2e-05,
+      "loss": 5.3672,
+      "step": 6933
+    },
+    {
+      "epoch": 0.4651037998457256,
+      "grad_norm": 0.14895246555611186,
+      "learning_rate": 2e-05,
+      "loss": 5.5306,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4651708756749505,
+      "grad_norm": 0.14576682536311208,
+      "learning_rate": 2e-05,
+      "loss": 5.4516,
+      "step": 6935
+    },
+    {
+      "epoch": 0.46523795150417546,
+      "grad_norm": 0.15044826716769136,
+      "learning_rate": 2e-05,
+      "loss": 5.352,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4653050273334004,
+      "grad_norm": 0.1499369976870163,
+      "learning_rate": 2e-05,
+      "loss": 5.4599,
+      "step": 6937
+    },
+    {
+      "epoch": 0.46537210316262534,
+      "grad_norm": 0.14619066504401848,
+      "learning_rate": 2e-05,
+      "loss": 5.4802,
+      "step": 6938
+    },
+    {
+      "epoch": 0.4654391789918503,
+      "grad_norm": 0.14499564909946372,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 6939
+    },
+    {
+      "epoch": 0.4655062548210752,
+      "grad_norm": 0.1450752512136105,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 6940
+    },
+    {
+      "epoch": 0.46557333065030015,
+      "grad_norm": 0.1521223880857344,
+      "learning_rate": 2e-05,
+      "loss": 5.3318,
+      "step": 6941
+    },
+    {
+      "epoch": 0.4656404064795251,
+      "grad_norm": 0.14509791375908196,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 6942
+    },
+    {
+      "epoch": 0.46570748230875003,
+      "grad_norm": 0.15544054107190958,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 6943
+    },
+    {
+      "epoch": 0.46577455813797497,
+      "grad_norm": 0.14575167730400948,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 6944
+    },
+    {
+      "epoch": 0.4658416339671999,
+      "grad_norm": 0.14807596060419498,
+      "learning_rate": 2e-05,
+      "loss": 5.4131,
+      "step": 6945
+    },
+    {
+      "epoch": 0.46590870979642485,
+      "grad_norm": 0.14322391638655688,
+      "learning_rate": 2e-05,
+      "loss": 5.3715,
+      "step": 6946
+    },
+    {
+      "epoch": 0.4659757856256498,
+      "grad_norm": 0.15063873837732356,
+      "learning_rate": 2e-05,
+      "loss": 5.5099,
+      "step": 6947
+    },
+    {
+      "epoch": 0.4660428614548747,
+      "grad_norm": 0.15105829587822217,
+      "learning_rate": 2e-05,
+      "loss": 5.2705,
+      "step": 6948
+    },
+    {
+      "epoch": 0.46610993728409966,
+      "grad_norm": 0.14459596212435813,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4661770131133246,
+      "grad_norm": 0.16080605490714095,
+      "learning_rate": 2e-05,
+      "loss": 5.4508,
+      "step": 6950
+    },
+    {
+      "epoch": 0.46624408894254954,
+      "grad_norm": 0.16460932887660304,
+      "learning_rate": 2e-05,
+      "loss": 5.3322,
+      "step": 6951
+    },
+    {
+      "epoch": 0.4663111647717745,
+      "grad_norm": 0.15574547189581986,
+      "learning_rate": 2e-05,
+      "loss": 5.4011,
+      "step": 6952
+    },
+    {
+      "epoch": 0.4663782406009994,
+      "grad_norm": 0.1577607376432456,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 6953
+    },
+    {
+      "epoch": 0.46644531643022435,
+      "grad_norm": 0.17555296342978122,
+      "learning_rate": 2e-05,
+      "loss": 5.2703,
+      "step": 6954
+    },
+    {
+      "epoch": 0.4665123922594493,
+      "grad_norm": 0.16838851330235774,
+      "learning_rate": 2e-05,
+      "loss": 5.2932,
+      "step": 6955
+    },
+    {
+      "epoch": 0.46657946808867423,
+      "grad_norm": 0.15901917232692345,
+      "learning_rate": 2e-05,
+      "loss": 5.5105,
+      "step": 6956
+    },
+    {
+      "epoch": 0.46664654391789917,
+      "grad_norm": 0.16861780933168086,
+      "learning_rate": 2e-05,
+      "loss": 5.3871,
+      "step": 6957
+    },
+    {
+      "epoch": 0.4667136197471241,
+      "grad_norm": 0.16326064206530153,
+      "learning_rate": 2e-05,
+      "loss": 5.4557,
+      "step": 6958
+    },
+    {
+      "epoch": 0.46678069557634905,
+      "grad_norm": 0.1633761599671977,
+      "learning_rate": 2e-05,
+      "loss": 5.3838,
+      "step": 6959
+    },
+    {
+      "epoch": 0.466847771405574,
+      "grad_norm": 0.1679975380080666,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4669148472347989,
+      "grad_norm": 0.158663602604193,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 6961
+    },
+    {
+      "epoch": 0.46698192306402386,
+      "grad_norm": 0.15678210495018338,
+      "learning_rate": 2e-05,
+      "loss": 5.388,
+      "step": 6962
+    },
+    {
+      "epoch": 0.4670489988932488,
+      "grad_norm": 0.17086768274088696,
+      "learning_rate": 2e-05,
+      "loss": 5.3663,
+      "step": 6963
+    },
+    {
+      "epoch": 0.46711607472247374,
+      "grad_norm": 0.162827416041489,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 6964
+    },
+    {
+      "epoch": 0.4671831505516987,
+      "grad_norm": 0.1550178347049582,
+      "learning_rate": 2e-05,
+      "loss": 5.4802,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4672502263809236,
+      "grad_norm": 0.15052288379174217,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 6966
+    },
+    {
+      "epoch": 0.46731730221014856,
+      "grad_norm": 0.1643149266747628,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4673843780393735,
+      "grad_norm": 0.15428797244819617,
+      "learning_rate": 2e-05,
+      "loss": 5.4965,
+      "step": 6968
+    },
+    {
+      "epoch": 0.46745145386859843,
+      "grad_norm": 0.15680098529623573,
+      "learning_rate": 2e-05,
+      "loss": 5.4042,
+      "step": 6969
+    },
+    {
+      "epoch": 0.46751852969782337,
+      "grad_norm": 0.15456396020306393,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4675856055270483,
+      "grad_norm": 0.15908333199389801,
+      "learning_rate": 2e-05,
+      "loss": 5.4843,
+      "step": 6971
+    },
+    {
+      "epoch": 0.46765268135627325,
+      "grad_norm": 0.15096196002979345,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 6972
+    },
+    {
+      "epoch": 0.4677197571854982,
+      "grad_norm": 0.14517318762685819,
+      "learning_rate": 2e-05,
+      "loss": 5.4378,
+      "step": 6973
+    },
+    {
+      "epoch": 0.4677868330147231,
+      "grad_norm": 0.14526678749809008,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 6974
+    },
+    {
+      "epoch": 0.46785390884394806,
+      "grad_norm": 0.14838438429484202,
+      "learning_rate": 2e-05,
+      "loss": 5.4366,
+      "step": 6975
+    },
+    {
+      "epoch": 0.467920984673173,
+      "grad_norm": 0.15083819579342492,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 6976
+    },
+    {
+      "epoch": 0.46798806050239794,
+      "grad_norm": 0.15078236004040363,
+      "learning_rate": 2e-05,
+      "loss": 5.3287,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4680551363316229,
+      "grad_norm": 0.14476192290056245,
+      "learning_rate": 2e-05,
+      "loss": 5.4115,
+      "step": 6978
+    },
+    {
+      "epoch": 0.4681222121608478,
+      "grad_norm": 0.1524476741219239,
+      "learning_rate": 2e-05,
+      "loss": 5.3283,
+      "step": 6979
+    },
+    {
+      "epoch": 0.46818928799007276,
+      "grad_norm": 0.15244660659648973,
+      "learning_rate": 2e-05,
+      "loss": 5.522,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4682563638192977,
+      "grad_norm": 0.14850735971232454,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 6981
+    },
+    {
+      "epoch": 0.46832343964852263,
+      "grad_norm": 0.14867585902672484,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 6982
+    },
+    {
+      "epoch": 0.46839051547774757,
+      "grad_norm": 0.15229088513187325,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 6983
+    },
+    {
+      "epoch": 0.4684575913069725,
+      "grad_norm": 0.15368260048624632,
+      "learning_rate": 2e-05,
+      "loss": 5.3887,
+      "step": 6984
+    },
+    {
+      "epoch": 0.46852466713619745,
+      "grad_norm": 0.14816642439735356,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4685917429654224,
+      "grad_norm": 0.15552532520012194,
+      "learning_rate": 2e-05,
+      "loss": 5.3546,
+      "step": 6986
+    },
+    {
+      "epoch": 0.4686588187946473,
+      "grad_norm": 0.15127746452766766,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 6987
+    },
+    {
+      "epoch": 0.46872589462387226,
+      "grad_norm": 0.1458201309497989,
+      "learning_rate": 2e-05,
+      "loss": 5.4824,
+      "step": 6988
+    },
+    {
+      "epoch": 0.4687929704530972,
+      "grad_norm": 0.142185068088717,
+      "learning_rate": 2e-05,
+      "loss": 5.3291,
+      "step": 6989
+    },
+    {
+      "epoch": 0.46886004628232214,
+      "grad_norm": 0.1528851349212613,
+      "learning_rate": 2e-05,
+      "loss": 5.6121,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4689271221115471,
+      "grad_norm": 0.1587520248158114,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 6991
+    },
+    {
+      "epoch": 0.468994197940772,
+      "grad_norm": 0.15525343762500834,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 6992
+    },
+    {
+      "epoch": 0.46906127376999696,
+      "grad_norm": 0.15060961543726586,
+      "learning_rate": 2e-05,
+      "loss": 5.3683,
+      "step": 6993
+    },
+    {
+      "epoch": 0.4691283495992219,
+      "grad_norm": 0.1536314619817886,
+      "learning_rate": 2e-05,
+      "loss": 5.3527,
+      "step": 6994
+    },
+    {
+      "epoch": 0.46919542542844683,
+      "grad_norm": 0.15283819098722823,
+      "learning_rate": 2e-05,
+      "loss": 5.4251,
+      "step": 6995
+    },
+    {
+      "epoch": 0.46926250125767177,
+      "grad_norm": 0.14469738123727,
+      "learning_rate": 2e-05,
+      "loss": 5.2665,
+      "step": 6996
+    },
+    {
+      "epoch": 0.4693295770868967,
+      "grad_norm": 0.15031851550471959,
+      "learning_rate": 2e-05,
+      "loss": 5.3827,
+      "step": 6997
+    },
+    {
+      "epoch": 0.46939665291612165,
+      "grad_norm": 0.15325799241283403,
+      "learning_rate": 2e-05,
+      "loss": 5.4197,
+      "step": 6998
+    },
+    {
+      "epoch": 0.4694637287453466,
+      "grad_norm": 0.1444738715388843,
+      "learning_rate": 2e-05,
+      "loss": 5.4171,
+      "step": 6999
+    },
+    {
+      "epoch": 0.4695308045745716,
+      "grad_norm": 0.15026752290282694,
+      "learning_rate": 2e-05,
+      "loss": 5.2907,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4695978804037965,
+      "grad_norm": 0.15124675083582748,
+      "learning_rate": 2e-05,
+      "loss": 5.3846,
+      "step": 7001
+    },
+    {
+      "epoch": 0.46966495623302146,
+      "grad_norm": 0.15140967716987938,
+      "learning_rate": 2e-05,
+      "loss": 5.5107,
+      "step": 7002
+    },
+    {
+      "epoch": 0.4697320320622464,
+      "grad_norm": 0.1580748981080875,
+      "learning_rate": 2e-05,
+      "loss": 5.4361,
+      "step": 7003
+    },
+    {
+      "epoch": 0.46979910789147133,
+      "grad_norm": 0.1523201867589836,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4698661837206963,
+      "grad_norm": 0.14733243919103625,
+      "learning_rate": 2e-05,
+      "loss": 5.3569,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4699332595499212,
+      "grad_norm": 0.14837503589564116,
+      "learning_rate": 2e-05,
+      "loss": 5.3707,
+      "step": 7006
+    },
+    {
+      "epoch": 0.47000033537914615,
+      "grad_norm": 0.1457263705325594,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 7007
+    },
+    {
+      "epoch": 0.4700674112083711,
+      "grad_norm": 0.14606040745015228,
+      "learning_rate": 2e-05,
+      "loss": 5.4696,
+      "step": 7008
+    },
+    {
+      "epoch": 0.470134487037596,
+      "grad_norm": 0.15692607254318305,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 7009
+    },
+    {
+      "epoch": 0.47020156286682097,
+      "grad_norm": 0.15575720115298333,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4702686386960459,
+      "grad_norm": 0.14326998630258567,
+      "learning_rate": 2e-05,
+      "loss": 5.4348,
+      "step": 7011
+    },
+    {
+      "epoch": 0.47033571452527084,
+      "grad_norm": 0.144821808548869,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 7012
+    },
+    {
+      "epoch": 0.4704027903544958,
+      "grad_norm": 0.15347878031280804,
+      "learning_rate": 2e-05,
+      "loss": 5.339,
+      "step": 7013
+    },
+    {
+      "epoch": 0.4704698661837207,
+      "grad_norm": 0.15847225713464538,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 7014
+    },
+    {
+      "epoch": 0.47053694201294566,
+      "grad_norm": 0.1455759737754067,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4706040178421706,
+      "grad_norm": 0.15040515834481022,
+      "learning_rate": 2e-05,
+      "loss": 5.4443,
+      "step": 7016
+    },
+    {
+      "epoch": 0.47067109367139554,
+      "grad_norm": 0.15125768525968966,
+      "learning_rate": 2e-05,
+      "loss": 5.4246,
+      "step": 7017
+    },
+    {
+      "epoch": 0.4707381695006205,
+      "grad_norm": 0.15180455601271675,
+      "learning_rate": 2e-05,
+      "loss": 5.4035,
+      "step": 7018
+    },
+    {
+      "epoch": 0.4708052453298454,
+      "grad_norm": 0.15198463968325512,
+      "learning_rate": 2e-05,
+      "loss": 5.5136,
+      "step": 7019
+    },
+    {
+      "epoch": 0.47087232115907035,
+      "grad_norm": 0.16574946399870077,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4709393969882953,
+      "grad_norm": 0.14894287388374852,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 7021
+    },
+    {
+      "epoch": 0.4710064728175202,
+      "grad_norm": 0.15069343913289585,
+      "learning_rate": 2e-05,
+      "loss": 5.3528,
+      "step": 7022
+    },
+    {
+      "epoch": 0.47107354864674517,
+      "grad_norm": 0.16120201934583922,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 7023
+    },
+    {
+      "epoch": 0.4711406244759701,
+      "grad_norm": 0.14591869113627476,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 7024
+    },
+    {
+      "epoch": 0.47120770030519504,
+      "grad_norm": 0.14396289683946004,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 7025
+    },
+    {
+      "epoch": 0.47127477613442,
+      "grad_norm": 0.1549584987498031,
+      "learning_rate": 2e-05,
+      "loss": 5.4564,
+      "step": 7026
+    },
+    {
+      "epoch": 0.4713418519636449,
+      "grad_norm": 0.14930450162423045,
+      "learning_rate": 2e-05,
+      "loss": 5.5726,
+      "step": 7027
+    },
+    {
+      "epoch": 0.47140892779286986,
+      "grad_norm": 0.14999222937203896,
+      "learning_rate": 2e-05,
+      "loss": 5.5259,
+      "step": 7028
+    },
+    {
+      "epoch": 0.4714760036220948,
+      "grad_norm": 0.14997826572325013,
+      "learning_rate": 2e-05,
+      "loss": 5.3575,
+      "step": 7029
+    },
+    {
+      "epoch": 0.47154307945131974,
+      "grad_norm": 0.15548288073005398,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4716101552805447,
+      "grad_norm": 0.14886411229145416,
+      "learning_rate": 2e-05,
+      "loss": 5.4284,
+      "step": 7031
+    },
+    {
+      "epoch": 0.4716772311097696,
+      "grad_norm": 0.15337719805985794,
+      "learning_rate": 2e-05,
+      "loss": 5.5198,
+      "step": 7032
+    },
+    {
+      "epoch": 0.47174430693899455,
+      "grad_norm": 0.1502071705543121,
+      "learning_rate": 2e-05,
+      "loss": 5.545,
+      "step": 7033
+    },
+    {
+      "epoch": 0.4718113827682195,
+      "grad_norm": 0.1476980118117674,
+      "learning_rate": 2e-05,
+      "loss": 5.3544,
+      "step": 7034
+    },
+    {
+      "epoch": 0.47187845859744443,
+      "grad_norm": 0.1457508672342308,
+      "learning_rate": 2e-05,
+      "loss": 5.5156,
+      "step": 7035
+    },
+    {
+      "epoch": 0.47194553442666937,
+      "grad_norm": 0.14546539959259364,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 7036
+    },
+    {
+      "epoch": 0.4720126102558943,
+      "grad_norm": 0.14592919466654572,
+      "learning_rate": 2e-05,
+      "loss": 5.4729,
+      "step": 7037
+    },
+    {
+      "epoch": 0.47207968608511924,
+      "grad_norm": 0.14557976831231637,
+      "learning_rate": 2e-05,
+      "loss": 5.3494,
+      "step": 7038
+    },
+    {
+      "epoch": 0.4721467619143442,
+      "grad_norm": 0.15221546779516065,
+      "learning_rate": 2e-05,
+      "loss": 5.2691,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4722138377435691,
+      "grad_norm": 0.15445237215868854,
+      "learning_rate": 2e-05,
+      "loss": 5.2303,
+      "step": 7040
+    },
+    {
+      "epoch": 0.47228091357279406,
+      "grad_norm": 0.1482400080966502,
+      "learning_rate": 2e-05,
+      "loss": 5.3332,
+      "step": 7041
+    },
+    {
+      "epoch": 0.472347989402019,
+      "grad_norm": 0.1453974738415693,
+      "learning_rate": 2e-05,
+      "loss": 5.5634,
+      "step": 7042
+    },
+    {
+      "epoch": 0.47241506523124394,
+      "grad_norm": 0.14333494847416883,
+      "learning_rate": 2e-05,
+      "loss": 5.5008,
+      "step": 7043
+    },
+    {
+      "epoch": 0.4724821410604689,
+      "grad_norm": 0.15001001145412718,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 7044
+    },
+    {
+      "epoch": 0.4725492168896938,
+      "grad_norm": 0.1476244881701856,
+      "learning_rate": 2e-05,
+      "loss": 5.3958,
+      "step": 7045
+    },
+    {
+      "epoch": 0.47261629271891875,
+      "grad_norm": 0.1545180779495855,
+      "learning_rate": 2e-05,
+      "loss": 5.4396,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4726833685481437,
+      "grad_norm": 0.14568541040636881,
+      "learning_rate": 2e-05,
+      "loss": 5.4488,
+      "step": 7047
+    },
+    {
+      "epoch": 0.47275044437736863,
+      "grad_norm": 0.15233894199417186,
+      "learning_rate": 2e-05,
+      "loss": 5.376,
+      "step": 7048
+    },
+    {
+      "epoch": 0.47281752020659357,
+      "grad_norm": 0.1525453208961433,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 7049
+    },
+    {
+      "epoch": 0.4728845960358185,
+      "grad_norm": 0.1501492999166353,
+      "learning_rate": 2e-05,
+      "loss": 5.4358,
+      "step": 7050
+    },
+    {
+      "epoch": 0.47295167186504344,
+      "grad_norm": 0.15407883866179395,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 7051
+    },
+    {
+      "epoch": 0.4730187476942684,
+      "grad_norm": 0.15416804151531227,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 7052
+    },
+    {
+      "epoch": 0.4730858235234933,
+      "grad_norm": 0.15495579803064236,
+      "learning_rate": 2e-05,
+      "loss": 5.4438,
+      "step": 7053
+    },
+    {
+      "epoch": 0.47315289935271826,
+      "grad_norm": 0.16426293962551494,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 7054
+    },
+    {
+      "epoch": 0.4732199751819432,
+      "grad_norm": 0.15300237598355898,
+      "learning_rate": 2e-05,
+      "loss": 5.3235,
+      "step": 7055
+    },
+    {
+      "epoch": 0.47328705101116814,
+      "grad_norm": 0.15259624293121657,
+      "learning_rate": 2e-05,
+      "loss": 5.3646,
+      "step": 7056
+    },
+    {
+      "epoch": 0.4733541268403931,
+      "grad_norm": 0.15888804921462327,
+      "learning_rate": 2e-05,
+      "loss": 5.334,
+      "step": 7057
+    },
+    {
+      "epoch": 0.473421202669618,
+      "grad_norm": 0.17105387336145111,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 7058
+    },
+    {
+      "epoch": 0.47348827849884295,
+      "grad_norm": 0.16155212181007028,
+      "learning_rate": 2e-05,
+      "loss": 5.5105,
+      "step": 7059
+    },
+    {
+      "epoch": 0.4735553543280679,
+      "grad_norm": 0.15102749878527363,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 7060
+    },
+    {
+      "epoch": 0.47362243015729283,
+      "grad_norm": 0.1496486713186681,
+      "learning_rate": 2e-05,
+      "loss": 5.2933,
+      "step": 7061
+    },
+    {
+      "epoch": 0.47368950598651777,
+      "grad_norm": 0.15676027513442467,
+      "learning_rate": 2e-05,
+      "loss": 5.5596,
+      "step": 7062
+    },
+    {
+      "epoch": 0.4737565818157427,
+      "grad_norm": 0.15441913640555863,
+      "learning_rate": 2e-05,
+      "loss": 5.3985,
+      "step": 7063
+    },
+    {
+      "epoch": 0.47382365764496764,
+      "grad_norm": 0.15608366308836696,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 7064
+    },
+    {
+      "epoch": 0.4738907334741926,
+      "grad_norm": 0.15554032002952767,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 7065
+    },
+    {
+      "epoch": 0.4739578093034175,
+      "grad_norm": 0.1565772864772116,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 7066
+    },
+    {
+      "epoch": 0.47402488513264246,
+      "grad_norm": 0.15530276000923693,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 7067
+    },
+    {
+      "epoch": 0.4740919609618674,
+      "grad_norm": 0.14587650640771035,
+      "learning_rate": 2e-05,
+      "loss": 5.4018,
+      "step": 7068
+    },
+    {
+      "epoch": 0.47415903679109234,
+      "grad_norm": 0.14556864064802366,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 7069
+    },
+    {
+      "epoch": 0.4742261126203173,
+      "grad_norm": 0.1563257736899565,
+      "learning_rate": 2e-05,
+      "loss": 5.4527,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4742931884495422,
+      "grad_norm": 0.1482914558110726,
+      "learning_rate": 2e-05,
+      "loss": 5.3431,
+      "step": 7071
+    },
+    {
+      "epoch": 0.47436026427876715,
+      "grad_norm": 0.1465860892842665,
+      "learning_rate": 2e-05,
+      "loss": 5.5619,
+      "step": 7072
+    },
+    {
+      "epoch": 0.4744273401079921,
+      "grad_norm": 0.1512103598922994,
+      "learning_rate": 2e-05,
+      "loss": 5.3591,
+      "step": 7073
+    },
+    {
+      "epoch": 0.47449441593721703,
+      "grad_norm": 0.1454430740876668,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 7074
+    },
+    {
+      "epoch": 0.47456149176644197,
+      "grad_norm": 0.1518782647031839,
+      "learning_rate": 2e-05,
+      "loss": 5.525,
+      "step": 7075
+    },
+    {
+      "epoch": 0.4746285675956669,
+      "grad_norm": 0.1589013008251621,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 7076
+    },
+    {
+      "epoch": 0.47469564342489184,
+      "grad_norm": 0.15092694733649542,
+      "learning_rate": 2e-05,
+      "loss": 5.4905,
+      "step": 7077
+    },
+    {
+      "epoch": 0.4747627192541168,
+      "grad_norm": 0.14860782855098018,
+      "learning_rate": 2e-05,
+      "loss": 5.5149,
+      "step": 7078
+    },
+    {
+      "epoch": 0.4748297950833417,
+      "grad_norm": 0.14835289670659169,
+      "learning_rate": 2e-05,
+      "loss": 5.4625,
+      "step": 7079
+    },
+    {
+      "epoch": 0.47489687091256666,
+      "grad_norm": 0.15095604646665606,
+      "learning_rate": 2e-05,
+      "loss": 5.4848,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4749639467417916,
+      "grad_norm": 0.15081421556807773,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 7081
+    },
+    {
+      "epoch": 0.47503102257101654,
+      "grad_norm": 0.15639563614816515,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 7082
+    },
+    {
+      "epoch": 0.4750980984002415,
+      "grad_norm": 0.14450567767434483,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 7083
+    },
+    {
+      "epoch": 0.4751651742294664,
+      "grad_norm": 0.15075090309577313,
+      "learning_rate": 2e-05,
+      "loss": 5.5447,
+      "step": 7084
+    },
+    {
+      "epoch": 0.47523225005869135,
+      "grad_norm": 0.17026457186549246,
+      "learning_rate": 2e-05,
+      "loss": 5.4593,
+      "step": 7085
+    },
+    {
+      "epoch": 0.4752993258879163,
+      "grad_norm": 0.151831800897495,
+      "learning_rate": 2e-05,
+      "loss": 5.3146,
+      "step": 7086
+    },
+    {
+      "epoch": 0.47536640171714123,
+      "grad_norm": 0.16474019473008378,
+      "learning_rate": 2e-05,
+      "loss": 5.3182,
+      "step": 7087
+    },
+    {
+      "epoch": 0.47543347754636617,
+      "grad_norm": 0.15141010858655685,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 7088
+    },
+    {
+      "epoch": 0.4755005533755911,
+      "grad_norm": 0.1495741489809151,
+      "learning_rate": 2e-05,
+      "loss": 5.5541,
+      "step": 7089
+    },
+    {
+      "epoch": 0.47556762920481604,
+      "grad_norm": 0.14818934058305516,
+      "learning_rate": 2e-05,
+      "loss": 5.374,
+      "step": 7090
+    },
+    {
+      "epoch": 0.475634705034041,
+      "grad_norm": 0.1505527596122846,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 7091
+    },
+    {
+      "epoch": 0.4757017808632659,
+      "grad_norm": 0.14557337942761822,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 7092
+    },
+    {
+      "epoch": 0.47576885669249086,
+      "grad_norm": 0.14912953077347485,
+      "learning_rate": 2e-05,
+      "loss": 5.4288,
+      "step": 7093
+    },
+    {
+      "epoch": 0.4758359325217158,
+      "grad_norm": 0.14725768142476706,
+      "learning_rate": 2e-05,
+      "loss": 5.4856,
+      "step": 7094
+    },
+    {
+      "epoch": 0.47590300835094074,
+      "grad_norm": 0.14971096477074916,
+      "learning_rate": 2e-05,
+      "loss": 5.3406,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4759700841801657,
+      "grad_norm": 0.15454710778429775,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 7096
+    },
+    {
+      "epoch": 0.4760371600093906,
+      "grad_norm": 0.15122767531304657,
+      "learning_rate": 2e-05,
+      "loss": 5.3907,
+      "step": 7097
+    },
+    {
+      "epoch": 0.47610423583861555,
+      "grad_norm": 0.1495880296343396,
+      "learning_rate": 2e-05,
+      "loss": 5.3662,
+      "step": 7098
+    },
+    {
+      "epoch": 0.4761713116678405,
+      "grad_norm": 0.14866414092183727,
+      "learning_rate": 2e-05,
+      "loss": 5.3875,
+      "step": 7099
+    },
+    {
+      "epoch": 0.47623838749706543,
+      "grad_norm": 0.1550179621638455,
+      "learning_rate": 2e-05,
+      "loss": 5.563,
+      "step": 7100
+    },
+    {
+      "epoch": 0.47630546332629037,
+      "grad_norm": 0.14794742047900147,
+      "learning_rate": 2e-05,
+      "loss": 5.4479,
+      "step": 7101
+    },
+    {
+      "epoch": 0.4763725391555153,
+      "grad_norm": 0.15181509113987407,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 7102
+    },
+    {
+      "epoch": 0.47643961498474025,
+      "grad_norm": 0.15324335247814216,
+      "learning_rate": 2e-05,
+      "loss": 5.4017,
+      "step": 7103
+    },
+    {
+      "epoch": 0.4765066908139652,
+      "grad_norm": 0.1503073529074295,
+      "learning_rate": 2e-05,
+      "loss": 5.4335,
+      "step": 7104
+    },
+    {
+      "epoch": 0.4765737666431901,
+      "grad_norm": 0.16218907575707472,
+      "learning_rate": 2e-05,
+      "loss": 5.353,
+      "step": 7105
+    },
+    {
+      "epoch": 0.47664084247241506,
+      "grad_norm": 0.16973919978090304,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 7106
+    },
+    {
+      "epoch": 0.47670791830164,
+      "grad_norm": 0.14840798072309921,
+      "learning_rate": 2e-05,
+      "loss": 5.5148,
+      "step": 7107
+    },
+    {
+      "epoch": 0.47677499413086494,
+      "grad_norm": 0.15403197375771643,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 7108
+    },
+    {
+      "epoch": 0.4768420699600899,
+      "grad_norm": 0.15853772750421377,
+      "learning_rate": 2e-05,
+      "loss": 5.3163,
+      "step": 7109
+    },
+    {
+      "epoch": 0.4769091457893148,
+      "grad_norm": 0.1539396463714218,
+      "learning_rate": 2e-05,
+      "loss": 5.3515,
+      "step": 7110
+    },
+    {
+      "epoch": 0.47697622161853975,
+      "grad_norm": 0.16022732707726917,
+      "learning_rate": 2e-05,
+      "loss": 5.3171,
+      "step": 7111
+    },
+    {
+      "epoch": 0.4770432974477647,
+      "grad_norm": 0.16481221026623305,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 7112
+    },
+    {
+      "epoch": 0.47711037327698963,
+      "grad_norm": 0.1540770060749335,
+      "learning_rate": 2e-05,
+      "loss": 5.5448,
+      "step": 7113
+    },
+    {
+      "epoch": 0.47717744910621457,
+      "grad_norm": 0.15249764515966305,
+      "learning_rate": 2e-05,
+      "loss": 5.4576,
+      "step": 7114
+    },
+    {
+      "epoch": 0.4772445249354395,
+      "grad_norm": 0.1606215745211595,
+      "learning_rate": 2e-05,
+      "loss": 5.4905,
+      "step": 7115
+    },
+    {
+      "epoch": 0.47731160076466445,
+      "grad_norm": 0.1580556603853395,
+      "learning_rate": 2e-05,
+      "loss": 5.5124,
+      "step": 7116
+    },
+    {
+      "epoch": 0.4773786765938894,
+      "grad_norm": 0.15495528731623578,
+      "learning_rate": 2e-05,
+      "loss": 5.5091,
+      "step": 7117
+    },
+    {
+      "epoch": 0.4774457524231143,
+      "grad_norm": 0.16303702654424315,
+      "learning_rate": 2e-05,
+      "loss": 5.3566,
+      "step": 7118
+    },
+    {
+      "epoch": 0.47751282825233926,
+      "grad_norm": 0.14742613795220103,
+      "learning_rate": 2e-05,
+      "loss": 5.4279,
+      "step": 7119
+    },
+    {
+      "epoch": 0.4775799040815642,
+      "grad_norm": 0.15846344948614238,
+      "learning_rate": 2e-05,
+      "loss": 5.3838,
+      "step": 7120
+    },
+    {
+      "epoch": 0.47764697991078914,
+      "grad_norm": 0.1551969568212782,
+      "learning_rate": 2e-05,
+      "loss": 5.3616,
+      "step": 7121
+    },
+    {
+      "epoch": 0.4777140557400141,
+      "grad_norm": 0.14348958252661287,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 7122
+    },
+    {
+      "epoch": 0.477781131569239,
+      "grad_norm": 0.15796100738070057,
+      "learning_rate": 2e-05,
+      "loss": 5.3398,
+      "step": 7123
+    },
+    {
+      "epoch": 0.47784820739846395,
+      "grad_norm": 0.1435400115671241,
+      "learning_rate": 2e-05,
+      "loss": 5.3768,
+      "step": 7124
+    },
+    {
+      "epoch": 0.4779152832276889,
+      "grad_norm": 0.14541366683966472,
+      "learning_rate": 2e-05,
+      "loss": 5.4386,
+      "step": 7125
+    },
+    {
+      "epoch": 0.47798235905691383,
+      "grad_norm": 0.14549469605691437,
+      "learning_rate": 2e-05,
+      "loss": 5.4135,
+      "step": 7126
+    },
+    {
+      "epoch": 0.47804943488613877,
+      "grad_norm": 0.14391941233102157,
+      "learning_rate": 2e-05,
+      "loss": 5.3327,
+      "step": 7127
+    },
+    {
+      "epoch": 0.4781165107153637,
+      "grad_norm": 0.15190921717238606,
+      "learning_rate": 2e-05,
+      "loss": 5.4388,
+      "step": 7128
+    },
+    {
+      "epoch": 0.47818358654458865,
+      "grad_norm": 0.1473232117612222,
+      "learning_rate": 2e-05,
+      "loss": 5.4679,
+      "step": 7129
+    },
+    {
+      "epoch": 0.4782506623738136,
+      "grad_norm": 0.1471886168172578,
+      "learning_rate": 2e-05,
+      "loss": 5.4574,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4783177382030385,
+      "grad_norm": 0.15140057146117514,
+      "learning_rate": 2e-05,
+      "loss": 5.4886,
+      "step": 7131
+    },
+    {
+      "epoch": 0.47838481403226346,
+      "grad_norm": 0.15301364638501258,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 7132
+    },
+    {
+      "epoch": 0.4784518898614884,
+      "grad_norm": 0.14662513096978805,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 7133
+    },
+    {
+      "epoch": 0.47851896569071334,
+      "grad_norm": 0.14906702573487288,
+      "learning_rate": 2e-05,
+      "loss": 5.4397,
+      "step": 7134
+    },
+    {
+      "epoch": 0.4785860415199383,
+      "grad_norm": 0.1522302810951062,
+      "learning_rate": 2e-05,
+      "loss": 5.4627,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4786531173491632,
+      "grad_norm": 0.15712710944592592,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 7136
+    },
+    {
+      "epoch": 0.47872019317838815,
+      "grad_norm": 0.1463798587314211,
+      "learning_rate": 2e-05,
+      "loss": 5.5135,
+      "step": 7137
+    },
+    {
+      "epoch": 0.4787872690076131,
+      "grad_norm": 0.14738389234853552,
+      "learning_rate": 2e-05,
+      "loss": 5.3001,
+      "step": 7138
+    },
+    {
+      "epoch": 0.47885434483683803,
+      "grad_norm": 0.15056120649457233,
+      "learning_rate": 2e-05,
+      "loss": 5.5988,
+      "step": 7139
+    },
+    {
+      "epoch": 0.47892142066606297,
+      "grad_norm": 0.14486173798406385,
+      "learning_rate": 2e-05,
+      "loss": 5.5992,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4789884964952879,
+      "grad_norm": 0.14951438366330638,
+      "learning_rate": 2e-05,
+      "loss": 5.3339,
+      "step": 7141
+    },
+    {
+      "epoch": 0.47905557232451285,
+      "grad_norm": 0.14523592284045908,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 7142
+    },
+    {
+      "epoch": 0.4791226481537378,
+      "grad_norm": 0.14637461306745242,
+      "learning_rate": 2e-05,
+      "loss": 5.3278,
+      "step": 7143
+    },
+    {
+      "epoch": 0.4791897239829627,
+      "grad_norm": 0.1471406674901263,
+      "learning_rate": 2e-05,
+      "loss": 5.239,
+      "step": 7144
+    },
+    {
+      "epoch": 0.47925679981218766,
+      "grad_norm": 0.15042059536662566,
+      "learning_rate": 2e-05,
+      "loss": 5.2991,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4793238756414126,
+      "grad_norm": 0.15012075157493426,
+      "learning_rate": 2e-05,
+      "loss": 5.4285,
+      "step": 7146
+    },
+    {
+      "epoch": 0.47939095147063754,
+      "grad_norm": 0.14980168960914758,
+      "learning_rate": 2e-05,
+      "loss": 5.4169,
+      "step": 7147
+    },
+    {
+      "epoch": 0.4794580272998625,
+      "grad_norm": 0.1486573783646942,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 7148
+    },
+    {
+      "epoch": 0.4795251031290874,
+      "grad_norm": 0.14805029127774072,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 7149
+    },
+    {
+      "epoch": 0.47959217895831235,
+      "grad_norm": 0.15585679469674993,
+      "learning_rate": 2e-05,
+      "loss": 5.3585,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4796592547875373,
+      "grad_norm": 0.14435586982336682,
+      "learning_rate": 2e-05,
+      "loss": 5.3387,
+      "step": 7151
+    },
+    {
+      "epoch": 0.47972633061676223,
+      "grad_norm": 0.14640287229526952,
+      "learning_rate": 2e-05,
+      "loss": 5.3128,
+      "step": 7152
+    },
+    {
+      "epoch": 0.47979340644598717,
+      "grad_norm": 0.14796852265759888,
+      "learning_rate": 2e-05,
+      "loss": 5.5741,
+      "step": 7153
+    },
+    {
+      "epoch": 0.4798604822752121,
+      "grad_norm": 0.14502091283514357,
+      "learning_rate": 2e-05,
+      "loss": 5.5194,
+      "step": 7154
+    },
+    {
+      "epoch": 0.47992755810443705,
+      "grad_norm": 0.15494615208648788,
+      "learning_rate": 2e-05,
+      "loss": 5.4171,
+      "step": 7155
+    },
+    {
+      "epoch": 0.479994633933662,
+      "grad_norm": 0.15149844425794795,
+      "learning_rate": 2e-05,
+      "loss": 5.3781,
+      "step": 7156
+    },
+    {
+      "epoch": 0.4800617097628869,
+      "grad_norm": 0.1466702461556769,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 7157
+    },
+    {
+      "epoch": 0.48012878559211186,
+      "grad_norm": 0.1588476330491587,
+      "learning_rate": 2e-05,
+      "loss": 5.4872,
+      "step": 7158
+    },
+    {
+      "epoch": 0.4801958614213368,
+      "grad_norm": 0.1515707869445218,
+      "learning_rate": 2e-05,
+      "loss": 5.4077,
+      "step": 7159
+    },
+    {
+      "epoch": 0.48026293725056174,
+      "grad_norm": 0.14864861657478365,
+      "learning_rate": 2e-05,
+      "loss": 5.4259,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4803300130797867,
+      "grad_norm": 0.15255478337652367,
+      "learning_rate": 2e-05,
+      "loss": 5.4176,
+      "step": 7161
+    },
+    {
+      "epoch": 0.4803970889090116,
+      "grad_norm": 0.15897453247855622,
+      "learning_rate": 2e-05,
+      "loss": 5.5404,
+      "step": 7162
+    },
+    {
+      "epoch": 0.48046416473823655,
+      "grad_norm": 0.15044540689810043,
+      "learning_rate": 2e-05,
+      "loss": 5.4386,
+      "step": 7163
+    },
+    {
+      "epoch": 0.4805312405674615,
+      "grad_norm": 0.1556626351240373,
+      "learning_rate": 2e-05,
+      "loss": 5.4885,
+      "step": 7164
+    },
+    {
+      "epoch": 0.48059831639668643,
+      "grad_norm": 0.14595894743633367,
+      "learning_rate": 2e-05,
+      "loss": 5.4048,
+      "step": 7165
+    },
+    {
+      "epoch": 0.48066539222591137,
+      "grad_norm": 0.15134208399259938,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 7166
+    },
+    {
+      "epoch": 0.4807324680551363,
+      "grad_norm": 0.15151241377607125,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 7167
+    },
+    {
+      "epoch": 0.48079954388436125,
+      "grad_norm": 0.15802932347009083,
+      "learning_rate": 2e-05,
+      "loss": 5.4681,
+      "step": 7168
+    },
+    {
+      "epoch": 0.4808666197135862,
+      "grad_norm": 0.15153621920174257,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 7169
+    },
+    {
+      "epoch": 0.4809336955428111,
+      "grad_norm": 0.14705603366068126,
+      "learning_rate": 2e-05,
+      "loss": 5.4962,
+      "step": 7170
+    },
+    {
+      "epoch": 0.48100077137203606,
+      "grad_norm": 0.156340159350827,
+      "learning_rate": 2e-05,
+      "loss": 5.4464,
+      "step": 7171
+    },
+    {
+      "epoch": 0.481067847201261,
+      "grad_norm": 0.15794460705235147,
+      "learning_rate": 2e-05,
+      "loss": 5.3045,
+      "step": 7172
+    },
+    {
+      "epoch": 0.48113492303048594,
+      "grad_norm": 0.1438930600130996,
+      "learning_rate": 2e-05,
+      "loss": 5.5131,
+      "step": 7173
+    },
+    {
+      "epoch": 0.4812019988597109,
+      "grad_norm": 0.14906007530156354,
+      "learning_rate": 2e-05,
+      "loss": 5.4346,
+      "step": 7174
+    },
+    {
+      "epoch": 0.4812690746889358,
+      "grad_norm": 0.14930351332128874,
+      "learning_rate": 2e-05,
+      "loss": 5.4465,
+      "step": 7175
+    },
+    {
+      "epoch": 0.48133615051816075,
+      "grad_norm": 0.15303461345419767,
+      "learning_rate": 2e-05,
+      "loss": 5.4803,
+      "step": 7176
+    },
+    {
+      "epoch": 0.4814032263473857,
+      "grad_norm": 0.14967384017253055,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 7177
+    },
+    {
+      "epoch": 0.48147030217661063,
+      "grad_norm": 0.1477154932622654,
+      "learning_rate": 2e-05,
+      "loss": 5.3413,
+      "step": 7178
+    },
+    {
+      "epoch": 0.48153737800583557,
+      "grad_norm": 0.1462549150576034,
+      "learning_rate": 2e-05,
+      "loss": 5.4788,
+      "step": 7179
+    },
+    {
+      "epoch": 0.4816044538350605,
+      "grad_norm": 0.15129461953756793,
+      "learning_rate": 2e-05,
+      "loss": 5.4202,
+      "step": 7180
+    },
+    {
+      "epoch": 0.48167152966428545,
+      "grad_norm": 0.15036248402378666,
+      "learning_rate": 2e-05,
+      "loss": 5.448,
+      "step": 7181
+    },
+    {
+      "epoch": 0.48173860549351044,
+      "grad_norm": 0.14607043923221408,
+      "learning_rate": 2e-05,
+      "loss": 5.4706,
+      "step": 7182
+    },
+    {
+      "epoch": 0.4818056813227354,
+      "grad_norm": 0.14673759201158743,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 7183
+    },
+    {
+      "epoch": 0.4818727571519603,
+      "grad_norm": 0.15376199553197617,
+      "learning_rate": 2e-05,
+      "loss": 5.3919,
+      "step": 7184
+    },
+    {
+      "epoch": 0.48193983298118526,
+      "grad_norm": 0.14706334524811973,
+      "learning_rate": 2e-05,
+      "loss": 5.3281,
+      "step": 7185
+    },
+    {
+      "epoch": 0.4820069088104102,
+      "grad_norm": 0.15865856994213615,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 7186
+    },
+    {
+      "epoch": 0.48207398463963513,
+      "grad_norm": 0.15233387825524852,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 7187
+    },
+    {
+      "epoch": 0.48214106046886007,
+      "grad_norm": 0.15114422030740565,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 7188
+    },
+    {
+      "epoch": 0.482208136298085,
+      "grad_norm": 0.1595546624424555,
+      "learning_rate": 2e-05,
+      "loss": 5.2434,
+      "step": 7189
+    },
+    {
+      "epoch": 0.48227521212730995,
+      "grad_norm": 0.15308210518761203,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4823422879565349,
+      "grad_norm": 0.15759850558057317,
+      "learning_rate": 2e-05,
+      "loss": 5.3783,
+      "step": 7191
+    },
+    {
+      "epoch": 0.4824093637857598,
+      "grad_norm": 0.15214376454907774,
+      "learning_rate": 2e-05,
+      "loss": 5.4918,
+      "step": 7192
+    },
+    {
+      "epoch": 0.48247643961498476,
+      "grad_norm": 0.15621428669641066,
+      "learning_rate": 2e-05,
+      "loss": 5.4321,
+      "step": 7193
+    },
+    {
+      "epoch": 0.4825435154442097,
+      "grad_norm": 0.14971181772536118,
+      "learning_rate": 2e-05,
+      "loss": 5.4541,
+      "step": 7194
+    },
+    {
+      "epoch": 0.48261059127343464,
+      "grad_norm": 0.14985748591935463,
+      "learning_rate": 2e-05,
+      "loss": 5.4902,
+      "step": 7195
+    },
+    {
+      "epoch": 0.4826776671026596,
+      "grad_norm": 0.1452995024340614,
+      "learning_rate": 2e-05,
+      "loss": 5.3554,
+      "step": 7196
+    },
+    {
+      "epoch": 0.4827447429318845,
+      "grad_norm": 0.1512344731888584,
+      "learning_rate": 2e-05,
+      "loss": 5.538,
+      "step": 7197
+    },
+    {
+      "epoch": 0.48281181876110946,
+      "grad_norm": 0.15236855763933382,
+      "learning_rate": 2e-05,
+      "loss": 5.59,
+      "step": 7198
+    },
+    {
+      "epoch": 0.4828788945903344,
+      "grad_norm": 0.14721440172251093,
+      "learning_rate": 2e-05,
+      "loss": 5.4057,
+      "step": 7199
+    },
+    {
+      "epoch": 0.48294597041955933,
+      "grad_norm": 0.15394570408204344,
+      "learning_rate": 2e-05,
+      "loss": 5.2487,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4830130462487843,
+      "grad_norm": 0.15833976955644558,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 7201
+    },
+    {
+      "epoch": 0.4830801220780092,
+      "grad_norm": 0.1460480833587102,
+      "learning_rate": 2e-05,
+      "loss": 5.5769,
+      "step": 7202
+    },
+    {
+      "epoch": 0.48314719790723415,
+      "grad_norm": 0.14901336553274924,
+      "learning_rate": 2e-05,
+      "loss": 5.4396,
+      "step": 7203
+    },
+    {
+      "epoch": 0.4832142737364591,
+      "grad_norm": 0.14973673641754803,
+      "learning_rate": 2e-05,
+      "loss": 5.3468,
+      "step": 7204
+    },
+    {
+      "epoch": 0.483281349565684,
+      "grad_norm": 0.14952930254309985,
+      "learning_rate": 2e-05,
+      "loss": 5.4307,
+      "step": 7205
+    },
+    {
+      "epoch": 0.48334842539490896,
+      "grad_norm": 0.15457975011053585,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 7206
+    },
+    {
+      "epoch": 0.4834155012241339,
+      "grad_norm": 0.15670827096402998,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 7207
+    },
+    {
+      "epoch": 0.48348257705335884,
+      "grad_norm": 0.14431913658098558,
+      "learning_rate": 2e-05,
+      "loss": 5.5043,
+      "step": 7208
+    },
+    {
+      "epoch": 0.4835496528825838,
+      "grad_norm": 0.14900481919289224,
+      "learning_rate": 2e-05,
+      "loss": 5.5302,
+      "step": 7209
+    },
+    {
+      "epoch": 0.4836167287118087,
+      "grad_norm": 0.16257939168451774,
+      "learning_rate": 2e-05,
+      "loss": 5.375,
+      "step": 7210
+    },
+    {
+      "epoch": 0.48368380454103366,
+      "grad_norm": 0.15706419470573438,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 7211
+    },
+    {
+      "epoch": 0.4837508803702586,
+      "grad_norm": 0.14698180286726303,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 7212
+    },
+    {
+      "epoch": 0.48381795619948353,
+      "grad_norm": 0.16144182112517527,
+      "learning_rate": 2e-05,
+      "loss": 5.5842,
+      "step": 7213
+    },
+    {
+      "epoch": 0.4838850320287085,
+      "grad_norm": 0.15160042566855805,
+      "learning_rate": 2e-05,
+      "loss": 5.2685,
+      "step": 7214
+    },
+    {
+      "epoch": 0.4839521078579334,
+      "grad_norm": 0.1479223062215044,
+      "learning_rate": 2e-05,
+      "loss": 5.4028,
+      "step": 7215
+    },
+    {
+      "epoch": 0.48401918368715835,
+      "grad_norm": 0.1723480593939872,
+      "learning_rate": 2e-05,
+      "loss": 5.4066,
+      "step": 7216
+    },
+    {
+      "epoch": 0.4840862595163833,
+      "grad_norm": 0.14556003276166202,
+      "learning_rate": 2e-05,
+      "loss": 5.4499,
+      "step": 7217
+    },
+    {
+      "epoch": 0.4841533353456082,
+      "grad_norm": 0.14727492038290915,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 7218
+    },
+    {
+      "epoch": 0.48422041117483317,
+      "grad_norm": 0.15191141240931258,
+      "learning_rate": 2e-05,
+      "loss": 5.3853,
+      "step": 7219
+    },
+    {
+      "epoch": 0.4842874870040581,
+      "grad_norm": 0.15748275258328945,
+      "learning_rate": 2e-05,
+      "loss": 5.3311,
+      "step": 7220
+    },
+    {
+      "epoch": 0.48435456283328304,
+      "grad_norm": 0.15367027542254708,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 7221
+    },
+    {
+      "epoch": 0.484421638662508,
+      "grad_norm": 0.15937339210677987,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 7222
+    },
+    {
+      "epoch": 0.4844887144917329,
+      "grad_norm": 0.1505050570673283,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 7223
+    },
+    {
+      "epoch": 0.48455579032095786,
+      "grad_norm": 0.1561622894509671,
+      "learning_rate": 2e-05,
+      "loss": 5.4289,
+      "step": 7224
+    },
+    {
+      "epoch": 0.4846228661501828,
+      "grad_norm": 0.1520399357717214,
+      "learning_rate": 2e-05,
+      "loss": 5.4604,
+      "step": 7225
+    },
+    {
+      "epoch": 0.48468994197940773,
+      "grad_norm": 0.16184145803396152,
+      "learning_rate": 2e-05,
+      "loss": 5.3956,
+      "step": 7226
+    },
+    {
+      "epoch": 0.4847570178086327,
+      "grad_norm": 0.153995835836526,
+      "learning_rate": 2e-05,
+      "loss": 5.3668,
+      "step": 7227
+    },
+    {
+      "epoch": 0.4848240936378576,
+      "grad_norm": 0.16102504275174043,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 7228
+    },
+    {
+      "epoch": 0.48489116946708255,
+      "grad_norm": 0.1514930306662499,
+      "learning_rate": 2e-05,
+      "loss": 5.3949,
+      "step": 7229
+    },
+    {
+      "epoch": 0.4849582452963075,
+      "grad_norm": 0.14792295346259657,
+      "learning_rate": 2e-05,
+      "loss": 5.3383,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4850253211255324,
+      "grad_norm": 0.14531363794565663,
+      "learning_rate": 2e-05,
+      "loss": 5.5121,
+      "step": 7231
+    },
+    {
+      "epoch": 0.48509239695475737,
+      "grad_norm": 0.15375885560313798,
+      "learning_rate": 2e-05,
+      "loss": 5.6094,
+      "step": 7232
+    },
+    {
+      "epoch": 0.4851594727839823,
+      "grad_norm": 0.14375431503557334,
+      "learning_rate": 2e-05,
+      "loss": 5.4084,
+      "step": 7233
+    },
+    {
+      "epoch": 0.48522654861320724,
+      "grad_norm": 0.1507104667862044,
+      "learning_rate": 2e-05,
+      "loss": 5.383,
+      "step": 7234
+    },
+    {
+      "epoch": 0.4852936244424322,
+      "grad_norm": 0.14909048413890083,
+      "learning_rate": 2e-05,
+      "loss": 5.4151,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4853607002716571,
+      "grad_norm": 0.15498586068943693,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 7236
+    },
+    {
+      "epoch": 0.48542777610088206,
+      "grad_norm": 0.1487593368095114,
+      "learning_rate": 2e-05,
+      "loss": 5.5252,
+      "step": 7237
+    },
+    {
+      "epoch": 0.485494851930107,
+      "grad_norm": 0.1441308767906382,
+      "learning_rate": 2e-05,
+      "loss": 5.3439,
+      "step": 7238
+    },
+    {
+      "epoch": 0.48556192775933193,
+      "grad_norm": 0.14471638911085416,
+      "learning_rate": 2e-05,
+      "loss": 5.3225,
+      "step": 7239
+    },
+    {
+      "epoch": 0.4856290035885569,
+      "grad_norm": 0.14794392887282126,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4856960794177818,
+      "grad_norm": 0.15655771164621232,
+      "learning_rate": 2e-05,
+      "loss": 5.3604,
+      "step": 7241
+    },
+    {
+      "epoch": 0.48576315524700675,
+      "grad_norm": 0.14686778225930835,
+      "learning_rate": 2e-05,
+      "loss": 5.5356,
+      "step": 7242
+    },
+    {
+      "epoch": 0.4858302310762317,
+      "grad_norm": 0.14688443094997333,
+      "learning_rate": 2e-05,
+      "loss": 5.2394,
+      "step": 7243
+    },
+    {
+      "epoch": 0.4858973069054566,
+      "grad_norm": 0.14975028423416792,
+      "learning_rate": 2e-05,
+      "loss": 5.5256,
+      "step": 7244
+    },
+    {
+      "epoch": 0.48596438273468157,
+      "grad_norm": 0.15171252965856863,
+      "learning_rate": 2e-05,
+      "loss": 5.5066,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4860314585639065,
+      "grad_norm": 0.15116870494138113,
+      "learning_rate": 2e-05,
+      "loss": 5.4447,
+      "step": 7246
+    },
+    {
+      "epoch": 0.48609853439313144,
+      "grad_norm": 0.1583084576506635,
+      "learning_rate": 2e-05,
+      "loss": 5.4208,
+      "step": 7247
+    },
+    {
+      "epoch": 0.4861656102223564,
+      "grad_norm": 0.14687487533256569,
+      "learning_rate": 2e-05,
+      "loss": 5.5114,
+      "step": 7248
+    },
+    {
+      "epoch": 0.4862326860515813,
+      "grad_norm": 0.14935406393678977,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 7249
+    },
+    {
+      "epoch": 0.48629976188080626,
+      "grad_norm": 0.14954897755542043,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4863668377100312,
+      "grad_norm": 0.1593018404290953,
+      "learning_rate": 2e-05,
+      "loss": 5.3228,
+      "step": 7251
+    },
+    {
+      "epoch": 0.48643391353925614,
+      "grad_norm": 0.15643533625858957,
+      "learning_rate": 2e-05,
+      "loss": 5.3544,
+      "step": 7252
+    },
+    {
+      "epoch": 0.4865009893684811,
+      "grad_norm": 0.15317024975845922,
+      "learning_rate": 2e-05,
+      "loss": 5.4816,
+      "step": 7253
+    },
+    {
+      "epoch": 0.486568065197706,
+      "grad_norm": 0.15804081139701784,
+      "learning_rate": 2e-05,
+      "loss": 5.2495,
+      "step": 7254
+    },
+    {
+      "epoch": 0.48663514102693095,
+      "grad_norm": 0.1512658096795007,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4867022168561559,
+      "grad_norm": 0.1604444991774679,
+      "learning_rate": 2e-05,
+      "loss": 5.5251,
+      "step": 7256
+    },
+    {
+      "epoch": 0.4867692926853808,
+      "grad_norm": 0.15487534591121785,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 7257
+    },
+    {
+      "epoch": 0.48683636851460577,
+      "grad_norm": 0.15744175053912654,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 7258
+    },
+    {
+      "epoch": 0.4869034443438307,
+      "grad_norm": 0.15064438494237076,
+      "learning_rate": 2e-05,
+      "loss": 5.3812,
+      "step": 7259
+    },
+    {
+      "epoch": 0.48697052017305564,
+      "grad_norm": 0.14759350471260566,
+      "learning_rate": 2e-05,
+      "loss": 5.3875,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4870375960022806,
+      "grad_norm": 0.15488651204879006,
+      "learning_rate": 2e-05,
+      "loss": 5.5139,
+      "step": 7261
+    },
+    {
+      "epoch": 0.4871046718315055,
+      "grad_norm": 0.1510709798287007,
+      "learning_rate": 2e-05,
+      "loss": 5.4949,
+      "step": 7262
+    },
+    {
+      "epoch": 0.48717174766073046,
+      "grad_norm": 0.15586635036988952,
+      "learning_rate": 2e-05,
+      "loss": 5.2981,
+      "step": 7263
+    },
+    {
+      "epoch": 0.4872388234899554,
+      "grad_norm": 0.14894841523593785,
+      "learning_rate": 2e-05,
+      "loss": 5.3099,
+      "step": 7264
+    },
+    {
+      "epoch": 0.48730589931918034,
+      "grad_norm": 0.16072686471492592,
+      "learning_rate": 2e-05,
+      "loss": 5.5359,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4873729751484053,
+      "grad_norm": 0.15898965332663906,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 7266
+    },
+    {
+      "epoch": 0.4874400509776302,
+      "grad_norm": 0.1586309921049504,
+      "learning_rate": 2e-05,
+      "loss": 5.4204,
+      "step": 7267
+    },
+    {
+      "epoch": 0.48750712680685515,
+      "grad_norm": 0.15640945612684462,
+      "learning_rate": 2e-05,
+      "loss": 5.5505,
+      "step": 7268
+    },
+    {
+      "epoch": 0.4875742026360801,
+      "grad_norm": 0.15190178550787586,
+      "learning_rate": 2e-05,
+      "loss": 5.4078,
+      "step": 7269
+    },
+    {
+      "epoch": 0.48764127846530503,
+      "grad_norm": 0.15615448861194242,
+      "learning_rate": 2e-05,
+      "loss": 5.4231,
+      "step": 7270
+    },
+    {
+      "epoch": 0.48770835429452997,
+      "grad_norm": 0.1526838129750418,
+      "learning_rate": 2e-05,
+      "loss": 5.5557,
+      "step": 7271
+    },
+    {
+      "epoch": 0.4877754301237549,
+      "grad_norm": 0.15652278016886093,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 7272
+    },
+    {
+      "epoch": 0.48784250595297984,
+      "grad_norm": 0.14567339673387988,
+      "learning_rate": 2e-05,
+      "loss": 5.35,
+      "step": 7273
+    },
+    {
+      "epoch": 0.4879095817822048,
+      "grad_norm": 0.15376818869708161,
+      "learning_rate": 2e-05,
+      "loss": 5.443,
+      "step": 7274
+    },
+    {
+      "epoch": 0.4879766576114297,
+      "grad_norm": 0.15669586971431973,
+      "learning_rate": 2e-05,
+      "loss": 5.4154,
+      "step": 7275
+    },
+    {
+      "epoch": 0.48804373344065466,
+      "grad_norm": 0.15883099207299614,
+      "learning_rate": 2e-05,
+      "loss": 5.4181,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4881108092698796,
+      "grad_norm": 0.15043008847153788,
+      "learning_rate": 2e-05,
+      "loss": 5.4429,
+      "step": 7277
+    },
+    {
+      "epoch": 0.48817788509910454,
+      "grad_norm": 0.15483432694847063,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 7278
+    },
+    {
+      "epoch": 0.4882449609283295,
+      "grad_norm": 0.1545682021886171,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4883120367575544,
+      "grad_norm": 0.15432544307291698,
+      "learning_rate": 2e-05,
+      "loss": 5.442,
+      "step": 7280
+    },
+    {
+      "epoch": 0.48837911258677935,
+      "grad_norm": 0.15102922866601076,
+      "learning_rate": 2e-05,
+      "loss": 5.4148,
+      "step": 7281
+    },
+    {
+      "epoch": 0.4884461884160043,
+      "grad_norm": 0.1511750886042475,
+      "learning_rate": 2e-05,
+      "loss": 5.3881,
+      "step": 7282
+    },
+    {
+      "epoch": 0.48851326424522923,
+      "grad_norm": 0.15092538841594275,
+      "learning_rate": 2e-05,
+      "loss": 5.497,
+      "step": 7283
+    },
+    {
+      "epoch": 0.48858034007445417,
+      "grad_norm": 0.1514836220463496,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 7284
+    },
+    {
+      "epoch": 0.4886474159036791,
+      "grad_norm": 0.14590567016284545,
+      "learning_rate": 2e-05,
+      "loss": 5.3227,
+      "step": 7285
+    },
+    {
+      "epoch": 0.48871449173290404,
+      "grad_norm": 0.14814018405574228,
+      "learning_rate": 2e-05,
+      "loss": 5.263,
+      "step": 7286
+    },
+    {
+      "epoch": 0.488781567562129,
+      "grad_norm": 0.16107482286755004,
+      "learning_rate": 2e-05,
+      "loss": 5.2532,
+      "step": 7287
+    },
+    {
+      "epoch": 0.4888486433913539,
+      "grad_norm": 0.16204589906385908,
+      "learning_rate": 2e-05,
+      "loss": 5.3667,
+      "step": 7288
+    },
+    {
+      "epoch": 0.48891571922057886,
+      "grad_norm": 0.14749550513383114,
+      "learning_rate": 2e-05,
+      "loss": 5.6149,
+      "step": 7289
+    },
+    {
+      "epoch": 0.4889827950498038,
+      "grad_norm": 0.1512847998639348,
+      "learning_rate": 2e-05,
+      "loss": 5.5244,
+      "step": 7290
+    },
+    {
+      "epoch": 0.48904987087902874,
+      "grad_norm": 0.1514335420552851,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 7291
+    },
+    {
+      "epoch": 0.4891169467082537,
+      "grad_norm": 0.14717525457234365,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 7292
+    },
+    {
+      "epoch": 0.4891840225374786,
+      "grad_norm": 0.16056604494878024,
+      "learning_rate": 2e-05,
+      "loss": 5.4254,
+      "step": 7293
+    },
+    {
+      "epoch": 0.48925109836670355,
+      "grad_norm": 0.1540764392954355,
+      "learning_rate": 2e-05,
+      "loss": 5.4145,
+      "step": 7294
+    },
+    {
+      "epoch": 0.4893181741959285,
+      "grad_norm": 0.14887749843084067,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 7295
+    },
+    {
+      "epoch": 0.48938525002515343,
+      "grad_norm": 0.15324048633624998,
+      "learning_rate": 2e-05,
+      "loss": 5.4373,
+      "step": 7296
+    },
+    {
+      "epoch": 0.48945232585437837,
+      "grad_norm": 0.15361168438499584,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 7297
+    },
+    {
+      "epoch": 0.4895194016836033,
+      "grad_norm": 0.14736234021738698,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 7298
+    },
+    {
+      "epoch": 0.48958647751282824,
+      "grad_norm": 0.15234083022592604,
+      "learning_rate": 2e-05,
+      "loss": 5.4698,
+      "step": 7299
+    },
+    {
+      "epoch": 0.4896535533420532,
+      "grad_norm": 0.17178232386592673,
+      "learning_rate": 2e-05,
+      "loss": 5.4816,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4897206291712781,
+      "grad_norm": 0.14889977147743444,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 7301
+    },
+    {
+      "epoch": 0.48978770500050306,
+      "grad_norm": 0.15351696885518062,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 7302
+    },
+    {
+      "epoch": 0.489854780829728,
+      "grad_norm": 0.16351921511092238,
+      "learning_rate": 2e-05,
+      "loss": 5.5137,
+      "step": 7303
+    },
+    {
+      "epoch": 0.48992185665895294,
+      "grad_norm": 0.1654184985618466,
+      "learning_rate": 2e-05,
+      "loss": 5.4756,
+      "step": 7304
+    },
+    {
+      "epoch": 0.4899889324881779,
+      "grad_norm": 0.15750446006244753,
+      "learning_rate": 2e-05,
+      "loss": 5.4447,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4900560083174028,
+      "grad_norm": 0.14765041868567377,
+      "learning_rate": 2e-05,
+      "loss": 5.4576,
+      "step": 7306
+    },
+    {
+      "epoch": 0.49012308414662775,
+      "grad_norm": 0.16307480625606408,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 7307
+    },
+    {
+      "epoch": 0.4901901599758527,
+      "grad_norm": 0.16886740398968939,
+      "learning_rate": 2e-05,
+      "loss": 5.5234,
+      "step": 7308
+    },
+    {
+      "epoch": 0.49025723580507763,
+      "grad_norm": 0.14715670891556135,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 7309
+    },
+    {
+      "epoch": 0.49032431163430257,
+      "grad_norm": 0.16467723103239904,
+      "learning_rate": 2e-05,
+      "loss": 5.3599,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4903913874635275,
+      "grad_norm": 0.1641650675084048,
+      "learning_rate": 2e-05,
+      "loss": 5.5028,
+      "step": 7311
+    },
+    {
+      "epoch": 0.49045846329275244,
+      "grad_norm": 0.160804365773499,
+      "learning_rate": 2e-05,
+      "loss": 5.2942,
+      "step": 7312
+    },
+    {
+      "epoch": 0.4905255391219774,
+      "grad_norm": 0.15114443611370876,
+      "learning_rate": 2e-05,
+      "loss": 5.5131,
+      "step": 7313
+    },
+    {
+      "epoch": 0.4905926149512023,
+      "grad_norm": 0.15351866253266877,
+      "learning_rate": 2e-05,
+      "loss": 5.3654,
+      "step": 7314
+    },
+    {
+      "epoch": 0.49065969078042726,
+      "grad_norm": 0.1491332456310917,
+      "learning_rate": 2e-05,
+      "loss": 5.471,
+      "step": 7315
+    },
+    {
+      "epoch": 0.4907267666096522,
+      "grad_norm": 0.15107609018904766,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 7316
+    },
+    {
+      "epoch": 0.49079384243887714,
+      "grad_norm": 0.15522129549452196,
+      "learning_rate": 2e-05,
+      "loss": 5.4005,
+      "step": 7317
+    },
+    {
+      "epoch": 0.4908609182681021,
+      "grad_norm": 0.15176919897160024,
+      "learning_rate": 2e-05,
+      "loss": 5.5291,
+      "step": 7318
+    },
+    {
+      "epoch": 0.490927994097327,
+      "grad_norm": 0.15745152959862943,
+      "learning_rate": 2e-05,
+      "loss": 5.4727,
+      "step": 7319
+    },
+    {
+      "epoch": 0.49099506992655195,
+      "grad_norm": 0.16024346162750774,
+      "learning_rate": 2e-05,
+      "loss": 5.4279,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4910621457557769,
+      "grad_norm": 0.1610090961026196,
+      "learning_rate": 2e-05,
+      "loss": 5.3595,
+      "step": 7321
+    },
+    {
+      "epoch": 0.49112922158500183,
+      "grad_norm": 0.14941672910416284,
+      "learning_rate": 2e-05,
+      "loss": 5.3921,
+      "step": 7322
+    },
+    {
+      "epoch": 0.49119629741422677,
+      "grad_norm": 0.16082172379639403,
+      "learning_rate": 2e-05,
+      "loss": 5.4394,
+      "step": 7323
+    },
+    {
+      "epoch": 0.4912633732434517,
+      "grad_norm": 0.16619122989563992,
+      "learning_rate": 2e-05,
+      "loss": 5.4383,
+      "step": 7324
+    },
+    {
+      "epoch": 0.49133044907267664,
+      "grad_norm": 0.15530321961486573,
+      "learning_rate": 2e-05,
+      "loss": 5.5008,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4913975249019016,
+      "grad_norm": 0.1633538832361557,
+      "learning_rate": 2e-05,
+      "loss": 5.3698,
+      "step": 7326
+    },
+    {
+      "epoch": 0.4914646007311265,
+      "grad_norm": 0.15519371061274828,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 7327
+    },
+    {
+      "epoch": 0.49153167656035146,
+      "grad_norm": 0.14880675116399666,
+      "learning_rate": 2e-05,
+      "loss": 5.219,
+      "step": 7328
+    },
+    {
+      "epoch": 0.4915987523895764,
+      "grad_norm": 0.1542097470398536,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 7329
+    },
+    {
+      "epoch": 0.49166582821880134,
+      "grad_norm": 0.16245376731204295,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4917329040480263,
+      "grad_norm": 0.15428889239198523,
+      "learning_rate": 2e-05,
+      "loss": 5.4762,
+      "step": 7331
+    },
+    {
+      "epoch": 0.4917999798772512,
+      "grad_norm": 0.16158572878443142,
+      "learning_rate": 2e-05,
+      "loss": 5.3672,
+      "step": 7332
+    },
+    {
+      "epoch": 0.49186705570647615,
+      "grad_norm": 0.16444598528504298,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 7333
+    },
+    {
+      "epoch": 0.4919341315357011,
+      "grad_norm": 0.15446604839917247,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 7334
+    },
+    {
+      "epoch": 0.49200120736492603,
+      "grad_norm": 0.15255504876955642,
+      "learning_rate": 2e-05,
+      "loss": 5.4209,
+      "step": 7335
+    },
+    {
+      "epoch": 0.49206828319415097,
+      "grad_norm": 0.17099724420502918,
+      "learning_rate": 2e-05,
+      "loss": 5.6624,
+      "step": 7336
+    },
+    {
+      "epoch": 0.4921353590233759,
+      "grad_norm": 0.15668427713119792,
+      "learning_rate": 2e-05,
+      "loss": 5.4815,
+      "step": 7337
+    },
+    {
+      "epoch": 0.49220243485260085,
+      "grad_norm": 0.16060892597048285,
+      "learning_rate": 2e-05,
+      "loss": 5.4941,
+      "step": 7338
+    },
+    {
+      "epoch": 0.4922695106818258,
+      "grad_norm": 0.15036976932564064,
+      "learning_rate": 2e-05,
+      "loss": 5.5813,
+      "step": 7339
+    },
+    {
+      "epoch": 0.4923365865110507,
+      "grad_norm": 0.16663014088537828,
+      "learning_rate": 2e-05,
+      "loss": 5.5168,
+      "step": 7340
+    },
+    {
+      "epoch": 0.49240366234027566,
+      "grad_norm": 0.15424923292352155,
+      "learning_rate": 2e-05,
+      "loss": 5.4477,
+      "step": 7341
+    },
+    {
+      "epoch": 0.4924707381695006,
+      "grad_norm": 0.15118684751994713,
+      "learning_rate": 2e-05,
+      "loss": 5.422,
+      "step": 7342
+    },
+    {
+      "epoch": 0.49253781399872554,
+      "grad_norm": 0.15504398032314665,
+      "learning_rate": 2e-05,
+      "loss": 5.3713,
+      "step": 7343
+    },
+    {
+      "epoch": 0.4926048898279505,
+      "grad_norm": 0.14348048036938035,
+      "learning_rate": 2e-05,
+      "loss": 5.3093,
+      "step": 7344
+    },
+    {
+      "epoch": 0.4926719656571754,
+      "grad_norm": 0.14715761135814384,
+      "learning_rate": 2e-05,
+      "loss": 5.5183,
+      "step": 7345
+    },
+    {
+      "epoch": 0.49273904148640035,
+      "grad_norm": 0.14802688271704167,
+      "learning_rate": 2e-05,
+      "loss": 5.4543,
+      "step": 7346
+    },
+    {
+      "epoch": 0.4928061173156253,
+      "grad_norm": 0.1637264293291327,
+      "learning_rate": 2e-05,
+      "loss": 5.3748,
+      "step": 7347
+    },
+    {
+      "epoch": 0.49287319314485023,
+      "grad_norm": 0.14902247554652967,
+      "learning_rate": 2e-05,
+      "loss": 5.2813,
+      "step": 7348
+    },
+    {
+      "epoch": 0.49294026897407517,
+      "grad_norm": 0.14992925160874318,
+      "learning_rate": 2e-05,
+      "loss": 5.2976,
+      "step": 7349
+    },
+    {
+      "epoch": 0.4930073448033001,
+      "grad_norm": 0.15076817263914208,
+      "learning_rate": 2e-05,
+      "loss": 5.4521,
+      "step": 7350
+    },
+    {
+      "epoch": 0.49307442063252505,
+      "grad_norm": 0.15181978518936617,
+      "learning_rate": 2e-05,
+      "loss": 5.3816,
+      "step": 7351
+    },
+    {
+      "epoch": 0.49314149646175,
+      "grad_norm": 0.16157486298243506,
+      "learning_rate": 2e-05,
+      "loss": 5.557,
+      "step": 7352
+    },
+    {
+      "epoch": 0.4932085722909749,
+      "grad_norm": 0.1470936428216905,
+      "learning_rate": 2e-05,
+      "loss": 5.4885,
+      "step": 7353
+    },
+    {
+      "epoch": 0.49327564812019986,
+      "grad_norm": 0.14869459341195967,
+      "learning_rate": 2e-05,
+      "loss": 5.3043,
+      "step": 7354
+    },
+    {
+      "epoch": 0.4933427239494248,
+      "grad_norm": 0.15416617890614712,
+      "learning_rate": 2e-05,
+      "loss": 5.296,
+      "step": 7355
+    },
+    {
+      "epoch": 0.49340979977864974,
+      "grad_norm": 0.14813384663982665,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 7356
+    },
+    {
+      "epoch": 0.4934768756078747,
+      "grad_norm": 0.14485615679068592,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 7357
+    },
+    {
+      "epoch": 0.4935439514370996,
+      "grad_norm": 0.16271656729039963,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 7358
+    },
+    {
+      "epoch": 0.49361102726632455,
+      "grad_norm": 0.1576225954400024,
+      "learning_rate": 2e-05,
+      "loss": 5.2975,
+      "step": 7359
+    },
+    {
+      "epoch": 0.4936781030955495,
+      "grad_norm": 0.14936831284340293,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 7360
+    },
+    {
+      "epoch": 0.49374517892477443,
+      "grad_norm": 0.15079549749380716,
+      "learning_rate": 2e-05,
+      "loss": 5.5759,
+      "step": 7361
+    },
+    {
+      "epoch": 0.49381225475399937,
+      "grad_norm": 0.15284468494725442,
+      "learning_rate": 2e-05,
+      "loss": 5.4067,
+      "step": 7362
+    },
+    {
+      "epoch": 0.4938793305832243,
+      "grad_norm": 0.14906736556558497,
+      "learning_rate": 2e-05,
+      "loss": 5.478,
+      "step": 7363
+    },
+    {
+      "epoch": 0.4939464064124493,
+      "grad_norm": 0.15180554364497692,
+      "learning_rate": 2e-05,
+      "loss": 5.5037,
+      "step": 7364
+    },
+    {
+      "epoch": 0.49401348224167424,
+      "grad_norm": 0.15852383605445566,
+      "learning_rate": 2e-05,
+      "loss": 5.4913,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4940805580708992,
+      "grad_norm": 0.1469196116340861,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 7366
+    },
+    {
+      "epoch": 0.4941476339001241,
+      "grad_norm": 0.15529352702521085,
+      "learning_rate": 2e-05,
+      "loss": 5.4973,
+      "step": 7367
+    },
+    {
+      "epoch": 0.49421470972934906,
+      "grad_norm": 0.15466601535986912,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 7368
+    },
+    {
+      "epoch": 0.494281785558574,
+      "grad_norm": 0.15574612564136242,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 7369
+    },
+    {
+      "epoch": 0.49434886138779893,
+      "grad_norm": 0.15092105507185322,
+      "learning_rate": 2e-05,
+      "loss": 5.3872,
+      "step": 7370
+    },
+    {
+      "epoch": 0.49441593721702387,
+      "grad_norm": 0.15201049110387202,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 7371
+    },
+    {
+      "epoch": 0.4944830130462488,
+      "grad_norm": 0.1491958637249756,
+      "learning_rate": 2e-05,
+      "loss": 5.416,
+      "step": 7372
+    },
+    {
+      "epoch": 0.49455008887547375,
+      "grad_norm": 0.15047858337264042,
+      "learning_rate": 2e-05,
+      "loss": 5.526,
+      "step": 7373
+    },
+    {
+      "epoch": 0.4946171647046987,
+      "grad_norm": 0.16167055546202805,
+      "learning_rate": 2e-05,
+      "loss": 5.3773,
+      "step": 7374
+    },
+    {
+      "epoch": 0.4946842405339236,
+      "grad_norm": 0.1487898324243125,
+      "learning_rate": 2e-05,
+      "loss": 5.4238,
+      "step": 7375
+    },
+    {
+      "epoch": 0.49475131636314856,
+      "grad_norm": 0.14987136475420948,
+      "learning_rate": 2e-05,
+      "loss": 5.2138,
+      "step": 7376
+    },
+    {
+      "epoch": 0.4948183921923735,
+      "grad_norm": 0.14438882208207665,
+      "learning_rate": 2e-05,
+      "loss": 5.4701,
+      "step": 7377
+    },
+    {
+      "epoch": 0.49488546802159844,
+      "grad_norm": 0.15058507796650691,
+      "learning_rate": 2e-05,
+      "loss": 5.5816,
+      "step": 7378
+    },
+    {
+      "epoch": 0.4949525438508234,
+      "grad_norm": 0.14828850986778236,
+      "learning_rate": 2e-05,
+      "loss": 5.394,
+      "step": 7379
+    },
+    {
+      "epoch": 0.4950196196800483,
+      "grad_norm": 0.15037040791649076,
+      "learning_rate": 2e-05,
+      "loss": 5.5762,
+      "step": 7380
+    },
+    {
+      "epoch": 0.49508669550927326,
+      "grad_norm": 0.14375111200201504,
+      "learning_rate": 2e-05,
+      "loss": 5.3719,
+      "step": 7381
+    },
+    {
+      "epoch": 0.4951537713384982,
+      "grad_norm": 0.14747365407318966,
+      "learning_rate": 2e-05,
+      "loss": 5.6286,
+      "step": 7382
+    },
+    {
+      "epoch": 0.49522084716772313,
+      "grad_norm": 0.1544799929180007,
+      "learning_rate": 2e-05,
+      "loss": 5.4175,
+      "step": 7383
+    },
+    {
+      "epoch": 0.49528792299694807,
+      "grad_norm": 0.15145863221557915,
+      "learning_rate": 2e-05,
+      "loss": 5.2984,
+      "step": 7384
+    },
+    {
+      "epoch": 0.495354998826173,
+      "grad_norm": 0.1504285231726418,
+      "learning_rate": 2e-05,
+      "loss": 5.6142,
+      "step": 7385
+    },
+    {
+      "epoch": 0.49542207465539795,
+      "grad_norm": 0.1504542847226113,
+      "learning_rate": 2e-05,
+      "loss": 5.4413,
+      "step": 7386
+    },
+    {
+      "epoch": 0.4954891504846229,
+      "grad_norm": 0.15654252390117626,
+      "learning_rate": 2e-05,
+      "loss": 5.5235,
+      "step": 7387
+    },
+    {
+      "epoch": 0.4955562263138478,
+      "grad_norm": 0.14656171902116488,
+      "learning_rate": 2e-05,
+      "loss": 5.2867,
+      "step": 7388
+    },
+    {
+      "epoch": 0.49562330214307276,
+      "grad_norm": 0.1408586372399371,
+      "learning_rate": 2e-05,
+      "loss": 5.4082,
+      "step": 7389
+    },
+    {
+      "epoch": 0.4956903779722977,
+      "grad_norm": 0.15875647223270764,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 7390
+    },
+    {
+      "epoch": 0.49575745380152264,
+      "grad_norm": 0.15090228138929665,
+      "learning_rate": 2e-05,
+      "loss": 5.459,
+      "step": 7391
+    },
+    {
+      "epoch": 0.4958245296307476,
+      "grad_norm": 0.14381644786301362,
+      "learning_rate": 2e-05,
+      "loss": 5.4971,
+      "step": 7392
+    },
+    {
+      "epoch": 0.4958916054599725,
+      "grad_norm": 0.14856094058700284,
+      "learning_rate": 2e-05,
+      "loss": 5.4548,
+      "step": 7393
+    },
+    {
+      "epoch": 0.49595868128919746,
+      "grad_norm": 0.14531309716895177,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 7394
+    },
+    {
+      "epoch": 0.4960257571184224,
+      "grad_norm": 0.15383101403535365,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 7395
+    },
+    {
+      "epoch": 0.49609283294764733,
+      "grad_norm": 0.15476480678891133,
+      "learning_rate": 2e-05,
+      "loss": 5.3693,
+      "step": 7396
+    },
+    {
+      "epoch": 0.49615990877687227,
+      "grad_norm": 0.1602932349819086,
+      "learning_rate": 2e-05,
+      "loss": 5.517,
+      "step": 7397
+    },
+    {
+      "epoch": 0.4962269846060972,
+      "grad_norm": 0.1434816171997583,
+      "learning_rate": 2e-05,
+      "loss": 5.3836,
+      "step": 7398
+    },
+    {
+      "epoch": 0.49629406043532215,
+      "grad_norm": 0.15329345308163647,
+      "learning_rate": 2e-05,
+      "loss": 5.3469,
+      "step": 7399
+    },
+    {
+      "epoch": 0.4963611362645471,
+      "grad_norm": 0.1585917060168978,
+      "learning_rate": 2e-05,
+      "loss": 5.4975,
+      "step": 7400
+    },
+    {
+      "epoch": 0.496428212093772,
+      "grad_norm": 0.16137478925880805,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 7401
+    },
+    {
+      "epoch": 0.49649528792299696,
+      "grad_norm": 0.15073666678885528,
+      "learning_rate": 2e-05,
+      "loss": 5.3102,
+      "step": 7402
+    },
+    {
+      "epoch": 0.4965623637522219,
+      "grad_norm": 0.15669872807964913,
+      "learning_rate": 2e-05,
+      "loss": 5.4322,
+      "step": 7403
+    },
+    {
+      "epoch": 0.49662943958144684,
+      "grad_norm": 0.14661242263164462,
+      "learning_rate": 2e-05,
+      "loss": 5.4173,
+      "step": 7404
+    },
+    {
+      "epoch": 0.4966965154106718,
+      "grad_norm": 0.15441676435937884,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 7405
+    },
+    {
+      "epoch": 0.4967635912398967,
+      "grad_norm": 0.16071996595779603,
+      "learning_rate": 2e-05,
+      "loss": 5.2499,
+      "step": 7406
+    },
+    {
+      "epoch": 0.49683066706912166,
+      "grad_norm": 0.15743580470117136,
+      "learning_rate": 2e-05,
+      "loss": 5.3126,
+      "step": 7407
+    },
+    {
+      "epoch": 0.4968977428983466,
+      "grad_norm": 0.15411161835432666,
+      "learning_rate": 2e-05,
+      "loss": 5.4146,
+      "step": 7408
+    },
+    {
+      "epoch": 0.49696481872757153,
+      "grad_norm": 0.1689865718608176,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 7409
+    },
+    {
+      "epoch": 0.49703189455679647,
+      "grad_norm": 0.16683067716775685,
+      "learning_rate": 2e-05,
+      "loss": 5.4864,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4970989703860214,
+      "grad_norm": 0.15409791750806667,
+      "learning_rate": 2e-05,
+      "loss": 5.2462,
+      "step": 7411
+    },
+    {
+      "epoch": 0.49716604621524635,
+      "grad_norm": 0.17872495096452615,
+      "learning_rate": 2e-05,
+      "loss": 5.3524,
+      "step": 7412
+    },
+    {
+      "epoch": 0.4972331220444713,
+      "grad_norm": 0.1638047627282164,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 7413
+    },
+    {
+      "epoch": 0.4973001978736962,
+      "grad_norm": 0.14708150480467316,
+      "learning_rate": 2e-05,
+      "loss": 5.4763,
+      "step": 7414
+    },
+    {
+      "epoch": 0.49736727370292116,
+      "grad_norm": 0.17311442235230826,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 7415
+    },
+    {
+      "epoch": 0.4974343495321461,
+      "grad_norm": 0.1551489731075624,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 7416
+    },
+    {
+      "epoch": 0.49750142536137104,
+      "grad_norm": 0.1532963874981527,
+      "learning_rate": 2e-05,
+      "loss": 5.2161,
+      "step": 7417
+    },
+    {
+      "epoch": 0.497568501190596,
+      "grad_norm": 0.15550569966947356,
+      "learning_rate": 2e-05,
+      "loss": 5.3474,
+      "step": 7418
+    },
+    {
+      "epoch": 0.4976355770198209,
+      "grad_norm": 0.151812488530657,
+      "learning_rate": 2e-05,
+      "loss": 5.3328,
+      "step": 7419
+    },
+    {
+      "epoch": 0.49770265284904586,
+      "grad_norm": 0.14911730902983716,
+      "learning_rate": 2e-05,
+      "loss": 5.355,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4977697286782708,
+      "grad_norm": 0.15374808901290318,
+      "learning_rate": 2e-05,
+      "loss": 5.2593,
+      "step": 7421
+    },
+    {
+      "epoch": 0.49783680450749573,
+      "grad_norm": 0.15459340247416756,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 7422
+    },
+    {
+      "epoch": 0.4979038803367207,
+      "grad_norm": 0.14881181808140675,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 7423
+    },
+    {
+      "epoch": 0.4979709561659456,
+      "grad_norm": 0.151215150313599,
+      "learning_rate": 2e-05,
+      "loss": 5.4635,
+      "step": 7424
+    },
+    {
+      "epoch": 0.49803803199517055,
+      "grad_norm": 0.15541645019531192,
+      "learning_rate": 2e-05,
+      "loss": 5.278,
+      "step": 7425
+    },
+    {
+      "epoch": 0.4981051078243955,
+      "grad_norm": 0.15876429010769288,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 7426
+    },
+    {
+      "epoch": 0.4981721836536204,
+      "grad_norm": 0.15270170925691304,
+      "learning_rate": 2e-05,
+      "loss": 5.3333,
+      "step": 7427
+    },
+    {
+      "epoch": 0.49823925948284536,
+      "grad_norm": 0.15207889511382583,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 7428
+    },
+    {
+      "epoch": 0.4983063353120703,
+      "grad_norm": 0.15060122726226408,
+      "learning_rate": 2e-05,
+      "loss": 5.401,
+      "step": 7429
+    },
+    {
+      "epoch": 0.49837341114129524,
+      "grad_norm": 0.16688460568266866,
+      "learning_rate": 2e-05,
+      "loss": 5.4604,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4984404869705202,
+      "grad_norm": 0.16014342537778195,
+      "learning_rate": 2e-05,
+      "loss": 5.3685,
+      "step": 7431
+    },
+    {
+      "epoch": 0.4985075627997451,
+      "grad_norm": 0.14719544350142807,
+      "learning_rate": 2e-05,
+      "loss": 5.4636,
+      "step": 7432
+    },
+    {
+      "epoch": 0.49857463862897006,
+      "grad_norm": 0.15738118857293631,
+      "learning_rate": 2e-05,
+      "loss": 5.4745,
+      "step": 7433
+    },
+    {
+      "epoch": 0.498641714458195,
+      "grad_norm": 0.16112020196640653,
+      "learning_rate": 2e-05,
+      "loss": 5.5399,
+      "step": 7434
+    },
+    {
+      "epoch": 0.49870879028741993,
+      "grad_norm": 0.15067087996804576,
+      "learning_rate": 2e-05,
+      "loss": 5.3715,
+      "step": 7435
+    },
+    {
+      "epoch": 0.4987758661166449,
+      "grad_norm": 0.1488473866798747,
+      "learning_rate": 2e-05,
+      "loss": 5.3963,
+      "step": 7436
+    },
+    {
+      "epoch": 0.4988429419458698,
+      "grad_norm": 0.1540006975750008,
+      "learning_rate": 2e-05,
+      "loss": 5.4914,
+      "step": 7437
+    },
+    {
+      "epoch": 0.49891001777509475,
+      "grad_norm": 0.16073220341683198,
+      "learning_rate": 2e-05,
+      "loss": 5.5453,
+      "step": 7438
+    },
+    {
+      "epoch": 0.4989770936043197,
+      "grad_norm": 0.14867142628417684,
+      "learning_rate": 2e-05,
+      "loss": 5.2785,
+      "step": 7439
+    },
+    {
+      "epoch": 0.4990441694335446,
+      "grad_norm": 0.15569247685686888,
+      "learning_rate": 2e-05,
+      "loss": 5.2703,
+      "step": 7440
+    },
+    {
+      "epoch": 0.49911124526276957,
+      "grad_norm": 0.14694011536293347,
+      "learning_rate": 2e-05,
+      "loss": 5.3905,
+      "step": 7441
+    },
+    {
+      "epoch": 0.4991783210919945,
+      "grad_norm": 0.15052813489249597,
+      "learning_rate": 2e-05,
+      "loss": 5.4102,
+      "step": 7442
+    },
+    {
+      "epoch": 0.49924539692121944,
+      "grad_norm": 0.1542810183907906,
+      "learning_rate": 2e-05,
+      "loss": 5.5069,
+      "step": 7443
+    },
+    {
+      "epoch": 0.4993124727504444,
+      "grad_norm": 0.15042989576473836,
+      "learning_rate": 2e-05,
+      "loss": 5.2159,
+      "step": 7444
+    },
+    {
+      "epoch": 0.4993795485796693,
+      "grad_norm": 0.14988137771919247,
+      "learning_rate": 2e-05,
+      "loss": 5.3857,
+      "step": 7445
+    },
+    {
+      "epoch": 0.49944662440889426,
+      "grad_norm": 0.15275640032052137,
+      "learning_rate": 2e-05,
+      "loss": 5.5608,
+      "step": 7446
+    },
+    {
+      "epoch": 0.4995137002381192,
+      "grad_norm": 0.15126945648402246,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 7447
+    },
+    {
+      "epoch": 0.49958077606734413,
+      "grad_norm": 0.15883782591422005,
+      "learning_rate": 2e-05,
+      "loss": 5.3209,
+      "step": 7448
+    },
+    {
+      "epoch": 0.4996478518965691,
+      "grad_norm": 0.15349342931470966,
+      "learning_rate": 2e-05,
+      "loss": 5.5526,
+      "step": 7449
+    },
+    {
+      "epoch": 0.499714927725794,
+      "grad_norm": 0.14638013764431043,
+      "learning_rate": 2e-05,
+      "loss": 5.5267,
+      "step": 7450
+    },
+    {
+      "epoch": 0.49978200355501895,
+      "grad_norm": 0.15136559821132822,
+      "learning_rate": 2e-05,
+      "loss": 5.3797,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4998490793842439,
+      "grad_norm": 0.14998787769067554,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 7452
+    },
+    {
+      "epoch": 0.4999161552134688,
+      "grad_norm": 0.15840799862304758,
+      "learning_rate": 2e-05,
+      "loss": 5.4894,
+      "step": 7453
+    },
+    {
+      "epoch": 0.49998323104269377,
+      "grad_norm": 0.14849769526703804,
+      "learning_rate": 2e-05,
+      "loss": 5.4163,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5000503068719188,
+      "grad_norm": 0.14468403785693854,
+      "learning_rate": 2e-05,
+      "loss": 5.4748,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5001173827011437,
+      "grad_norm": 0.1536513854431842,
+      "learning_rate": 2e-05,
+      "loss": 5.4774,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5001844585303686,
+      "grad_norm": 0.1551653800762011,
+      "learning_rate": 2e-05,
+      "loss": 5.5548,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5002515343595936,
+      "grad_norm": 0.1460361939763679,
+      "learning_rate": 2e-05,
+      "loss": 5.5239,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5003186101888185,
+      "grad_norm": 0.15328549245028047,
+      "learning_rate": 2e-05,
+      "loss": 5.3016,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5003856860180435,
+      "grad_norm": 0.15898958587691303,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5004527618472684,
+      "grad_norm": 0.14853568944815865,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5005198376764933,
+      "grad_norm": 0.14242239516661745,
+      "learning_rate": 2e-05,
+      "loss": 5.5353,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5005869135057183,
+      "grad_norm": 0.15860296455282025,
+      "learning_rate": 2e-05,
+      "loss": 5.3804,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5006539893349432,
+      "grad_norm": 0.14894528368643964,
+      "learning_rate": 2e-05,
+      "loss": 5.3532,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5007210651641681,
+      "grad_norm": 0.15036953853873208,
+      "learning_rate": 2e-05,
+      "loss": 5.4405,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5007881409933931,
+      "grad_norm": 0.1470899224497803,
+      "learning_rate": 2e-05,
+      "loss": 5.5348,
+      "step": 7466
+    },
+    {
+      "epoch": 0.500855216822618,
+      "grad_norm": 0.14559703289358347,
+      "learning_rate": 2e-05,
+      "loss": 5.2816,
+      "step": 7467
+    },
+    {
+      "epoch": 0.500922292651843,
+      "grad_norm": 0.1474493534721766,
+      "learning_rate": 2e-05,
+      "loss": 5.5469,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5009893684810679,
+      "grad_norm": 0.14403289299810737,
+      "learning_rate": 2e-05,
+      "loss": 5.4195,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5010564443102928,
+      "grad_norm": 0.14993877040447684,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5011235201395178,
+      "grad_norm": 0.1581128719959514,
+      "learning_rate": 2e-05,
+      "loss": 5.5002,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5011905959687427,
+      "grad_norm": 0.1465733119617993,
+      "learning_rate": 2e-05,
+      "loss": 5.5134,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5012576717979677,
+      "grad_norm": 0.14569129751067164,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 7473
+    },
+    {
+      "epoch": 0.5013247476271926,
+      "grad_norm": 0.15336130598634462,
+      "learning_rate": 2e-05,
+      "loss": 5.475,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5013918234564175,
+      "grad_norm": 0.15042078315036397,
+      "learning_rate": 2e-05,
+      "loss": 5.2311,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5014588992856425,
+      "grad_norm": 0.14840586359556102,
+      "learning_rate": 2e-05,
+      "loss": 5.3188,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5015259751148674,
+      "grad_norm": 0.1503979229280777,
+      "learning_rate": 2e-05,
+      "loss": 5.4255,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5015930509440923,
+      "grad_norm": 0.14575791892980236,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5016601267733173,
+      "grad_norm": 0.1535817836787605,
+      "learning_rate": 2e-05,
+      "loss": 5.4903,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5017272026025422,
+      "grad_norm": 0.15282269498577516,
+      "learning_rate": 2e-05,
+      "loss": 5.4876,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5017942784317672,
+      "grad_norm": 0.14605284985272918,
+      "learning_rate": 2e-05,
+      "loss": 5.4032,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5018613542609921,
+      "grad_norm": 0.1470085488483259,
+      "learning_rate": 2e-05,
+      "loss": 5.4573,
+      "step": 7482
+    },
+    {
+      "epoch": 0.501928430090217,
+      "grad_norm": 0.15016859367679483,
+      "learning_rate": 2e-05,
+      "loss": 5.3463,
+      "step": 7483
+    },
+    {
+      "epoch": 0.501995505919442,
+      "grad_norm": 0.15668470595044134,
+      "learning_rate": 2e-05,
+      "loss": 5.4228,
+      "step": 7484
+    },
+    {
+      "epoch": 0.5020625817486669,
+      "grad_norm": 0.15199804067890302,
+      "learning_rate": 2e-05,
+      "loss": 5.5068,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5021296575778919,
+      "grad_norm": 0.15039841388477043,
+      "learning_rate": 2e-05,
+      "loss": 5.293,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5021967334071168,
+      "grad_norm": 0.14749382168170322,
+      "learning_rate": 2e-05,
+      "loss": 5.4478,
+      "step": 7487
+    },
+    {
+      "epoch": 0.5022638092363417,
+      "grad_norm": 0.1582346818306294,
+      "learning_rate": 2e-05,
+      "loss": 5.3132,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5023308850655667,
+      "grad_norm": 0.1545445294288114,
+      "learning_rate": 2e-05,
+      "loss": 5.4659,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5023979608947916,
+      "grad_norm": 0.15415239218859283,
+      "learning_rate": 2e-05,
+      "loss": 5.4686,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5024650367240165,
+      "grad_norm": 0.15372309857508143,
+      "learning_rate": 2e-05,
+      "loss": 5.4709,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5025321125532415,
+      "grad_norm": 0.14849575550976413,
+      "learning_rate": 2e-05,
+      "loss": 5.4922,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5025991883824664,
+      "grad_norm": 0.150202069607838,
+      "learning_rate": 2e-05,
+      "loss": 5.3911,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5026662642116914,
+      "grad_norm": 0.15858805114438046,
+      "learning_rate": 2e-05,
+      "loss": 5.421,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5027333400409163,
+      "grad_norm": 0.15120216648640586,
+      "learning_rate": 2e-05,
+      "loss": 5.1626,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5028004158701412,
+      "grad_norm": 0.16220349942644346,
+      "learning_rate": 2e-05,
+      "loss": 5.4816,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5028674916993662,
+      "grad_norm": 0.15301888991750312,
+      "learning_rate": 2e-05,
+      "loss": 5.6546,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5029345675285911,
+      "grad_norm": 0.15581420242804653,
+      "learning_rate": 2e-05,
+      "loss": 5.4836,
+      "step": 7498
+    },
+    {
+      "epoch": 0.503001643357816,
+      "grad_norm": 0.16409706201197455,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 7499
+    },
+    {
+      "epoch": 0.503068719187041,
+      "grad_norm": 0.16316110661818853,
+      "learning_rate": 2e-05,
+      "loss": 5.5302,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5031357950162659,
+      "grad_norm": 0.1552723440105388,
+      "learning_rate": 2e-05,
+      "loss": 5.3801,
+      "step": 7501
+    },
+    {
+      "epoch": 0.5032028708454909,
+      "grad_norm": 0.15904496100285218,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5032699466747158,
+      "grad_norm": 0.16172437947438217,
+      "learning_rate": 2e-05,
+      "loss": 5.3544,
+      "step": 7503
+    },
+    {
+      "epoch": 0.5033370225039407,
+      "grad_norm": 0.1555977406966109,
+      "learning_rate": 2e-05,
+      "loss": 5.565,
+      "step": 7504
+    },
+    {
+      "epoch": 0.5034040983331657,
+      "grad_norm": 0.1512855758698087,
+      "learning_rate": 2e-05,
+      "loss": 5.6012,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5034711741623906,
+      "grad_norm": 0.16741341175430866,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5035382499916156,
+      "grad_norm": 0.15421272991232857,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 7507
+    },
+    {
+      "epoch": 0.5036053258208405,
+      "grad_norm": 0.14945417798386607,
+      "learning_rate": 2e-05,
+      "loss": 5.3949,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5036724016500654,
+      "grad_norm": 0.15499159571218507,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5037394774792904,
+      "grad_norm": 0.16135046871618616,
+      "learning_rate": 2e-05,
+      "loss": 5.491,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5038065533085153,
+      "grad_norm": 0.1442685783308196,
+      "learning_rate": 2e-05,
+      "loss": 5.5971,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5038736291377403,
+      "grad_norm": 0.1549940509401805,
+      "learning_rate": 2e-05,
+      "loss": 5.4764,
+      "step": 7512
+    },
+    {
+      "epoch": 0.5039407049669652,
+      "grad_norm": 0.15443209251802545,
+      "learning_rate": 2e-05,
+      "loss": 5.494,
+      "step": 7513
+    },
+    {
+      "epoch": 0.5040077807961901,
+      "grad_norm": 0.15769478747414042,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5040748566254151,
+      "grad_norm": 0.14724824822733562,
+      "learning_rate": 2e-05,
+      "loss": 5.3798,
+      "step": 7515
+    },
+    {
+      "epoch": 0.50414193245464,
+      "grad_norm": 0.16200403724468118,
+      "learning_rate": 2e-05,
+      "loss": 5.3543,
+      "step": 7516
+    },
+    {
+      "epoch": 0.504209008283865,
+      "grad_norm": 0.15966422013384904,
+      "learning_rate": 2e-05,
+      "loss": 5.3369,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5042760841130899,
+      "grad_norm": 0.15860087424768224,
+      "learning_rate": 2e-05,
+      "loss": 5.3875,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5043431599423148,
+      "grad_norm": 0.1701809595720194,
+      "learning_rate": 2e-05,
+      "loss": 5.5964,
+      "step": 7519
+    },
+    {
+      "epoch": 0.5044102357715398,
+      "grad_norm": 0.14543982872551395,
+      "learning_rate": 2e-05,
+      "loss": 5.3797,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5044773116007647,
+      "grad_norm": 0.15402973294431732,
+      "learning_rate": 2e-05,
+      "loss": 5.3252,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5045443874299896,
+      "grad_norm": 0.15622226430320704,
+      "learning_rate": 2e-05,
+      "loss": 5.4391,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5046114632592146,
+      "grad_norm": 0.15622885906127595,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 7523
+    },
+    {
+      "epoch": 0.5046785390884395,
+      "grad_norm": 0.1488878580951184,
+      "learning_rate": 2e-05,
+      "loss": 5.4351,
+      "step": 7524
+    },
+    {
+      "epoch": 0.5047456149176645,
+      "grad_norm": 0.14865825809103014,
+      "learning_rate": 2e-05,
+      "loss": 5.3247,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5048126907468894,
+      "grad_norm": 0.1524148310653133,
+      "learning_rate": 2e-05,
+      "loss": 5.3722,
+      "step": 7526
+    },
+    {
+      "epoch": 0.5048797665761143,
+      "grad_norm": 0.15064548386176987,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5049468424053393,
+      "grad_norm": 0.15336677885681033,
+      "learning_rate": 2e-05,
+      "loss": 5.4941,
+      "step": 7528
+    },
+    {
+      "epoch": 0.5050139182345642,
+      "grad_norm": 0.15571494998708832,
+      "learning_rate": 2e-05,
+      "loss": 5.3418,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5050809940637891,
+      "grad_norm": 0.14579573042145175,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5051480698930141,
+      "grad_norm": 0.15597170315336029,
+      "learning_rate": 2e-05,
+      "loss": 5.3774,
+      "step": 7531
+    },
+    {
+      "epoch": 0.505215145722239,
+      "grad_norm": 0.15051703713472628,
+      "learning_rate": 2e-05,
+      "loss": 5.4618,
+      "step": 7532
+    },
+    {
+      "epoch": 0.505282221551464,
+      "grad_norm": 0.15581982642043402,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5053492973806889,
+      "grad_norm": 0.15255052382456633,
+      "learning_rate": 2e-05,
+      "loss": 5.3646,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5054163732099138,
+      "grad_norm": 0.15561757272734764,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5054834490391388,
+      "grad_norm": 0.14595237775738276,
+      "learning_rate": 2e-05,
+      "loss": 5.5249,
+      "step": 7536
+    },
+    {
+      "epoch": 0.5055505248683637,
+      "grad_norm": 0.15246835065024367,
+      "learning_rate": 2e-05,
+      "loss": 5.4411,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5056176006975887,
+      "grad_norm": 0.15462230404000918,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5056846765268136,
+      "grad_norm": 0.14925308527166486,
+      "learning_rate": 2e-05,
+      "loss": 5.5144,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5057517523560385,
+      "grad_norm": 0.1477493951387361,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5058188281852635,
+      "grad_norm": 0.15520885657615124,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5058859040144884,
+      "grad_norm": 0.14238976743383475,
+      "learning_rate": 2e-05,
+      "loss": 5.3297,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5059529798437133,
+      "grad_norm": 0.15294774764066438,
+      "learning_rate": 2e-05,
+      "loss": 5.4485,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5060200556729383,
+      "grad_norm": 0.14647035411921716,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 7544
+    },
+    {
+      "epoch": 0.5060871315021632,
+      "grad_norm": 0.1609507220392526,
+      "learning_rate": 2e-05,
+      "loss": 5.3018,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5061542073313882,
+      "grad_norm": 0.1540920904053237,
+      "learning_rate": 2e-05,
+      "loss": 5.3038,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5062212831606131,
+      "grad_norm": 0.14572142642548347,
+      "learning_rate": 2e-05,
+      "loss": 5.3856,
+      "step": 7547
+    },
+    {
+      "epoch": 0.506288358989838,
+      "grad_norm": 0.15462267685004072,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 7548
+    },
+    {
+      "epoch": 0.506355434819063,
+      "grad_norm": 0.15287491033591236,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 7549
+    },
+    {
+      "epoch": 0.5064225106482879,
+      "grad_norm": 0.14631982073646543,
+      "learning_rate": 2e-05,
+      "loss": 5.3563,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5064895864775129,
+      "grad_norm": 0.14938858592745075,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 7551
+    },
+    {
+      "epoch": 0.5065566623067378,
+      "grad_norm": 0.14804948860152078,
+      "learning_rate": 2e-05,
+      "loss": 5.3713,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5066237381359627,
+      "grad_norm": 0.14608953982602568,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 7553
+    },
+    {
+      "epoch": 0.5066908139651877,
+      "grad_norm": 0.14860303200834343,
+      "learning_rate": 2e-05,
+      "loss": 5.4153,
+      "step": 7554
+    },
+    {
+      "epoch": 0.5067578897944126,
+      "grad_norm": 0.14746463124460668,
+      "learning_rate": 2e-05,
+      "loss": 5.5013,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5068249656236375,
+      "grad_norm": 0.14792704423918232,
+      "learning_rate": 2e-05,
+      "loss": 5.3648,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5068920414528625,
+      "grad_norm": 0.1571752402921104,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5069591172820874,
+      "grad_norm": 0.1517817598756393,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5070261931113124,
+      "grad_norm": 0.148922437795159,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5070932689405373,
+      "grad_norm": 0.14973462643150195,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5071603447697622,
+      "grad_norm": 0.16104454047550829,
+      "learning_rate": 2e-05,
+      "loss": 5.5939,
+      "step": 7561
+    },
+    {
+      "epoch": 0.5072274205989872,
+      "grad_norm": 0.14717330435305756,
+      "learning_rate": 2e-05,
+      "loss": 5.3391,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5072944964282121,
+      "grad_norm": 0.14919223523818573,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 7563
+    },
+    {
+      "epoch": 0.507361572257437,
+      "grad_norm": 0.15406037726872496,
+      "learning_rate": 2e-05,
+      "loss": 5.3623,
+      "step": 7564
+    },
+    {
+      "epoch": 0.507428648086662,
+      "grad_norm": 0.15006822839063239,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5074957239158869,
+      "grad_norm": 0.14956077846074434,
+      "learning_rate": 2e-05,
+      "loss": 5.5132,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5075627997451119,
+      "grad_norm": 0.15181966359698698,
+      "learning_rate": 2e-05,
+      "loss": 5.178,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5076298755743368,
+      "grad_norm": 0.14718386882615855,
+      "learning_rate": 2e-05,
+      "loss": 5.4282,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5076969514035617,
+      "grad_norm": 0.14563893946392228,
+      "learning_rate": 2e-05,
+      "loss": 5.2596,
+      "step": 7569
+    },
+    {
+      "epoch": 0.5077640272327867,
+      "grad_norm": 0.15052146440477016,
+      "learning_rate": 2e-05,
+      "loss": 5.4976,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5078311030620116,
+      "grad_norm": 0.14938538793459796,
+      "learning_rate": 2e-05,
+      "loss": 5.3431,
+      "step": 7571
+    },
+    {
+      "epoch": 0.5078981788912366,
+      "grad_norm": 0.147062637690144,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5079652547204615,
+      "grad_norm": 0.15014932738276432,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 7573
+    },
+    {
+      "epoch": 0.5080323305496864,
+      "grad_norm": 0.1489978517965839,
+      "learning_rate": 2e-05,
+      "loss": 5.4018,
+      "step": 7574
+    },
+    {
+      "epoch": 0.5080994063789114,
+      "grad_norm": 0.15063499453367937,
+      "learning_rate": 2e-05,
+      "loss": 5.4362,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5081664822081363,
+      "grad_norm": 0.150408440855287,
+      "learning_rate": 2e-05,
+      "loss": 5.465,
+      "step": 7576
+    },
+    {
+      "epoch": 0.5082335580373613,
+      "grad_norm": 0.15103741755638933,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 7577
+    },
+    {
+      "epoch": 0.5083006338665862,
+      "grad_norm": 0.1509799651660148,
+      "learning_rate": 2e-05,
+      "loss": 5.2888,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5083677096958111,
+      "grad_norm": 0.14840615309117128,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5084347855250361,
+      "grad_norm": 0.15018406027837014,
+      "learning_rate": 2e-05,
+      "loss": 5.3081,
+      "step": 7580
+    },
+    {
+      "epoch": 0.508501861354261,
+      "grad_norm": 0.15923753262360887,
+      "learning_rate": 2e-05,
+      "loss": 5.4035,
+      "step": 7581
+    },
+    {
+      "epoch": 0.508568937183486,
+      "grad_norm": 0.1456558606257999,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5086360130127109,
+      "grad_norm": 0.14965769461585673,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 7583
+    },
+    {
+      "epoch": 0.5087030888419358,
+      "grad_norm": 0.15422933604087044,
+      "learning_rate": 2e-05,
+      "loss": 5.4153,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5087701646711608,
+      "grad_norm": 0.14910146634428328,
+      "learning_rate": 2e-05,
+      "loss": 5.3739,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5088372405003857,
+      "grad_norm": 0.1443853942229689,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5089043163296106,
+      "grad_norm": 0.15334256271399696,
+      "learning_rate": 2e-05,
+      "loss": 5.3126,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5089713921588356,
+      "grad_norm": 0.15072423909399402,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 7588
+    },
+    {
+      "epoch": 0.5090384679880605,
+      "grad_norm": 0.14727678453373363,
+      "learning_rate": 2e-05,
+      "loss": 5.5347,
+      "step": 7589
+    },
+    {
+      "epoch": 0.5091055438172855,
+      "grad_norm": 0.15863148233431099,
+      "learning_rate": 2e-05,
+      "loss": 5.357,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5091726196465104,
+      "grad_norm": 0.15373381307326667,
+      "learning_rate": 2e-05,
+      "loss": 5.5135,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5092396954757353,
+      "grad_norm": 0.13931670169627325,
+      "learning_rate": 2e-05,
+      "loss": 5.439,
+      "step": 7592
+    },
+    {
+      "epoch": 0.5093067713049603,
+      "grad_norm": 0.15801862930125396,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 7593
+    },
+    {
+      "epoch": 0.5093738471341852,
+      "grad_norm": 0.1573108988707798,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 7594
+    },
+    {
+      "epoch": 0.5094409229634101,
+      "grad_norm": 0.1453394338526487,
+      "learning_rate": 2e-05,
+      "loss": 5.4832,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5095079987926351,
+      "grad_norm": 0.14730489405289945,
+      "learning_rate": 2e-05,
+      "loss": 5.4851,
+      "step": 7596
+    },
+    {
+      "epoch": 0.50957507462186,
+      "grad_norm": 0.15397280382726442,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 7597
+    },
+    {
+      "epoch": 0.509642150451085,
+      "grad_norm": 0.15546870645047167,
+      "learning_rate": 2e-05,
+      "loss": 5.4355,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5097092262803099,
+      "grad_norm": 0.1484884084723075,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 7599
+    },
+    {
+      "epoch": 0.5097763021095348,
+      "grad_norm": 0.1514570436352918,
+      "learning_rate": 2e-05,
+      "loss": 5.5365,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5098433779387598,
+      "grad_norm": 0.153990255441471,
+      "learning_rate": 2e-05,
+      "loss": 5.5479,
+      "step": 7601
+    },
+    {
+      "epoch": 0.5099104537679847,
+      "grad_norm": 0.15788832680558837,
+      "learning_rate": 2e-05,
+      "loss": 5.3309,
+      "step": 7602
+    },
+    {
+      "epoch": 0.5099775295972097,
+      "grad_norm": 0.157094327313736,
+      "learning_rate": 2e-05,
+      "loss": 5.4606,
+      "step": 7603
+    },
+    {
+      "epoch": 0.5100446054264346,
+      "grad_norm": 0.15307942020264406,
+      "learning_rate": 2e-05,
+      "loss": 5.3572,
+      "step": 7604
+    },
+    {
+      "epoch": 0.5101116812556595,
+      "grad_norm": 0.15079700323832285,
+      "learning_rate": 2e-05,
+      "loss": 5.42,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5101787570848845,
+      "grad_norm": 0.15662299469034038,
+      "learning_rate": 2e-05,
+      "loss": 5.4682,
+      "step": 7606
+    },
+    {
+      "epoch": 0.5102458329141094,
+      "grad_norm": 0.15609398035392588,
+      "learning_rate": 2e-05,
+      "loss": 5.3695,
+      "step": 7607
+    },
+    {
+      "epoch": 0.5103129087433343,
+      "grad_norm": 0.1492633766170121,
+      "learning_rate": 2e-05,
+      "loss": 5.353,
+      "step": 7608
+    },
+    {
+      "epoch": 0.5103799845725593,
+      "grad_norm": 0.14941435342987225,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 7609
+    },
+    {
+      "epoch": 0.5104470604017842,
+      "grad_norm": 0.15620502254308494,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5105141362310092,
+      "grad_norm": 0.14499461845251094,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 7611
+    },
+    {
+      "epoch": 0.5105812120602341,
+      "grad_norm": 0.14766362521951215,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 7612
+    },
+    {
+      "epoch": 0.510648287889459,
+      "grad_norm": 0.14472576882281735,
+      "learning_rate": 2e-05,
+      "loss": 5.5275,
+      "step": 7613
+    },
+    {
+      "epoch": 0.510715363718684,
+      "grad_norm": 0.14912541925591422,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 7614
+    },
+    {
+      "epoch": 0.5107824395479089,
+      "grad_norm": 0.15234222343508735,
+      "learning_rate": 2e-05,
+      "loss": 5.4509,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5108495153771339,
+      "grad_norm": 0.14868414212318373,
+      "learning_rate": 2e-05,
+      "loss": 5.5894,
+      "step": 7616
+    },
+    {
+      "epoch": 0.5109165912063588,
+      "grad_norm": 0.15085819596684,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 7617
+    },
+    {
+      "epoch": 0.5109836670355837,
+      "grad_norm": 0.14476936667735474,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 7618
+    },
+    {
+      "epoch": 0.5110507428648087,
+      "grad_norm": 0.14401685248645452,
+      "learning_rate": 2e-05,
+      "loss": 5.5535,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5111178186940336,
+      "grad_norm": 0.15172274667696406,
+      "learning_rate": 2e-05,
+      "loss": 5.4994,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5111848945232585,
+      "grad_norm": 0.1591335148530309,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5112519703524835,
+      "grad_norm": 0.15656667069772207,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5113190461817084,
+      "grad_norm": 0.15326980473018242,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 7623
+    },
+    {
+      "epoch": 0.5113861220109334,
+      "grad_norm": 0.14989742764568398,
+      "learning_rate": 2e-05,
+      "loss": 5.4836,
+      "step": 7624
+    },
+    {
+      "epoch": 0.5114531978401583,
+      "grad_norm": 0.15302909877947926,
+      "learning_rate": 2e-05,
+      "loss": 5.4993,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5115202736693832,
+      "grad_norm": 0.14476993387232606,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5115873494986082,
+      "grad_norm": 0.14607021395160572,
+      "learning_rate": 2e-05,
+      "loss": 5.4959,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5116544253278331,
+      "grad_norm": 0.15997244055005225,
+      "learning_rate": 2e-05,
+      "loss": 5.3844,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5117215011570581,
+      "grad_norm": 0.15881026810411028,
+      "learning_rate": 2e-05,
+      "loss": 5.5858,
+      "step": 7629
+    },
+    {
+      "epoch": 0.511788576986283,
+      "grad_norm": 0.15695121400969286,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5118556528155079,
+      "grad_norm": 0.15887182533019192,
+      "learning_rate": 2e-05,
+      "loss": 5.4359,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5119227286447329,
+      "grad_norm": 0.15475040428560927,
+      "learning_rate": 2e-05,
+      "loss": 5.4574,
+      "step": 7632
+    },
+    {
+      "epoch": 0.5119898044739578,
+      "grad_norm": 0.15442883795291307,
+      "learning_rate": 2e-05,
+      "loss": 5.4706,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5120568803031827,
+      "grad_norm": 0.15162997438110584,
+      "learning_rate": 2e-05,
+      "loss": 5.4955,
+      "step": 7634
+    },
+    {
+      "epoch": 0.5121239561324077,
+      "grad_norm": 0.15573136519131445,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5121910319616326,
+      "grad_norm": 0.14705516620108924,
+      "learning_rate": 2e-05,
+      "loss": 5.5971,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5122581077908576,
+      "grad_norm": 0.1437975587992663,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5123251836200825,
+      "grad_norm": 0.1507593343983091,
+      "learning_rate": 2e-05,
+      "loss": 5.439,
+      "step": 7638
+    },
+    {
+      "epoch": 0.5123922594493074,
+      "grad_norm": 0.15213452640145486,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5124593352785324,
+      "grad_norm": 0.15975794290191753,
+      "learning_rate": 2e-05,
+      "loss": 5.3974,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5125264111077573,
+      "grad_norm": 0.15000335106080795,
+      "learning_rate": 2e-05,
+      "loss": 5.438,
+      "step": 7641
+    },
+    {
+      "epoch": 0.5125934869369823,
+      "grad_norm": 0.15284599122181805,
+      "learning_rate": 2e-05,
+      "loss": 5.4216,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5126605627662072,
+      "grad_norm": 0.16316873210352706,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5127276385954321,
+      "grad_norm": 0.15250168852958268,
+      "learning_rate": 2e-05,
+      "loss": 5.3392,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5127947144246571,
+      "grad_norm": 0.14562764627007063,
+      "learning_rate": 2e-05,
+      "loss": 5.3206,
+      "step": 7645
+    },
+    {
+      "epoch": 0.512861790253882,
+      "grad_norm": 0.1479518019326918,
+      "learning_rate": 2e-05,
+      "loss": 5.5282,
+      "step": 7646
+    },
+    {
+      "epoch": 0.512928866083107,
+      "grad_norm": 0.16029993146307445,
+      "learning_rate": 2e-05,
+      "loss": 5.4764,
+      "step": 7647
+    },
+    {
+      "epoch": 0.5129959419123319,
+      "grad_norm": 0.15639745502381192,
+      "learning_rate": 2e-05,
+      "loss": 5.3985,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5130630177415568,
+      "grad_norm": 0.14423703904796029,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 7649
+    },
+    {
+      "epoch": 0.5131300935707818,
+      "grad_norm": 0.14629728231502157,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5131971694000067,
+      "grad_norm": 0.1658535509047625,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 7651
+    },
+    {
+      "epoch": 0.5132642452292316,
+      "grad_norm": 0.1481675659466687,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5133313210584566,
+      "grad_norm": 0.15892548494603626,
+      "learning_rate": 2e-05,
+      "loss": 5.3836,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5133983968876815,
+      "grad_norm": 0.16006703389268298,
+      "learning_rate": 2e-05,
+      "loss": 5.527,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5134654727169065,
+      "grad_norm": 0.16533057819683486,
+      "learning_rate": 2e-05,
+      "loss": 5.4979,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5135325485461314,
+      "grad_norm": 0.15021866358211147,
+      "learning_rate": 2e-05,
+      "loss": 5.4679,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5135996243753563,
+      "grad_norm": 0.1608442447319529,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5136667002045813,
+      "grad_norm": 0.16017250648408266,
+      "learning_rate": 2e-05,
+      "loss": 5.3705,
+      "step": 7658
+    },
+    {
+      "epoch": 0.5137337760338062,
+      "grad_norm": 0.1615898978727251,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 7659
+    },
+    {
+      "epoch": 0.5138008518630311,
+      "grad_norm": 0.15245707512707762,
+      "learning_rate": 2e-05,
+      "loss": 5.2335,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5138679276922561,
+      "grad_norm": 0.150608780898398,
+      "learning_rate": 2e-05,
+      "loss": 5.3518,
+      "step": 7661
+    },
+    {
+      "epoch": 0.513935003521481,
+      "grad_norm": 0.16658344348129966,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 7662
+    },
+    {
+      "epoch": 0.514002079350706,
+      "grad_norm": 0.15414837865642642,
+      "learning_rate": 2e-05,
+      "loss": 5.3994,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5140691551799309,
+      "grad_norm": 0.14710594458071233,
+      "learning_rate": 2e-05,
+      "loss": 5.6021,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5141362310091558,
+      "grad_norm": 0.15883633776531575,
+      "learning_rate": 2e-05,
+      "loss": 5.4396,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5142033068383808,
+      "grad_norm": 0.15656448728474712,
+      "learning_rate": 2e-05,
+      "loss": 5.3889,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5142703826676057,
+      "grad_norm": 0.15708723653061363,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5143374584968307,
+      "grad_norm": 0.1512681668550003,
+      "learning_rate": 2e-05,
+      "loss": 5.4363,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5144045343260556,
+      "grad_norm": 0.15001780911602486,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 7669
+    },
+    {
+      "epoch": 0.5144716101552805,
+      "grad_norm": 0.1526575649139081,
+      "learning_rate": 2e-05,
+      "loss": 5.4889,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5145386859845055,
+      "grad_norm": 0.15287851987678633,
+      "learning_rate": 2e-05,
+      "loss": 5.442,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5146057618137304,
+      "grad_norm": 0.15589549987610213,
+      "learning_rate": 2e-05,
+      "loss": 5.3592,
+      "step": 7672
+    },
+    {
+      "epoch": 0.5146728376429553,
+      "grad_norm": 0.14883967822712496,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5147399134721803,
+      "grad_norm": 0.14671914412309597,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 7674
+    },
+    {
+      "epoch": 0.5148069893014052,
+      "grad_norm": 0.14921673064995838,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5148740651306302,
+      "grad_norm": 0.16230092916474514,
+      "learning_rate": 2e-05,
+      "loss": 5.5898,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5149411409598551,
+      "grad_norm": 0.15442752652597477,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 7677
+    },
+    {
+      "epoch": 0.51500821678908,
+      "grad_norm": 0.15378506819838372,
+      "learning_rate": 2e-05,
+      "loss": 5.3752,
+      "step": 7678
+    },
+    {
+      "epoch": 0.515075292618305,
+      "grad_norm": 0.1558073644226669,
+      "learning_rate": 2e-05,
+      "loss": 5.4768,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5151423684475299,
+      "grad_norm": 0.1583105554294859,
+      "learning_rate": 2e-05,
+      "loss": 5.398,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5152094442767549,
+      "grad_norm": 0.14661098547558882,
+      "learning_rate": 2e-05,
+      "loss": 5.4678,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5152765201059798,
+      "grad_norm": 0.17733575805153223,
+      "learning_rate": 2e-05,
+      "loss": 5.4574,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5153435959352047,
+      "grad_norm": 0.1600050222764656,
+      "learning_rate": 2e-05,
+      "loss": 5.341,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5154106717644297,
+      "grad_norm": 0.1549602610877884,
+      "learning_rate": 2e-05,
+      "loss": 5.3858,
+      "step": 7684
+    },
+    {
+      "epoch": 0.5154777475936546,
+      "grad_norm": 0.15267494414411276,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5155448234228796,
+      "grad_norm": 0.16480339005872374,
+      "learning_rate": 2e-05,
+      "loss": 5.3944,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5156118992521045,
+      "grad_norm": 0.1536399577533238,
+      "learning_rate": 2e-05,
+      "loss": 5.3053,
+      "step": 7687
+    },
+    {
+      "epoch": 0.5156789750813294,
+      "grad_norm": 0.15476920740158034,
+      "learning_rate": 2e-05,
+      "loss": 5.3604,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5157460509105544,
+      "grad_norm": 0.15908825798849247,
+      "learning_rate": 2e-05,
+      "loss": 5.3364,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5158131267397793,
+      "grad_norm": 0.14820788959885473,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5158802025690042,
+      "grad_norm": 0.1469596227520547,
+      "learning_rate": 2e-05,
+      "loss": 5.3922,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5159472783982292,
+      "grad_norm": 0.1499449168157889,
+      "learning_rate": 2e-05,
+      "loss": 5.3436,
+      "step": 7692
+    },
+    {
+      "epoch": 0.5160143542274541,
+      "grad_norm": 0.15300477427799156,
+      "learning_rate": 2e-05,
+      "loss": 5.5045,
+      "step": 7693
+    },
+    {
+      "epoch": 0.5160814300566791,
+      "grad_norm": 0.1448400497149715,
+      "learning_rate": 2e-05,
+      "loss": 5.5308,
+      "step": 7694
+    },
+    {
+      "epoch": 0.516148505885904,
+      "grad_norm": 0.14829417926863836,
+      "learning_rate": 2e-05,
+      "loss": 5.4247,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5162155817151289,
+      "grad_norm": 0.15171919349912233,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 7696
+    },
+    {
+      "epoch": 0.5162826575443539,
+      "grad_norm": 0.14647916072400471,
+      "learning_rate": 2e-05,
+      "loss": 5.3832,
+      "step": 7697
+    },
+    {
+      "epoch": 0.5163497333735788,
+      "grad_norm": 0.14662961808995686,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 7698
+    },
+    {
+      "epoch": 0.5164168092028038,
+      "grad_norm": 0.14818305159158227,
+      "learning_rate": 2e-05,
+      "loss": 5.4611,
+      "step": 7699
+    },
+    {
+      "epoch": 0.5164838850320287,
+      "grad_norm": 0.1472703916531397,
+      "learning_rate": 2e-05,
+      "loss": 5.4866,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5165509608612536,
+      "grad_norm": 0.1498648294507365,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 7701
+    },
+    {
+      "epoch": 0.5166180366904786,
+      "grad_norm": 0.1547913262560126,
+      "learning_rate": 2e-05,
+      "loss": 5.3999,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5166851125197035,
+      "grad_norm": 0.14985115160033355,
+      "learning_rate": 2e-05,
+      "loss": 5.4727,
+      "step": 7703
+    },
+    {
+      "epoch": 0.5167521883489284,
+      "grad_norm": 0.1491213591920621,
+      "learning_rate": 2e-05,
+      "loss": 5.5009,
+      "step": 7704
+    },
+    {
+      "epoch": 0.5168192641781534,
+      "grad_norm": 0.1484014863360101,
+      "learning_rate": 2e-05,
+      "loss": 5.4409,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5168863400073783,
+      "grad_norm": 0.14767020121203944,
+      "learning_rate": 2e-05,
+      "loss": 5.3215,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5169534158366033,
+      "grad_norm": 0.14523954830326963,
+      "learning_rate": 2e-05,
+      "loss": 5.3752,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5170204916658282,
+      "grad_norm": 0.1480729836567034,
+      "learning_rate": 2e-05,
+      "loss": 5.3981,
+      "step": 7708
+    },
+    {
+      "epoch": 0.5170875674950531,
+      "grad_norm": 0.14965903400382483,
+      "learning_rate": 2e-05,
+      "loss": 5.4133,
+      "step": 7709
+    },
+    {
+      "epoch": 0.5171546433242781,
+      "grad_norm": 0.15254793415495618,
+      "learning_rate": 2e-05,
+      "loss": 5.5497,
+      "step": 7710
+    },
+    {
+      "epoch": 0.517221719153503,
+      "grad_norm": 0.1458417599610557,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 7711
+    },
+    {
+      "epoch": 0.517288794982728,
+      "grad_norm": 0.15053496708476682,
+      "learning_rate": 2e-05,
+      "loss": 5.3102,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5173558708119529,
+      "grad_norm": 0.14679675421741548,
+      "learning_rate": 2e-05,
+      "loss": 5.3301,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5174229466411778,
+      "grad_norm": 0.15303785797929498,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 7714
+    },
+    {
+      "epoch": 0.5174900224704028,
+      "grad_norm": 0.15522401878999745,
+      "learning_rate": 2e-05,
+      "loss": 5.4346,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5175570982996277,
+      "grad_norm": 0.14778025186612223,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5176241741288526,
+      "grad_norm": 0.15256367397035484,
+      "learning_rate": 2e-05,
+      "loss": 5.507,
+      "step": 7717
+    },
+    {
+      "epoch": 0.5176912499580776,
+      "grad_norm": 0.15352111225142825,
+      "learning_rate": 2e-05,
+      "loss": 5.3384,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5177583257873025,
+      "grad_norm": 0.1459458002368136,
+      "learning_rate": 2e-05,
+      "loss": 5.3999,
+      "step": 7719
+    },
+    {
+      "epoch": 0.5178254016165275,
+      "grad_norm": 0.15463651723491362,
+      "learning_rate": 2e-05,
+      "loss": 5.4925,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5178924774457524,
+      "grad_norm": 0.1519103050633396,
+      "learning_rate": 2e-05,
+      "loss": 5.4338,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5179595532749773,
+      "grad_norm": 0.1518184449778245,
+      "learning_rate": 2e-05,
+      "loss": 5.4893,
+      "step": 7722
+    },
+    {
+      "epoch": 0.5180266291042023,
+      "grad_norm": 0.1512378277351227,
+      "learning_rate": 2e-05,
+      "loss": 5.7071,
+      "step": 7723
+    },
+    {
+      "epoch": 0.5180937049334272,
+      "grad_norm": 0.15384512355507957,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 7724
+    },
+    {
+      "epoch": 0.5181607807626522,
+      "grad_norm": 0.15415309905112679,
+      "learning_rate": 2e-05,
+      "loss": 5.4563,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5182278565918771,
+      "grad_norm": 0.1600368102916811,
+      "learning_rate": 2e-05,
+      "loss": 5.385,
+      "step": 7726
+    },
+    {
+      "epoch": 0.518294932421102,
+      "grad_norm": 0.16303720932024757,
+      "learning_rate": 2e-05,
+      "loss": 5.3924,
+      "step": 7727
+    },
+    {
+      "epoch": 0.518362008250327,
+      "grad_norm": 0.15420793331311367,
+      "learning_rate": 2e-05,
+      "loss": 5.27,
+      "step": 7728
+    },
+    {
+      "epoch": 0.5184290840795519,
+      "grad_norm": 0.15500285062120248,
+      "learning_rate": 2e-05,
+      "loss": 5.355,
+      "step": 7729
+    },
+    {
+      "epoch": 0.5184961599087768,
+      "grad_norm": 0.1635744722453185,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5185632357380018,
+      "grad_norm": 0.1471330943568756,
+      "learning_rate": 2e-05,
+      "loss": 5.4967,
+      "step": 7731
+    },
+    {
+      "epoch": 0.5186303115672267,
+      "grad_norm": 0.15240709052636142,
+      "learning_rate": 2e-05,
+      "loss": 5.4287,
+      "step": 7732
+    },
+    {
+      "epoch": 0.5186973873964517,
+      "grad_norm": 0.15720918772652734,
+      "learning_rate": 2e-05,
+      "loss": 5.3122,
+      "step": 7733
+    },
+    {
+      "epoch": 0.5187644632256766,
+      "grad_norm": 0.14925517069121788,
+      "learning_rate": 2e-05,
+      "loss": 5.5154,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5188315390549015,
+      "grad_norm": 0.1501882984802569,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5188986148841265,
+      "grad_norm": 0.15364358821597138,
+      "learning_rate": 2e-05,
+      "loss": 5.3612,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5189656907133514,
+      "grad_norm": 0.1557392986372127,
+      "learning_rate": 2e-05,
+      "loss": 5.3272,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5190327665425764,
+      "grad_norm": 0.15570753754171898,
+      "learning_rate": 2e-05,
+      "loss": 5.5156,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5190998423718013,
+      "grad_norm": 0.16017658011749175,
+      "learning_rate": 2e-05,
+      "loss": 5.6207,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5191669182010262,
+      "grad_norm": 0.1549384950662759,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5192339940302512,
+      "grad_norm": 0.1501984805434172,
+      "learning_rate": 2e-05,
+      "loss": 5.3577,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5193010698594761,
+      "grad_norm": 0.1502579822025191,
+      "learning_rate": 2e-05,
+      "loss": 5.39,
+      "step": 7742
+    },
+    {
+      "epoch": 0.519368145688701,
+      "grad_norm": 0.15477423334600743,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 7743
+    },
+    {
+      "epoch": 0.519435221517926,
+      "grad_norm": 0.15785626158768254,
+      "learning_rate": 2e-05,
+      "loss": 5.4976,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5195022973471509,
+      "grad_norm": 0.1487412840523818,
+      "learning_rate": 2e-05,
+      "loss": 5.3023,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5195693731763759,
+      "grad_norm": 0.14653341245293147,
+      "learning_rate": 2e-05,
+      "loss": 5.5266,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5196364490056008,
+      "grad_norm": 0.147534331267081,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5197035248348257,
+      "grad_norm": 0.14573432825398613,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5197706006640507,
+      "grad_norm": 0.1538001515832809,
+      "learning_rate": 2e-05,
+      "loss": 5.4002,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5198376764932756,
+      "grad_norm": 0.14984352418119395,
+      "learning_rate": 2e-05,
+      "loss": 5.4749,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5199047523225006,
+      "grad_norm": 0.1522305471893058,
+      "learning_rate": 2e-05,
+      "loss": 5.4602,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5199718281517255,
+      "grad_norm": 0.15407570889706987,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5200389039809504,
+      "grad_norm": 0.15100129091019632,
+      "learning_rate": 2e-05,
+      "loss": 5.3494,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5201059798101754,
+      "grad_norm": 0.15459092976610744,
+      "learning_rate": 2e-05,
+      "loss": 5.4935,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5201730556394003,
+      "grad_norm": 0.1500051508247163,
+      "learning_rate": 2e-05,
+      "loss": 5.5554,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5202401314686252,
+      "grad_norm": 0.15526610859586082,
+      "learning_rate": 2e-05,
+      "loss": 5.3313,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5203072072978502,
+      "grad_norm": 0.15499508480194826,
+      "learning_rate": 2e-05,
+      "loss": 5.5955,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5203742831270751,
+      "grad_norm": 0.15130242337698946,
+      "learning_rate": 2e-05,
+      "loss": 5.2889,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5204413589563001,
+      "grad_norm": 0.15830696765801408,
+      "learning_rate": 2e-05,
+      "loss": 5.2442,
+      "step": 7759
+    },
+    {
+      "epoch": 0.520508434785525,
+      "grad_norm": 0.1572949744548908,
+      "learning_rate": 2e-05,
+      "loss": 5.338,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5205755106147499,
+      "grad_norm": 0.1459604654815613,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5206425864439749,
+      "grad_norm": 0.1572204425976535,
+      "learning_rate": 2e-05,
+      "loss": 5.5209,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5207096622731998,
+      "grad_norm": 0.15645183130435852,
+      "learning_rate": 2e-05,
+      "loss": 5.4577,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5207767381024248,
+      "grad_norm": 0.16117027429622713,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5208438139316497,
+      "grad_norm": 0.16002656280473487,
+      "learning_rate": 2e-05,
+      "loss": 5.3289,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5209108897608746,
+      "grad_norm": 0.14938308454589772,
+      "learning_rate": 2e-05,
+      "loss": 5.3241,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5209779655900996,
+      "grad_norm": 0.16377694854659905,
+      "learning_rate": 2e-05,
+      "loss": 5.4073,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5210450414193245,
+      "grad_norm": 0.15809894600669008,
+      "learning_rate": 2e-05,
+      "loss": 5.423,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5211121172485494,
+      "grad_norm": 0.15365266085238227,
+      "learning_rate": 2e-05,
+      "loss": 5.3012,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5211791930777744,
+      "grad_norm": 0.16448464538097085,
+      "learning_rate": 2e-05,
+      "loss": 5.3253,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5212462689069993,
+      "grad_norm": 0.1409923326631212,
+      "learning_rate": 2e-05,
+      "loss": 5.452,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5213133447362243,
+      "grad_norm": 0.1579726933985827,
+      "learning_rate": 2e-05,
+      "loss": 5.4354,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5213804205654492,
+      "grad_norm": 0.16555858419571556,
+      "learning_rate": 2e-05,
+      "loss": 5.3472,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5214474963946741,
+      "grad_norm": 0.1459448136360117,
+      "learning_rate": 2e-05,
+      "loss": 5.4907,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5215145722238991,
+      "grad_norm": 0.15188370354923134,
+      "learning_rate": 2e-05,
+      "loss": 5.4607,
+      "step": 7775
+    },
+    {
+      "epoch": 0.521581648053124,
+      "grad_norm": 0.15212481234973688,
+      "learning_rate": 2e-05,
+      "loss": 5.4669,
+      "step": 7776
+    },
+    {
+      "epoch": 0.521648723882349,
+      "grad_norm": 0.15097703167170445,
+      "learning_rate": 2e-05,
+      "loss": 5.3932,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5217157997115739,
+      "grad_norm": 0.1543832973542512,
+      "learning_rate": 2e-05,
+      "loss": 5.4101,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5217828755407988,
+      "grad_norm": 0.148657251230234,
+      "learning_rate": 2e-05,
+      "loss": 5.4027,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5218499513700238,
+      "grad_norm": 0.1653826438530842,
+      "learning_rate": 2e-05,
+      "loss": 5.5198,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5219170271992487,
+      "grad_norm": 0.15310882108095947,
+      "learning_rate": 2e-05,
+      "loss": 5.4086,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5219841030284736,
+      "grad_norm": 0.1498807652890514,
+      "learning_rate": 2e-05,
+      "loss": 5.395,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5220511788576986,
+      "grad_norm": 0.16374002547562255,
+      "learning_rate": 2e-05,
+      "loss": 5.3071,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5221182546869235,
+      "grad_norm": 0.15395650423817206,
+      "learning_rate": 2e-05,
+      "loss": 5.6125,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5221853305161485,
+      "grad_norm": 0.15217238720462195,
+      "learning_rate": 2e-05,
+      "loss": 5.498,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5222524063453734,
+      "grad_norm": 0.1534732159214014,
+      "learning_rate": 2e-05,
+      "loss": 5.5392,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5223194821745983,
+      "grad_norm": 0.15098889329075169,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5223865580038233,
+      "grad_norm": 0.15377524795660452,
+      "learning_rate": 2e-05,
+      "loss": 5.5268,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5224536338330482,
+      "grad_norm": 0.15364221847653528,
+      "learning_rate": 2e-05,
+      "loss": 5.5734,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5225207096622732,
+      "grad_norm": 0.1497946149788449,
+      "learning_rate": 2e-05,
+      "loss": 5.3909,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5225877854914981,
+      "grad_norm": 0.15364734674130867,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 7791
+    },
+    {
+      "epoch": 0.522654861320723,
+      "grad_norm": 0.1531406917025649,
+      "learning_rate": 2e-05,
+      "loss": 5.5191,
+      "step": 7792
+    },
+    {
+      "epoch": 0.522721937149948,
+      "grad_norm": 0.15186789005735263,
+      "learning_rate": 2e-05,
+      "loss": 5.3441,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5227890129791729,
+      "grad_norm": 0.15951888120385435,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5228560888083978,
+      "grad_norm": 0.16031197151125873,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5229231646376228,
+      "grad_norm": 0.15132770087623668,
+      "learning_rate": 2e-05,
+      "loss": 5.282,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5229902404668477,
+      "grad_norm": 0.15789909655824738,
+      "learning_rate": 2e-05,
+      "loss": 5.3604,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5230573162960727,
+      "grad_norm": 0.15463590774857486,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5231243921252976,
+      "grad_norm": 0.15486518902818117,
+      "learning_rate": 2e-05,
+      "loss": 5.3898,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5231914679545225,
+      "grad_norm": 0.15597035418653515,
+      "learning_rate": 2e-05,
+      "loss": 5.5954,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5232585437837475,
+      "grad_norm": 0.1603367239589493,
+      "learning_rate": 2e-05,
+      "loss": 5.4429,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5233256196129724,
+      "grad_norm": 0.1529500392340027,
+      "learning_rate": 2e-05,
+      "loss": 5.4102,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5233926954421974,
+      "grad_norm": 0.1558738934671131,
+      "learning_rate": 2e-05,
+      "loss": 5.5235,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5234597712714223,
+      "grad_norm": 0.15152993729348724,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5235268471006472,
+      "grad_norm": 0.1564671248754073,
+      "learning_rate": 2e-05,
+      "loss": 5.3462,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5235939229298722,
+      "grad_norm": 0.15485769699665825,
+      "learning_rate": 2e-05,
+      "loss": 5.4917,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5236609987590971,
+      "grad_norm": 0.14599405527663023,
+      "learning_rate": 2e-05,
+      "loss": 5.3993,
+      "step": 7807
+    },
+    {
+      "epoch": 0.523728074588322,
+      "grad_norm": 0.15647151540812226,
+      "learning_rate": 2e-05,
+      "loss": 5.3653,
+      "step": 7808
+    },
+    {
+      "epoch": 0.523795150417547,
+      "grad_norm": 0.16256560821975638,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5238622262467719,
+      "grad_norm": 0.1553898096004137,
+      "learning_rate": 2e-05,
+      "loss": 5.4629,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5239293020759969,
+      "grad_norm": 0.1543514317210756,
+      "learning_rate": 2e-05,
+      "loss": 5.6014,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5239963779052218,
+      "grad_norm": 0.15074802944389154,
+      "learning_rate": 2e-05,
+      "loss": 5.5521,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5240634537344467,
+      "grad_norm": 0.1603864159794326,
+      "learning_rate": 2e-05,
+      "loss": 5.6569,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5241305295636717,
+      "grad_norm": 0.1681061152296203,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5241976053928966,
+      "grad_norm": 0.1522040028856795,
+      "learning_rate": 2e-05,
+      "loss": 5.2686,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5242646812221216,
+      "grad_norm": 0.16210137011498088,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5243317570513465,
+      "grad_norm": 0.15739068549212606,
+      "learning_rate": 2e-05,
+      "loss": 5.2828,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5243988328805715,
+      "grad_norm": 0.1508840476567197,
+      "learning_rate": 2e-05,
+      "loss": 5.3579,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5244659087097965,
+      "grad_norm": 0.16056766936238379,
+      "learning_rate": 2e-05,
+      "loss": 5.5143,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5245329845390214,
+      "grad_norm": 0.16517815226555924,
+      "learning_rate": 2e-05,
+      "loss": 5.3724,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5246000603682464,
+      "grad_norm": 0.1445178206225716,
+      "learning_rate": 2e-05,
+      "loss": 5.4414,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5246671361974713,
+      "grad_norm": 0.16146435331071488,
+      "learning_rate": 2e-05,
+      "loss": 5.3663,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5247342120266962,
+      "grad_norm": 0.16237203551192098,
+      "learning_rate": 2e-05,
+      "loss": 5.3787,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5248012878559212,
+      "grad_norm": 0.15259733943546275,
+      "learning_rate": 2e-05,
+      "loss": 5.3969,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5248683636851461,
+      "grad_norm": 0.16035610205891926,
+      "learning_rate": 2e-05,
+      "loss": 5.4286,
+      "step": 7825
+    },
+    {
+      "epoch": 0.524935439514371,
+      "grad_norm": 0.1588089922542072,
+      "learning_rate": 2e-05,
+      "loss": 5.5123,
+      "step": 7826
+    },
+    {
+      "epoch": 0.525002515343596,
+      "grad_norm": 0.15189237402575362,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5250695911728209,
+      "grad_norm": 0.1504514240318182,
+      "learning_rate": 2e-05,
+      "loss": 5.4887,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5251366670020459,
+      "grad_norm": 0.16249341479688614,
+      "learning_rate": 2e-05,
+      "loss": 5.3911,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5252037428312708,
+      "grad_norm": 0.15485748258399187,
+      "learning_rate": 2e-05,
+      "loss": 5.3825,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5252708186604957,
+      "grad_norm": 0.14396389116924976,
+      "learning_rate": 2e-05,
+      "loss": 5.4722,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5253378944897207,
+      "grad_norm": 0.1513086466726961,
+      "learning_rate": 2e-05,
+      "loss": 5.4683,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5254049703189456,
+      "grad_norm": 0.16275975148802158,
+      "learning_rate": 2e-05,
+      "loss": 5.4511,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5254720461481706,
+      "grad_norm": 0.15906142427706754,
+      "learning_rate": 2e-05,
+      "loss": 5.4092,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5255391219773955,
+      "grad_norm": 0.1483855660360458,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5256061978066204,
+      "grad_norm": 0.1576068884336193,
+      "learning_rate": 2e-05,
+      "loss": 5.2484,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5256732736358454,
+      "grad_norm": 0.1446482493243385,
+      "learning_rate": 2e-05,
+      "loss": 5.4381,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5257403494650703,
+      "grad_norm": 0.15831069105958231,
+      "learning_rate": 2e-05,
+      "loss": 5.3721,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5258074252942952,
+      "grad_norm": 0.15505023263710777,
+      "learning_rate": 2e-05,
+      "loss": 5.494,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5258745011235202,
+      "grad_norm": 0.15494663192980804,
+      "learning_rate": 2e-05,
+      "loss": 5.4239,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5259415769527451,
+      "grad_norm": 0.14975795730330585,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5260086527819701,
+      "grad_norm": 0.14949036987480338,
+      "learning_rate": 2e-05,
+      "loss": 5.494,
+      "step": 7842
+    },
+    {
+      "epoch": 0.526075728611195,
+      "grad_norm": 0.16413491952728157,
+      "learning_rate": 2e-05,
+      "loss": 5.3754,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5261428044404199,
+      "grad_norm": 0.1578570545198146,
+      "learning_rate": 2e-05,
+      "loss": 5.4968,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5262098802696449,
+      "grad_norm": 0.14967031446786255,
+      "learning_rate": 2e-05,
+      "loss": 5.4014,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5262769560988698,
+      "grad_norm": 0.15354209293757426,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5263440319280948,
+      "grad_norm": 0.14481298752104105,
+      "learning_rate": 2e-05,
+      "loss": 5.3982,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5264111077573197,
+      "grad_norm": 0.15274358620122883,
+      "learning_rate": 2e-05,
+      "loss": 5.5467,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5264781835865446,
+      "grad_norm": 0.15948715434912336,
+      "learning_rate": 2e-05,
+      "loss": 5.536,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5265452594157696,
+      "grad_norm": 0.1522068348628448,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5266123352449945,
+      "grad_norm": 0.15791434620011074,
+      "learning_rate": 2e-05,
+      "loss": 5.5366,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5266794110742195,
+      "grad_norm": 0.15160395307427085,
+      "learning_rate": 2e-05,
+      "loss": 5.5165,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5267464869034444,
+      "grad_norm": 0.14822012691440492,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5268135627326693,
+      "grad_norm": 0.15351617555129007,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5268806385618943,
+      "grad_norm": 0.14680986794717935,
+      "learning_rate": 2e-05,
+      "loss": 5.4987,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5269477143911192,
+      "grad_norm": 0.1493705038981016,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5270147902203441,
+      "grad_norm": 0.15961843699136083,
+      "learning_rate": 2e-05,
+      "loss": 5.342,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5270818660495691,
+      "grad_norm": 0.14693991215459498,
+      "learning_rate": 2e-05,
+      "loss": 5.3445,
+      "step": 7858
+    },
+    {
+      "epoch": 0.527148941878794,
+      "grad_norm": 0.15188486447124286,
+      "learning_rate": 2e-05,
+      "loss": 5.341,
+      "step": 7859
+    },
+    {
+      "epoch": 0.527216017708019,
+      "grad_norm": 0.15070780808172662,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5272830935372439,
+      "grad_norm": 0.15119726761021757,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5273501693664688,
+      "grad_norm": 0.15157601749676133,
+      "learning_rate": 2e-05,
+      "loss": 5.3832,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5274172451956938,
+      "grad_norm": 0.15086121524210236,
+      "learning_rate": 2e-05,
+      "loss": 5.4054,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5274843210249187,
+      "grad_norm": 0.15145609431685852,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5275513968541437,
+      "grad_norm": 0.15160043171071863,
+      "learning_rate": 2e-05,
+      "loss": 5.4426,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5276184726833686,
+      "grad_norm": 0.1529645795378329,
+      "learning_rate": 2e-05,
+      "loss": 5.4927,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5276855485125935,
+      "grad_norm": 0.1576090606745968,
+      "learning_rate": 2e-05,
+      "loss": 5.4699,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5277526243418185,
+      "grad_norm": 0.15059675719104398,
+      "learning_rate": 2e-05,
+      "loss": 5.4771,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5278197001710434,
+      "grad_norm": 0.1511953023826405,
+      "learning_rate": 2e-05,
+      "loss": 5.4957,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5278867760002683,
+      "grad_norm": 0.14667731669488052,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5279538518294933,
+      "grad_norm": 0.15309711871119253,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5280209276587182,
+      "grad_norm": 0.15075148633909893,
+      "learning_rate": 2e-05,
+      "loss": 5.3789,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5280880034879432,
+      "grad_norm": 0.14575499640162426,
+      "learning_rate": 2e-05,
+      "loss": 5.2921,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5281550793171681,
+      "grad_norm": 0.1449291803627925,
+      "learning_rate": 2e-05,
+      "loss": 5.4261,
+      "step": 7874
+    },
+    {
+      "epoch": 0.528222155146393,
+      "grad_norm": 0.1503161622343501,
+      "learning_rate": 2e-05,
+      "loss": 5.2234,
+      "step": 7875
+    },
+    {
+      "epoch": 0.528289230975618,
+      "grad_norm": 0.18182115584865982,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5283563068048429,
+      "grad_norm": 0.14779429329085458,
+      "learning_rate": 2e-05,
+      "loss": 5.5049,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5284233826340679,
+      "grad_norm": 0.15235973760347493,
+      "learning_rate": 2e-05,
+      "loss": 5.35,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5284904584632928,
+      "grad_norm": 0.15513366196432363,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5285575342925177,
+      "grad_norm": 0.14097231600881868,
+      "learning_rate": 2e-05,
+      "loss": 5.3838,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5286246101217427,
+      "grad_norm": 0.14594959992687043,
+      "learning_rate": 2e-05,
+      "loss": 5.3846,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5286916859509676,
+      "grad_norm": 0.14643193819553307,
+      "learning_rate": 2e-05,
+      "loss": 5.3655,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5287587617801925,
+      "grad_norm": 0.15135195501404244,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5288258376094175,
+      "grad_norm": 0.14861402876769497,
+      "learning_rate": 2e-05,
+      "loss": 5.3818,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5288929134386424,
+      "grad_norm": 0.14422475675919114,
+      "learning_rate": 2e-05,
+      "loss": 5.4842,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5289599892678674,
+      "grad_norm": 0.14215302437467428,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5290270650970923,
+      "grad_norm": 0.15540526075359407,
+      "learning_rate": 2e-05,
+      "loss": 5.3806,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5290941409263172,
+      "grad_norm": 0.14790738700693545,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5291612167555422,
+      "grad_norm": 0.14849144180636384,
+      "learning_rate": 2e-05,
+      "loss": 5.4674,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5292282925847671,
+      "grad_norm": 0.14968086101398353,
+      "learning_rate": 2e-05,
+      "loss": 5.4107,
+      "step": 7890
+    },
+    {
+      "epoch": 0.529295368413992,
+      "grad_norm": 0.15276096303040979,
+      "learning_rate": 2e-05,
+      "loss": 5.5651,
+      "step": 7891
+    },
+    {
+      "epoch": 0.529362444243217,
+      "grad_norm": 0.14441667432730676,
+      "learning_rate": 2e-05,
+      "loss": 5.3955,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5294295200724419,
+      "grad_norm": 0.14429171024104792,
+      "learning_rate": 2e-05,
+      "loss": 5.5665,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5294965959016669,
+      "grad_norm": 0.148871768656105,
+      "learning_rate": 2e-05,
+      "loss": 5.3817,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5295636717308918,
+      "grad_norm": 0.15680387886423344,
+      "learning_rate": 2e-05,
+      "loss": 5.5562,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5296307475601167,
+      "grad_norm": 0.14604122239184,
+      "learning_rate": 2e-05,
+      "loss": 5.4688,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5296978233893417,
+      "grad_norm": 0.15080588247782387,
+      "learning_rate": 2e-05,
+      "loss": 5.4653,
+      "step": 7897
+    },
+    {
+      "epoch": 0.5297648992185666,
+      "grad_norm": 0.15086532791092067,
+      "learning_rate": 2e-05,
+      "loss": 5.5307,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5298319750477916,
+      "grad_norm": 0.1537087554713722,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5298990508770165,
+      "grad_norm": 0.1459453669440902,
+      "learning_rate": 2e-05,
+      "loss": 5.3264,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5299661267062414,
+      "grad_norm": 0.15186682814365923,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5300332025354664,
+      "grad_norm": 0.1569843275812385,
+      "learning_rate": 2e-05,
+      "loss": 5.4647,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5301002783646913,
+      "grad_norm": 0.14704962098228108,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5301673541939163,
+      "grad_norm": 0.14897570927334883,
+      "learning_rate": 2e-05,
+      "loss": 5.4429,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5302344300231412,
+      "grad_norm": 0.14606767718283856,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5303015058523661,
+      "grad_norm": 0.15336883031280332,
+      "learning_rate": 2e-05,
+      "loss": 5.3738,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5303685816815911,
+      "grad_norm": 0.14508427375589303,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 7907
+    },
+    {
+      "epoch": 0.530435657510816,
+      "grad_norm": 0.15645292003232347,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5305027333400409,
+      "grad_norm": 0.14693338303129982,
+      "learning_rate": 2e-05,
+      "loss": 5.4305,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5305698091692659,
+      "grad_norm": 0.14494568217681658,
+      "learning_rate": 2e-05,
+      "loss": 5.3446,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5306368849984908,
+      "grad_norm": 0.14231065744974158,
+      "learning_rate": 2e-05,
+      "loss": 5.2931,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5307039608277158,
+      "grad_norm": 0.1457340402389788,
+      "learning_rate": 2e-05,
+      "loss": 5.462,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5307710366569407,
+      "grad_norm": 0.1496568332198432,
+      "learning_rate": 2e-05,
+      "loss": 5.3345,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5308381124861656,
+      "grad_norm": 0.14589791965052368,
+      "learning_rate": 2e-05,
+      "loss": 5.3382,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5309051883153906,
+      "grad_norm": 0.15249542826758267,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5309722641446155,
+      "grad_norm": 0.14900659779162892,
+      "learning_rate": 2e-05,
+      "loss": 5.4844,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5310393399738405,
+      "grad_norm": 0.14548110990228677,
+      "learning_rate": 2e-05,
+      "loss": 5.2137,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5311064158030654,
+      "grad_norm": 0.14990601093818076,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5311734916322903,
+      "grad_norm": 0.15077531304582845,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5312405674615153,
+      "grad_norm": 0.15174763827801516,
+      "learning_rate": 2e-05,
+      "loss": 5.5039,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5313076432907402,
+      "grad_norm": 0.14949591091211567,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5313747191199651,
+      "grad_norm": 0.1497589192643047,
+      "learning_rate": 2e-05,
+      "loss": 5.4879,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5314417949491901,
+      "grad_norm": 0.14945421183043242,
+      "learning_rate": 2e-05,
+      "loss": 5.5259,
+      "step": 7923
+    },
+    {
+      "epoch": 0.531508870778415,
+      "grad_norm": 0.14569434092195951,
+      "learning_rate": 2e-05,
+      "loss": 5.5597,
+      "step": 7924
+    },
+    {
+      "epoch": 0.53157594660764,
+      "grad_norm": 0.14633826309716944,
+      "learning_rate": 2e-05,
+      "loss": 5.3578,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5316430224368649,
+      "grad_norm": 0.14333339774900059,
+      "learning_rate": 2e-05,
+      "loss": 5.5712,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5317100982660898,
+      "grad_norm": 0.14677909236818223,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5317771740953148,
+      "grad_norm": 0.15542752014304007,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5318442499245397,
+      "grad_norm": 0.15321877455537508,
+      "learning_rate": 2e-05,
+      "loss": 5.3525,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5319113257537647,
+      "grad_norm": 0.1477680450953325,
+      "learning_rate": 2e-05,
+      "loss": 5.4177,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5319784015829896,
+      "grad_norm": 0.15434788277623696,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5320454774122145,
+      "grad_norm": 0.14692984470464854,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5321125532414395,
+      "grad_norm": 0.1455853991491499,
+      "learning_rate": 2e-05,
+      "loss": 5.4851,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5321796290706644,
+      "grad_norm": 0.14982438689837918,
+      "learning_rate": 2e-05,
+      "loss": 5.3759,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5322467048998893,
+      "grad_norm": 0.15452908302703156,
+      "learning_rate": 2e-05,
+      "loss": 5.5032,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5323137807291143,
+      "grad_norm": 0.152943379022605,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5323808565583392,
+      "grad_norm": 0.15695355288606078,
+      "learning_rate": 2e-05,
+      "loss": 5.3668,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5324479323875642,
+      "grad_norm": 0.14812575917556958,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5325150082167891,
+      "grad_norm": 0.14460321811420837,
+      "learning_rate": 2e-05,
+      "loss": 5.3153,
+      "step": 7939
+    },
+    {
+      "epoch": 0.532582084046014,
+      "grad_norm": 0.1449861846669442,
+      "learning_rate": 2e-05,
+      "loss": 5.4862,
+      "step": 7940
+    },
+    {
+      "epoch": 0.532649159875239,
+      "grad_norm": 0.1548270145744908,
+      "learning_rate": 2e-05,
+      "loss": 5.4111,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5327162357044639,
+      "grad_norm": 0.15336700886252208,
+      "learning_rate": 2e-05,
+      "loss": 5.4583,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5327833115336889,
+      "grad_norm": 0.14994074587122827,
+      "learning_rate": 2e-05,
+      "loss": 5.5246,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5328503873629138,
+      "grad_norm": 0.15556537339921953,
+      "learning_rate": 2e-05,
+      "loss": 5.3805,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5329174631921387,
+      "grad_norm": 0.15206201415128698,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5329845390213637,
+      "grad_norm": 0.15308613157511577,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 7946
+    },
+    {
+      "epoch": 0.5330516148505886,
+      "grad_norm": 0.1571447933618705,
+      "learning_rate": 2e-05,
+      "loss": 5.4948,
+      "step": 7947
+    },
+    {
+      "epoch": 0.5331186906798135,
+      "grad_norm": 0.15011956928288958,
+      "learning_rate": 2e-05,
+      "loss": 5.4854,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5331857665090385,
+      "grad_norm": 0.1511146839762173,
+      "learning_rate": 2e-05,
+      "loss": 5.3504,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5332528423382634,
+      "grad_norm": 0.1545162558795858,
+      "learning_rate": 2e-05,
+      "loss": 5.4493,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5333199181674884,
+      "grad_norm": 0.14829154322495153,
+      "learning_rate": 2e-05,
+      "loss": 5.3984,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5333869939967133,
+      "grad_norm": 0.15413931895175936,
+      "learning_rate": 2e-05,
+      "loss": 5.3753,
+      "step": 7952
+    },
+    {
+      "epoch": 0.5334540698259382,
+      "grad_norm": 0.1568340631863365,
+      "learning_rate": 2e-05,
+      "loss": 5.5781,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5335211456551632,
+      "grad_norm": 0.14581521179803322,
+      "learning_rate": 2e-05,
+      "loss": 5.3851,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5335882214843881,
+      "grad_norm": 0.15187226082732014,
+      "learning_rate": 2e-05,
+      "loss": 5.2566,
+      "step": 7955
+    },
+    {
+      "epoch": 0.533655297313613,
+      "grad_norm": 0.18073211922856136,
+      "learning_rate": 2e-05,
+      "loss": 5.4412,
+      "step": 7956
+    },
+    {
+      "epoch": 0.533722373142838,
+      "grad_norm": 0.1528307214483082,
+      "learning_rate": 2e-05,
+      "loss": 5.3595,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5337894489720629,
+      "grad_norm": 0.1577671427810294,
+      "learning_rate": 2e-05,
+      "loss": 5.481,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5338565248012879,
+      "grad_norm": 0.15924944435553243,
+      "learning_rate": 2e-05,
+      "loss": 5.5028,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5339236006305128,
+      "grad_norm": 0.16123419308038286,
+      "learning_rate": 2e-05,
+      "loss": 5.4124,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5339906764597377,
+      "grad_norm": 0.16212535270988446,
+      "learning_rate": 2e-05,
+      "loss": 5.5024,
+      "step": 7961
+    },
+    {
+      "epoch": 0.5340577522889627,
+      "grad_norm": 0.15736833132201752,
+      "learning_rate": 2e-05,
+      "loss": 5.5497,
+      "step": 7962
+    },
+    {
+      "epoch": 0.5341248281181876,
+      "grad_norm": 0.1614228533723106,
+      "learning_rate": 2e-05,
+      "loss": 5.335,
+      "step": 7963
+    },
+    {
+      "epoch": 0.5341919039474126,
+      "grad_norm": 0.14980678229321498,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 7964
+    },
+    {
+      "epoch": 0.5342589797766375,
+      "grad_norm": 0.15116800141862827,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5343260556058624,
+      "grad_norm": 0.16126262730468194,
+      "learning_rate": 2e-05,
+      "loss": 5.3035,
+      "step": 7966
+    },
+    {
+      "epoch": 0.5343931314350874,
+      "grad_norm": 0.15889674802627435,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5344602072643123,
+      "grad_norm": 0.1565712383787536,
+      "learning_rate": 2e-05,
+      "loss": 5.552,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5345272830935373,
+      "grad_norm": 0.1642500182203417,
+      "learning_rate": 2e-05,
+      "loss": 5.3089,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5345943589227622,
+      "grad_norm": 0.15493214848103143,
+      "learning_rate": 2e-05,
+      "loss": 5.3969,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5346614347519871,
+      "grad_norm": 0.15692139701810226,
+      "learning_rate": 2e-05,
+      "loss": 5.4865,
+      "step": 7971
+    },
+    {
+      "epoch": 0.5347285105812121,
+      "grad_norm": 0.15563480188553255,
+      "learning_rate": 2e-05,
+      "loss": 5.3802,
+      "step": 7972
+    },
+    {
+      "epoch": 0.534795586410437,
+      "grad_norm": 0.157492833845838,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 7973
+    },
+    {
+      "epoch": 0.534862662239662,
+      "grad_norm": 0.1525544679990963,
+      "learning_rate": 2e-05,
+      "loss": 5.3433,
+      "step": 7974
+    },
+    {
+      "epoch": 0.5349297380688869,
+      "grad_norm": 0.15197576690970116,
+      "learning_rate": 2e-05,
+      "loss": 5.4264,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5349968138981118,
+      "grad_norm": 0.1517858425916738,
+      "learning_rate": 2e-05,
+      "loss": 5.3028,
+      "step": 7976
+    },
+    {
+      "epoch": 0.5350638897273368,
+      "grad_norm": 0.15468470426097242,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 7977
+    },
+    {
+      "epoch": 0.5351309655565617,
+      "grad_norm": 0.14998016167282283,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 7978
+    },
+    {
+      "epoch": 0.5351980413857866,
+      "grad_norm": 0.148889982391403,
+      "learning_rate": 2e-05,
+      "loss": 5.3721,
+      "step": 7979
+    },
+    {
+      "epoch": 0.5352651172150116,
+      "grad_norm": 0.15481407273321532,
+      "learning_rate": 2e-05,
+      "loss": 5.4826,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5353321930442365,
+      "grad_norm": 0.15926796764556994,
+      "learning_rate": 2e-05,
+      "loss": 5.3966,
+      "step": 7981
+    },
+    {
+      "epoch": 0.5353992688734615,
+      "grad_norm": 0.16217781369087095,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 7982
+    },
+    {
+      "epoch": 0.5354663447026864,
+      "grad_norm": 0.14532243565215963,
+      "learning_rate": 2e-05,
+      "loss": 5.4107,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5355334205319113,
+      "grad_norm": 0.1571901947721094,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5356004963611363,
+      "grad_norm": 0.1603157605383176,
+      "learning_rate": 2e-05,
+      "loss": 5.3533,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5356675721903612,
+      "grad_norm": 0.15387055840831695,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5357346480195861,
+      "grad_norm": 0.1554204669923519,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 7987
+    },
+    {
+      "epoch": 0.5358017238488111,
+      "grad_norm": 0.172204387885398,
+      "learning_rate": 2e-05,
+      "loss": 5.5707,
+      "step": 7988
+    },
+    {
+      "epoch": 0.535868799678036,
+      "grad_norm": 0.1571156938620936,
+      "learning_rate": 2e-05,
+      "loss": 5.3108,
+      "step": 7989
+    },
+    {
+      "epoch": 0.535935875507261,
+      "grad_norm": 0.1544235628398669,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5360029513364859,
+      "grad_norm": 0.15066702917641026,
+      "learning_rate": 2e-05,
+      "loss": 5.3601,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5360700271657108,
+      "grad_norm": 0.15227133868412107,
+      "learning_rate": 2e-05,
+      "loss": 5.5127,
+      "step": 7992
+    },
+    {
+      "epoch": 0.5361371029949358,
+      "grad_norm": 0.15116870951332156,
+      "learning_rate": 2e-05,
+      "loss": 5.4102,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5362041788241607,
+      "grad_norm": 0.15787341664399057,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5362712546533857,
+      "grad_norm": 0.1604584382520857,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5363383304826106,
+      "grad_norm": 0.15536905556543887,
+      "learning_rate": 2e-05,
+      "loss": 5.4345,
+      "step": 7996
+    },
+    {
+      "epoch": 0.5364054063118355,
+      "grad_norm": 0.1650013245091036,
+      "learning_rate": 2e-05,
+      "loss": 5.3215,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5364724821410605,
+      "grad_norm": 0.14889799733081943,
+      "learning_rate": 2e-05,
+      "loss": 5.3881,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5365395579702854,
+      "grad_norm": 0.146719626306035,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5366066337995103,
+      "grad_norm": 0.16311873578180833,
+      "learning_rate": 2e-05,
+      "loss": 5.4489,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5366737096287353,
+      "grad_norm": 0.1592923970825995,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 8001
+    },
+    {
+      "epoch": 0.5367407854579602,
+      "grad_norm": 0.15029018207457187,
+      "learning_rate": 2e-05,
+      "loss": 5.5518,
+      "step": 8002
+    },
+    {
+      "epoch": 0.5368078612871852,
+      "grad_norm": 0.1576857221011176,
+      "learning_rate": 2e-05,
+      "loss": 5.4873,
+      "step": 8003
+    },
+    {
+      "epoch": 0.5368749371164101,
+      "grad_norm": 0.16605944061720984,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 8004
+    },
+    {
+      "epoch": 0.536942012945635,
+      "grad_norm": 0.15218321986193542,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 8005
+    },
+    {
+      "epoch": 0.53700908877486,
+      "grad_norm": 0.14931808586868622,
+      "learning_rate": 2e-05,
+      "loss": 5.4343,
+      "step": 8006
+    },
+    {
+      "epoch": 0.5370761646040849,
+      "grad_norm": 0.15591041795073587,
+      "learning_rate": 2e-05,
+      "loss": 5.2907,
+      "step": 8007
+    },
+    {
+      "epoch": 0.5371432404333099,
+      "grad_norm": 0.15107953660282145,
+      "learning_rate": 2e-05,
+      "loss": 5.3844,
+      "step": 8008
+    },
+    {
+      "epoch": 0.5372103162625348,
+      "grad_norm": 0.15611783947836552,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 8009
+    },
+    {
+      "epoch": 0.5372773920917597,
+      "grad_norm": 0.16509538674029575,
+      "learning_rate": 2e-05,
+      "loss": 5.4243,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5373444679209847,
+      "grad_norm": 0.15167097556229278,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 8011
+    },
+    {
+      "epoch": 0.5374115437502096,
+      "grad_norm": 0.16175926646371033,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 8012
+    },
+    {
+      "epoch": 0.5374786195794345,
+      "grad_norm": 0.15207447669545077,
+      "learning_rate": 2e-05,
+      "loss": 5.5854,
+      "step": 8013
+    },
+    {
+      "epoch": 0.5375456954086595,
+      "grad_norm": 0.1538779074577012,
+      "learning_rate": 2e-05,
+      "loss": 5.3303,
+      "step": 8014
+    },
+    {
+      "epoch": 0.5376127712378844,
+      "grad_norm": 0.1586552295952071,
+      "learning_rate": 2e-05,
+      "loss": 5.3269,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5376798470671094,
+      "grad_norm": 0.14977918655066744,
+      "learning_rate": 2e-05,
+      "loss": 5.351,
+      "step": 8016
+    },
+    {
+      "epoch": 0.5377469228963343,
+      "grad_norm": 0.1629294176877139,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 8017
+    },
+    {
+      "epoch": 0.5378139987255592,
+      "grad_norm": 0.15521655584071947,
+      "learning_rate": 2e-05,
+      "loss": 5.4783,
+      "step": 8018
+    },
+    {
+      "epoch": 0.5378810745547842,
+      "grad_norm": 0.1547733858381867,
+      "learning_rate": 2e-05,
+      "loss": 5.5949,
+      "step": 8019
+    },
+    {
+      "epoch": 0.5379481503840091,
+      "grad_norm": 0.15571969094846416,
+      "learning_rate": 2e-05,
+      "loss": 5.3261,
+      "step": 8020
+    },
+    {
+      "epoch": 0.538015226213234,
+      "grad_norm": 0.1590526815823308,
+      "learning_rate": 2e-05,
+      "loss": 5.4134,
+      "step": 8021
+    },
+    {
+      "epoch": 0.538082302042459,
+      "grad_norm": 0.1544878333533228,
+      "learning_rate": 2e-05,
+      "loss": 5.3658,
+      "step": 8022
+    },
+    {
+      "epoch": 0.5381493778716839,
+      "grad_norm": 0.16313952939117046,
+      "learning_rate": 2e-05,
+      "loss": 5.5156,
+      "step": 8023
+    },
+    {
+      "epoch": 0.5382164537009089,
+      "grad_norm": 0.15649626289625718,
+      "learning_rate": 2e-05,
+      "loss": 5.6011,
+      "step": 8024
+    },
+    {
+      "epoch": 0.5382835295301338,
+      "grad_norm": 0.15687723421261496,
+      "learning_rate": 2e-05,
+      "loss": 5.404,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5383506053593587,
+      "grad_norm": 0.15871873539009274,
+      "learning_rate": 2e-05,
+      "loss": 5.3097,
+      "step": 8026
+    },
+    {
+      "epoch": 0.5384176811885837,
+      "grad_norm": 0.1461845716512738,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 8027
+    },
+    {
+      "epoch": 0.5384847570178086,
+      "grad_norm": 0.15853344768988964,
+      "learning_rate": 2e-05,
+      "loss": 5.3299,
+      "step": 8028
+    },
+    {
+      "epoch": 0.5385518328470336,
+      "grad_norm": 0.15213628887606623,
+      "learning_rate": 2e-05,
+      "loss": 5.4738,
+      "step": 8029
+    },
+    {
+      "epoch": 0.5386189086762585,
+      "grad_norm": 0.15221433134102988,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5386859845054834,
+      "grad_norm": 0.1573328532853536,
+      "learning_rate": 2e-05,
+      "loss": 5.5523,
+      "step": 8031
+    },
+    {
+      "epoch": 0.5387530603347084,
+      "grad_norm": 0.1520564792589964,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 8032
+    },
+    {
+      "epoch": 0.5388201361639333,
+      "grad_norm": 0.16625355474504297,
+      "learning_rate": 2e-05,
+      "loss": 5.4332,
+      "step": 8033
+    },
+    {
+      "epoch": 0.5388872119931583,
+      "grad_norm": 0.1614241539250218,
+      "learning_rate": 2e-05,
+      "loss": 5.498,
+      "step": 8034
+    },
+    {
+      "epoch": 0.5389542878223832,
+      "grad_norm": 0.15492934220212962,
+      "learning_rate": 2e-05,
+      "loss": 5.4339,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5390213636516081,
+      "grad_norm": 0.1520958852244175,
+      "learning_rate": 2e-05,
+      "loss": 5.3841,
+      "step": 8036
+    },
+    {
+      "epoch": 0.5390884394808331,
+      "grad_norm": 0.1533364660919665,
+      "learning_rate": 2e-05,
+      "loss": 5.3306,
+      "step": 8037
+    },
+    {
+      "epoch": 0.539155515310058,
+      "grad_norm": 0.15815556345552076,
+      "learning_rate": 2e-05,
+      "loss": 5.2948,
+      "step": 8038
+    },
+    {
+      "epoch": 0.539222591139283,
+      "grad_norm": 0.14851044912874253,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 8039
+    },
+    {
+      "epoch": 0.5392896669685079,
+      "grad_norm": 0.1488950584503278,
+      "learning_rate": 2e-05,
+      "loss": 5.3752,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5393567427977328,
+      "grad_norm": 0.1518850086834417,
+      "learning_rate": 2e-05,
+      "loss": 5.4446,
+      "step": 8041
+    },
+    {
+      "epoch": 0.5394238186269578,
+      "grad_norm": 0.14591696973590323,
+      "learning_rate": 2e-05,
+      "loss": 5.5253,
+      "step": 8042
+    },
+    {
+      "epoch": 0.5394908944561827,
+      "grad_norm": 0.15193375463898878,
+      "learning_rate": 2e-05,
+      "loss": 5.2323,
+      "step": 8043
+    },
+    {
+      "epoch": 0.5395579702854076,
+      "grad_norm": 0.14801333283368776,
+      "learning_rate": 2e-05,
+      "loss": 5.3993,
+      "step": 8044
+    },
+    {
+      "epoch": 0.5396250461146326,
+      "grad_norm": 0.1484799852460922,
+      "learning_rate": 2e-05,
+      "loss": 5.3766,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5396921219438575,
+      "grad_norm": 0.16742072210835376,
+      "learning_rate": 2e-05,
+      "loss": 5.5727,
+      "step": 8046
+    },
+    {
+      "epoch": 0.5397591977730825,
+      "grad_norm": 0.14829870714418228,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 8047
+    },
+    {
+      "epoch": 0.5398262736023074,
+      "grad_norm": 0.1451079863596441,
+      "learning_rate": 2e-05,
+      "loss": 5.3851,
+      "step": 8048
+    },
+    {
+      "epoch": 0.5398933494315323,
+      "grad_norm": 0.14888419339404596,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 8049
+    },
+    {
+      "epoch": 0.5399604252607573,
+      "grad_norm": 0.15109512798190733,
+      "learning_rate": 2e-05,
+      "loss": 5.4603,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5400275010899822,
+      "grad_norm": 0.15355653676438805,
+      "learning_rate": 2e-05,
+      "loss": 5.3933,
+      "step": 8051
+    },
+    {
+      "epoch": 0.5400945769192071,
+      "grad_norm": 0.14620387446120525,
+      "learning_rate": 2e-05,
+      "loss": 5.447,
+      "step": 8052
+    },
+    {
+      "epoch": 0.5401616527484321,
+      "grad_norm": 0.14980421536752347,
+      "learning_rate": 2e-05,
+      "loss": 5.393,
+      "step": 8053
+    },
+    {
+      "epoch": 0.540228728577657,
+      "grad_norm": 0.15350446492877712,
+      "learning_rate": 2e-05,
+      "loss": 5.4479,
+      "step": 8054
+    },
+    {
+      "epoch": 0.540295804406882,
+      "grad_norm": 0.14878839586578896,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5403628802361069,
+      "grad_norm": 0.14763412220713984,
+      "learning_rate": 2e-05,
+      "loss": 5.5953,
+      "step": 8056
+    },
+    {
+      "epoch": 0.5404299560653318,
+      "grad_norm": 0.15291176875534915,
+      "learning_rate": 2e-05,
+      "loss": 5.4157,
+      "step": 8057
+    },
+    {
+      "epoch": 0.5404970318945568,
+      "grad_norm": 0.15239373384783234,
+      "learning_rate": 2e-05,
+      "loss": 5.502,
+      "step": 8058
+    },
+    {
+      "epoch": 0.5405641077237817,
+      "grad_norm": 0.15733221740177689,
+      "learning_rate": 2e-05,
+      "loss": 5.3445,
+      "step": 8059
+    },
+    {
+      "epoch": 0.5406311835530067,
+      "grad_norm": 0.15254205561318263,
+      "learning_rate": 2e-05,
+      "loss": 5.3182,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5406982593822316,
+      "grad_norm": 0.1533211593519062,
+      "learning_rate": 2e-05,
+      "loss": 5.3429,
+      "step": 8061
+    },
+    {
+      "epoch": 0.5407653352114565,
+      "grad_norm": 0.14820029849654903,
+      "learning_rate": 2e-05,
+      "loss": 5.2802,
+      "step": 8062
+    },
+    {
+      "epoch": 0.5408324110406815,
+      "grad_norm": 0.15274645840545903,
+      "learning_rate": 2e-05,
+      "loss": 5.4427,
+      "step": 8063
+    },
+    {
+      "epoch": 0.5408994868699064,
+      "grad_norm": 0.15240707965909359,
+      "learning_rate": 2e-05,
+      "loss": 5.4171,
+      "step": 8064
+    },
+    {
+      "epoch": 0.5409665626991313,
+      "grad_norm": 0.14828500100309236,
+      "learning_rate": 2e-05,
+      "loss": 5.4344,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5410336385283563,
+      "grad_norm": 0.1583207034406194,
+      "learning_rate": 2e-05,
+      "loss": 5.3632,
+      "step": 8066
+    },
+    {
+      "epoch": 0.5411007143575812,
+      "grad_norm": 0.15066016369146115,
+      "learning_rate": 2e-05,
+      "loss": 5.3073,
+      "step": 8067
+    },
+    {
+      "epoch": 0.5411677901868062,
+      "grad_norm": 0.15216978106922296,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 8068
+    },
+    {
+      "epoch": 0.5412348660160311,
+      "grad_norm": 0.14689546133809245,
+      "learning_rate": 2e-05,
+      "loss": 5.5097,
+      "step": 8069
+    },
+    {
+      "epoch": 0.541301941845256,
+      "grad_norm": 0.1545220509742145,
+      "learning_rate": 2e-05,
+      "loss": 5.4384,
+      "step": 8070
+    },
+    {
+      "epoch": 0.541369017674481,
+      "grad_norm": 0.15963050556558567,
+      "learning_rate": 2e-05,
+      "loss": 5.3772,
+      "step": 8071
+    },
+    {
+      "epoch": 0.5414360935037059,
+      "grad_norm": 0.14905368035222133,
+      "learning_rate": 2e-05,
+      "loss": 5.2755,
+      "step": 8072
+    },
+    {
+      "epoch": 0.5415031693329309,
+      "grad_norm": 0.15278808188202547,
+      "learning_rate": 2e-05,
+      "loss": 5.3032,
+      "step": 8073
+    },
+    {
+      "epoch": 0.5415702451621558,
+      "grad_norm": 0.1600204067364068,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 8074
+    },
+    {
+      "epoch": 0.5416373209913807,
+      "grad_norm": 0.15468289392471843,
+      "learning_rate": 2e-05,
+      "loss": 5.4028,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5417043968206057,
+      "grad_norm": 0.15024405988841008,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 8076
+    },
+    {
+      "epoch": 0.5417714726498306,
+      "grad_norm": 0.15097516736842412,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 8077
+    },
+    {
+      "epoch": 0.5418385484790555,
+      "grad_norm": 0.16924045640161492,
+      "learning_rate": 2e-05,
+      "loss": 5.4962,
+      "step": 8078
+    },
+    {
+      "epoch": 0.5419056243082805,
+      "grad_norm": 0.16557193942858897,
+      "learning_rate": 2e-05,
+      "loss": 5.3839,
+      "step": 8079
+    },
+    {
+      "epoch": 0.5419727001375054,
+      "grad_norm": 0.15298784424758777,
+      "learning_rate": 2e-05,
+      "loss": 5.3785,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5420397759667304,
+      "grad_norm": 0.15785318577445906,
+      "learning_rate": 2e-05,
+      "loss": 5.3705,
+      "step": 8081
+    },
+    {
+      "epoch": 0.5421068517959553,
+      "grad_norm": 0.14998916198639103,
+      "learning_rate": 2e-05,
+      "loss": 5.477,
+      "step": 8082
+    },
+    {
+      "epoch": 0.5421739276251802,
+      "grad_norm": 0.15568489485702502,
+      "learning_rate": 2e-05,
+      "loss": 5.326,
+      "step": 8083
+    },
+    {
+      "epoch": 0.5422410034544052,
+      "grad_norm": 0.15661898645225955,
+      "learning_rate": 2e-05,
+      "loss": 5.4782,
+      "step": 8084
+    },
+    {
+      "epoch": 0.5423080792836301,
+      "grad_norm": 0.15230057528988178,
+      "learning_rate": 2e-05,
+      "loss": 5.6172,
+      "step": 8085
+    },
+    {
+      "epoch": 0.5423751551128551,
+      "grad_norm": 0.14997101659835607,
+      "learning_rate": 2e-05,
+      "loss": 5.313,
+      "step": 8086
+    },
+    {
+      "epoch": 0.54244223094208,
+      "grad_norm": 0.16562703887980654,
+      "learning_rate": 2e-05,
+      "loss": 5.4864,
+      "step": 8087
+    },
+    {
+      "epoch": 0.5425093067713049,
+      "grad_norm": 0.15059162106226817,
+      "learning_rate": 2e-05,
+      "loss": 5.4423,
+      "step": 8088
+    },
+    {
+      "epoch": 0.5425763826005299,
+      "grad_norm": 0.15668259444938953,
+      "learning_rate": 2e-05,
+      "loss": 5.4288,
+      "step": 8089
+    },
+    {
+      "epoch": 0.5426434584297548,
+      "grad_norm": 0.15118820991837176,
+      "learning_rate": 2e-05,
+      "loss": 5.5856,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5427105342589797,
+      "grad_norm": 0.14994576698168324,
+      "learning_rate": 2e-05,
+      "loss": 5.4157,
+      "step": 8091
+    },
+    {
+      "epoch": 0.5427776100882047,
+      "grad_norm": 0.15495407728836108,
+      "learning_rate": 2e-05,
+      "loss": 5.4173,
+      "step": 8092
+    },
+    {
+      "epoch": 0.5428446859174296,
+      "grad_norm": 0.1510685631626895,
+      "learning_rate": 2e-05,
+      "loss": 5.5098,
+      "step": 8093
+    },
+    {
+      "epoch": 0.5429117617466546,
+      "grad_norm": 0.1657283497097139,
+      "learning_rate": 2e-05,
+      "loss": 5.4083,
+      "step": 8094
+    },
+    {
+      "epoch": 0.5429788375758795,
+      "grad_norm": 0.16906851070298348,
+      "learning_rate": 2e-05,
+      "loss": 5.4614,
+      "step": 8095
+    },
+    {
+      "epoch": 0.5430459134051044,
+      "grad_norm": 0.15378970545017168,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 8096
+    },
+    {
+      "epoch": 0.5431129892343294,
+      "grad_norm": 0.15654581865145553,
+      "learning_rate": 2e-05,
+      "loss": 5.4156,
+      "step": 8097
+    },
+    {
+      "epoch": 0.5431800650635543,
+      "grad_norm": 0.1492311410941927,
+      "learning_rate": 2e-05,
+      "loss": 5.3512,
+      "step": 8098
+    },
+    {
+      "epoch": 0.5432471408927793,
+      "grad_norm": 0.14887528921579526,
+      "learning_rate": 2e-05,
+      "loss": 5.5123,
+      "step": 8099
+    },
+    {
+      "epoch": 0.5433142167220042,
+      "grad_norm": 0.14920987736337352,
+      "learning_rate": 2e-05,
+      "loss": 5.271,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5433812925512291,
+      "grad_norm": 0.15643018155938498,
+      "learning_rate": 2e-05,
+      "loss": 5.2471,
+      "step": 8101
+    },
+    {
+      "epoch": 0.5434483683804541,
+      "grad_norm": 0.15306642677179472,
+      "learning_rate": 2e-05,
+      "loss": 5.4609,
+      "step": 8102
+    },
+    {
+      "epoch": 0.543515444209679,
+      "grad_norm": 0.1483767014915664,
+      "learning_rate": 2e-05,
+      "loss": 5.2796,
+      "step": 8103
+    },
+    {
+      "epoch": 0.543582520038904,
+      "grad_norm": 0.1499940666347463,
+      "learning_rate": 2e-05,
+      "loss": 5.405,
+      "step": 8104
+    },
+    {
+      "epoch": 0.5436495958681289,
+      "grad_norm": 0.15290890286557696,
+      "learning_rate": 2e-05,
+      "loss": 5.3103,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5437166716973538,
+      "grad_norm": 0.16063755354122247,
+      "learning_rate": 2e-05,
+      "loss": 5.384,
+      "step": 8106
+    },
+    {
+      "epoch": 0.5437837475265788,
+      "grad_norm": 0.14938905549750517,
+      "learning_rate": 2e-05,
+      "loss": 5.4839,
+      "step": 8107
+    },
+    {
+      "epoch": 0.5438508233558037,
+      "grad_norm": 0.15442063589595864,
+      "learning_rate": 2e-05,
+      "loss": 5.4766,
+      "step": 8108
+    },
+    {
+      "epoch": 0.5439178991850286,
+      "grad_norm": 0.14560229957882181,
+      "learning_rate": 2e-05,
+      "loss": 5.5143,
+      "step": 8109
+    },
+    {
+      "epoch": 0.5439849750142536,
+      "grad_norm": 0.1604898632889769,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5440520508434785,
+      "grad_norm": 0.15566707066141264,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 8111
+    },
+    {
+      "epoch": 0.5441191266727035,
+      "grad_norm": 0.1535656729824476,
+      "learning_rate": 2e-05,
+      "loss": 5.6032,
+      "step": 8112
+    },
+    {
+      "epoch": 0.5441862025019284,
+      "grad_norm": 0.17266386734283076,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 8113
+    },
+    {
+      "epoch": 0.5442532783311533,
+      "grad_norm": 0.146029868382708,
+      "learning_rate": 2e-05,
+      "loss": 5.327,
+      "step": 8114
+    },
+    {
+      "epoch": 0.5443203541603783,
+      "grad_norm": 0.15206809889530903,
+      "learning_rate": 2e-05,
+      "loss": 5.3191,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5443874299896032,
+      "grad_norm": 0.15552164728474935,
+      "learning_rate": 2e-05,
+      "loss": 5.3918,
+      "step": 8116
+    },
+    {
+      "epoch": 0.5444545058188281,
+      "grad_norm": 0.16470811074839356,
+      "learning_rate": 2e-05,
+      "loss": 5.3635,
+      "step": 8117
+    },
+    {
+      "epoch": 0.5445215816480531,
+      "grad_norm": 0.14775884432873293,
+      "learning_rate": 2e-05,
+      "loss": 5.5761,
+      "step": 8118
+    },
+    {
+      "epoch": 0.544588657477278,
+      "grad_norm": 0.1509437940762139,
+      "learning_rate": 2e-05,
+      "loss": 5.4374,
+      "step": 8119
+    },
+    {
+      "epoch": 0.544655733306503,
+      "grad_norm": 0.15460998561761677,
+      "learning_rate": 2e-05,
+      "loss": 5.3768,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5447228091357279,
+      "grad_norm": 0.15710065782948024,
+      "learning_rate": 2e-05,
+      "loss": 5.5801,
+      "step": 8121
+    },
+    {
+      "epoch": 0.5447898849649528,
+      "grad_norm": 0.15266494961957405,
+      "learning_rate": 2e-05,
+      "loss": 5.5092,
+      "step": 8122
+    },
+    {
+      "epoch": 0.5448569607941778,
+      "grad_norm": 0.14844959275823674,
+      "learning_rate": 2e-05,
+      "loss": 5.3352,
+      "step": 8123
+    },
+    {
+      "epoch": 0.5449240366234027,
+      "grad_norm": 0.15587385065065248,
+      "learning_rate": 2e-05,
+      "loss": 5.461,
+      "step": 8124
+    },
+    {
+      "epoch": 0.5449911124526277,
+      "grad_norm": 0.1506947721820962,
+      "learning_rate": 2e-05,
+      "loss": 5.4432,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5450581882818526,
+      "grad_norm": 0.15031844817494744,
+      "learning_rate": 2e-05,
+      "loss": 5.414,
+      "step": 8126
+    },
+    {
+      "epoch": 0.5451252641110775,
+      "grad_norm": 0.15349317070741456,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 8127
+    },
+    {
+      "epoch": 0.5451923399403025,
+      "grad_norm": 0.15114717028831717,
+      "learning_rate": 2e-05,
+      "loss": 5.3763,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5452594157695274,
+      "grad_norm": 0.15179821769452584,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 8129
+    },
+    {
+      "epoch": 0.5453264915987523,
+      "grad_norm": 0.1562933660114895,
+      "learning_rate": 2e-05,
+      "loss": 5.2784,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5453935674279773,
+      "grad_norm": 0.15813847465953904,
+      "learning_rate": 2e-05,
+      "loss": 5.6686,
+      "step": 8131
+    },
+    {
+      "epoch": 0.5454606432572022,
+      "grad_norm": 0.15312471722271292,
+      "learning_rate": 2e-05,
+      "loss": 5.3608,
+      "step": 8132
+    },
+    {
+      "epoch": 0.5455277190864272,
+      "grad_norm": 0.1661455270457996,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 8133
+    },
+    {
+      "epoch": 0.5455947949156521,
+      "grad_norm": 0.14587072945003215,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 8134
+    },
+    {
+      "epoch": 0.545661870744877,
+      "grad_norm": 0.15418472759202576,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 8135
+    },
+    {
+      "epoch": 0.545728946574102,
+      "grad_norm": 0.16970131618348905,
+      "learning_rate": 2e-05,
+      "loss": 5.3539,
+      "step": 8136
+    },
+    {
+      "epoch": 0.5457960224033269,
+      "grad_norm": 0.1540771483055607,
+      "learning_rate": 2e-05,
+      "loss": 5.3474,
+      "step": 8137
+    },
+    {
+      "epoch": 0.5458630982325519,
+      "grad_norm": 0.15523817101300477,
+      "learning_rate": 2e-05,
+      "loss": 5.3407,
+      "step": 8138
+    },
+    {
+      "epoch": 0.5459301740617768,
+      "grad_norm": 0.15390005360158374,
+      "learning_rate": 2e-05,
+      "loss": 5.386,
+      "step": 8139
+    },
+    {
+      "epoch": 0.5459972498910017,
+      "grad_norm": 0.16012925256326177,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5460643257202267,
+      "grad_norm": 0.1456365592647378,
+      "learning_rate": 2e-05,
+      "loss": 5.3015,
+      "step": 8141
+    },
+    {
+      "epoch": 0.5461314015494516,
+      "grad_norm": 0.1590706853943673,
+      "learning_rate": 2e-05,
+      "loss": 5.3675,
+      "step": 8142
+    },
+    {
+      "epoch": 0.5461984773786766,
+      "grad_norm": 0.1464833346173416,
+      "learning_rate": 2e-05,
+      "loss": 5.3842,
+      "step": 8143
+    },
+    {
+      "epoch": 0.5462655532079015,
+      "grad_norm": 0.14794352178348644,
+      "learning_rate": 2e-05,
+      "loss": 5.4027,
+      "step": 8144
+    },
+    {
+      "epoch": 0.5463326290371264,
+      "grad_norm": 0.15527105483825107,
+      "learning_rate": 2e-05,
+      "loss": 5.3568,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5463997048663514,
+      "grad_norm": 0.16006541181685022,
+      "learning_rate": 2e-05,
+      "loss": 5.484,
+      "step": 8146
+    },
+    {
+      "epoch": 0.5464667806955763,
+      "grad_norm": 0.14863858390997708,
+      "learning_rate": 2e-05,
+      "loss": 5.4718,
+      "step": 8147
+    },
+    {
+      "epoch": 0.5465338565248012,
+      "grad_norm": 0.14464507032823687,
+      "learning_rate": 2e-05,
+      "loss": 5.5182,
+      "step": 8148
+    },
+    {
+      "epoch": 0.5466009323540262,
+      "grad_norm": 0.1549506323625637,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 8149
+    },
+    {
+      "epoch": 0.5466680081832511,
+      "grad_norm": 0.1500227293738333,
+      "learning_rate": 2e-05,
+      "loss": 5.4301,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5467350840124761,
+      "grad_norm": 0.1426058472266355,
+      "learning_rate": 2e-05,
+      "loss": 5.4838,
+      "step": 8151
+    },
+    {
+      "epoch": 0.546802159841701,
+      "grad_norm": 0.15250598125754558,
+      "learning_rate": 2e-05,
+      "loss": 5.4623,
+      "step": 8152
+    },
+    {
+      "epoch": 0.5468692356709259,
+      "grad_norm": 0.1517587565530394,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 8153
+    },
+    {
+      "epoch": 0.5469363115001509,
+      "grad_norm": 0.14926574354981506,
+      "learning_rate": 2e-05,
+      "loss": 5.2828,
+      "step": 8154
+    },
+    {
+      "epoch": 0.5470033873293758,
+      "grad_norm": 0.14488102400689595,
+      "learning_rate": 2e-05,
+      "loss": 5.4292,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5470704631586008,
+      "grad_norm": 0.1602347930540904,
+      "learning_rate": 2e-05,
+      "loss": 5.3992,
+      "step": 8156
+    },
+    {
+      "epoch": 0.5471375389878257,
+      "grad_norm": 0.15557251567187577,
+      "learning_rate": 2e-05,
+      "loss": 5.3454,
+      "step": 8157
+    },
+    {
+      "epoch": 0.5472046148170506,
+      "grad_norm": 0.1497712054390959,
+      "learning_rate": 2e-05,
+      "loss": 5.3676,
+      "step": 8158
+    },
+    {
+      "epoch": 0.5472716906462756,
+      "grad_norm": 0.1595922797846491,
+      "learning_rate": 2e-05,
+      "loss": 5.3894,
+      "step": 8159
+    },
+    {
+      "epoch": 0.5473387664755005,
+      "grad_norm": 0.1549989749586518,
+      "learning_rate": 2e-05,
+      "loss": 5.4743,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5474058423047254,
+      "grad_norm": 0.1486848074985591,
+      "learning_rate": 2e-05,
+      "loss": 5.3591,
+      "step": 8161
+    },
+    {
+      "epoch": 0.5474729181339504,
+      "grad_norm": 0.16112149452106658,
+      "learning_rate": 2e-05,
+      "loss": 5.3831,
+      "step": 8162
+    },
+    {
+      "epoch": 0.5475399939631753,
+      "grad_norm": 0.15327707386091471,
+      "learning_rate": 2e-05,
+      "loss": 5.4831,
+      "step": 8163
+    },
+    {
+      "epoch": 0.5476070697924003,
+      "grad_norm": 0.15052748103176328,
+      "learning_rate": 2e-05,
+      "loss": 5.3488,
+      "step": 8164
+    },
+    {
+      "epoch": 0.5476741456216252,
+      "grad_norm": 0.15703251706676932,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5477412214508501,
+      "grad_norm": 0.1650491344354078,
+      "learning_rate": 2e-05,
+      "loss": 5.3235,
+      "step": 8166
+    },
+    {
+      "epoch": 0.5478082972800751,
+      "grad_norm": 0.14484872291207357,
+      "learning_rate": 2e-05,
+      "loss": 5.4289,
+      "step": 8167
+    },
+    {
+      "epoch": 0.5478753731093,
+      "grad_norm": 0.1481687325235352,
+      "learning_rate": 2e-05,
+      "loss": 5.3546,
+      "step": 8168
+    },
+    {
+      "epoch": 0.547942448938525,
+      "grad_norm": 0.15610630947208215,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 8169
+    },
+    {
+      "epoch": 0.5480095247677499,
+      "grad_norm": 0.14815872625750243,
+      "learning_rate": 2e-05,
+      "loss": 5.3481,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5480766005969748,
+      "grad_norm": 0.150955230791515,
+      "learning_rate": 2e-05,
+      "loss": 5.2898,
+      "step": 8171
+    },
+    {
+      "epoch": 0.5481436764261998,
+      "grad_norm": 0.14895838790967147,
+      "learning_rate": 2e-05,
+      "loss": 5.4531,
+      "step": 8172
+    },
+    {
+      "epoch": 0.5482107522554247,
+      "grad_norm": 0.16348223757242036,
+      "learning_rate": 2e-05,
+      "loss": 5.3862,
+      "step": 8173
+    },
+    {
+      "epoch": 0.5482778280846496,
+      "grad_norm": 0.14337758223595606,
+      "learning_rate": 2e-05,
+      "loss": 5.3856,
+      "step": 8174
+    },
+    {
+      "epoch": 0.5483449039138746,
+      "grad_norm": 0.1478521685484305,
+      "learning_rate": 2e-05,
+      "loss": 5.3814,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5484119797430995,
+      "grad_norm": 0.1549844324485233,
+      "learning_rate": 2e-05,
+      "loss": 5.3314,
+      "step": 8176
+    },
+    {
+      "epoch": 0.5484790555723245,
+      "grad_norm": 0.14953250814702862,
+      "learning_rate": 2e-05,
+      "loss": 5.5114,
+      "step": 8177
+    },
+    {
+      "epoch": 0.5485461314015494,
+      "grad_norm": 0.15483221294368704,
+      "learning_rate": 2e-05,
+      "loss": 5.3505,
+      "step": 8178
+    },
+    {
+      "epoch": 0.5486132072307743,
+      "grad_norm": 0.1539648789606409,
+      "learning_rate": 2e-05,
+      "loss": 5.5057,
+      "step": 8179
+    },
+    {
+      "epoch": 0.5486802830599993,
+      "grad_norm": 0.14910185771064308,
+      "learning_rate": 2e-05,
+      "loss": 5.4815,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5487473588892242,
+      "grad_norm": 0.14982734098255227,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 8181
+    },
+    {
+      "epoch": 0.5488144347184493,
+      "grad_norm": 0.17184903898059561,
+      "learning_rate": 2e-05,
+      "loss": 5.4937,
+      "step": 8182
+    },
+    {
+      "epoch": 0.5488815105476742,
+      "grad_norm": 0.1463617048644366,
+      "learning_rate": 2e-05,
+      "loss": 5.4791,
+      "step": 8183
+    },
+    {
+      "epoch": 0.5489485863768991,
+      "grad_norm": 0.15369862481624572,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 8184
+    },
+    {
+      "epoch": 0.5490156622061241,
+      "grad_norm": 0.16368334421590436,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 8185
+    },
+    {
+      "epoch": 0.549082738035349,
+      "grad_norm": 0.14788503373488934,
+      "learning_rate": 2e-05,
+      "loss": 5.555,
+      "step": 8186
+    },
+    {
+      "epoch": 0.549149813864574,
+      "grad_norm": 0.1525385302816832,
+      "learning_rate": 2e-05,
+      "loss": 5.4089,
+      "step": 8187
+    },
+    {
+      "epoch": 0.5492168896937989,
+      "grad_norm": 0.15208865182691986,
+      "learning_rate": 2e-05,
+      "loss": 5.3385,
+      "step": 8188
+    },
+    {
+      "epoch": 0.5492839655230238,
+      "grad_norm": 0.15180239317362337,
+      "learning_rate": 2e-05,
+      "loss": 5.4101,
+      "step": 8189
+    },
+    {
+      "epoch": 0.5493510413522488,
+      "grad_norm": 0.1484478447517482,
+      "learning_rate": 2e-05,
+      "loss": 5.4341,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5494181171814737,
+      "grad_norm": 0.14680706166323843,
+      "learning_rate": 2e-05,
+      "loss": 5.5426,
+      "step": 8191
+    },
+    {
+      "epoch": 0.5494851930106986,
+      "grad_norm": 0.1490990136476689,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 8192
+    },
+    {
+      "epoch": 0.5495522688399236,
+      "grad_norm": 0.1579471987845354,
+      "learning_rate": 2e-05,
+      "loss": 5.5765,
+      "step": 8193
+    },
+    {
+      "epoch": 0.5496193446691485,
+      "grad_norm": 0.1528914595796624,
+      "learning_rate": 2e-05,
+      "loss": 5.374,
+      "step": 8194
+    },
+    {
+      "epoch": 0.5496864204983735,
+      "grad_norm": 0.14540377351459188,
+      "learning_rate": 2e-05,
+      "loss": 5.4293,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5497534963275984,
+      "grad_norm": 0.14183129013897722,
+      "learning_rate": 2e-05,
+      "loss": 5.3963,
+      "step": 8196
+    },
+    {
+      "epoch": 0.5498205721568233,
+      "grad_norm": 0.15339978111377944,
+      "learning_rate": 2e-05,
+      "loss": 5.2794,
+      "step": 8197
+    },
+    {
+      "epoch": 0.5498876479860483,
+      "grad_norm": 0.15254842969028898,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 8198
+    },
+    {
+      "epoch": 0.5499547238152732,
+      "grad_norm": 0.1483506084635696,
+      "learning_rate": 2e-05,
+      "loss": 5.4594,
+      "step": 8199
+    },
+    {
+      "epoch": 0.5500217996444982,
+      "grad_norm": 0.14946801898456127,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5500888754737231,
+      "grad_norm": 0.14675115486100937,
+      "learning_rate": 2e-05,
+      "loss": 5.4044,
+      "step": 8201
+    },
+    {
+      "epoch": 0.550155951302948,
+      "grad_norm": 0.1489944640255545,
+      "learning_rate": 2e-05,
+      "loss": 5.4236,
+      "step": 8202
+    },
+    {
+      "epoch": 0.550223027132173,
+      "grad_norm": 0.15119799988277488,
+      "learning_rate": 2e-05,
+      "loss": 5.4785,
+      "step": 8203
+    },
+    {
+      "epoch": 0.5502901029613979,
+      "grad_norm": 0.14945116237678122,
+      "learning_rate": 2e-05,
+      "loss": 5.5602,
+      "step": 8204
+    },
+    {
+      "epoch": 0.5503571787906228,
+      "grad_norm": 0.1521983044813861,
+      "learning_rate": 2e-05,
+      "loss": 5.4416,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5504242546198478,
+      "grad_norm": 0.14728473282127102,
+      "learning_rate": 2e-05,
+      "loss": 5.4194,
+      "step": 8206
+    },
+    {
+      "epoch": 0.5504913304490727,
+      "grad_norm": 0.14928465291333395,
+      "learning_rate": 2e-05,
+      "loss": 5.3189,
+      "step": 8207
+    },
+    {
+      "epoch": 0.5505584062782977,
+      "grad_norm": 0.15158721241806966,
+      "learning_rate": 2e-05,
+      "loss": 5.3966,
+      "step": 8208
+    },
+    {
+      "epoch": 0.5506254821075226,
+      "grad_norm": 0.15126568607216742,
+      "learning_rate": 2e-05,
+      "loss": 5.5323,
+      "step": 8209
+    },
+    {
+      "epoch": 0.5506925579367475,
+      "grad_norm": 0.14691165683649607,
+      "learning_rate": 2e-05,
+      "loss": 5.4141,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5507596337659725,
+      "grad_norm": 0.15129359742048487,
+      "learning_rate": 2e-05,
+      "loss": 5.3357,
+      "step": 8211
+    },
+    {
+      "epoch": 0.5508267095951974,
+      "grad_norm": 0.14617677393442768,
+      "learning_rate": 2e-05,
+      "loss": 5.4249,
+      "step": 8212
+    },
+    {
+      "epoch": 0.5508937854244224,
+      "grad_norm": 0.14284685255561347,
+      "learning_rate": 2e-05,
+      "loss": 5.3304,
+      "step": 8213
+    },
+    {
+      "epoch": 0.5509608612536473,
+      "grad_norm": 0.14941189274652064,
+      "learning_rate": 2e-05,
+      "loss": 5.4578,
+      "step": 8214
+    },
+    {
+      "epoch": 0.5510279370828722,
+      "grad_norm": 0.1508830961339486,
+      "learning_rate": 2e-05,
+      "loss": 5.4534,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5510950129120972,
+      "grad_norm": 0.14579965931340863,
+      "learning_rate": 2e-05,
+      "loss": 5.349,
+      "step": 8216
+    },
+    {
+      "epoch": 0.5511620887413221,
+      "grad_norm": 0.14300751379554089,
+      "learning_rate": 2e-05,
+      "loss": 5.3539,
+      "step": 8217
+    },
+    {
+      "epoch": 0.551229164570547,
+      "grad_norm": 0.15067144250405778,
+      "learning_rate": 2e-05,
+      "loss": 5.2961,
+      "step": 8218
+    },
+    {
+      "epoch": 0.551296240399772,
+      "grad_norm": 0.15524502208483504,
+      "learning_rate": 2e-05,
+      "loss": 5.4327,
+      "step": 8219
+    },
+    {
+      "epoch": 0.5513633162289969,
+      "grad_norm": 0.15077011770767532,
+      "learning_rate": 2e-05,
+      "loss": 5.4114,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5514303920582219,
+      "grad_norm": 0.14478195315584663,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 8221
+    },
+    {
+      "epoch": 0.5514974678874468,
+      "grad_norm": 0.1461726399062049,
+      "learning_rate": 2e-05,
+      "loss": 5.3847,
+      "step": 8222
+    },
+    {
+      "epoch": 0.5515645437166717,
+      "grad_norm": 0.15221291978975018,
+      "learning_rate": 2e-05,
+      "loss": 5.4812,
+      "step": 8223
+    },
+    {
+      "epoch": 0.5516316195458967,
+      "grad_norm": 0.15278929647322048,
+      "learning_rate": 2e-05,
+      "loss": 5.4735,
+      "step": 8224
+    },
+    {
+      "epoch": 0.5516986953751216,
+      "grad_norm": 0.14997534605997337,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5517657712043466,
+      "grad_norm": 0.1579703071644443,
+      "learning_rate": 2e-05,
+      "loss": 5.2951,
+      "step": 8226
+    },
+    {
+      "epoch": 0.5518328470335715,
+      "grad_norm": 0.1495339522761811,
+      "learning_rate": 2e-05,
+      "loss": 5.2821,
+      "step": 8227
+    },
+    {
+      "epoch": 0.5518999228627964,
+      "grad_norm": 0.1596901287283497,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 8228
+    },
+    {
+      "epoch": 0.5519669986920214,
+      "grad_norm": 0.15699932299888067,
+      "learning_rate": 2e-05,
+      "loss": 5.37,
+      "step": 8229
+    },
+    {
+      "epoch": 0.5520340745212463,
+      "grad_norm": 0.14965354798167987,
+      "learning_rate": 2e-05,
+      "loss": 5.4503,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5521011503504712,
+      "grad_norm": 0.15275309768006534,
+      "learning_rate": 2e-05,
+      "loss": 5.5034,
+      "step": 8231
+    },
+    {
+      "epoch": 0.5521682261796962,
+      "grad_norm": 0.15575828094070776,
+      "learning_rate": 2e-05,
+      "loss": 5.3314,
+      "step": 8232
+    },
+    {
+      "epoch": 0.5522353020089211,
+      "grad_norm": 0.14773924391673862,
+      "learning_rate": 2e-05,
+      "loss": 5.3273,
+      "step": 8233
+    },
+    {
+      "epoch": 0.5523023778381461,
+      "grad_norm": 0.1513193969739345,
+      "learning_rate": 2e-05,
+      "loss": 5.3217,
+      "step": 8234
+    },
+    {
+      "epoch": 0.552369453667371,
+      "grad_norm": 0.14943311126573827,
+      "learning_rate": 2e-05,
+      "loss": 5.341,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5524365294965959,
+      "grad_norm": 0.14541158227237122,
+      "learning_rate": 2e-05,
+      "loss": 5.4136,
+      "step": 8236
+    },
+    {
+      "epoch": 0.5525036053258209,
+      "grad_norm": 0.1464255414523395,
+      "learning_rate": 2e-05,
+      "loss": 5.4113,
+      "step": 8237
+    },
+    {
+      "epoch": 0.5525706811550458,
+      "grad_norm": 0.1453793890002281,
+      "learning_rate": 2e-05,
+      "loss": 5.4202,
+      "step": 8238
+    },
+    {
+      "epoch": 0.5526377569842708,
+      "grad_norm": 0.1513486855254119,
+      "learning_rate": 2e-05,
+      "loss": 5.4883,
+      "step": 8239
+    },
+    {
+      "epoch": 0.5527048328134957,
+      "grad_norm": 0.1454040955413562,
+      "learning_rate": 2e-05,
+      "loss": 5.5173,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5527719086427206,
+      "grad_norm": 0.147686139395543,
+      "learning_rate": 2e-05,
+      "loss": 5.4986,
+      "step": 8241
+    },
+    {
+      "epoch": 0.5528389844719456,
+      "grad_norm": 0.1484648534839245,
+      "learning_rate": 2e-05,
+      "loss": 5.3067,
+      "step": 8242
+    },
+    {
+      "epoch": 0.5529060603011705,
+      "grad_norm": 0.15236231785040016,
+      "learning_rate": 2e-05,
+      "loss": 5.4138,
+      "step": 8243
+    },
+    {
+      "epoch": 0.5529731361303954,
+      "grad_norm": 0.1464202313646898,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 8244
+    },
+    {
+      "epoch": 0.5530402119596204,
+      "grad_norm": 0.14753626401732203,
+      "learning_rate": 2e-05,
+      "loss": 5.4309,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5531072877888453,
+      "grad_norm": 0.15829020618244374,
+      "learning_rate": 2e-05,
+      "loss": 5.3262,
+      "step": 8246
+    },
+    {
+      "epoch": 0.5531743636180703,
+      "grad_norm": 0.149583975330147,
+      "learning_rate": 2e-05,
+      "loss": 5.3222,
+      "step": 8247
+    },
+    {
+      "epoch": 0.5532414394472952,
+      "grad_norm": 0.15430850607343888,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 8248
+    },
+    {
+      "epoch": 0.5533085152765201,
+      "grad_norm": 0.15205473966919333,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 8249
+    },
+    {
+      "epoch": 0.5533755911057451,
+      "grad_norm": 0.1483926032493202,
+      "learning_rate": 2e-05,
+      "loss": 5.4356,
+      "step": 8250
+    },
+    {
+      "epoch": 0.55344266693497,
+      "grad_norm": 0.1559469419891767,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 8251
+    },
+    {
+      "epoch": 0.553509742764195,
+      "grad_norm": 0.14933209162370703,
+      "learning_rate": 2e-05,
+      "loss": 5.3549,
+      "step": 8252
+    },
+    {
+      "epoch": 0.5535768185934199,
+      "grad_norm": 0.15315375370619175,
+      "learning_rate": 2e-05,
+      "loss": 5.3308,
+      "step": 8253
+    },
+    {
+      "epoch": 0.5536438944226448,
+      "grad_norm": 0.14989681946708774,
+      "learning_rate": 2e-05,
+      "loss": 5.3834,
+      "step": 8254
+    },
+    {
+      "epoch": 0.5537109702518698,
+      "grad_norm": 0.15115901346165336,
+      "learning_rate": 2e-05,
+      "loss": 5.426,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5537780460810947,
+      "grad_norm": 0.1539068700282336,
+      "learning_rate": 2e-05,
+      "loss": 5.3763,
+      "step": 8256
+    },
+    {
+      "epoch": 0.5538451219103196,
+      "grad_norm": 0.14956411001562736,
+      "learning_rate": 2e-05,
+      "loss": 5.5353,
+      "step": 8257
+    },
+    {
+      "epoch": 0.5539121977395446,
+      "grad_norm": 0.15525538034032874,
+      "learning_rate": 2e-05,
+      "loss": 5.3347,
+      "step": 8258
+    },
+    {
+      "epoch": 0.5539792735687695,
+      "grad_norm": 0.14941868725469717,
+      "learning_rate": 2e-05,
+      "loss": 5.537,
+      "step": 8259
+    },
+    {
+      "epoch": 0.5540463493979945,
+      "grad_norm": 0.14818609298542654,
+      "learning_rate": 2e-05,
+      "loss": 5.4207,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5541134252272194,
+      "grad_norm": 0.14316490909584592,
+      "learning_rate": 2e-05,
+      "loss": 5.3665,
+      "step": 8261
+    },
+    {
+      "epoch": 0.5541805010564443,
+      "grad_norm": 0.16113802157895382,
+      "learning_rate": 2e-05,
+      "loss": 5.5101,
+      "step": 8262
+    },
+    {
+      "epoch": 0.5542475768856693,
+      "grad_norm": 0.15035223073019627,
+      "learning_rate": 2e-05,
+      "loss": 5.4965,
+      "step": 8263
+    },
+    {
+      "epoch": 0.5543146527148942,
+      "grad_norm": 0.14626139442926084,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 8264
+    },
+    {
+      "epoch": 0.5543817285441192,
+      "grad_norm": 0.1548105072371759,
+      "learning_rate": 2e-05,
+      "loss": 5.2848,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5544488043733441,
+      "grad_norm": 0.15593069901071968,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 8266
+    },
+    {
+      "epoch": 0.554515880202569,
+      "grad_norm": 0.14611449997760229,
+      "learning_rate": 2e-05,
+      "loss": 5.4102,
+      "step": 8267
+    },
+    {
+      "epoch": 0.554582956031794,
+      "grad_norm": 0.1526076497103781,
+      "learning_rate": 2e-05,
+      "loss": 5.3591,
+      "step": 8268
+    },
+    {
+      "epoch": 0.5546500318610189,
+      "grad_norm": 0.15429474734773252,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 8269
+    },
+    {
+      "epoch": 0.5547171076902438,
+      "grad_norm": 0.1584446431419016,
+      "learning_rate": 2e-05,
+      "loss": 5.5587,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5547841835194688,
+      "grad_norm": 0.14676391101395134,
+      "learning_rate": 2e-05,
+      "loss": 5.435,
+      "step": 8271
+    },
+    {
+      "epoch": 0.5548512593486937,
+      "grad_norm": 0.16126542449570175,
+      "learning_rate": 2e-05,
+      "loss": 5.4226,
+      "step": 8272
+    },
+    {
+      "epoch": 0.5549183351779187,
+      "grad_norm": 0.15686653578830856,
+      "learning_rate": 2e-05,
+      "loss": 5.3427,
+      "step": 8273
+    },
+    {
+      "epoch": 0.5549854110071436,
+      "grad_norm": 0.14716210903599206,
+      "learning_rate": 2e-05,
+      "loss": 5.4408,
+      "step": 8274
+    },
+    {
+      "epoch": 0.5550524868363685,
+      "grad_norm": 0.15180405802146757,
+      "learning_rate": 2e-05,
+      "loss": 5.5865,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5551195626655935,
+      "grad_norm": 0.15613321663256705,
+      "learning_rate": 2e-05,
+      "loss": 5.4316,
+      "step": 8276
+    },
+    {
+      "epoch": 0.5551866384948184,
+      "grad_norm": 0.15745664692647957,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5552537143240434,
+      "grad_norm": 0.16100177673687216,
+      "learning_rate": 2e-05,
+      "loss": 5.3655,
+      "step": 8278
+    },
+    {
+      "epoch": 0.5553207901532683,
+      "grad_norm": 0.15187883197244395,
+      "learning_rate": 2e-05,
+      "loss": 5.6164,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5553878659824932,
+      "grad_norm": 0.15357272727349028,
+      "learning_rate": 2e-05,
+      "loss": 5.4188,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5554549418117182,
+      "grad_norm": 0.1576566593108162,
+      "learning_rate": 2e-05,
+      "loss": 5.3755,
+      "step": 8281
+    },
+    {
+      "epoch": 0.5555220176409431,
+      "grad_norm": 0.1512132381204248,
+      "learning_rate": 2e-05,
+      "loss": 5.3036,
+      "step": 8282
+    },
+    {
+      "epoch": 0.555589093470168,
+      "grad_norm": 0.15307899824236904,
+      "learning_rate": 2e-05,
+      "loss": 5.4666,
+      "step": 8283
+    },
+    {
+      "epoch": 0.555656169299393,
+      "grad_norm": 0.14970703253937534,
+      "learning_rate": 2e-05,
+      "loss": 5.4582,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5557232451286179,
+      "grad_norm": 0.15828164938874664,
+      "learning_rate": 2e-05,
+      "loss": 5.3389,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5557903209578429,
+      "grad_norm": 0.15175198013400706,
+      "learning_rate": 2e-05,
+      "loss": 5.4564,
+      "step": 8286
+    },
+    {
+      "epoch": 0.5558573967870678,
+      "grad_norm": 0.15658384048941204,
+      "learning_rate": 2e-05,
+      "loss": 5.3999,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5559244726162927,
+      "grad_norm": 0.15828614524175227,
+      "learning_rate": 2e-05,
+      "loss": 5.458,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5559915484455177,
+      "grad_norm": 0.1510477471455415,
+      "learning_rate": 2e-05,
+      "loss": 5.4798,
+      "step": 8289
+    },
+    {
+      "epoch": 0.5560586242747426,
+      "grad_norm": 0.15588807290749543,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5561257001039676,
+      "grad_norm": 0.14640978231112872,
+      "learning_rate": 2e-05,
+      "loss": 5.4714,
+      "step": 8291
+    },
+    {
+      "epoch": 0.5561927759331925,
+      "grad_norm": 0.1519501752513945,
+      "learning_rate": 2e-05,
+      "loss": 5.3937,
+      "step": 8292
+    },
+    {
+      "epoch": 0.5562598517624174,
+      "grad_norm": 0.1532297331404758,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 8293
+    },
+    {
+      "epoch": 0.5563269275916424,
+      "grad_norm": 0.15261630576950933,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5563940034208673,
+      "grad_norm": 0.14890349447593038,
+      "learning_rate": 2e-05,
+      "loss": 5.5614,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5564610792500922,
+      "grad_norm": 0.16363340978450694,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 8296
+    },
+    {
+      "epoch": 0.5565281550793172,
+      "grad_norm": 0.14895165284470765,
+      "learning_rate": 2e-05,
+      "loss": 5.3843,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5565952309085421,
+      "grad_norm": 0.1519447507001162,
+      "learning_rate": 2e-05,
+      "loss": 5.3611,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5566623067377671,
+      "grad_norm": 0.15421082823425947,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 8299
+    },
+    {
+      "epoch": 0.556729382566992,
+      "grad_norm": 0.15259306204149434,
+      "learning_rate": 2e-05,
+      "loss": 5.5,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5567964583962169,
+      "grad_norm": 0.15386453067300931,
+      "learning_rate": 2e-05,
+      "loss": 5.6181,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5568635342254419,
+      "grad_norm": 0.14750306075807867,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5569306100546668,
+      "grad_norm": 0.15748544212857554,
+      "learning_rate": 2e-05,
+      "loss": 5.5321,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5569976858838918,
+      "grad_norm": 0.15376834004399553,
+      "learning_rate": 2e-05,
+      "loss": 5.4912,
+      "step": 8304
+    },
+    {
+      "epoch": 0.5570647617131167,
+      "grad_norm": 0.15451406242634405,
+      "learning_rate": 2e-05,
+      "loss": 5.5553,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5571318375423416,
+      "grad_norm": 0.1569305574527607,
+      "learning_rate": 2e-05,
+      "loss": 5.3952,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5571989133715666,
+      "grad_norm": 0.16330563204600257,
+      "learning_rate": 2e-05,
+      "loss": 5.4003,
+      "step": 8307
+    },
+    {
+      "epoch": 0.5572659892007915,
+      "grad_norm": 0.14796863515420938,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 8308
+    },
+    {
+      "epoch": 0.5573330650300165,
+      "grad_norm": 0.14303294047330292,
+      "learning_rate": 2e-05,
+      "loss": 5.4505,
+      "step": 8309
+    },
+    {
+      "epoch": 0.5574001408592414,
+      "grad_norm": 0.15296427914280272,
+      "learning_rate": 2e-05,
+      "loss": 5.4195,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5574672166884663,
+      "grad_norm": 0.15969487402125518,
+      "learning_rate": 2e-05,
+      "loss": 5.2911,
+      "step": 8311
+    },
+    {
+      "epoch": 0.5575342925176913,
+      "grad_norm": 0.15138503008053225,
+      "learning_rate": 2e-05,
+      "loss": 5.4948,
+      "step": 8312
+    },
+    {
+      "epoch": 0.5576013683469162,
+      "grad_norm": 0.15477826481456702,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 8313
+    },
+    {
+      "epoch": 0.5576684441761411,
+      "grad_norm": 0.15153825051347877,
+      "learning_rate": 2e-05,
+      "loss": 5.5398,
+      "step": 8314
+    },
+    {
+      "epoch": 0.5577355200053661,
+      "grad_norm": 0.15199609807307748,
+      "learning_rate": 2e-05,
+      "loss": 5.2844,
+      "step": 8315
+    },
+    {
+      "epoch": 0.557802595834591,
+      "grad_norm": 0.1533456392429621,
+      "learning_rate": 2e-05,
+      "loss": 5.4406,
+      "step": 8316
+    },
+    {
+      "epoch": 0.557869671663816,
+      "grad_norm": 0.1528296217589683,
+      "learning_rate": 2e-05,
+      "loss": 5.3578,
+      "step": 8317
+    },
+    {
+      "epoch": 0.5579367474930409,
+      "grad_norm": 0.14757477522552476,
+      "learning_rate": 2e-05,
+      "loss": 5.348,
+      "step": 8318
+    },
+    {
+      "epoch": 0.5580038233222658,
+      "grad_norm": 0.15504064087558125,
+      "learning_rate": 2e-05,
+      "loss": 5.4367,
+      "step": 8319
+    },
+    {
+      "epoch": 0.5580708991514908,
+      "grad_norm": 0.15519028766770956,
+      "learning_rate": 2e-05,
+      "loss": 5.501,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5581379749807157,
+      "grad_norm": 0.15102130424441665,
+      "learning_rate": 2e-05,
+      "loss": 5.492,
+      "step": 8321
+    },
+    {
+      "epoch": 0.5582050508099407,
+      "grad_norm": 0.16009591837059045,
+      "learning_rate": 2e-05,
+      "loss": 5.5231,
+      "step": 8322
+    },
+    {
+      "epoch": 0.5582721266391656,
+      "grad_norm": 0.1480009096881547,
+      "learning_rate": 2e-05,
+      "loss": 5.327,
+      "step": 8323
+    },
+    {
+      "epoch": 0.5583392024683905,
+      "grad_norm": 0.15753859418171093,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 8324
+    },
+    {
+      "epoch": 0.5584062782976155,
+      "grad_norm": 0.1551262021942478,
+      "learning_rate": 2e-05,
+      "loss": 5.4152,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5584733541268404,
+      "grad_norm": 0.1510748817423966,
+      "learning_rate": 2e-05,
+      "loss": 5.4139,
+      "step": 8326
+    },
+    {
+      "epoch": 0.5585404299560653,
+      "grad_norm": 0.15135840424785282,
+      "learning_rate": 2e-05,
+      "loss": 5.3907,
+      "step": 8327
+    },
+    {
+      "epoch": 0.5586075057852903,
+      "grad_norm": 0.15306869976661122,
+      "learning_rate": 2e-05,
+      "loss": 5.5642,
+      "step": 8328
+    },
+    {
+      "epoch": 0.5586745816145152,
+      "grad_norm": 0.1460569902637703,
+      "learning_rate": 2e-05,
+      "loss": 5.4209,
+      "step": 8329
+    },
+    {
+      "epoch": 0.5587416574437402,
+      "grad_norm": 0.14997450105466273,
+      "learning_rate": 2e-05,
+      "loss": 5.4753,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5588087332729651,
+      "grad_norm": 0.15327190038422833,
+      "learning_rate": 2e-05,
+      "loss": 5.4622,
+      "step": 8331
+    },
+    {
+      "epoch": 0.55887580910219,
+      "grad_norm": 0.15336808672343402,
+      "learning_rate": 2e-05,
+      "loss": 5.5805,
+      "step": 8332
+    },
+    {
+      "epoch": 0.558942884931415,
+      "grad_norm": 0.14844237368002117,
+      "learning_rate": 2e-05,
+      "loss": 5.6053,
+      "step": 8333
+    },
+    {
+      "epoch": 0.5590099607606399,
+      "grad_norm": 0.1540317997029307,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 8334
+    },
+    {
+      "epoch": 0.5590770365898649,
+      "grad_norm": 0.1517265298965081,
+      "learning_rate": 2e-05,
+      "loss": 5.4335,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5591441124190898,
+      "grad_norm": 0.15147732903925684,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 8336
+    },
+    {
+      "epoch": 0.5592111882483147,
+      "grad_norm": 0.16075828219236538,
+      "learning_rate": 2e-05,
+      "loss": 5.5868,
+      "step": 8337
+    },
+    {
+      "epoch": 0.5592782640775397,
+      "grad_norm": 0.15112382061187563,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 8338
+    },
+    {
+      "epoch": 0.5593453399067646,
+      "grad_norm": 0.14966050543324746,
+      "learning_rate": 2e-05,
+      "loss": 5.3903,
+      "step": 8339
+    },
+    {
+      "epoch": 0.5594124157359895,
+      "grad_norm": 0.14442142871261798,
+      "learning_rate": 2e-05,
+      "loss": 5.431,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5594794915652145,
+      "grad_norm": 0.15155222164790577,
+      "learning_rate": 2e-05,
+      "loss": 5.3823,
+      "step": 8341
+    },
+    {
+      "epoch": 0.5595465673944394,
+      "grad_norm": 0.14763425168887864,
+      "learning_rate": 2e-05,
+      "loss": 5.2399,
+      "step": 8342
+    },
+    {
+      "epoch": 0.5596136432236644,
+      "grad_norm": 0.1566247706134838,
+      "learning_rate": 2e-05,
+      "loss": 5.2657,
+      "step": 8343
+    },
+    {
+      "epoch": 0.5596807190528893,
+      "grad_norm": 0.15172232818093487,
+      "learning_rate": 2e-05,
+      "loss": 5.4817,
+      "step": 8344
+    },
+    {
+      "epoch": 0.5597477948821142,
+      "grad_norm": 0.1540314560578181,
+      "learning_rate": 2e-05,
+      "loss": 5.3569,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5598148707113392,
+      "grad_norm": 0.1506744315853682,
+      "learning_rate": 2e-05,
+      "loss": 5.3014,
+      "step": 8346
+    },
+    {
+      "epoch": 0.5598819465405641,
+      "grad_norm": 0.14760550155047691,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 8347
+    },
+    {
+      "epoch": 0.559949022369789,
+      "grad_norm": 0.15015271195795335,
+      "learning_rate": 2e-05,
+      "loss": 5.5835,
+      "step": 8348
+    },
+    {
+      "epoch": 0.560016098199014,
+      "grad_norm": 0.15972147955388066,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 8349
+    },
+    {
+      "epoch": 0.5600831740282389,
+      "grad_norm": 0.15091135596017882,
+      "learning_rate": 2e-05,
+      "loss": 5.4072,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5601502498574639,
+      "grad_norm": 0.1527351784464482,
+      "learning_rate": 2e-05,
+      "loss": 5.399,
+      "step": 8351
+    },
+    {
+      "epoch": 0.5602173256866888,
+      "grad_norm": 0.14967316863810318,
+      "learning_rate": 2e-05,
+      "loss": 5.5019,
+      "step": 8352
+    },
+    {
+      "epoch": 0.5602844015159137,
+      "grad_norm": 0.1566380364050661,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 8353
+    },
+    {
+      "epoch": 0.5603514773451387,
+      "grad_norm": 0.15100020669556558,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 8354
+    },
+    {
+      "epoch": 0.5604185531743636,
+      "grad_norm": 0.1619616917331649,
+      "learning_rate": 2e-05,
+      "loss": 5.4539,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5604856290035886,
+      "grad_norm": 0.1503963485529452,
+      "learning_rate": 2e-05,
+      "loss": 5.4231,
+      "step": 8356
+    },
+    {
+      "epoch": 0.5605527048328135,
+      "grad_norm": 0.15473229706002933,
+      "learning_rate": 2e-05,
+      "loss": 5.3786,
+      "step": 8357
+    },
+    {
+      "epoch": 0.5606197806620384,
+      "grad_norm": 0.1518823569142259,
+      "learning_rate": 2e-05,
+      "loss": 5.3305,
+      "step": 8358
+    },
+    {
+      "epoch": 0.5606868564912634,
+      "grad_norm": 0.15099532128879317,
+      "learning_rate": 2e-05,
+      "loss": 5.4291,
+      "step": 8359
+    },
+    {
+      "epoch": 0.5607539323204883,
+      "grad_norm": 0.154266932631846,
+      "learning_rate": 2e-05,
+      "loss": 5.5347,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5608210081497133,
+      "grad_norm": 0.15533611304582307,
+      "learning_rate": 2e-05,
+      "loss": 5.3042,
+      "step": 8361
+    },
+    {
+      "epoch": 0.5608880839789382,
+      "grad_norm": 0.15060355062219735,
+      "learning_rate": 2e-05,
+      "loss": 5.4909,
+      "step": 8362
+    },
+    {
+      "epoch": 0.5609551598081631,
+      "grad_norm": 0.15795026446548519,
+      "learning_rate": 2e-05,
+      "loss": 5.2894,
+      "step": 8363
+    },
+    {
+      "epoch": 0.5610222356373881,
+      "grad_norm": 0.14633987572405066,
+      "learning_rate": 2e-05,
+      "loss": 5.4609,
+      "step": 8364
+    },
+    {
+      "epoch": 0.561089311466613,
+      "grad_norm": 0.15038676496904652,
+      "learning_rate": 2e-05,
+      "loss": 5.5203,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5611563872958379,
+      "grad_norm": 0.1595298496524941,
+      "learning_rate": 2e-05,
+      "loss": 5.3551,
+      "step": 8366
+    },
+    {
+      "epoch": 0.5612234631250629,
+      "grad_norm": 0.14888604149280033,
+      "learning_rate": 2e-05,
+      "loss": 5.5212,
+      "step": 8367
+    },
+    {
+      "epoch": 0.5612905389542878,
+      "grad_norm": 0.15128021504539538,
+      "learning_rate": 2e-05,
+      "loss": 5.3581,
+      "step": 8368
+    },
+    {
+      "epoch": 0.5613576147835128,
+      "grad_norm": 0.14697353930359078,
+      "learning_rate": 2e-05,
+      "loss": 5.6004,
+      "step": 8369
+    },
+    {
+      "epoch": 0.5614246906127377,
+      "grad_norm": 0.1527827272727508,
+      "learning_rate": 2e-05,
+      "loss": 5.3725,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5614917664419626,
+      "grad_norm": 0.15682145793322352,
+      "learning_rate": 2e-05,
+      "loss": 5.4536,
+      "step": 8371
+    },
+    {
+      "epoch": 0.5615588422711876,
+      "grad_norm": 0.15865152406863064,
+      "learning_rate": 2e-05,
+      "loss": 5.3382,
+      "step": 8372
+    },
+    {
+      "epoch": 0.5616259181004125,
+      "grad_norm": 0.15137824471286823,
+      "learning_rate": 2e-05,
+      "loss": 5.3982,
+      "step": 8373
+    },
+    {
+      "epoch": 0.5616929939296375,
+      "grad_norm": 0.15776370373826257,
+      "learning_rate": 2e-05,
+      "loss": 5.5403,
+      "step": 8374
+    },
+    {
+      "epoch": 0.5617600697588624,
+      "grad_norm": 0.15052358763622256,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5618271455880873,
+      "grad_norm": 0.15090561321261087,
+      "learning_rate": 2e-05,
+      "loss": 5.5386,
+      "step": 8376
+    },
+    {
+      "epoch": 0.5618942214173123,
+      "grad_norm": 0.15004043059487152,
+      "learning_rate": 2e-05,
+      "loss": 5.4433,
+      "step": 8377
+    },
+    {
+      "epoch": 0.5619612972465372,
+      "grad_norm": 0.14438066679810113,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 8378
+    },
+    {
+      "epoch": 0.5620283730757621,
+      "grad_norm": 0.1520898745400064,
+      "learning_rate": 2e-05,
+      "loss": 5.3517,
+      "step": 8379
+    },
+    {
+      "epoch": 0.5620954489049871,
+      "grad_norm": 0.1477204860574268,
+      "learning_rate": 2e-05,
+      "loss": 5.2988,
+      "step": 8380
+    },
+    {
+      "epoch": 0.562162524734212,
+      "grad_norm": 0.15270897204231845,
+      "learning_rate": 2e-05,
+      "loss": 5.5864,
+      "step": 8381
+    },
+    {
+      "epoch": 0.562229600563437,
+      "grad_norm": 0.15564392818529,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 8382
+    },
+    {
+      "epoch": 0.5622966763926619,
+      "grad_norm": 0.1572526515341801,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 8383
+    },
+    {
+      "epoch": 0.5623637522218868,
+      "grad_norm": 0.17177555573677625,
+      "learning_rate": 2e-05,
+      "loss": 5.3495,
+      "step": 8384
+    },
+    {
+      "epoch": 0.5624308280511118,
+      "grad_norm": 0.14694975489229783,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 8385
+    },
+    {
+      "epoch": 0.5624979038803367,
+      "grad_norm": 0.16437681213598365,
+      "learning_rate": 2e-05,
+      "loss": 5.5052,
+      "step": 8386
+    },
+    {
+      "epoch": 0.5625649797095617,
+      "grad_norm": 0.15901036436633817,
+      "learning_rate": 2e-05,
+      "loss": 5.3422,
+      "step": 8387
+    },
+    {
+      "epoch": 0.5626320555387866,
+      "grad_norm": 0.15006739719843395,
+      "learning_rate": 2e-05,
+      "loss": 5.4071,
+      "step": 8388
+    },
+    {
+      "epoch": 0.5626991313680115,
+      "grad_norm": 0.15803539800186256,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 8389
+    },
+    {
+      "epoch": 0.5627662071972365,
+      "grad_norm": 0.1598848944290701,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5628332830264614,
+      "grad_norm": 0.1498341090428145,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 8391
+    },
+    {
+      "epoch": 0.5629003588556863,
+      "grad_norm": 0.1583015242037235,
+      "learning_rate": 2e-05,
+      "loss": 5.4313,
+      "step": 8392
+    },
+    {
+      "epoch": 0.5629674346849113,
+      "grad_norm": 0.16476038253448347,
+      "learning_rate": 2e-05,
+      "loss": 5.3549,
+      "step": 8393
+    },
+    {
+      "epoch": 0.5630345105141362,
+      "grad_norm": 0.14955668999017047,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 8394
+    },
+    {
+      "epoch": 0.5631015863433612,
+      "grad_norm": 0.15268080141388873,
+      "learning_rate": 2e-05,
+      "loss": 5.4415,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5631686621725861,
+      "grad_norm": 0.15903865049506544,
+      "learning_rate": 2e-05,
+      "loss": 5.4189,
+      "step": 8396
+    },
+    {
+      "epoch": 0.563235738001811,
+      "grad_norm": 0.15459186892131302,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 8397
+    },
+    {
+      "epoch": 0.563302813831036,
+      "grad_norm": 0.14541495621357595,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 8398
+    },
+    {
+      "epoch": 0.5633698896602609,
+      "grad_norm": 0.1458868796149539,
+      "learning_rate": 2e-05,
+      "loss": 5.3527,
+      "step": 8399
+    },
+    {
+      "epoch": 0.5634369654894859,
+      "grad_norm": 0.16608744018766378,
+      "learning_rate": 2e-05,
+      "loss": 5.3655,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5635040413187108,
+      "grad_norm": 0.15378546428106524,
+      "learning_rate": 2e-05,
+      "loss": 5.3385,
+      "step": 8401
+    },
+    {
+      "epoch": 0.5635711171479357,
+      "grad_norm": 0.15220201418548943,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 8402
+    },
+    {
+      "epoch": 0.5636381929771607,
+      "grad_norm": 0.1530914075935242,
+      "learning_rate": 2e-05,
+      "loss": 5.4459,
+      "step": 8403
+    },
+    {
+      "epoch": 0.5637052688063856,
+      "grad_norm": 0.14944556643990206,
+      "learning_rate": 2e-05,
+      "loss": 5.4765,
+      "step": 8404
+    },
+    {
+      "epoch": 0.5637723446356105,
+      "grad_norm": 0.1527858542958006,
+      "learning_rate": 2e-05,
+      "loss": 5.4436,
+      "step": 8405
+    },
+    {
+      "epoch": 0.5638394204648355,
+      "grad_norm": 0.1489460886452453,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 8406
+    },
+    {
+      "epoch": 0.5639064962940604,
+      "grad_norm": 0.1488990513356671,
+      "learning_rate": 2e-05,
+      "loss": 5.541,
+      "step": 8407
+    },
+    {
+      "epoch": 0.5639735721232854,
+      "grad_norm": 0.1495540587991848,
+      "learning_rate": 2e-05,
+      "loss": 5.4878,
+      "step": 8408
+    },
+    {
+      "epoch": 0.5640406479525103,
+      "grad_norm": 0.15604963713016998,
+      "learning_rate": 2e-05,
+      "loss": 5.3188,
+      "step": 8409
+    },
+    {
+      "epoch": 0.5641077237817352,
+      "grad_norm": 0.14515552657441133,
+      "learning_rate": 2e-05,
+      "loss": 5.2816,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5641747996109602,
+      "grad_norm": 0.1456144617312568,
+      "learning_rate": 2e-05,
+      "loss": 5.5187,
+      "step": 8411
+    },
+    {
+      "epoch": 0.5642418754401851,
+      "grad_norm": 0.14910032149783178,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 8412
+    },
+    {
+      "epoch": 0.56430895126941,
+      "grad_norm": 0.1506880382358528,
+      "learning_rate": 2e-05,
+      "loss": 5.4183,
+      "step": 8413
+    },
+    {
+      "epoch": 0.564376027098635,
+      "grad_norm": 0.15452549278894273,
+      "learning_rate": 2e-05,
+      "loss": 5.3992,
+      "step": 8414
+    },
+    {
+      "epoch": 0.5644431029278599,
+      "grad_norm": 0.15850898832514562,
+      "learning_rate": 2e-05,
+      "loss": 5.4203,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5645101787570849,
+      "grad_norm": 0.15489653897171982,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 8416
+    },
+    {
+      "epoch": 0.5645772545863098,
+      "grad_norm": 0.15608648445536705,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 8417
+    },
+    {
+      "epoch": 0.5646443304155347,
+      "grad_norm": 0.14779946552260964,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 8418
+    },
+    {
+      "epoch": 0.5647114062447597,
+      "grad_norm": 0.1444922558697645,
+      "learning_rate": 2e-05,
+      "loss": 5.3888,
+      "step": 8419
+    },
+    {
+      "epoch": 0.5647784820739846,
+      "grad_norm": 0.1522173170663822,
+      "learning_rate": 2e-05,
+      "loss": 5.4694,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5648455579032096,
+      "grad_norm": 0.14874645412890672,
+      "learning_rate": 2e-05,
+      "loss": 5.4278,
+      "step": 8421
+    },
+    {
+      "epoch": 0.5649126337324345,
+      "grad_norm": 0.1482926367684781,
+      "learning_rate": 2e-05,
+      "loss": 5.3545,
+      "step": 8422
+    },
+    {
+      "epoch": 0.5649797095616594,
+      "grad_norm": 0.15230848934103225,
+      "learning_rate": 2e-05,
+      "loss": 5.5115,
+      "step": 8423
+    },
+    {
+      "epoch": 0.5650467853908844,
+      "grad_norm": 0.15292766247489034,
+      "learning_rate": 2e-05,
+      "loss": 5.5586,
+      "step": 8424
+    },
+    {
+      "epoch": 0.5651138612201093,
+      "grad_norm": 0.1538813957250163,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5651809370493343,
+      "grad_norm": 0.15784646542422423,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 8426
+    },
+    {
+      "epoch": 0.5652480128785592,
+      "grad_norm": 0.16696873485580177,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 8427
+    },
+    {
+      "epoch": 0.5653150887077841,
+      "grad_norm": 0.15656208415397616,
+      "learning_rate": 2e-05,
+      "loss": 5.3664,
+      "step": 8428
+    },
+    {
+      "epoch": 0.5653821645370091,
+      "grad_norm": 0.15635242290140475,
+      "learning_rate": 2e-05,
+      "loss": 5.5388,
+      "step": 8429
+    },
+    {
+      "epoch": 0.565449240366234,
+      "grad_norm": 0.1586134275186656,
+      "learning_rate": 2e-05,
+      "loss": 5.416,
+      "step": 8430
+    },
+    {
+      "epoch": 0.565516316195459,
+      "grad_norm": 0.16847096406947315,
+      "learning_rate": 2e-05,
+      "loss": 5.3797,
+      "step": 8431
+    },
+    {
+      "epoch": 0.5655833920246839,
+      "grad_norm": 0.14789299605653566,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 8432
+    },
+    {
+      "epoch": 0.5656504678539088,
+      "grad_norm": 0.15797099555905078,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 8433
+    },
+    {
+      "epoch": 0.5657175436831338,
+      "grad_norm": 0.16602562971830442,
+      "learning_rate": 2e-05,
+      "loss": 5.5051,
+      "step": 8434
+    },
+    {
+      "epoch": 0.5657846195123587,
+      "grad_norm": 0.1503582033213871,
+      "learning_rate": 2e-05,
+      "loss": 5.4733,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5658516953415836,
+      "grad_norm": 0.1726786449498398,
+      "learning_rate": 2e-05,
+      "loss": 5.4292,
+      "step": 8436
+    },
+    {
+      "epoch": 0.5659187711708086,
+      "grad_norm": 0.1548586247893178,
+      "learning_rate": 2e-05,
+      "loss": 5.4334,
+      "step": 8437
+    },
+    {
+      "epoch": 0.5659858470000335,
+      "grad_norm": 0.1458126575746454,
+      "learning_rate": 2e-05,
+      "loss": 5.2854,
+      "step": 8438
+    },
+    {
+      "epoch": 0.5660529228292585,
+      "grad_norm": 0.1589711132605877,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 8439
+    },
+    {
+      "epoch": 0.5661199986584834,
+      "grad_norm": 0.15772034709294838,
+      "learning_rate": 2e-05,
+      "loss": 5.3893,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5661870744877083,
+      "grad_norm": 0.1559549534280849,
+      "learning_rate": 2e-05,
+      "loss": 5.5151,
+      "step": 8441
+    },
+    {
+      "epoch": 0.5662541503169333,
+      "grad_norm": 0.1575522491870645,
+      "learning_rate": 2e-05,
+      "loss": 5.4659,
+      "step": 8442
+    },
+    {
+      "epoch": 0.5663212261461582,
+      "grad_norm": 0.1517250867815135,
+      "learning_rate": 2e-05,
+      "loss": 5.463,
+      "step": 8443
+    },
+    {
+      "epoch": 0.5663883019753831,
+      "grad_norm": 0.14753555682757594,
+      "learning_rate": 2e-05,
+      "loss": 5.5071,
+      "step": 8444
+    },
+    {
+      "epoch": 0.5664553778046081,
+      "grad_norm": 0.16741883419211065,
+      "learning_rate": 2e-05,
+      "loss": 5.385,
+      "step": 8445
+    },
+    {
+      "epoch": 0.566522453633833,
+      "grad_norm": 0.15371629917850108,
+      "learning_rate": 2e-05,
+      "loss": 5.6194,
+      "step": 8446
+    },
+    {
+      "epoch": 0.566589529463058,
+      "grad_norm": 0.15199427486085162,
+      "learning_rate": 2e-05,
+      "loss": 5.533,
+      "step": 8447
+    },
+    {
+      "epoch": 0.5666566052922829,
+      "grad_norm": 0.1578834943405736,
+      "learning_rate": 2e-05,
+      "loss": 5.4897,
+      "step": 8448
+    },
+    {
+      "epoch": 0.5667236811215078,
+      "grad_norm": 0.15027163355541792,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 8449
+    },
+    {
+      "epoch": 0.5667907569507328,
+      "grad_norm": 0.14923000798474464,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5668578327799577,
+      "grad_norm": 0.15290069146592325,
+      "learning_rate": 2e-05,
+      "loss": 5.4233,
+      "step": 8451
+    },
+    {
+      "epoch": 0.5669249086091827,
+      "grad_norm": 0.14664679119161939,
+      "learning_rate": 2e-05,
+      "loss": 5.4052,
+      "step": 8452
+    },
+    {
+      "epoch": 0.5669919844384076,
+      "grad_norm": 0.15501071779918568,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 8453
+    },
+    {
+      "epoch": 0.5670590602676325,
+      "grad_norm": 0.15428853737421733,
+      "learning_rate": 2e-05,
+      "loss": 5.3263,
+      "step": 8454
+    },
+    {
+      "epoch": 0.5671261360968575,
+      "grad_norm": 0.1572244546462615,
+      "learning_rate": 2e-05,
+      "loss": 5.3401,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5671932119260824,
+      "grad_norm": 0.16084650499984968,
+      "learning_rate": 2e-05,
+      "loss": 5.46,
+      "step": 8456
+    },
+    {
+      "epoch": 0.5672602877553073,
+      "grad_norm": 0.1507006618676138,
+      "learning_rate": 2e-05,
+      "loss": 5.2004,
+      "step": 8457
+    },
+    {
+      "epoch": 0.5673273635845323,
+      "grad_norm": 0.15300876112799863,
+      "learning_rate": 2e-05,
+      "loss": 5.2884,
+      "step": 8458
+    },
+    {
+      "epoch": 0.5673944394137572,
+      "grad_norm": 0.15434474423423553,
+      "learning_rate": 2e-05,
+      "loss": 5.3252,
+      "step": 8459
+    },
+    {
+      "epoch": 0.5674615152429822,
+      "grad_norm": 0.15908972640332844,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5675285910722071,
+      "grad_norm": 0.1523515632045846,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 8461
+    },
+    {
+      "epoch": 0.567595666901432,
+      "grad_norm": 0.16371953814948748,
+      "learning_rate": 2e-05,
+      "loss": 5.4375,
+      "step": 8462
+    },
+    {
+      "epoch": 0.567662742730657,
+      "grad_norm": 0.14901318738443828,
+      "learning_rate": 2e-05,
+      "loss": 5.5109,
+      "step": 8463
+    },
+    {
+      "epoch": 0.5677298185598819,
+      "grad_norm": 0.15503997977357425,
+      "learning_rate": 2e-05,
+      "loss": 5.3269,
+      "step": 8464
+    },
+    {
+      "epoch": 0.5677968943891069,
+      "grad_norm": 0.14968441527340132,
+      "learning_rate": 2e-05,
+      "loss": 5.3473,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5678639702183318,
+      "grad_norm": 0.15511608741623054,
+      "learning_rate": 2e-05,
+      "loss": 5.2636,
+      "step": 8466
+    },
+    {
+      "epoch": 0.5679310460475567,
+      "grad_norm": 0.1581949967285679,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 8467
+    },
+    {
+      "epoch": 0.5679981218767817,
+      "grad_norm": 0.1515832914460592,
+      "learning_rate": 2e-05,
+      "loss": 5.501,
+      "step": 8468
+    },
+    {
+      "epoch": 0.5680651977060066,
+      "grad_norm": 0.16486691888531466,
+      "learning_rate": 2e-05,
+      "loss": 5.542,
+      "step": 8469
+    },
+    {
+      "epoch": 0.5681322735352315,
+      "grad_norm": 0.15213133963866501,
+      "learning_rate": 2e-05,
+      "loss": 5.2827,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5681993493644565,
+      "grad_norm": 0.1493452317876701,
+      "learning_rate": 2e-05,
+      "loss": 5.4395,
+      "step": 8471
+    },
+    {
+      "epoch": 0.5682664251936814,
+      "grad_norm": 0.15784873141233988,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 8472
+    },
+    {
+      "epoch": 0.5683335010229064,
+      "grad_norm": 0.15493931492366828,
+      "learning_rate": 2e-05,
+      "loss": 5.2869,
+      "step": 8473
+    },
+    {
+      "epoch": 0.5684005768521313,
+      "grad_norm": 0.14723216986482565,
+      "learning_rate": 2e-05,
+      "loss": 5.4557,
+      "step": 8474
+    },
+    {
+      "epoch": 0.5684676526813562,
+      "grad_norm": 0.1580244169345664,
+      "learning_rate": 2e-05,
+      "loss": 5.4025,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5685347285105812,
+      "grad_norm": 0.15891621061785166,
+      "learning_rate": 2e-05,
+      "loss": 5.2926,
+      "step": 8476
+    },
+    {
+      "epoch": 0.5686018043398061,
+      "grad_norm": 0.14390966227127866,
+      "learning_rate": 2e-05,
+      "loss": 5.5695,
+      "step": 8477
+    },
+    {
+      "epoch": 0.568668880169031,
+      "grad_norm": 0.15767102118778262,
+      "learning_rate": 2e-05,
+      "loss": 5.3172,
+      "step": 8478
+    },
+    {
+      "epoch": 0.568735955998256,
+      "grad_norm": 0.1588579827262618,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 8479
+    },
+    {
+      "epoch": 0.5688030318274809,
+      "grad_norm": 0.14857056061278134,
+      "learning_rate": 2e-05,
+      "loss": 5.4922,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5688701076567059,
+      "grad_norm": 0.15157267489578333,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 8481
+    },
+    {
+      "epoch": 0.5689371834859308,
+      "grad_norm": 0.15297987045363698,
+      "learning_rate": 2e-05,
+      "loss": 5.3408,
+      "step": 8482
+    },
+    {
+      "epoch": 0.5690042593151557,
+      "grad_norm": 0.15751013514600856,
+      "learning_rate": 2e-05,
+      "loss": 5.3757,
+      "step": 8483
+    },
+    {
+      "epoch": 0.5690713351443807,
+      "grad_norm": 0.15419700114481355,
+      "learning_rate": 2e-05,
+      "loss": 5.3453,
+      "step": 8484
+    },
+    {
+      "epoch": 0.5691384109736056,
+      "grad_norm": 0.15644408889275718,
+      "learning_rate": 2e-05,
+      "loss": 5.4378,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5692054868028306,
+      "grad_norm": 0.16054423409571883,
+      "learning_rate": 2e-05,
+      "loss": 5.4879,
+      "step": 8486
+    },
+    {
+      "epoch": 0.5692725626320555,
+      "grad_norm": 0.15267500762674455,
+      "learning_rate": 2e-05,
+      "loss": 5.5964,
+      "step": 8487
+    },
+    {
+      "epoch": 0.5693396384612804,
+      "grad_norm": 0.14748741782925634,
+      "learning_rate": 2e-05,
+      "loss": 5.156,
+      "step": 8488
+    },
+    {
+      "epoch": 0.5694067142905054,
+      "grad_norm": 0.15116660431178192,
+      "learning_rate": 2e-05,
+      "loss": 5.5364,
+      "step": 8489
+    },
+    {
+      "epoch": 0.5694737901197303,
+      "grad_norm": 0.15148032766952277,
+      "learning_rate": 2e-05,
+      "loss": 5.1452,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5695408659489553,
+      "grad_norm": 0.1514017563987818,
+      "learning_rate": 2e-05,
+      "loss": 5.4147,
+      "step": 8491
+    },
+    {
+      "epoch": 0.5696079417781802,
+      "grad_norm": 0.1557151317300217,
+      "learning_rate": 2e-05,
+      "loss": 5.3564,
+      "step": 8492
+    },
+    {
+      "epoch": 0.5696750176074051,
+      "grad_norm": 0.15952474183624973,
+      "learning_rate": 2e-05,
+      "loss": 5.4136,
+      "step": 8493
+    },
+    {
+      "epoch": 0.5697420934366301,
+      "grad_norm": 0.15775733799663494,
+      "learning_rate": 2e-05,
+      "loss": 5.6499,
+      "step": 8494
+    },
+    {
+      "epoch": 0.569809169265855,
+      "grad_norm": 0.1510979571135918,
+      "learning_rate": 2e-05,
+      "loss": 5.3946,
+      "step": 8495
+    },
+    {
+      "epoch": 0.56987624509508,
+      "grad_norm": 0.15538522493662282,
+      "learning_rate": 2e-05,
+      "loss": 5.4862,
+      "step": 8496
+    },
+    {
+      "epoch": 0.5699433209243049,
+      "grad_norm": 0.1603480323535246,
+      "learning_rate": 2e-05,
+      "loss": 5.4807,
+      "step": 8497
+    },
+    {
+      "epoch": 0.5700103967535298,
+      "grad_norm": 0.15633978044561814,
+      "learning_rate": 2e-05,
+      "loss": 5.4549,
+      "step": 8498
+    },
+    {
+      "epoch": 0.5700774725827548,
+      "grad_norm": 0.16337335933401934,
+      "learning_rate": 2e-05,
+      "loss": 5.4423,
+      "step": 8499
+    },
+    {
+      "epoch": 0.5701445484119797,
+      "grad_norm": 0.15766861441914842,
+      "learning_rate": 2e-05,
+      "loss": 5.4161,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5702116242412046,
+      "grad_norm": 0.16530274158268696,
+      "learning_rate": 2e-05,
+      "loss": 5.3399,
+      "step": 8501
+    },
+    {
+      "epoch": 0.5702787000704296,
+      "grad_norm": 0.1532731372985545,
+      "learning_rate": 2e-05,
+      "loss": 5.473,
+      "step": 8502
+    },
+    {
+      "epoch": 0.5703457758996545,
+      "grad_norm": 0.15409150440702696,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 8503
+    },
+    {
+      "epoch": 0.5704128517288795,
+      "grad_norm": 0.15630817655691787,
+      "learning_rate": 2e-05,
+      "loss": 5.5776,
+      "step": 8504
+    },
+    {
+      "epoch": 0.5704799275581044,
+      "grad_norm": 0.14676220544163973,
+      "learning_rate": 2e-05,
+      "loss": 5.3869,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5705470033873293,
+      "grad_norm": 0.15285598136087647,
+      "learning_rate": 2e-05,
+      "loss": 5.4683,
+      "step": 8506
+    },
+    {
+      "epoch": 0.5706140792165543,
+      "grad_norm": 0.15380555839496768,
+      "learning_rate": 2e-05,
+      "loss": 5.3762,
+      "step": 8507
+    },
+    {
+      "epoch": 0.5706811550457792,
+      "grad_norm": 0.1557795096847412,
+      "learning_rate": 2e-05,
+      "loss": 5.4825,
+      "step": 8508
+    },
+    {
+      "epoch": 0.5707482308750041,
+      "grad_norm": 0.1501139312854184,
+      "learning_rate": 2e-05,
+      "loss": 5.4632,
+      "step": 8509
+    },
+    {
+      "epoch": 0.5708153067042291,
+      "grad_norm": 0.15295836990648143,
+      "learning_rate": 2e-05,
+      "loss": 5.497,
+      "step": 8510
+    },
+    {
+      "epoch": 0.570882382533454,
+      "grad_norm": 0.15368620779342276,
+      "learning_rate": 2e-05,
+      "loss": 5.4627,
+      "step": 8511
+    },
+    {
+      "epoch": 0.570949458362679,
+      "grad_norm": 0.15584504688813106,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 8512
+    },
+    {
+      "epoch": 0.5710165341919039,
+      "grad_norm": 0.1555409479516046,
+      "learning_rate": 2e-05,
+      "loss": 5.3694,
+      "step": 8513
+    },
+    {
+      "epoch": 0.5710836100211288,
+      "grad_norm": 0.15630855512733477,
+      "learning_rate": 2e-05,
+      "loss": 5.3322,
+      "step": 8514
+    },
+    {
+      "epoch": 0.5711506858503538,
+      "grad_norm": 0.16269288951280308,
+      "learning_rate": 2e-05,
+      "loss": 5.5705,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5712177616795787,
+      "grad_norm": 0.15286646185953673,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 8516
+    },
+    {
+      "epoch": 0.5712848375088037,
+      "grad_norm": 0.17165244986244296,
+      "learning_rate": 2e-05,
+      "loss": 5.5666,
+      "step": 8517
+    },
+    {
+      "epoch": 0.5713519133380286,
+      "grad_norm": 0.16310191933496454,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 8518
+    },
+    {
+      "epoch": 0.5714189891672535,
+      "grad_norm": 0.15491588992705904,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 8519
+    },
+    {
+      "epoch": 0.5714860649964785,
+      "grad_norm": 0.1488184652997587,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5715531408257034,
+      "grad_norm": 0.16076806729375948,
+      "learning_rate": 2e-05,
+      "loss": 5.4759,
+      "step": 8521
+    },
+    {
+      "epoch": 0.5716202166549283,
+      "grad_norm": 0.15784546323669266,
+      "learning_rate": 2e-05,
+      "loss": 5.3787,
+      "step": 8522
+    },
+    {
+      "epoch": 0.5716872924841533,
+      "grad_norm": 0.1475270129091651,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 8523
+    },
+    {
+      "epoch": 0.5717543683133782,
+      "grad_norm": 0.15743756391300776,
+      "learning_rate": 2e-05,
+      "loss": 5.483,
+      "step": 8524
+    },
+    {
+      "epoch": 0.5718214441426032,
+      "grad_norm": 0.16045943231644696,
+      "learning_rate": 2e-05,
+      "loss": 5.3571,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5718885199718281,
+      "grad_norm": 0.1479814786630079,
+      "learning_rate": 2e-05,
+      "loss": 5.467,
+      "step": 8526
+    },
+    {
+      "epoch": 0.571955595801053,
+      "grad_norm": 0.15463178127161237,
+      "learning_rate": 2e-05,
+      "loss": 5.4312,
+      "step": 8527
+    },
+    {
+      "epoch": 0.572022671630278,
+      "grad_norm": 0.1592374962219593,
+      "learning_rate": 2e-05,
+      "loss": 5.4586,
+      "step": 8528
+    },
+    {
+      "epoch": 0.5720897474595029,
+      "grad_norm": 0.15059445433567878,
+      "learning_rate": 2e-05,
+      "loss": 5.4677,
+      "step": 8529
+    },
+    {
+      "epoch": 0.5721568232887279,
+      "grad_norm": 0.14823185112809012,
+      "learning_rate": 2e-05,
+      "loss": 5.3409,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5722238991179528,
+      "grad_norm": 0.16273872962740665,
+      "learning_rate": 2e-05,
+      "loss": 5.3582,
+      "step": 8531
+    },
+    {
+      "epoch": 0.5722909749471777,
+      "grad_norm": 0.15932168156229648,
+      "learning_rate": 2e-05,
+      "loss": 5.3014,
+      "step": 8532
+    },
+    {
+      "epoch": 0.5723580507764027,
+      "grad_norm": 0.1492249650865691,
+      "learning_rate": 2e-05,
+      "loss": 5.3487,
+      "step": 8533
+    },
+    {
+      "epoch": 0.5724251266056276,
+      "grad_norm": 0.16128652619700662,
+      "learning_rate": 2e-05,
+      "loss": 5.4622,
+      "step": 8534
+    },
+    {
+      "epoch": 0.5724922024348525,
+      "grad_norm": 0.1635237451219097,
+      "learning_rate": 2e-05,
+      "loss": 5.422,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5725592782640775,
+      "grad_norm": 0.15529691031742313,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 8536
+    },
+    {
+      "epoch": 0.5726263540933024,
+      "grad_norm": 0.1639068347936482,
+      "learning_rate": 2e-05,
+      "loss": 5.4737,
+      "step": 8537
+    },
+    {
+      "epoch": 0.5726934299225274,
+      "grad_norm": 0.1585655661841891,
+      "learning_rate": 2e-05,
+      "loss": 5.4307,
+      "step": 8538
+    },
+    {
+      "epoch": 0.5727605057517523,
+      "grad_norm": 0.15051279854780178,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 8539
+    },
+    {
+      "epoch": 0.5728275815809772,
+      "grad_norm": 0.15578586082237403,
+      "learning_rate": 2e-05,
+      "loss": 5.5098,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5728946574102022,
+      "grad_norm": 0.1542641127007967,
+      "learning_rate": 2e-05,
+      "loss": 5.4233,
+      "step": 8541
+    },
+    {
+      "epoch": 0.5729617332394271,
+      "grad_norm": 0.15054599561949678,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 8542
+    },
+    {
+      "epoch": 0.5730288090686521,
+      "grad_norm": 0.14666911641427152,
+      "learning_rate": 2e-05,
+      "loss": 5.3379,
+      "step": 8543
+    },
+    {
+      "epoch": 0.573095884897877,
+      "grad_norm": 0.16717758313532485,
+      "learning_rate": 2e-05,
+      "loss": 5.4923,
+      "step": 8544
+    },
+    {
+      "epoch": 0.5731629607271019,
+      "grad_norm": 0.16081305091621845,
+      "learning_rate": 2e-05,
+      "loss": 5.4666,
+      "step": 8545
+    },
+    {
+      "epoch": 0.573230036556327,
+      "grad_norm": 0.16024898429530782,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 8546
+    },
+    {
+      "epoch": 0.5732971123855519,
+      "grad_norm": 0.16987856079958674,
+      "learning_rate": 2e-05,
+      "loss": 5.3511,
+      "step": 8547
+    },
+    {
+      "epoch": 0.5733641882147769,
+      "grad_norm": 0.1498617095733163,
+      "learning_rate": 2e-05,
+      "loss": 5.516,
+      "step": 8548
+    },
+    {
+      "epoch": 0.5734312640440018,
+      "grad_norm": 0.1565912790156857,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 8549
+    },
+    {
+      "epoch": 0.5734983398732267,
+      "grad_norm": 0.16361976266563688,
+      "learning_rate": 2e-05,
+      "loss": 5.3551,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5735654157024517,
+      "grad_norm": 0.15707510527519383,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 8551
+    },
+    {
+      "epoch": 0.5736324915316766,
+      "grad_norm": 0.15251535615294545,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 8552
+    },
+    {
+      "epoch": 0.5736995673609016,
+      "grad_norm": 0.16398521839653482,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 8553
+    },
+    {
+      "epoch": 0.5737666431901265,
+      "grad_norm": 0.15799944073792763,
+      "learning_rate": 2e-05,
+      "loss": 5.3714,
+      "step": 8554
+    },
+    {
+      "epoch": 0.5738337190193514,
+      "grad_norm": 0.15290206294637326,
+      "learning_rate": 2e-05,
+      "loss": 5.5313,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5739007948485764,
+      "grad_norm": 0.15040267859944667,
+      "learning_rate": 2e-05,
+      "loss": 5.4401,
+      "step": 8556
+    },
+    {
+      "epoch": 0.5739678706778013,
+      "grad_norm": 0.14681179804478442,
+      "learning_rate": 2e-05,
+      "loss": 5.3452,
+      "step": 8557
+    },
+    {
+      "epoch": 0.5740349465070262,
+      "grad_norm": 0.15055088134145153,
+      "learning_rate": 2e-05,
+      "loss": 5.3601,
+      "step": 8558
+    },
+    {
+      "epoch": 0.5741020223362512,
+      "grad_norm": 0.15029719716438483,
+      "learning_rate": 2e-05,
+      "loss": 5.405,
+      "step": 8559
+    },
+    {
+      "epoch": 0.5741690981654761,
+      "grad_norm": 0.16052055120401004,
+      "learning_rate": 2e-05,
+      "loss": 5.2887,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5742361739947011,
+      "grad_norm": 0.1508775473341432,
+      "learning_rate": 2e-05,
+      "loss": 5.5038,
+      "step": 8561
+    },
+    {
+      "epoch": 0.574303249823926,
+      "grad_norm": 0.16323111230011217,
+      "learning_rate": 2e-05,
+      "loss": 5.3105,
+      "step": 8562
+    },
+    {
+      "epoch": 0.5743703256531509,
+      "grad_norm": 0.15489066263149817,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 8563
+    },
+    {
+      "epoch": 0.5744374014823759,
+      "grad_norm": 0.15041064069152693,
+      "learning_rate": 2e-05,
+      "loss": 5.44,
+      "step": 8564
+    },
+    {
+      "epoch": 0.5745044773116008,
+      "grad_norm": 0.15730774688290047,
+      "learning_rate": 2e-05,
+      "loss": 5.2418,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5745715531408258,
+      "grad_norm": 0.1628971024497423,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 8566
+    },
+    {
+      "epoch": 0.5746386289700507,
+      "grad_norm": 0.15149898117403487,
+      "learning_rate": 2e-05,
+      "loss": 5.4767,
+      "step": 8567
+    },
+    {
+      "epoch": 0.5747057047992756,
+      "grad_norm": 0.14334772115318917,
+      "learning_rate": 2e-05,
+      "loss": 5.3136,
+      "step": 8568
+    },
+    {
+      "epoch": 0.5747727806285006,
+      "grad_norm": 0.15686673628737427,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 8569
+    },
+    {
+      "epoch": 0.5748398564577255,
+      "grad_norm": 0.15839883574390537,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5749069322869504,
+      "grad_norm": 0.16050051176774535,
+      "learning_rate": 2e-05,
+      "loss": 5.2628,
+      "step": 8571
+    },
+    {
+      "epoch": 0.5749740081161754,
+      "grad_norm": 0.15280334030375128,
+      "learning_rate": 2e-05,
+      "loss": 5.4925,
+      "step": 8572
+    },
+    {
+      "epoch": 0.5750410839454003,
+      "grad_norm": 0.1520709732564147,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 8573
+    },
+    {
+      "epoch": 0.5751081597746253,
+      "grad_norm": 0.14802965789023728,
+      "learning_rate": 2e-05,
+      "loss": 5.4911,
+      "step": 8574
+    },
+    {
+      "epoch": 0.5751752356038502,
+      "grad_norm": 0.14425334381879507,
+      "learning_rate": 2e-05,
+      "loss": 5.3619,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5752423114330751,
+      "grad_norm": 0.15255286720317432,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 8576
+    },
+    {
+      "epoch": 0.5753093872623001,
+      "grad_norm": 0.15321988213453622,
+      "learning_rate": 2e-05,
+      "loss": 5.453,
+      "step": 8577
+    },
+    {
+      "epoch": 0.575376463091525,
+      "grad_norm": 0.15833801092736124,
+      "learning_rate": 2e-05,
+      "loss": 5.4506,
+      "step": 8578
+    },
+    {
+      "epoch": 0.57544353892075,
+      "grad_norm": 0.15036446336810788,
+      "learning_rate": 2e-05,
+      "loss": 5.5276,
+      "step": 8579
+    },
+    {
+      "epoch": 0.5755106147499749,
+      "grad_norm": 0.15233910350584418,
+      "learning_rate": 2e-05,
+      "loss": 5.5283,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5755776905791998,
+      "grad_norm": 0.14175506173138822,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 8581
+    },
+    {
+      "epoch": 0.5756447664084248,
+      "grad_norm": 0.15527185601578114,
+      "learning_rate": 2e-05,
+      "loss": 5.356,
+      "step": 8582
+    },
+    {
+      "epoch": 0.5757118422376497,
+      "grad_norm": 0.14898327096475977,
+      "learning_rate": 2e-05,
+      "loss": 5.5281,
+      "step": 8583
+    },
+    {
+      "epoch": 0.5757789180668746,
+      "grad_norm": 0.14517031645244763,
+      "learning_rate": 2e-05,
+      "loss": 5.3346,
+      "step": 8584
+    },
+    {
+      "epoch": 0.5758459938960996,
+      "grad_norm": 0.1415778180833047,
+      "learning_rate": 2e-05,
+      "loss": 5.3058,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5759130697253245,
+      "grad_norm": 0.1531117015643341,
+      "learning_rate": 2e-05,
+      "loss": 5.3807,
+      "step": 8586
+    },
+    {
+      "epoch": 0.5759801455545495,
+      "grad_norm": 0.1461505099057545,
+      "learning_rate": 2e-05,
+      "loss": 5.4977,
+      "step": 8587
+    },
+    {
+      "epoch": 0.5760472213837744,
+      "grad_norm": 0.1537902322602564,
+      "learning_rate": 2e-05,
+      "loss": 5.4225,
+      "step": 8588
+    },
+    {
+      "epoch": 0.5761142972129993,
+      "grad_norm": 0.15028844356808027,
+      "learning_rate": 2e-05,
+      "loss": 5.4883,
+      "step": 8589
+    },
+    {
+      "epoch": 0.5761813730422243,
+      "grad_norm": 0.1444747808819393,
+      "learning_rate": 2e-05,
+      "loss": 5.596,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5762484488714492,
+      "grad_norm": 0.14569189852379952,
+      "learning_rate": 2e-05,
+      "loss": 5.3621,
+      "step": 8591
+    },
+    {
+      "epoch": 0.5763155247006742,
+      "grad_norm": 0.1524322260566053,
+      "learning_rate": 2e-05,
+      "loss": 5.5452,
+      "step": 8592
+    },
+    {
+      "epoch": 0.5763826005298991,
+      "grad_norm": 0.14779099342711913,
+      "learning_rate": 2e-05,
+      "loss": 5.4562,
+      "step": 8593
+    },
+    {
+      "epoch": 0.576449676359124,
+      "grad_norm": 0.15461821949930718,
+      "learning_rate": 2e-05,
+      "loss": 5.3905,
+      "step": 8594
+    },
+    {
+      "epoch": 0.576516752188349,
+      "grad_norm": 0.142557323648621,
+      "learning_rate": 2e-05,
+      "loss": 5.4335,
+      "step": 8595
+    },
+    {
+      "epoch": 0.5765838280175739,
+      "grad_norm": 0.15027204505906497,
+      "learning_rate": 2e-05,
+      "loss": 5.4049,
+      "step": 8596
+    },
+    {
+      "epoch": 0.5766509038467988,
+      "grad_norm": 0.13927610972580665,
+      "learning_rate": 2e-05,
+      "loss": 5.1896,
+      "step": 8597
+    },
+    {
+      "epoch": 0.5767179796760238,
+      "grad_norm": 0.15902615280669757,
+      "learning_rate": 2e-05,
+      "loss": 5.3393,
+      "step": 8598
+    },
+    {
+      "epoch": 0.5767850555052487,
+      "grad_norm": 0.15482513946240933,
+      "learning_rate": 2e-05,
+      "loss": 5.3666,
+      "step": 8599
+    },
+    {
+      "epoch": 0.5768521313344737,
+      "grad_norm": 0.14653756416063993,
+      "learning_rate": 2e-05,
+      "loss": 5.4787,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5769192071636986,
+      "grad_norm": 0.1433803862190507,
+      "learning_rate": 2e-05,
+      "loss": 5.4357,
+      "step": 8601
+    },
+    {
+      "epoch": 0.5769862829929235,
+      "grad_norm": 0.14490224691380985,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 8602
+    },
+    {
+      "epoch": 0.5770533588221485,
+      "grad_norm": 0.15282621982609265,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 8603
+    },
+    {
+      "epoch": 0.5771204346513734,
+      "grad_norm": 0.14984675740941827,
+      "learning_rate": 2e-05,
+      "loss": 5.5763,
+      "step": 8604
+    },
+    {
+      "epoch": 0.5771875104805984,
+      "grad_norm": 0.14949590911483107,
+      "learning_rate": 2e-05,
+      "loss": 5.4656,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5772545863098233,
+      "grad_norm": 0.1484272757111534,
+      "learning_rate": 2e-05,
+      "loss": 5.4043,
+      "step": 8606
+    },
+    {
+      "epoch": 0.5773216621390482,
+      "grad_norm": 0.14758357442180775,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 8607
+    },
+    {
+      "epoch": 0.5773887379682732,
+      "grad_norm": 0.15083880612986408,
+      "learning_rate": 2e-05,
+      "loss": 5.3378,
+      "step": 8608
+    },
+    {
+      "epoch": 0.5774558137974981,
+      "grad_norm": 0.14962234772763613,
+      "learning_rate": 2e-05,
+      "loss": 5.5276,
+      "step": 8609
+    },
+    {
+      "epoch": 0.577522889626723,
+      "grad_norm": 0.14619116676332924,
+      "learning_rate": 2e-05,
+      "loss": 5.5217,
+      "step": 8610
+    },
+    {
+      "epoch": 0.577589965455948,
+      "grad_norm": 0.14711263195515265,
+      "learning_rate": 2e-05,
+      "loss": 5.2833,
+      "step": 8611
+    },
+    {
+      "epoch": 0.5776570412851729,
+      "grad_norm": 0.16135554917101225,
+      "learning_rate": 2e-05,
+      "loss": 5.5664,
+      "step": 8612
+    },
+    {
+      "epoch": 0.5777241171143979,
+      "grad_norm": 0.15221389999030016,
+      "learning_rate": 2e-05,
+      "loss": 5.4288,
+      "step": 8613
+    },
+    {
+      "epoch": 0.5777911929436228,
+      "grad_norm": 0.15023349568311253,
+      "learning_rate": 2e-05,
+      "loss": 5.5613,
+      "step": 8614
+    },
+    {
+      "epoch": 0.5778582687728477,
+      "grad_norm": 0.14969010724494222,
+      "learning_rate": 2e-05,
+      "loss": 5.4634,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5779253446020727,
+      "grad_norm": 0.15772319366570378,
+      "learning_rate": 2e-05,
+      "loss": 5.5317,
+      "step": 8616
+    },
+    {
+      "epoch": 0.5779924204312976,
+      "grad_norm": 0.148003407220582,
+      "learning_rate": 2e-05,
+      "loss": 5.447,
+      "step": 8617
+    },
+    {
+      "epoch": 0.5780594962605226,
+      "grad_norm": 0.16058615096951714,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 8618
+    },
+    {
+      "epoch": 0.5781265720897475,
+      "grad_norm": 0.15766728898582819,
+      "learning_rate": 2e-05,
+      "loss": 5.3589,
+      "step": 8619
+    },
+    {
+      "epoch": 0.5781936479189724,
+      "grad_norm": 0.15602206491506523,
+      "learning_rate": 2e-05,
+      "loss": 5.323,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5782607237481974,
+      "grad_norm": 0.15691304385972746,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 8621
+    },
+    {
+      "epoch": 0.5783277995774223,
+      "grad_norm": 0.1478039285197623,
+      "learning_rate": 2e-05,
+      "loss": 5.2851,
+      "step": 8622
+    },
+    {
+      "epoch": 0.5783948754066472,
+      "grad_norm": 0.154952299368246,
+      "learning_rate": 2e-05,
+      "loss": 5.3887,
+      "step": 8623
+    },
+    {
+      "epoch": 0.5784619512358722,
+      "grad_norm": 0.15429174089777317,
+      "learning_rate": 2e-05,
+      "loss": 5.4761,
+      "step": 8624
+    },
+    {
+      "epoch": 0.5785290270650971,
+      "grad_norm": 0.1509883726878346,
+      "learning_rate": 2e-05,
+      "loss": 5.3181,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5785961028943221,
+      "grad_norm": 0.15099209334923214,
+      "learning_rate": 2e-05,
+      "loss": 5.5404,
+      "step": 8626
+    },
+    {
+      "epoch": 0.578663178723547,
+      "grad_norm": 0.14956864424670965,
+      "learning_rate": 2e-05,
+      "loss": 5.5031,
+      "step": 8627
+    },
+    {
+      "epoch": 0.5787302545527719,
+      "grad_norm": 0.155790562427099,
+      "learning_rate": 2e-05,
+      "loss": 5.5585,
+      "step": 8628
+    },
+    {
+      "epoch": 0.5787973303819969,
+      "grad_norm": 0.15434029704785007,
+      "learning_rate": 2e-05,
+      "loss": 5.3874,
+      "step": 8629
+    },
+    {
+      "epoch": 0.5788644062112218,
+      "grad_norm": 0.15200817713377052,
+      "learning_rate": 2e-05,
+      "loss": 5.4689,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5789314820404468,
+      "grad_norm": 0.15357500808042576,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 8631
+    },
+    {
+      "epoch": 0.5789985578696717,
+      "grad_norm": 0.14919901220009632,
+      "learning_rate": 2e-05,
+      "loss": 5.4885,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5790656336988966,
+      "grad_norm": 0.14974535375990275,
+      "learning_rate": 2e-05,
+      "loss": 5.3111,
+      "step": 8633
+    },
+    {
+      "epoch": 0.5791327095281216,
+      "grad_norm": 0.15833436303461257,
+      "learning_rate": 2e-05,
+      "loss": 5.507,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5791997853573465,
+      "grad_norm": 0.14707138972761813,
+      "learning_rate": 2e-05,
+      "loss": 5.4786,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5792668611865714,
+      "grad_norm": 0.15452260438981197,
+      "learning_rate": 2e-05,
+      "loss": 5.2219,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5793339370157964,
+      "grad_norm": 0.14935433747806964,
+      "learning_rate": 2e-05,
+      "loss": 5.5127,
+      "step": 8637
+    },
+    {
+      "epoch": 0.5794010128450213,
+      "grad_norm": 0.15464981549661022,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5794680886742463,
+      "grad_norm": 0.157334632264988,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 8639
+    },
+    {
+      "epoch": 0.5795351645034712,
+      "grad_norm": 0.1625576708103907,
+      "learning_rate": 2e-05,
+      "loss": 5.3488,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5796022403326961,
+      "grad_norm": 0.1517623277498487,
+      "learning_rate": 2e-05,
+      "loss": 5.5131,
+      "step": 8641
+    },
+    {
+      "epoch": 0.5796693161619211,
+      "grad_norm": 0.1466993544355269,
+      "learning_rate": 2e-05,
+      "loss": 5.3377,
+      "step": 8642
+    },
+    {
+      "epoch": 0.579736391991146,
+      "grad_norm": 0.14956118539136706,
+      "learning_rate": 2e-05,
+      "loss": 5.4379,
+      "step": 8643
+    },
+    {
+      "epoch": 0.579803467820371,
+      "grad_norm": 0.1539279684452843,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 8644
+    },
+    {
+      "epoch": 0.5798705436495959,
+      "grad_norm": 0.1495728035299672,
+      "learning_rate": 2e-05,
+      "loss": 5.5413,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5799376194788208,
+      "grad_norm": 0.15107788579180761,
+      "learning_rate": 2e-05,
+      "loss": 5.5205,
+      "step": 8646
+    },
+    {
+      "epoch": 0.5800046953080458,
+      "grad_norm": 0.1500061225590325,
+      "learning_rate": 2e-05,
+      "loss": 5.3582,
+      "step": 8647
+    },
+    {
+      "epoch": 0.5800717711372707,
+      "grad_norm": 0.15095223200443705,
+      "learning_rate": 2e-05,
+      "loss": 5.464,
+      "step": 8648
+    },
+    {
+      "epoch": 0.5801388469664956,
+      "grad_norm": 0.14775013286545224,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 8649
+    },
+    {
+      "epoch": 0.5802059227957206,
+      "grad_norm": 0.15089382311223734,
+      "learning_rate": 2e-05,
+      "loss": 5.3442,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5802729986249455,
+      "grad_norm": 0.15078686325396487,
+      "learning_rate": 2e-05,
+      "loss": 5.4344,
+      "step": 8651
+    },
+    {
+      "epoch": 0.5803400744541705,
+      "grad_norm": 0.15881997566108474,
+      "learning_rate": 2e-05,
+      "loss": 5.4802,
+      "step": 8652
+    },
+    {
+      "epoch": 0.5804071502833954,
+      "grad_norm": 0.1580852621934443,
+      "learning_rate": 2e-05,
+      "loss": 5.4006,
+      "step": 8653
+    },
+    {
+      "epoch": 0.5804742261126203,
+      "grad_norm": 0.15142093912325894,
+      "learning_rate": 2e-05,
+      "loss": 5.5089,
+      "step": 8654
+    },
+    {
+      "epoch": 0.5805413019418453,
+      "grad_norm": 0.15605577785628091,
+      "learning_rate": 2e-05,
+      "loss": 5.5202,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5806083777710702,
+      "grad_norm": 0.15867931007048944,
+      "learning_rate": 2e-05,
+      "loss": 5.473,
+      "step": 8656
+    },
+    {
+      "epoch": 0.5806754536002952,
+      "grad_norm": 0.14622167957139617,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 8657
+    },
+    {
+      "epoch": 0.5807425294295201,
+      "grad_norm": 0.1495081301178369,
+      "learning_rate": 2e-05,
+      "loss": 5.5081,
+      "step": 8658
+    },
+    {
+      "epoch": 0.580809605258745,
+      "grad_norm": 0.1449010424639129,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 8659
+    },
+    {
+      "epoch": 0.58087668108797,
+      "grad_norm": 0.14978753373818687,
+      "learning_rate": 2e-05,
+      "loss": 5.2951,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5809437569171949,
+      "grad_norm": 0.15037946526583185,
+      "learning_rate": 2e-05,
+      "loss": 5.4581,
+      "step": 8661
+    },
+    {
+      "epoch": 0.5810108327464198,
+      "grad_norm": 0.15034252101459084,
+      "learning_rate": 2e-05,
+      "loss": 5.3461,
+      "step": 8662
+    },
+    {
+      "epoch": 0.5810779085756448,
+      "grad_norm": 0.14746755649560025,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 8663
+    },
+    {
+      "epoch": 0.5811449844048697,
+      "grad_norm": 0.1515810716102731,
+      "learning_rate": 2e-05,
+      "loss": 5.4526,
+      "step": 8664
+    },
+    {
+      "epoch": 0.5812120602340947,
+      "grad_norm": 0.15246472060819383,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5812791360633196,
+      "grad_norm": 0.14619643246612837,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 8666
+    },
+    {
+      "epoch": 0.5813462118925445,
+      "grad_norm": 0.14463563724501308,
+      "learning_rate": 2e-05,
+      "loss": 5.4345,
+      "step": 8667
+    },
+    {
+      "epoch": 0.5814132877217695,
+      "grad_norm": 0.14967025330107098,
+      "learning_rate": 2e-05,
+      "loss": 5.3727,
+      "step": 8668
+    },
+    {
+      "epoch": 0.5814803635509944,
+      "grad_norm": 0.15323025678471655,
+      "learning_rate": 2e-05,
+      "loss": 5.3254,
+      "step": 8669
+    },
+    {
+      "epoch": 0.5815474393802194,
+      "grad_norm": 0.15488387855973437,
+      "learning_rate": 2e-05,
+      "loss": 5.4894,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5816145152094443,
+      "grad_norm": 0.15079064272231602,
+      "learning_rate": 2e-05,
+      "loss": 5.5202,
+      "step": 8671
+    },
+    {
+      "epoch": 0.5816815910386692,
+      "grad_norm": 0.1500582287344305,
+      "learning_rate": 2e-05,
+      "loss": 5.4252,
+      "step": 8672
+    },
+    {
+      "epoch": 0.5817486668678942,
+      "grad_norm": 0.15904504669681427,
+      "learning_rate": 2e-05,
+      "loss": 5.4369,
+      "step": 8673
+    },
+    {
+      "epoch": 0.5818157426971191,
+      "grad_norm": 0.1555620133684657,
+      "learning_rate": 2e-05,
+      "loss": 5.3578,
+      "step": 8674
+    },
+    {
+      "epoch": 0.581882818526344,
+      "grad_norm": 0.14649108300471184,
+      "learning_rate": 2e-05,
+      "loss": 5.512,
+      "step": 8675
+    },
+    {
+      "epoch": 0.581949894355569,
+      "grad_norm": 0.14941318925975142,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 8676
+    },
+    {
+      "epoch": 0.5820169701847939,
+      "grad_norm": 0.1566208517127794,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 8677
+    },
+    {
+      "epoch": 0.5820840460140189,
+      "grad_norm": 0.15297291471239258,
+      "learning_rate": 2e-05,
+      "loss": 5.4503,
+      "step": 8678
+    },
+    {
+      "epoch": 0.5821511218432438,
+      "grad_norm": 0.1482053595166656,
+      "learning_rate": 2e-05,
+      "loss": 5.3572,
+      "step": 8679
+    },
+    {
+      "epoch": 0.5822181976724687,
+      "grad_norm": 0.16247910154933398,
+      "learning_rate": 2e-05,
+      "loss": 5.3421,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5822852735016937,
+      "grad_norm": 0.16278085119046273,
+      "learning_rate": 2e-05,
+      "loss": 5.4595,
+      "step": 8681
+    },
+    {
+      "epoch": 0.5823523493309186,
+      "grad_norm": 0.14970950342309064,
+      "learning_rate": 2e-05,
+      "loss": 5.5275,
+      "step": 8682
+    },
+    {
+      "epoch": 0.5824194251601436,
+      "grad_norm": 0.15248440455876264,
+      "learning_rate": 2e-05,
+      "loss": 5.3943,
+      "step": 8683
+    },
+    {
+      "epoch": 0.5824865009893685,
+      "grad_norm": 0.14877721359589688,
+      "learning_rate": 2e-05,
+      "loss": 5.4713,
+      "step": 8684
+    },
+    {
+      "epoch": 0.5825535768185934,
+      "grad_norm": 0.14670771232136393,
+      "learning_rate": 2e-05,
+      "loss": 5.3215,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5826206526478184,
+      "grad_norm": 0.15845894577513872,
+      "learning_rate": 2e-05,
+      "loss": 5.2497,
+      "step": 8686
+    },
+    {
+      "epoch": 0.5826877284770433,
+      "grad_norm": 0.1493083590699222,
+      "learning_rate": 2e-05,
+      "loss": 5.6382,
+      "step": 8687
+    },
+    {
+      "epoch": 0.5827548043062682,
+      "grad_norm": 0.15163362557220086,
+      "learning_rate": 2e-05,
+      "loss": 5.3369,
+      "step": 8688
+    },
+    {
+      "epoch": 0.5828218801354932,
+      "grad_norm": 0.14690026845085968,
+      "learning_rate": 2e-05,
+      "loss": 5.3159,
+      "step": 8689
+    },
+    {
+      "epoch": 0.5828889559647181,
+      "grad_norm": 0.15041990751849008,
+      "learning_rate": 2e-05,
+      "loss": 5.3997,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5829560317939431,
+      "grad_norm": 0.14637649412162704,
+      "learning_rate": 2e-05,
+      "loss": 5.3079,
+      "step": 8691
+    },
+    {
+      "epoch": 0.583023107623168,
+      "grad_norm": 0.15368592469053938,
+      "learning_rate": 2e-05,
+      "loss": 5.4518,
+      "step": 8692
+    },
+    {
+      "epoch": 0.5830901834523929,
+      "grad_norm": 0.15008431725827764,
+      "learning_rate": 2e-05,
+      "loss": 5.4894,
+      "step": 8693
+    },
+    {
+      "epoch": 0.5831572592816179,
+      "grad_norm": 0.14811427073337313,
+      "learning_rate": 2e-05,
+      "loss": 5.3398,
+      "step": 8694
+    },
+    {
+      "epoch": 0.5832243351108428,
+      "grad_norm": 0.15298513137330197,
+      "learning_rate": 2e-05,
+      "loss": 5.4544,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5832914109400678,
+      "grad_norm": 0.15913387367719692,
+      "learning_rate": 2e-05,
+      "loss": 5.3372,
+      "step": 8696
+    },
+    {
+      "epoch": 0.5833584867692927,
+      "grad_norm": 0.14905406864271634,
+      "learning_rate": 2e-05,
+      "loss": 5.3861,
+      "step": 8697
+    },
+    {
+      "epoch": 0.5834255625985176,
+      "grad_norm": 0.14702890394814533,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 8698
+    },
+    {
+      "epoch": 0.5834926384277426,
+      "grad_norm": 0.14747104948148848,
+      "learning_rate": 2e-05,
+      "loss": 5.4621,
+      "step": 8699
+    },
+    {
+      "epoch": 0.5835597142569675,
+      "grad_norm": 0.14838175103241177,
+      "learning_rate": 2e-05,
+      "loss": 5.3451,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5836267900861924,
+      "grad_norm": 0.16273353203939042,
+      "learning_rate": 2e-05,
+      "loss": 5.3402,
+      "step": 8701
+    },
+    {
+      "epoch": 0.5836938659154174,
+      "grad_norm": 0.1703747235686989,
+      "learning_rate": 2e-05,
+      "loss": 5.4204,
+      "step": 8702
+    },
+    {
+      "epoch": 0.5837609417446423,
+      "grad_norm": 0.14762153161104624,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 8703
+    },
+    {
+      "epoch": 0.5838280175738673,
+      "grad_norm": 0.15304392382718993,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 8704
+    },
+    {
+      "epoch": 0.5838950934030922,
+      "grad_norm": 0.1649276109848706,
+      "learning_rate": 2e-05,
+      "loss": 5.3596,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5839621692323171,
+      "grad_norm": 0.1492339953187639,
+      "learning_rate": 2e-05,
+      "loss": 5.2345,
+      "step": 8706
+    },
+    {
+      "epoch": 0.5840292450615421,
+      "grad_norm": 0.14827794142748124,
+      "learning_rate": 2e-05,
+      "loss": 5.4714,
+      "step": 8707
+    },
+    {
+      "epoch": 0.584096320890767,
+      "grad_norm": 0.15557255666461126,
+      "learning_rate": 2e-05,
+      "loss": 5.3502,
+      "step": 8708
+    },
+    {
+      "epoch": 0.584163396719992,
+      "grad_norm": 0.15878419621344114,
+      "learning_rate": 2e-05,
+      "loss": 5.4628,
+      "step": 8709
+    },
+    {
+      "epoch": 0.5842304725492169,
+      "grad_norm": 0.14632259881723292,
+      "learning_rate": 2e-05,
+      "loss": 5.419,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5842975483784418,
+      "grad_norm": 0.15554312893471306,
+      "learning_rate": 2e-05,
+      "loss": 5.5588,
+      "step": 8711
+    },
+    {
+      "epoch": 0.5843646242076668,
+      "grad_norm": 0.15573423450809643,
+      "learning_rate": 2e-05,
+      "loss": 5.4488,
+      "step": 8712
+    },
+    {
+      "epoch": 0.5844317000368917,
+      "grad_norm": 0.14899995973325983,
+      "learning_rate": 2e-05,
+      "loss": 5.6146,
+      "step": 8713
+    },
+    {
+      "epoch": 0.5844987758661166,
+      "grad_norm": 0.14768121962819405,
+      "learning_rate": 2e-05,
+      "loss": 5.3994,
+      "step": 8714
+    },
+    {
+      "epoch": 0.5845658516953416,
+      "grad_norm": 0.1478431721826911,
+      "learning_rate": 2e-05,
+      "loss": 5.3581,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5846329275245665,
+      "grad_norm": 0.15806888094975308,
+      "learning_rate": 2e-05,
+      "loss": 5.3943,
+      "step": 8716
+    },
+    {
+      "epoch": 0.5847000033537915,
+      "grad_norm": 0.1543160938434998,
+      "learning_rate": 2e-05,
+      "loss": 5.2983,
+      "step": 8717
+    },
+    {
+      "epoch": 0.5847670791830164,
+      "grad_norm": 0.15163897528374848,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 8718
+    },
+    {
+      "epoch": 0.5848341550122413,
+      "grad_norm": 0.1509395297787899,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 8719
+    },
+    {
+      "epoch": 0.5849012308414663,
+      "grad_norm": 0.14345263874149802,
+      "learning_rate": 2e-05,
+      "loss": 5.5621,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5849683066706912,
+      "grad_norm": 0.15016928341448763,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 8721
+    },
+    {
+      "epoch": 0.5850353824999162,
+      "grad_norm": 0.1498635382124269,
+      "learning_rate": 2e-05,
+      "loss": 5.4653,
+      "step": 8722
+    },
+    {
+      "epoch": 0.5851024583291411,
+      "grad_norm": 0.1466433415935404,
+      "learning_rate": 2e-05,
+      "loss": 5.5192,
+      "step": 8723
+    },
+    {
+      "epoch": 0.585169534158366,
+      "grad_norm": 0.1533085633863625,
+      "learning_rate": 2e-05,
+      "loss": 5.4353,
+      "step": 8724
+    },
+    {
+      "epoch": 0.585236609987591,
+      "grad_norm": 0.1476152400685941,
+      "learning_rate": 2e-05,
+      "loss": 5.3866,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5853036858168159,
+      "grad_norm": 0.14277235124397147,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 8726
+    },
+    {
+      "epoch": 0.5853707616460408,
+      "grad_norm": 0.1464658126254063,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 8727
+    },
+    {
+      "epoch": 0.5854378374752658,
+      "grad_norm": 0.15652674781349785,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 8728
+    },
+    {
+      "epoch": 0.5855049133044907,
+      "grad_norm": 0.15785700928576116,
+      "learning_rate": 2e-05,
+      "loss": 5.4652,
+      "step": 8729
+    },
+    {
+      "epoch": 0.5855719891337157,
+      "grad_norm": 0.15586874209524912,
+      "learning_rate": 2e-05,
+      "loss": 5.4796,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5856390649629406,
+      "grad_norm": 0.14574553079645114,
+      "learning_rate": 2e-05,
+      "loss": 5.3797,
+      "step": 8731
+    },
+    {
+      "epoch": 0.5857061407921655,
+      "grad_norm": 0.15099975368312132,
+      "learning_rate": 2e-05,
+      "loss": 5.59,
+      "step": 8732
+    },
+    {
+      "epoch": 0.5857732166213905,
+      "grad_norm": 0.14902805613316478,
+      "learning_rate": 2e-05,
+      "loss": 5.3218,
+      "step": 8733
+    },
+    {
+      "epoch": 0.5858402924506154,
+      "grad_norm": 0.1661730264304653,
+      "learning_rate": 2e-05,
+      "loss": 5.4354,
+      "step": 8734
+    },
+    {
+      "epoch": 0.5859073682798404,
+      "grad_norm": 0.1508564499564276,
+      "learning_rate": 2e-05,
+      "loss": 5.4786,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5859744441090653,
+      "grad_norm": 0.14311010699980223,
+      "learning_rate": 2e-05,
+      "loss": 5.3874,
+      "step": 8736
+    },
+    {
+      "epoch": 0.5860415199382902,
+      "grad_norm": 0.14904420789835232,
+      "learning_rate": 2e-05,
+      "loss": 5.5228,
+      "step": 8737
+    },
+    {
+      "epoch": 0.5861085957675152,
+      "grad_norm": 0.1556357334506694,
+      "learning_rate": 2e-05,
+      "loss": 5.3244,
+      "step": 8738
+    },
+    {
+      "epoch": 0.5861756715967401,
+      "grad_norm": 0.15140347221123093,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 8739
+    },
+    {
+      "epoch": 0.586242747425965,
+      "grad_norm": 0.15320729691699353,
+      "learning_rate": 2e-05,
+      "loss": 5.3884,
+      "step": 8740
+    },
+    {
+      "epoch": 0.58630982325519,
+      "grad_norm": 0.14548429643239827,
+      "learning_rate": 2e-05,
+      "loss": 5.3524,
+      "step": 8741
+    },
+    {
+      "epoch": 0.5863768990844149,
+      "grad_norm": 0.15058650940862708,
+      "learning_rate": 2e-05,
+      "loss": 5.4926,
+      "step": 8742
+    },
+    {
+      "epoch": 0.5864439749136399,
+      "grad_norm": 0.15358892042143654,
+      "learning_rate": 2e-05,
+      "loss": 5.4409,
+      "step": 8743
+    },
+    {
+      "epoch": 0.5865110507428648,
+      "grad_norm": 0.1602322171621314,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 8744
+    },
+    {
+      "epoch": 0.5865781265720897,
+      "grad_norm": 0.1525516449533458,
+      "learning_rate": 2e-05,
+      "loss": 5.3323,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5866452024013147,
+      "grad_norm": 0.1420611975441344,
+      "learning_rate": 2e-05,
+      "loss": 5.3502,
+      "step": 8746
+    },
+    {
+      "epoch": 0.5867122782305396,
+      "grad_norm": 0.15317911303668252,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 8747
+    },
+    {
+      "epoch": 0.5867793540597646,
+      "grad_norm": 0.1463971184920436,
+      "learning_rate": 2e-05,
+      "loss": 5.4125,
+      "step": 8748
+    },
+    {
+      "epoch": 0.5868464298889895,
+      "grad_norm": 0.15568549715294291,
+      "learning_rate": 2e-05,
+      "loss": 5.543,
+      "step": 8749
+    },
+    {
+      "epoch": 0.5869135057182144,
+      "grad_norm": 0.16221370647917868,
+      "learning_rate": 2e-05,
+      "loss": 5.333,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5869805815474394,
+      "grad_norm": 0.14504002772650904,
+      "learning_rate": 2e-05,
+      "loss": 5.5276,
+      "step": 8751
+    },
+    {
+      "epoch": 0.5870476573766643,
+      "grad_norm": 0.1505336552532555,
+      "learning_rate": 2e-05,
+      "loss": 5.3534,
+      "step": 8752
+    },
+    {
+      "epoch": 0.5871147332058892,
+      "grad_norm": 0.15831358197969014,
+      "learning_rate": 2e-05,
+      "loss": 5.4686,
+      "step": 8753
+    },
+    {
+      "epoch": 0.5871818090351142,
+      "grad_norm": 0.1508450940589923,
+      "learning_rate": 2e-05,
+      "loss": 5.4349,
+      "step": 8754
+    },
+    {
+      "epoch": 0.5872488848643391,
+      "grad_norm": 0.14435226319272487,
+      "learning_rate": 2e-05,
+      "loss": 5.4297,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5873159606935641,
+      "grad_norm": 0.14693773458207238,
+      "learning_rate": 2e-05,
+      "loss": 5.3976,
+      "step": 8756
+    },
+    {
+      "epoch": 0.587383036522789,
+      "grad_norm": 0.14880970374627545,
+      "learning_rate": 2e-05,
+      "loss": 5.316,
+      "step": 8757
+    },
+    {
+      "epoch": 0.5874501123520139,
+      "grad_norm": 0.15068365364221886,
+      "learning_rate": 2e-05,
+      "loss": 5.5078,
+      "step": 8758
+    },
+    {
+      "epoch": 0.5875171881812389,
+      "grad_norm": 0.1473978343011179,
+      "learning_rate": 2e-05,
+      "loss": 5.5645,
+      "step": 8759
+    },
+    {
+      "epoch": 0.5875842640104638,
+      "grad_norm": 0.1491389426823022,
+      "learning_rate": 2e-05,
+      "loss": 5.4446,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5876513398396888,
+      "grad_norm": 0.15211512279357692,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 8761
+    },
+    {
+      "epoch": 0.5877184156689137,
+      "grad_norm": 0.1528041917664228,
+      "learning_rate": 2e-05,
+      "loss": 5.2886,
+      "step": 8762
+    },
+    {
+      "epoch": 0.5877854914981386,
+      "grad_norm": 0.16073926672025496,
+      "learning_rate": 2e-05,
+      "loss": 5.4855,
+      "step": 8763
+    },
+    {
+      "epoch": 0.5878525673273636,
+      "grad_norm": 0.14880235881383952,
+      "learning_rate": 2e-05,
+      "loss": 5.3312,
+      "step": 8764
+    },
+    {
+      "epoch": 0.5879196431565885,
+      "grad_norm": 0.16835233979054567,
+      "learning_rate": 2e-05,
+      "loss": 5.4955,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5879867189858135,
+      "grad_norm": 0.16639674098825077,
+      "learning_rate": 2e-05,
+      "loss": 5.4484,
+      "step": 8766
+    },
+    {
+      "epoch": 0.5880537948150384,
+      "grad_norm": 0.1538497668588462,
+      "learning_rate": 2e-05,
+      "loss": 5.575,
+      "step": 8767
+    },
+    {
+      "epoch": 0.5881208706442633,
+      "grad_norm": 0.15890129546206222,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 8768
+    },
+    {
+      "epoch": 0.5881879464734883,
+      "grad_norm": 0.17211123174197812,
+      "learning_rate": 2e-05,
+      "loss": 5.4138,
+      "step": 8769
+    },
+    {
+      "epoch": 0.5882550223027132,
+      "grad_norm": 0.15443322983470872,
+      "learning_rate": 2e-05,
+      "loss": 5.3607,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5883220981319381,
+      "grad_norm": 0.15251224409526148,
+      "learning_rate": 2e-05,
+      "loss": 5.4108,
+      "step": 8771
+    },
+    {
+      "epoch": 0.5883891739611631,
+      "grad_norm": 0.15057479568942625,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 8772
+    },
+    {
+      "epoch": 0.588456249790388,
+      "grad_norm": 0.15578061987911376,
+      "learning_rate": 2e-05,
+      "loss": 5.3484,
+      "step": 8773
+    },
+    {
+      "epoch": 0.588523325619613,
+      "grad_norm": 0.1574270684163253,
+      "learning_rate": 2e-05,
+      "loss": 5.4587,
+      "step": 8774
+    },
+    {
+      "epoch": 0.5885904014488379,
+      "grad_norm": 0.16912723265986712,
+      "learning_rate": 2e-05,
+      "loss": 5.4822,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5886574772780628,
+      "grad_norm": 0.15426908464973407,
+      "learning_rate": 2e-05,
+      "loss": 5.393,
+      "step": 8776
+    },
+    {
+      "epoch": 0.5887245531072878,
+      "grad_norm": 0.1564141620867365,
+      "learning_rate": 2e-05,
+      "loss": 5.4802,
+      "step": 8777
+    },
+    {
+      "epoch": 0.5887916289365127,
+      "grad_norm": 0.151244797757271,
+      "learning_rate": 2e-05,
+      "loss": 5.3199,
+      "step": 8778
+    },
+    {
+      "epoch": 0.5888587047657377,
+      "grad_norm": 0.15100083114036214,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 8779
+    },
+    {
+      "epoch": 0.5889257805949626,
+      "grad_norm": 0.15086341577375117,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5889928564241875,
+      "grad_norm": 0.15337600486921457,
+      "learning_rate": 2e-05,
+      "loss": 5.529,
+      "step": 8781
+    },
+    {
+      "epoch": 0.5890599322534125,
+      "grad_norm": 0.15275766290777798,
+      "learning_rate": 2e-05,
+      "loss": 5.5029,
+      "step": 8782
+    },
+    {
+      "epoch": 0.5891270080826374,
+      "grad_norm": 0.15633750940380195,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 8783
+    },
+    {
+      "epoch": 0.5891940839118623,
+      "grad_norm": 0.15456946383232334,
+      "learning_rate": 2e-05,
+      "loss": 5.3549,
+      "step": 8784
+    },
+    {
+      "epoch": 0.5892611597410873,
+      "grad_norm": 0.1505622510952308,
+      "learning_rate": 2e-05,
+      "loss": 5.4961,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5893282355703122,
+      "grad_norm": 0.14800569858307644,
+      "learning_rate": 2e-05,
+      "loss": 5.5091,
+      "step": 8786
+    },
+    {
+      "epoch": 0.5893953113995372,
+      "grad_norm": 0.15303397532274587,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 8787
+    },
+    {
+      "epoch": 0.5894623872287621,
+      "grad_norm": 0.14953130495786032,
+      "learning_rate": 2e-05,
+      "loss": 5.3636,
+      "step": 8788
+    },
+    {
+      "epoch": 0.589529463057987,
+      "grad_norm": 0.150699678709202,
+      "learning_rate": 2e-05,
+      "loss": 5.5591,
+      "step": 8789
+    },
+    {
+      "epoch": 0.589596538887212,
+      "grad_norm": 0.15293690226004542,
+      "learning_rate": 2e-05,
+      "loss": 5.2997,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5896636147164369,
+      "grad_norm": 0.15376145945413505,
+      "learning_rate": 2e-05,
+      "loss": 5.4271,
+      "step": 8791
+    },
+    {
+      "epoch": 0.5897306905456619,
+      "grad_norm": 0.15453766894650206,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 8792
+    },
+    {
+      "epoch": 0.5897977663748868,
+      "grad_norm": 0.15467264753043383,
+      "learning_rate": 2e-05,
+      "loss": 5.3035,
+      "step": 8793
+    },
+    {
+      "epoch": 0.5898648422041117,
+      "grad_norm": 0.1601982612780676,
+      "learning_rate": 2e-05,
+      "loss": 5.364,
+      "step": 8794
+    },
+    {
+      "epoch": 0.5899319180333367,
+      "grad_norm": 0.14701155877034677,
+      "learning_rate": 2e-05,
+      "loss": 5.4126,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5899989938625616,
+      "grad_norm": 0.1582543274108118,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 8796
+    },
+    {
+      "epoch": 0.5900660696917865,
+      "grad_norm": 0.15092202583774714,
+      "learning_rate": 2e-05,
+      "loss": 5.4175,
+      "step": 8797
+    },
+    {
+      "epoch": 0.5901331455210115,
+      "grad_norm": 0.15008316999657986,
+      "learning_rate": 2e-05,
+      "loss": 5.371,
+      "step": 8798
+    },
+    {
+      "epoch": 0.5902002213502364,
+      "grad_norm": 0.1488604708816072,
+      "learning_rate": 2e-05,
+      "loss": 5.3322,
+      "step": 8799
+    },
+    {
+      "epoch": 0.5902672971794614,
+      "grad_norm": 0.1455049610103134,
+      "learning_rate": 2e-05,
+      "loss": 5.4288,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5903343730086863,
+      "grad_norm": 0.15095212032579314,
+      "learning_rate": 2e-05,
+      "loss": 5.3333,
+      "step": 8801
+    },
+    {
+      "epoch": 0.5904014488379112,
+      "grad_norm": 0.15006345416432376,
+      "learning_rate": 2e-05,
+      "loss": 5.2272,
+      "step": 8802
+    },
+    {
+      "epoch": 0.5904685246671362,
+      "grad_norm": 0.15471554698154746,
+      "learning_rate": 2e-05,
+      "loss": 5.4045,
+      "step": 8803
+    },
+    {
+      "epoch": 0.5905356004963611,
+      "grad_norm": 0.15882903950296492,
+      "learning_rate": 2e-05,
+      "loss": 5.5098,
+      "step": 8804
+    },
+    {
+      "epoch": 0.590602676325586,
+      "grad_norm": 0.1502529446843442,
+      "learning_rate": 2e-05,
+      "loss": 5.466,
+      "step": 8805
+    },
+    {
+      "epoch": 0.590669752154811,
+      "grad_norm": 0.15422559697711233,
+      "learning_rate": 2e-05,
+      "loss": 5.3502,
+      "step": 8806
+    },
+    {
+      "epoch": 0.5907368279840359,
+      "grad_norm": 0.1466931563386025,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 8807
+    },
+    {
+      "epoch": 0.5908039038132609,
+      "grad_norm": 0.1463584303765083,
+      "learning_rate": 2e-05,
+      "loss": 5.5239,
+      "step": 8808
+    },
+    {
+      "epoch": 0.5908709796424858,
+      "grad_norm": 0.1474392999956731,
+      "learning_rate": 2e-05,
+      "loss": 5.4275,
+      "step": 8809
+    },
+    {
+      "epoch": 0.5909380554717107,
+      "grad_norm": 0.14681279975545108,
+      "learning_rate": 2e-05,
+      "loss": 5.4834,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5910051313009357,
+      "grad_norm": 0.14878097057856787,
+      "learning_rate": 2e-05,
+      "loss": 5.4846,
+      "step": 8811
+    },
+    {
+      "epoch": 0.5910722071301606,
+      "grad_norm": 0.14755433805499688,
+      "learning_rate": 2e-05,
+      "loss": 5.3902,
+      "step": 8812
+    },
+    {
+      "epoch": 0.5911392829593856,
+      "grad_norm": 0.15002076284300978,
+      "learning_rate": 2e-05,
+      "loss": 5.4606,
+      "step": 8813
+    },
+    {
+      "epoch": 0.5912063587886105,
+      "grad_norm": 0.16097257801349588,
+      "learning_rate": 2e-05,
+      "loss": 5.5569,
+      "step": 8814
+    },
+    {
+      "epoch": 0.5912734346178354,
+      "grad_norm": 0.15143243367999465,
+      "learning_rate": 2e-05,
+      "loss": 5.4039,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5913405104470604,
+      "grad_norm": 0.14902178212856623,
+      "learning_rate": 2e-05,
+      "loss": 5.4693,
+      "step": 8816
+    },
+    {
+      "epoch": 0.5914075862762853,
+      "grad_norm": 0.15303101804086944,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 8817
+    },
+    {
+      "epoch": 0.5914746621055103,
+      "grad_norm": 0.14822877066437226,
+      "learning_rate": 2e-05,
+      "loss": 5.371,
+      "step": 8818
+    },
+    {
+      "epoch": 0.5915417379347352,
+      "grad_norm": 0.14620393842620572,
+      "learning_rate": 2e-05,
+      "loss": 5.5829,
+      "step": 8819
+    },
+    {
+      "epoch": 0.5916088137639601,
+      "grad_norm": 0.15042747283183533,
+      "learning_rate": 2e-05,
+      "loss": 5.5114,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5916758895931851,
+      "grad_norm": 0.14848912753475052,
+      "learning_rate": 2e-05,
+      "loss": 5.3085,
+      "step": 8821
+    },
+    {
+      "epoch": 0.59174296542241,
+      "grad_norm": 0.1558023135492045,
+      "learning_rate": 2e-05,
+      "loss": 5.344,
+      "step": 8822
+    },
+    {
+      "epoch": 0.5918100412516349,
+      "grad_norm": 0.15142136410637244,
+      "learning_rate": 2e-05,
+      "loss": 5.5198,
+      "step": 8823
+    },
+    {
+      "epoch": 0.5918771170808599,
+      "grad_norm": 0.14378647422472746,
+      "learning_rate": 2e-05,
+      "loss": 5.4238,
+      "step": 8824
+    },
+    {
+      "epoch": 0.5919441929100848,
+      "grad_norm": 0.14477620586075624,
+      "learning_rate": 2e-05,
+      "loss": 5.3464,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5920112687393098,
+      "grad_norm": 0.15062018799494054,
+      "learning_rate": 2e-05,
+      "loss": 5.3695,
+      "step": 8826
+    },
+    {
+      "epoch": 0.5920783445685347,
+      "grad_norm": 0.1472723782177999,
+      "learning_rate": 2e-05,
+      "loss": 5.2854,
+      "step": 8827
+    },
+    {
+      "epoch": 0.5921454203977596,
+      "grad_norm": 0.15590281269151698,
+      "learning_rate": 2e-05,
+      "loss": 5.3876,
+      "step": 8828
+    },
+    {
+      "epoch": 0.5922124962269846,
+      "grad_norm": 0.16134416191514084,
+      "learning_rate": 2e-05,
+      "loss": 5.4697,
+      "step": 8829
+    },
+    {
+      "epoch": 0.5922795720562095,
+      "grad_norm": 0.1497795154290064,
+      "learning_rate": 2e-05,
+      "loss": 5.3853,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5923466478854345,
+      "grad_norm": 0.14519322025989478,
+      "learning_rate": 2e-05,
+      "loss": 5.5394,
+      "step": 8831
+    },
+    {
+      "epoch": 0.5924137237146594,
+      "grad_norm": 0.14891145226364666,
+      "learning_rate": 2e-05,
+      "loss": 5.4205,
+      "step": 8832
+    },
+    {
+      "epoch": 0.5924807995438843,
+      "grad_norm": 0.1449104003968067,
+      "learning_rate": 2e-05,
+      "loss": 5.524,
+      "step": 8833
+    },
+    {
+      "epoch": 0.5925478753731093,
+      "grad_norm": 0.15972137450439053,
+      "learning_rate": 2e-05,
+      "loss": 5.2747,
+      "step": 8834
+    },
+    {
+      "epoch": 0.5926149512023342,
+      "grad_norm": 0.14647847425655983,
+      "learning_rate": 2e-05,
+      "loss": 5.3486,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5926820270315591,
+      "grad_norm": 0.14692885704065697,
+      "learning_rate": 2e-05,
+      "loss": 5.3099,
+      "step": 8836
+    },
+    {
+      "epoch": 0.5927491028607841,
+      "grad_norm": 0.1466938688729281,
+      "learning_rate": 2e-05,
+      "loss": 5.3028,
+      "step": 8837
+    },
+    {
+      "epoch": 0.592816178690009,
+      "grad_norm": 0.15140888071202277,
+      "learning_rate": 2e-05,
+      "loss": 5.4157,
+      "step": 8838
+    },
+    {
+      "epoch": 0.592883254519234,
+      "grad_norm": 0.14976122617388915,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 8839
+    },
+    {
+      "epoch": 0.5929503303484589,
+      "grad_norm": 0.1463176207441066,
+      "learning_rate": 2e-05,
+      "loss": 5.4584,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5930174061776838,
+      "grad_norm": 0.15123461824090273,
+      "learning_rate": 2e-05,
+      "loss": 5.3717,
+      "step": 8841
+    },
+    {
+      "epoch": 0.5930844820069088,
+      "grad_norm": 0.15420596881265727,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 8842
+    },
+    {
+      "epoch": 0.5931515578361337,
+      "grad_norm": 0.15191458839036573,
+      "learning_rate": 2e-05,
+      "loss": 5.5254,
+      "step": 8843
+    },
+    {
+      "epoch": 0.5932186336653587,
+      "grad_norm": 0.1584990979277572,
+      "learning_rate": 2e-05,
+      "loss": 5.4646,
+      "step": 8844
+    },
+    {
+      "epoch": 0.5932857094945836,
+      "grad_norm": 0.15741512259944881,
+      "learning_rate": 2e-05,
+      "loss": 5.3565,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5933527853238085,
+      "grad_norm": 0.15129449148215446,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 8846
+    },
+    {
+      "epoch": 0.5934198611530335,
+      "grad_norm": 0.1536966984038622,
+      "learning_rate": 2e-05,
+      "loss": 5.3251,
+      "step": 8847
+    },
+    {
+      "epoch": 0.5934869369822584,
+      "grad_norm": 0.1512710435396921,
+      "learning_rate": 2e-05,
+      "loss": 5.468,
+      "step": 8848
+    },
+    {
+      "epoch": 0.5935540128114833,
+      "grad_norm": 0.14920483829222977,
+      "learning_rate": 2e-05,
+      "loss": 5.4675,
+      "step": 8849
+    },
+    {
+      "epoch": 0.5936210886407083,
+      "grad_norm": 0.14694906259879723,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5936881644699332,
+      "grad_norm": 0.15723538333804743,
+      "learning_rate": 2e-05,
+      "loss": 5.5063,
+      "step": 8851
+    },
+    {
+      "epoch": 0.5937552402991582,
+      "grad_norm": 0.15494196955677642,
+      "learning_rate": 2e-05,
+      "loss": 5.271,
+      "step": 8852
+    },
+    {
+      "epoch": 0.5938223161283831,
+      "grad_norm": 0.1422266801981949,
+      "learning_rate": 2e-05,
+      "loss": 5.4324,
+      "step": 8853
+    },
+    {
+      "epoch": 0.593889391957608,
+      "grad_norm": 0.15197762811219703,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 8854
+    },
+    {
+      "epoch": 0.593956467786833,
+      "grad_norm": 0.15494950475552513,
+      "learning_rate": 2e-05,
+      "loss": 5.4748,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5940235436160579,
+      "grad_norm": 0.14821978094517224,
+      "learning_rate": 2e-05,
+      "loss": 5.4709,
+      "step": 8856
+    },
+    {
+      "epoch": 0.5940906194452829,
+      "grad_norm": 0.15077508705203266,
+      "learning_rate": 2e-05,
+      "loss": 5.3598,
+      "step": 8857
+    },
+    {
+      "epoch": 0.5941576952745078,
+      "grad_norm": 0.1565816179765634,
+      "learning_rate": 2e-05,
+      "loss": 5.5734,
+      "step": 8858
+    },
+    {
+      "epoch": 0.5942247711037327,
+      "grad_norm": 0.15365586134073278,
+      "learning_rate": 2e-05,
+      "loss": 5.4582,
+      "step": 8859
+    },
+    {
+      "epoch": 0.5942918469329577,
+      "grad_norm": 0.15048139359099189,
+      "learning_rate": 2e-05,
+      "loss": 5.3969,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5943589227621826,
+      "grad_norm": 0.150811568999591,
+      "learning_rate": 2e-05,
+      "loss": 5.3276,
+      "step": 8861
+    },
+    {
+      "epoch": 0.5944259985914075,
+      "grad_norm": 0.15092886287780755,
+      "learning_rate": 2e-05,
+      "loss": 5.4561,
+      "step": 8862
+    },
+    {
+      "epoch": 0.5944930744206325,
+      "grad_norm": 0.14549703481940948,
+      "learning_rate": 2e-05,
+      "loss": 5.4942,
+      "step": 8863
+    },
+    {
+      "epoch": 0.5945601502498574,
+      "grad_norm": 0.14336584877253794,
+      "learning_rate": 2e-05,
+      "loss": 5.378,
+      "step": 8864
+    },
+    {
+      "epoch": 0.5946272260790824,
+      "grad_norm": 0.15517953947836868,
+      "learning_rate": 2e-05,
+      "loss": 5.4235,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5946943019083073,
+      "grad_norm": 0.1428738626246211,
+      "learning_rate": 2e-05,
+      "loss": 5.3731,
+      "step": 8866
+    },
+    {
+      "epoch": 0.5947613777375322,
+      "grad_norm": 0.146434521644361,
+      "learning_rate": 2e-05,
+      "loss": 5.2707,
+      "step": 8867
+    },
+    {
+      "epoch": 0.5948284535667572,
+      "grad_norm": 0.14955164488681366,
+      "learning_rate": 2e-05,
+      "loss": 5.3268,
+      "step": 8868
+    },
+    {
+      "epoch": 0.5948955293959821,
+      "grad_norm": 0.15687633087552846,
+      "learning_rate": 2e-05,
+      "loss": 5.3419,
+      "step": 8869
+    },
+    {
+      "epoch": 0.594962605225207,
+      "grad_norm": 0.1479935358929816,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 8870
+    },
+    {
+      "epoch": 0.595029681054432,
+      "grad_norm": 0.1487618535371918,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 8871
+    },
+    {
+      "epoch": 0.5950967568836569,
+      "grad_norm": 0.15928673058803572,
+      "learning_rate": 2e-05,
+      "loss": 5.4098,
+      "step": 8872
+    },
+    {
+      "epoch": 0.5951638327128819,
+      "grad_norm": 0.14457748767101677,
+      "learning_rate": 2e-05,
+      "loss": 5.25,
+      "step": 8873
+    },
+    {
+      "epoch": 0.5952309085421068,
+      "grad_norm": 0.15600442750653748,
+      "learning_rate": 2e-05,
+      "loss": 5.3895,
+      "step": 8874
+    },
+    {
+      "epoch": 0.5952979843713317,
+      "grad_norm": 0.16991893907091926,
+      "learning_rate": 2e-05,
+      "loss": 5.4992,
+      "step": 8875
+    },
+    {
+      "epoch": 0.5953650602005567,
+      "grad_norm": 0.14544147552320166,
+      "learning_rate": 2e-05,
+      "loss": 5.261,
+      "step": 8876
+    },
+    {
+      "epoch": 0.5954321360297816,
+      "grad_norm": 0.15875974068444304,
+      "learning_rate": 2e-05,
+      "loss": 5.2855,
+      "step": 8877
+    },
+    {
+      "epoch": 0.5954992118590066,
+      "grad_norm": 0.15245151896879838,
+      "learning_rate": 2e-05,
+      "loss": 5.2538,
+      "step": 8878
+    },
+    {
+      "epoch": 0.5955662876882315,
+      "grad_norm": 0.15972850062409083,
+      "learning_rate": 2e-05,
+      "loss": 5.3397,
+      "step": 8879
+    },
+    {
+      "epoch": 0.5956333635174564,
+      "grad_norm": 0.14231089919871664,
+      "learning_rate": 2e-05,
+      "loss": 5.5062,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5957004393466814,
+      "grad_norm": 0.15443227688556793,
+      "learning_rate": 2e-05,
+      "loss": 5.5852,
+      "step": 8881
+    },
+    {
+      "epoch": 0.5957675151759063,
+      "grad_norm": 0.15525189850159074,
+      "learning_rate": 2e-05,
+      "loss": 5.4149,
+      "step": 8882
+    },
+    {
+      "epoch": 0.5958345910051313,
+      "grad_norm": 0.15696008000451464,
+      "learning_rate": 2e-05,
+      "loss": 5.5641,
+      "step": 8883
+    },
+    {
+      "epoch": 0.5959016668343562,
+      "grad_norm": 0.16725399544822428,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 8884
+    },
+    {
+      "epoch": 0.5959687426635811,
+      "grad_norm": 0.1525040201455231,
+      "learning_rate": 2e-05,
+      "loss": 5.4119,
+      "step": 8885
+    },
+    {
+      "epoch": 0.5960358184928061,
+      "grad_norm": 0.1529089919821779,
+      "learning_rate": 2e-05,
+      "loss": 5.4538,
+      "step": 8886
+    },
+    {
+      "epoch": 0.596102894322031,
+      "grad_norm": 0.15093118441903522,
+      "learning_rate": 2e-05,
+      "loss": 5.4142,
+      "step": 8887
+    },
+    {
+      "epoch": 0.596169970151256,
+      "grad_norm": 0.14637428960971616,
+      "learning_rate": 2e-05,
+      "loss": 5.5788,
+      "step": 8888
+    },
+    {
+      "epoch": 0.5962370459804809,
+      "grad_norm": 0.14642373207553308,
+      "learning_rate": 2e-05,
+      "loss": 5.4267,
+      "step": 8889
+    },
+    {
+      "epoch": 0.5963041218097058,
+      "grad_norm": 0.1486679721175502,
+      "learning_rate": 2e-05,
+      "loss": 5.4528,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5963711976389308,
+      "grad_norm": 0.15194954236013772,
+      "learning_rate": 2e-05,
+      "loss": 5.4519,
+      "step": 8891
+    },
+    {
+      "epoch": 0.5964382734681557,
+      "grad_norm": 0.14245107867738388,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 8892
+    },
+    {
+      "epoch": 0.5965053492973806,
+      "grad_norm": 0.15475412675159164,
+      "learning_rate": 2e-05,
+      "loss": 5.55,
+      "step": 8893
+    },
+    {
+      "epoch": 0.5965724251266056,
+      "grad_norm": 0.14823897285588536,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 8894
+    },
+    {
+      "epoch": 0.5966395009558305,
+      "grad_norm": 0.15126370052147312,
+      "learning_rate": 2e-05,
+      "loss": 5.2158,
+      "step": 8895
+    },
+    {
+      "epoch": 0.5967065767850555,
+      "grad_norm": 0.1451987252409673,
+      "learning_rate": 2e-05,
+      "loss": 5.3464,
+      "step": 8896
+    },
+    {
+      "epoch": 0.5967736526142804,
+      "grad_norm": 0.15031167868281378,
+      "learning_rate": 2e-05,
+      "loss": 5.3617,
+      "step": 8897
+    },
+    {
+      "epoch": 0.5968407284435053,
+      "grad_norm": 0.1499711073306696,
+      "learning_rate": 2e-05,
+      "loss": 5.3728,
+      "step": 8898
+    },
+    {
+      "epoch": 0.5969078042727303,
+      "grad_norm": 0.14863845802135792,
+      "learning_rate": 2e-05,
+      "loss": 5.4947,
+      "step": 8899
+    },
+    {
+      "epoch": 0.5969748801019552,
+      "grad_norm": 0.1451730208039769,
+      "learning_rate": 2e-05,
+      "loss": 5.3627,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5970419559311801,
+      "grad_norm": 0.14232818708456565,
+      "learning_rate": 2e-05,
+      "loss": 5.2539,
+      "step": 8901
+    },
+    {
+      "epoch": 0.5971090317604051,
+      "grad_norm": 0.15301233643819712,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 8902
+    },
+    {
+      "epoch": 0.59717610758963,
+      "grad_norm": 0.15493625714660492,
+      "learning_rate": 2e-05,
+      "loss": 5.5294,
+      "step": 8903
+    },
+    {
+      "epoch": 0.597243183418855,
+      "grad_norm": 0.1492409459278405,
+      "learning_rate": 2e-05,
+      "loss": 5.3244,
+      "step": 8904
+    },
+    {
+      "epoch": 0.5973102592480799,
+      "grad_norm": 0.15179562103103253,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5973773350773048,
+      "grad_norm": 0.1559493357643404,
+      "learning_rate": 2e-05,
+      "loss": 5.4124,
+      "step": 8906
+    },
+    {
+      "epoch": 0.5974444109065298,
+      "grad_norm": 0.1515639706775748,
+      "learning_rate": 2e-05,
+      "loss": 5.4017,
+      "step": 8907
+    },
+    {
+      "epoch": 0.5975114867357547,
+      "grad_norm": 0.1582043120088637,
+      "learning_rate": 2e-05,
+      "loss": 5.3281,
+      "step": 8908
+    },
+    {
+      "epoch": 0.5975785625649798,
+      "grad_norm": 0.16273746537546904,
+      "learning_rate": 2e-05,
+      "loss": 5.2426,
+      "step": 8909
+    },
+    {
+      "epoch": 0.5976456383942047,
+      "grad_norm": 0.14320682599348275,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5977127142234296,
+      "grad_norm": 0.1551317255514292,
+      "learning_rate": 2e-05,
+      "loss": 5.493,
+      "step": 8911
+    },
+    {
+      "epoch": 0.5977797900526546,
+      "grad_norm": 0.15097490445583012,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 8912
+    },
+    {
+      "epoch": 0.5978468658818795,
+      "grad_norm": 0.15075016233572527,
+      "learning_rate": 2e-05,
+      "loss": 5.4953,
+      "step": 8913
+    },
+    {
+      "epoch": 0.5979139417111045,
+      "grad_norm": 0.15144316849451236,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 8914
+    },
+    {
+      "epoch": 0.5979810175403294,
+      "grad_norm": 0.16954722654102236,
+      "learning_rate": 2e-05,
+      "loss": 5.3834,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5980480933695543,
+      "grad_norm": 0.15327777413444815,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 8916
+    },
+    {
+      "epoch": 0.5981151691987793,
+      "grad_norm": 0.15399343943509464,
+      "learning_rate": 2e-05,
+      "loss": 5.308,
+      "step": 8917
+    },
+    {
+      "epoch": 0.5981822450280042,
+      "grad_norm": 0.16477241042965346,
+      "learning_rate": 2e-05,
+      "loss": 5.3287,
+      "step": 8918
+    },
+    {
+      "epoch": 0.5982493208572291,
+      "grad_norm": 0.14939703489727796,
+      "learning_rate": 2e-05,
+      "loss": 5.3944,
+      "step": 8919
+    },
+    {
+      "epoch": 0.5983163966864541,
+      "grad_norm": 0.15550363792641597,
+      "learning_rate": 2e-05,
+      "loss": 5.4456,
+      "step": 8920
+    },
+    {
+      "epoch": 0.598383472515679,
+      "grad_norm": 0.149789436956052,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 8921
+    },
+    {
+      "epoch": 0.598450548344904,
+      "grad_norm": 0.154481091032482,
+      "learning_rate": 2e-05,
+      "loss": 5.3816,
+      "step": 8922
+    },
+    {
+      "epoch": 0.5985176241741289,
+      "grad_norm": 0.1471984083604363,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 8923
+    },
+    {
+      "epoch": 0.5985847000033538,
+      "grad_norm": 0.1499960494633731,
+      "learning_rate": 2e-05,
+      "loss": 5.327,
+      "step": 8924
+    },
+    {
+      "epoch": 0.5986517758325788,
+      "grad_norm": 0.1514858658530323,
+      "learning_rate": 2e-05,
+      "loss": 5.4259,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5987188516618037,
+      "grad_norm": 0.15397640652636038,
+      "learning_rate": 2e-05,
+      "loss": 5.2469,
+      "step": 8926
+    },
+    {
+      "epoch": 0.5987859274910287,
+      "grad_norm": 0.14888293374851314,
+      "learning_rate": 2e-05,
+      "loss": 5.4897,
+      "step": 8927
+    },
+    {
+      "epoch": 0.5988530033202536,
+      "grad_norm": 0.15499072903526667,
+      "learning_rate": 2e-05,
+      "loss": 5.5377,
+      "step": 8928
+    },
+    {
+      "epoch": 0.5989200791494785,
+      "grad_norm": 0.14915319380950565,
+      "learning_rate": 2e-05,
+      "loss": 5.3622,
+      "step": 8929
+    },
+    {
+      "epoch": 0.5989871549787035,
+      "grad_norm": 0.1488767623706366,
+      "learning_rate": 2e-05,
+      "loss": 5.2749,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5990542308079284,
+      "grad_norm": 0.1543742664705043,
+      "learning_rate": 2e-05,
+      "loss": 5.4106,
+      "step": 8931
+    },
+    {
+      "epoch": 0.5991213066371534,
+      "grad_norm": 0.1581348833177915,
+      "learning_rate": 2e-05,
+      "loss": 5.4469,
+      "step": 8932
+    },
+    {
+      "epoch": 0.5991883824663783,
+      "grad_norm": 0.15434236512304,
+      "learning_rate": 2e-05,
+      "loss": 5.315,
+      "step": 8933
+    },
+    {
+      "epoch": 0.5992554582956032,
+      "grad_norm": 0.15668518433819012,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 8934
+    },
+    {
+      "epoch": 0.5993225341248282,
+      "grad_norm": 0.15123681198636502,
+      "learning_rate": 2e-05,
+      "loss": 5.627,
+      "step": 8935
+    },
+    {
+      "epoch": 0.5993896099540531,
+      "grad_norm": 0.15573952265536892,
+      "learning_rate": 2e-05,
+      "loss": 5.5373,
+      "step": 8936
+    },
+    {
+      "epoch": 0.599456685783278,
+      "grad_norm": 0.1434198130682002,
+      "learning_rate": 2e-05,
+      "loss": 5.4687,
+      "step": 8937
+    },
+    {
+      "epoch": 0.599523761612503,
+      "grad_norm": 0.1505180301689619,
+      "learning_rate": 2e-05,
+      "loss": 5.2637,
+      "step": 8938
+    },
+    {
+      "epoch": 0.5995908374417279,
+      "grad_norm": 0.15136053683321693,
+      "learning_rate": 2e-05,
+      "loss": 5.4481,
+      "step": 8939
+    },
+    {
+      "epoch": 0.5996579132709529,
+      "grad_norm": 0.14919847398430724,
+      "learning_rate": 2e-05,
+      "loss": 5.3648,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5997249891001778,
+      "grad_norm": 0.15543809135227818,
+      "learning_rate": 2e-05,
+      "loss": 5.4377,
+      "step": 8941
+    },
+    {
+      "epoch": 0.5997920649294027,
+      "grad_norm": 0.14770963959610758,
+      "learning_rate": 2e-05,
+      "loss": 5.3781,
+      "step": 8942
+    },
+    {
+      "epoch": 0.5998591407586277,
+      "grad_norm": 0.1511334872050148,
+      "learning_rate": 2e-05,
+      "loss": 5.3772,
+      "step": 8943
+    },
+    {
+      "epoch": 0.5999262165878526,
+      "grad_norm": 0.15218741822313644,
+      "learning_rate": 2e-05,
+      "loss": 5.3406,
+      "step": 8944
+    },
+    {
+      "epoch": 0.5999932924170776,
+      "grad_norm": 0.1542486836345395,
+      "learning_rate": 2e-05,
+      "loss": 5.3126,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6000603682463025,
+      "grad_norm": 0.15451660383886048,
+      "learning_rate": 2e-05,
+      "loss": 5.4012,
+      "step": 8946
+    },
+    {
+      "epoch": 0.6001274440755274,
+      "grad_norm": 0.16138440751668123,
+      "learning_rate": 2e-05,
+      "loss": 5.2503,
+      "step": 8947
+    },
+    {
+      "epoch": 0.6001945199047524,
+      "grad_norm": 0.1492753627709322,
+      "learning_rate": 2e-05,
+      "loss": 5.363,
+      "step": 8948
+    },
+    {
+      "epoch": 0.6002615957339773,
+      "grad_norm": 0.15169511522524723,
+      "learning_rate": 2e-05,
+      "loss": 5.4841,
+      "step": 8949
+    },
+    {
+      "epoch": 0.6003286715632022,
+      "grad_norm": 0.15038400126487497,
+      "learning_rate": 2e-05,
+      "loss": 5.4553,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6003957473924272,
+      "grad_norm": 0.1500153997312142,
+      "learning_rate": 2e-05,
+      "loss": 5.442,
+      "step": 8951
+    },
+    {
+      "epoch": 0.6004628232216521,
+      "grad_norm": 0.15408022739030752,
+      "learning_rate": 2e-05,
+      "loss": 5.3068,
+      "step": 8952
+    },
+    {
+      "epoch": 0.6005298990508771,
+      "grad_norm": 0.15076596353841898,
+      "learning_rate": 2e-05,
+      "loss": 5.3745,
+      "step": 8953
+    },
+    {
+      "epoch": 0.600596974880102,
+      "grad_norm": 0.14442188605308076,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 8954
+    },
+    {
+      "epoch": 0.6006640507093269,
+      "grad_norm": 0.1543946648116208,
+      "learning_rate": 2e-05,
+      "loss": 5.6082,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6007311265385519,
+      "grad_norm": 0.1556852252036636,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 8956
+    },
+    {
+      "epoch": 0.6007982023677768,
+      "grad_norm": 0.15877478210477042,
+      "learning_rate": 2e-05,
+      "loss": 5.4809,
+      "step": 8957
+    },
+    {
+      "epoch": 0.6008652781970018,
+      "grad_norm": 0.15112758104741317,
+      "learning_rate": 2e-05,
+      "loss": 5.3632,
+      "step": 8958
+    },
+    {
+      "epoch": 0.6009323540262267,
+      "grad_norm": 0.15153221356382965,
+      "learning_rate": 2e-05,
+      "loss": 5.4139,
+      "step": 8959
+    },
+    {
+      "epoch": 0.6009994298554516,
+      "grad_norm": 0.15484897207698378,
+      "learning_rate": 2e-05,
+      "loss": 5.5857,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6010665056846766,
+      "grad_norm": 0.1579073279975335,
+      "learning_rate": 2e-05,
+      "loss": 5.5867,
+      "step": 8961
+    },
+    {
+      "epoch": 0.6011335815139015,
+      "grad_norm": 0.15465219090088217,
+      "learning_rate": 2e-05,
+      "loss": 5.4069,
+      "step": 8962
+    },
+    {
+      "epoch": 0.6012006573431264,
+      "grad_norm": 0.1531967369602585,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 8963
+    },
+    {
+      "epoch": 0.6012677331723514,
+      "grad_norm": 0.15759248277407034,
+      "learning_rate": 2e-05,
+      "loss": 5.4964,
+      "step": 8964
+    },
+    {
+      "epoch": 0.6013348090015763,
+      "grad_norm": 0.14787314299165863,
+      "learning_rate": 2e-05,
+      "loss": 5.4404,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6014018848308013,
+      "grad_norm": 0.15122246578523943,
+      "learning_rate": 2e-05,
+      "loss": 5.3929,
+      "step": 8966
+    },
+    {
+      "epoch": 0.6014689606600262,
+      "grad_norm": 0.1661789610460339,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 8967
+    },
+    {
+      "epoch": 0.6015360364892511,
+      "grad_norm": 0.14737555893194235,
+      "learning_rate": 2e-05,
+      "loss": 5.4786,
+      "step": 8968
+    },
+    {
+      "epoch": 0.6016031123184761,
+      "grad_norm": 0.15464646754633127,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 8969
+    },
+    {
+      "epoch": 0.601670188147701,
+      "grad_norm": 0.15411467148849903,
+      "learning_rate": 2e-05,
+      "loss": 5.5472,
+      "step": 8970
+    },
+    {
+      "epoch": 0.601737263976926,
+      "grad_norm": 0.15377254433098472,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 8971
+    },
+    {
+      "epoch": 0.6018043398061509,
+      "grad_norm": 0.1548168442946976,
+      "learning_rate": 2e-05,
+      "loss": 5.3979,
+      "step": 8972
+    },
+    {
+      "epoch": 0.6018714156353758,
+      "grad_norm": 0.14305001840246215,
+      "learning_rate": 2e-05,
+      "loss": 5.5346,
+      "step": 8973
+    },
+    {
+      "epoch": 0.6019384914646008,
+      "grad_norm": 0.14413154924886673,
+      "learning_rate": 2e-05,
+      "loss": 5.3403,
+      "step": 8974
+    },
+    {
+      "epoch": 0.6020055672938257,
+      "grad_norm": 0.14838643282752045,
+      "learning_rate": 2e-05,
+      "loss": 5.2971,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6020726431230506,
+      "grad_norm": 0.14871205345921573,
+      "learning_rate": 2e-05,
+      "loss": 5.366,
+      "step": 8976
+    },
+    {
+      "epoch": 0.6021397189522756,
+      "grad_norm": 0.15082078703633256,
+      "learning_rate": 2e-05,
+      "loss": 5.4716,
+      "step": 8977
+    },
+    {
+      "epoch": 0.6022067947815005,
+      "grad_norm": 0.14928597363732266,
+      "learning_rate": 2e-05,
+      "loss": 5.4194,
+      "step": 8978
+    },
+    {
+      "epoch": 0.6022738706107255,
+      "grad_norm": 0.15023228404067324,
+      "learning_rate": 2e-05,
+      "loss": 5.4024,
+      "step": 8979
+    },
+    {
+      "epoch": 0.6023409464399504,
+      "grad_norm": 0.15025713005870747,
+      "learning_rate": 2e-05,
+      "loss": 5.2615,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6024080222691753,
+      "grad_norm": 0.15403690207945142,
+      "learning_rate": 2e-05,
+      "loss": 5.3488,
+      "step": 8981
+    },
+    {
+      "epoch": 0.6024750980984003,
+      "grad_norm": 0.148328267223966,
+      "learning_rate": 2e-05,
+      "loss": 5.3357,
+      "step": 8982
+    },
+    {
+      "epoch": 0.6025421739276252,
+      "grad_norm": 0.1553964314173716,
+      "learning_rate": 2e-05,
+      "loss": 5.428,
+      "step": 8983
+    },
+    {
+      "epoch": 0.6026092497568502,
+      "grad_norm": 0.14356550963384257,
+      "learning_rate": 2e-05,
+      "loss": 5.4449,
+      "step": 8984
+    },
+    {
+      "epoch": 0.6026763255860751,
+      "grad_norm": 0.142019170935836,
+      "learning_rate": 2e-05,
+      "loss": 5.4277,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6027434014153,
+      "grad_norm": 0.14850268751193504,
+      "learning_rate": 2e-05,
+      "loss": 5.5613,
+      "step": 8986
+    },
+    {
+      "epoch": 0.602810477244525,
+      "grad_norm": 0.1507454832402468,
+      "learning_rate": 2e-05,
+      "loss": 5.4407,
+      "step": 8987
+    },
+    {
+      "epoch": 0.6028775530737499,
+      "grad_norm": 0.15543196132438,
+      "learning_rate": 2e-05,
+      "loss": 5.4988,
+      "step": 8988
+    },
+    {
+      "epoch": 0.6029446289029748,
+      "grad_norm": 0.14764897724193848,
+      "learning_rate": 2e-05,
+      "loss": 5.3523,
+      "step": 8989
+    },
+    {
+      "epoch": 0.6030117047321998,
+      "grad_norm": 0.14991705591401405,
+      "learning_rate": 2e-05,
+      "loss": 5.3284,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6030787805614247,
+      "grad_norm": 0.1553581354160704,
+      "learning_rate": 2e-05,
+      "loss": 5.2727,
+      "step": 8991
+    },
+    {
+      "epoch": 0.6031458563906497,
+      "grad_norm": 0.15500576069988062,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 8992
+    },
+    {
+      "epoch": 0.6032129322198746,
+      "grad_norm": 0.1463429364609225,
+      "learning_rate": 2e-05,
+      "loss": 5.4765,
+      "step": 8993
+    },
+    {
+      "epoch": 0.6032800080490995,
+      "grad_norm": 0.15837228189100128,
+      "learning_rate": 2e-05,
+      "loss": 5.3945,
+      "step": 8994
+    },
+    {
+      "epoch": 0.6033470838783245,
+      "grad_norm": 0.14725921696478694,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6034141597075494,
+      "grad_norm": 0.1525603640052087,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 8996
+    },
+    {
+      "epoch": 0.6034812355367744,
+      "grad_norm": 0.15198840903220093,
+      "learning_rate": 2e-05,
+      "loss": 5.3688,
+      "step": 8997
+    },
+    {
+      "epoch": 0.6035483113659993,
+      "grad_norm": 0.15108819788733924,
+      "learning_rate": 2e-05,
+      "loss": 5.3679,
+      "step": 8998
+    },
+    {
+      "epoch": 0.6036153871952242,
+      "grad_norm": 0.16321541558801106,
+      "learning_rate": 2e-05,
+      "loss": 5.4714,
+      "step": 8999
+    },
+    {
+      "epoch": 0.6036824630244492,
+      "grad_norm": 0.15236251150740707,
+      "learning_rate": 2e-05,
+      "loss": 5.5963,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6037495388536741,
+      "grad_norm": 0.15983853687725857,
+      "learning_rate": 2e-05,
+      "loss": 5.4828,
+      "step": 9001
+    },
+    {
+      "epoch": 0.603816614682899,
+      "grad_norm": 0.15819974061098582,
+      "learning_rate": 2e-05,
+      "loss": 5.4186,
+      "step": 9002
+    },
+    {
+      "epoch": 0.603883690512124,
+      "grad_norm": 0.15216511872137053,
+      "learning_rate": 2e-05,
+      "loss": 5.3866,
+      "step": 9003
+    },
+    {
+      "epoch": 0.6039507663413489,
+      "grad_norm": 0.15276083808059057,
+      "learning_rate": 2e-05,
+      "loss": 5.4266,
+      "step": 9004
+    },
+    {
+      "epoch": 0.6040178421705739,
+      "grad_norm": 0.15965280581317243,
+      "learning_rate": 2e-05,
+      "loss": 5.329,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6040849179997988,
+      "grad_norm": 0.15310023115550733,
+      "learning_rate": 2e-05,
+      "loss": 5.5107,
+      "step": 9006
+    },
+    {
+      "epoch": 0.6041519938290237,
+      "grad_norm": 0.14387517632609015,
+      "learning_rate": 2e-05,
+      "loss": 5.6101,
+      "step": 9007
+    },
+    {
+      "epoch": 0.6042190696582487,
+      "grad_norm": 0.15517298947954805,
+      "learning_rate": 2e-05,
+      "loss": 5.4462,
+      "step": 9008
+    },
+    {
+      "epoch": 0.6042861454874736,
+      "grad_norm": 0.17232547131228793,
+      "learning_rate": 2e-05,
+      "loss": 5.4662,
+      "step": 9009
+    },
+    {
+      "epoch": 0.6043532213166986,
+      "grad_norm": 0.14630755391061998,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6044202971459235,
+      "grad_norm": 0.15062601186984317,
+      "learning_rate": 2e-05,
+      "loss": 5.3533,
+      "step": 9011
+    },
+    {
+      "epoch": 0.6044873729751484,
+      "grad_norm": 0.15772659558633342,
+      "learning_rate": 2e-05,
+      "loss": 5.3823,
+      "step": 9012
+    },
+    {
+      "epoch": 0.6045544488043734,
+      "grad_norm": 0.1500009887502321,
+      "learning_rate": 2e-05,
+      "loss": 5.4222,
+      "step": 9013
+    },
+    {
+      "epoch": 0.6046215246335983,
+      "grad_norm": 0.14655572844289408,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 9014
+    },
+    {
+      "epoch": 0.6046886004628232,
+      "grad_norm": 0.15090520771682703,
+      "learning_rate": 2e-05,
+      "loss": 5.2534,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6047556762920482,
+      "grad_norm": 0.15573891238215898,
+      "learning_rate": 2e-05,
+      "loss": 5.4138,
+      "step": 9016
+    },
+    {
+      "epoch": 0.6048227521212731,
+      "grad_norm": 0.15175656419306613,
+      "learning_rate": 2e-05,
+      "loss": 5.4869,
+      "step": 9017
+    },
+    {
+      "epoch": 0.6048898279504981,
+      "grad_norm": 0.150313502108073,
+      "learning_rate": 2e-05,
+      "loss": 5.3718,
+      "step": 9018
+    },
+    {
+      "epoch": 0.604956903779723,
+      "grad_norm": 0.15603766247727635,
+      "learning_rate": 2e-05,
+      "loss": 5.2364,
+      "step": 9019
+    },
+    {
+      "epoch": 0.6050239796089479,
+      "grad_norm": 0.14757537344924523,
+      "learning_rate": 2e-05,
+      "loss": 5.5894,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6050910554381729,
+      "grad_norm": 0.17045406922620399,
+      "learning_rate": 2e-05,
+      "loss": 5.6016,
+      "step": 9021
+    },
+    {
+      "epoch": 0.6051581312673978,
+      "grad_norm": 0.15951765435310425,
+      "learning_rate": 2e-05,
+      "loss": 5.4626,
+      "step": 9022
+    },
+    {
+      "epoch": 0.6052252070966228,
+      "grad_norm": 0.15615472549591994,
+      "learning_rate": 2e-05,
+      "loss": 5.3404,
+      "step": 9023
+    },
+    {
+      "epoch": 0.6052922829258477,
+      "grad_norm": 0.16352008072224333,
+      "learning_rate": 2e-05,
+      "loss": 5.5314,
+      "step": 9024
+    },
+    {
+      "epoch": 0.6053593587550726,
+      "grad_norm": 0.15262101690059107,
+      "learning_rate": 2e-05,
+      "loss": 5.4742,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6054264345842976,
+      "grad_norm": 0.14995412868326066,
+      "learning_rate": 2e-05,
+      "loss": 5.4165,
+      "step": 9026
+    },
+    {
+      "epoch": 0.6054935104135225,
+      "grad_norm": 0.16396824953415823,
+      "learning_rate": 2e-05,
+      "loss": 5.4779,
+      "step": 9027
+    },
+    {
+      "epoch": 0.6055605862427474,
+      "grad_norm": 0.15367947844926366,
+      "learning_rate": 2e-05,
+      "loss": 5.4574,
+      "step": 9028
+    },
+    {
+      "epoch": 0.6056276620719724,
+      "grad_norm": 0.14993829770969036,
+      "learning_rate": 2e-05,
+      "loss": 5.4521,
+      "step": 9029
+    },
+    {
+      "epoch": 0.6056947379011973,
+      "grad_norm": 0.1486552251973269,
+      "learning_rate": 2e-05,
+      "loss": 5.3296,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6057618137304223,
+      "grad_norm": 0.15252689634945368,
+      "learning_rate": 2e-05,
+      "loss": 5.3456,
+      "step": 9031
+    },
+    {
+      "epoch": 0.6058288895596472,
+      "grad_norm": 0.160997564263249,
+      "learning_rate": 2e-05,
+      "loss": 5.2987,
+      "step": 9032
+    },
+    {
+      "epoch": 0.6058959653888721,
+      "grad_norm": 0.14736269658009427,
+      "learning_rate": 2e-05,
+      "loss": 5.327,
+      "step": 9033
+    },
+    {
+      "epoch": 0.6059630412180971,
+      "grad_norm": 0.15722534869701743,
+      "learning_rate": 2e-05,
+      "loss": 5.4551,
+      "step": 9034
+    },
+    {
+      "epoch": 0.606030117047322,
+      "grad_norm": 0.1437140830910672,
+      "learning_rate": 2e-05,
+      "loss": 5.3408,
+      "step": 9035
+    },
+    {
+      "epoch": 0.606097192876547,
+      "grad_norm": 0.14674147253267747,
+      "learning_rate": 2e-05,
+      "loss": 5.5675,
+      "step": 9036
+    },
+    {
+      "epoch": 0.6061642687057719,
+      "grad_norm": 0.16135903868618987,
+      "learning_rate": 2e-05,
+      "loss": 5.5075,
+      "step": 9037
+    },
+    {
+      "epoch": 0.6062313445349968,
+      "grad_norm": 0.15170535853221465,
+      "learning_rate": 2e-05,
+      "loss": 5.2254,
+      "step": 9038
+    },
+    {
+      "epoch": 0.6062984203642218,
+      "grad_norm": 0.15061563452745205,
+      "learning_rate": 2e-05,
+      "loss": 5.5477,
+      "step": 9039
+    },
+    {
+      "epoch": 0.6063654961934467,
+      "grad_norm": 0.14941987758071648,
+      "learning_rate": 2e-05,
+      "loss": 5.3743,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6064325720226716,
+      "grad_norm": 0.14905957219784888,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 9041
+    },
+    {
+      "epoch": 0.6064996478518966,
+      "grad_norm": 0.15236135677264073,
+      "learning_rate": 2e-05,
+      "loss": 5.5152,
+      "step": 9042
+    },
+    {
+      "epoch": 0.6065667236811215,
+      "grad_norm": 0.15204669861875156,
+      "learning_rate": 2e-05,
+      "loss": 5.5051,
+      "step": 9043
+    },
+    {
+      "epoch": 0.6066337995103465,
+      "grad_norm": 0.1501744019718792,
+      "learning_rate": 2e-05,
+      "loss": 5.2498,
+      "step": 9044
+    },
+    {
+      "epoch": 0.6067008753395714,
+      "grad_norm": 0.15835659108354702,
+      "learning_rate": 2e-05,
+      "loss": 5.3579,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6067679511687963,
+      "grad_norm": 0.15329364468904505,
+      "learning_rate": 2e-05,
+      "loss": 5.5705,
+      "step": 9046
+    },
+    {
+      "epoch": 0.6068350269980213,
+      "grad_norm": 0.1480475226210665,
+      "learning_rate": 2e-05,
+      "loss": 5.5419,
+      "step": 9047
+    },
+    {
+      "epoch": 0.6069021028272462,
+      "grad_norm": 0.1610067679666325,
+      "learning_rate": 2e-05,
+      "loss": 5.3716,
+      "step": 9048
+    },
+    {
+      "epoch": 0.6069691786564712,
+      "grad_norm": 0.14792985191093305,
+      "learning_rate": 2e-05,
+      "loss": 5.4031,
+      "step": 9049
+    },
+    {
+      "epoch": 0.6070362544856961,
+      "grad_norm": 0.15084485940956005,
+      "learning_rate": 2e-05,
+      "loss": 5.3718,
+      "step": 9050
+    },
+    {
+      "epoch": 0.607103330314921,
+      "grad_norm": 0.154226644095758,
+      "learning_rate": 2e-05,
+      "loss": 5.4864,
+      "step": 9051
+    },
+    {
+      "epoch": 0.607170406144146,
+      "grad_norm": 0.16141146658108096,
+      "learning_rate": 2e-05,
+      "loss": 5.524,
+      "step": 9052
+    },
+    {
+      "epoch": 0.6072374819733709,
+      "grad_norm": 0.14835895925580098,
+      "learning_rate": 2e-05,
+      "loss": 5.4596,
+      "step": 9053
+    },
+    {
+      "epoch": 0.6073045578025958,
+      "grad_norm": 0.14989467166588258,
+      "learning_rate": 2e-05,
+      "loss": 5.2762,
+      "step": 9054
+    },
+    {
+      "epoch": 0.6073716336318208,
+      "grad_norm": 0.16458166988151135,
+      "learning_rate": 2e-05,
+      "loss": 5.4244,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6074387094610457,
+      "grad_norm": 0.15441941999247885,
+      "learning_rate": 2e-05,
+      "loss": 5.3784,
+      "step": 9056
+    },
+    {
+      "epoch": 0.6075057852902707,
+      "grad_norm": 0.1624701689271173,
+      "learning_rate": 2e-05,
+      "loss": 5.421,
+      "step": 9057
+    },
+    {
+      "epoch": 0.6075728611194956,
+      "grad_norm": 0.1643811406943147,
+      "learning_rate": 2e-05,
+      "loss": 5.3702,
+      "step": 9058
+    },
+    {
+      "epoch": 0.6076399369487205,
+      "grad_norm": 0.15388195277696853,
+      "learning_rate": 2e-05,
+      "loss": 5.5132,
+      "step": 9059
+    },
+    {
+      "epoch": 0.6077070127779455,
+      "grad_norm": 0.15918925104587828,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6077740886071704,
+      "grad_norm": 0.15355643987004347,
+      "learning_rate": 2e-05,
+      "loss": 5.4012,
+      "step": 9061
+    },
+    {
+      "epoch": 0.6078411644363954,
+      "grad_norm": 0.1560296435700642,
+      "learning_rate": 2e-05,
+      "loss": 5.4704,
+      "step": 9062
+    },
+    {
+      "epoch": 0.6079082402656203,
+      "grad_norm": 0.15423556707250996,
+      "learning_rate": 2e-05,
+      "loss": 5.3943,
+      "step": 9063
+    },
+    {
+      "epoch": 0.6079753160948452,
+      "grad_norm": 0.16701324471122955,
+      "learning_rate": 2e-05,
+      "loss": 5.3997,
+      "step": 9064
+    },
+    {
+      "epoch": 0.6080423919240702,
+      "grad_norm": 0.1587009407201048,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6081094677532951,
+      "grad_norm": 0.14788047646421326,
+      "learning_rate": 2e-05,
+      "loss": 5.3695,
+      "step": 9066
+    },
+    {
+      "epoch": 0.60817654358252,
+      "grad_norm": 0.15149529238947582,
+      "learning_rate": 2e-05,
+      "loss": 5.3517,
+      "step": 9067
+    },
+    {
+      "epoch": 0.608243619411745,
+      "grad_norm": 0.1539677940905713,
+      "learning_rate": 2e-05,
+      "loss": 5.4893,
+      "step": 9068
+    },
+    {
+      "epoch": 0.6083106952409699,
+      "grad_norm": 0.156240157114063,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 9069
+    },
+    {
+      "epoch": 0.6083777710701949,
+      "grad_norm": 0.15466474697001698,
+      "learning_rate": 2e-05,
+      "loss": 5.3091,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6084448468994198,
+      "grad_norm": 0.14992377458102238,
+      "learning_rate": 2e-05,
+      "loss": 5.3718,
+      "step": 9071
+    },
+    {
+      "epoch": 0.6085119227286447,
+      "grad_norm": 0.16306069738613704,
+      "learning_rate": 2e-05,
+      "loss": 5.5052,
+      "step": 9072
+    },
+    {
+      "epoch": 0.6085789985578697,
+      "grad_norm": 0.161290662916394,
+      "learning_rate": 2e-05,
+      "loss": 5.5034,
+      "step": 9073
+    },
+    {
+      "epoch": 0.6086460743870946,
+      "grad_norm": 0.1542979000357331,
+      "learning_rate": 2e-05,
+      "loss": 5.3364,
+      "step": 9074
+    },
+    {
+      "epoch": 0.6087131502163196,
+      "grad_norm": 0.1560862790102712,
+      "learning_rate": 2e-05,
+      "loss": 5.4304,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6087802260455445,
+      "grad_norm": 0.15260175428044737,
+      "learning_rate": 2e-05,
+      "loss": 5.3699,
+      "step": 9076
+    },
+    {
+      "epoch": 0.6088473018747694,
+      "grad_norm": 0.15692990653286534,
+      "learning_rate": 2e-05,
+      "loss": 5.3974,
+      "step": 9077
+    },
+    {
+      "epoch": 0.6089143777039944,
+      "grad_norm": 0.15118076823407234,
+      "learning_rate": 2e-05,
+      "loss": 5.4479,
+      "step": 9078
+    },
+    {
+      "epoch": 0.6089814535332193,
+      "grad_norm": 0.1545019673284196,
+      "learning_rate": 2e-05,
+      "loss": 5.4213,
+      "step": 9079
+    },
+    {
+      "epoch": 0.6090485293624442,
+      "grad_norm": 0.15645428787318477,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6091156051916692,
+      "grad_norm": 0.15160053131839996,
+      "learning_rate": 2e-05,
+      "loss": 5.4273,
+      "step": 9081
+    },
+    {
+      "epoch": 0.6091826810208941,
+      "grad_norm": 0.14995496418303117,
+      "learning_rate": 2e-05,
+      "loss": 5.4542,
+      "step": 9082
+    },
+    {
+      "epoch": 0.6092497568501191,
+      "grad_norm": 0.15509185340000212,
+      "learning_rate": 2e-05,
+      "loss": 5.4644,
+      "step": 9083
+    },
+    {
+      "epoch": 0.609316832679344,
+      "grad_norm": 0.15752633776929348,
+      "learning_rate": 2e-05,
+      "loss": 5.4208,
+      "step": 9084
+    },
+    {
+      "epoch": 0.6093839085085689,
+      "grad_norm": 0.15180853440087552,
+      "learning_rate": 2e-05,
+      "loss": 5.4667,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6094509843377939,
+      "grad_norm": 0.14387976076719428,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 9086
+    },
+    {
+      "epoch": 0.6095180601670188,
+      "grad_norm": 0.15054827385069564,
+      "learning_rate": 2e-05,
+      "loss": 5.4252,
+      "step": 9087
+    },
+    {
+      "epoch": 0.6095851359962438,
+      "grad_norm": 0.15611207883751796,
+      "learning_rate": 2e-05,
+      "loss": 5.4336,
+      "step": 9088
+    },
+    {
+      "epoch": 0.6096522118254687,
+      "grad_norm": 0.1492779842193079,
+      "learning_rate": 2e-05,
+      "loss": 5.4079,
+      "step": 9089
+    },
+    {
+      "epoch": 0.6097192876546936,
+      "grad_norm": 0.15388225655769122,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6097863634839186,
+      "grad_norm": 0.15421826092032268,
+      "learning_rate": 2e-05,
+      "loss": 5.4495,
+      "step": 9091
+    },
+    {
+      "epoch": 0.6098534393131435,
+      "grad_norm": 0.1507844303454805,
+      "learning_rate": 2e-05,
+      "loss": 5.4441,
+      "step": 9092
+    },
+    {
+      "epoch": 0.6099205151423684,
+      "grad_norm": 0.145165364685143,
+      "learning_rate": 2e-05,
+      "loss": 5.5461,
+      "step": 9093
+    },
+    {
+      "epoch": 0.6099875909715934,
+      "grad_norm": 0.14635622020401207,
+      "learning_rate": 2e-05,
+      "loss": 5.4008,
+      "step": 9094
+    },
+    {
+      "epoch": 0.6100546668008183,
+      "grad_norm": 0.14776912734415257,
+      "learning_rate": 2e-05,
+      "loss": 5.2758,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6101217426300433,
+      "grad_norm": 0.14971931721422518,
+      "learning_rate": 2e-05,
+      "loss": 5.6557,
+      "step": 9096
+    },
+    {
+      "epoch": 0.6101888184592682,
+      "grad_norm": 0.15018281944937972,
+      "learning_rate": 2e-05,
+      "loss": 5.4215,
+      "step": 9097
+    },
+    {
+      "epoch": 0.6102558942884931,
+      "grad_norm": 0.14733589829931923,
+      "learning_rate": 2e-05,
+      "loss": 5.4903,
+      "step": 9098
+    },
+    {
+      "epoch": 0.6103229701177181,
+      "grad_norm": 0.1464679490964221,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 9099
+    },
+    {
+      "epoch": 0.610390045946943,
+      "grad_norm": 0.14702674792363746,
+      "learning_rate": 2e-05,
+      "loss": 5.3436,
+      "step": 9100
+    },
+    {
+      "epoch": 0.610457121776168,
+      "grad_norm": 0.14915972550948736,
+      "learning_rate": 2e-05,
+      "loss": 5.3824,
+      "step": 9101
+    },
+    {
+      "epoch": 0.6105241976053929,
+      "grad_norm": 0.15245428394217023,
+      "learning_rate": 2e-05,
+      "loss": 5.4076,
+      "step": 9102
+    },
+    {
+      "epoch": 0.6105912734346178,
+      "grad_norm": 0.15590748206559785,
+      "learning_rate": 2e-05,
+      "loss": 5.3859,
+      "step": 9103
+    },
+    {
+      "epoch": 0.6106583492638428,
+      "grad_norm": 0.14536956924912117,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 9104
+    },
+    {
+      "epoch": 0.6107254250930677,
+      "grad_norm": 0.1522827465972116,
+      "learning_rate": 2e-05,
+      "loss": 5.4313,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6107925009222926,
+      "grad_norm": 0.14678616950949264,
+      "learning_rate": 2e-05,
+      "loss": 5.5454,
+      "step": 9106
+    },
+    {
+      "epoch": 0.6108595767515176,
+      "grad_norm": 0.15178208334602963,
+      "learning_rate": 2e-05,
+      "loss": 5.3848,
+      "step": 9107
+    },
+    {
+      "epoch": 0.6109266525807425,
+      "grad_norm": 0.14824846503691663,
+      "learning_rate": 2e-05,
+      "loss": 5.2603,
+      "step": 9108
+    },
+    {
+      "epoch": 0.6109937284099675,
+      "grad_norm": 0.14810020022294135,
+      "learning_rate": 2e-05,
+      "loss": 5.4601,
+      "step": 9109
+    },
+    {
+      "epoch": 0.6110608042391924,
+      "grad_norm": 0.15698330254477863,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6111278800684173,
+      "grad_norm": 0.1523687535254334,
+      "learning_rate": 2e-05,
+      "loss": 5.3757,
+      "step": 9111
+    },
+    {
+      "epoch": 0.6111949558976423,
+      "grad_norm": 0.15110261553240234,
+      "learning_rate": 2e-05,
+      "loss": 5.2791,
+      "step": 9112
+    },
+    {
+      "epoch": 0.6112620317268672,
+      "grad_norm": 0.1560759476046622,
+      "learning_rate": 2e-05,
+      "loss": 5.546,
+      "step": 9113
+    },
+    {
+      "epoch": 0.6113291075560922,
+      "grad_norm": 0.1551675260142818,
+      "learning_rate": 2e-05,
+      "loss": 5.3756,
+      "step": 9114
+    },
+    {
+      "epoch": 0.6113961833853171,
+      "grad_norm": 0.15069460926644276,
+      "learning_rate": 2e-05,
+      "loss": 5.3119,
+      "step": 9115
+    },
+    {
+      "epoch": 0.611463259214542,
+      "grad_norm": 0.15663351341485507,
+      "learning_rate": 2e-05,
+      "loss": 5.5164,
+      "step": 9116
+    },
+    {
+      "epoch": 0.611530335043767,
+      "grad_norm": 0.14700169215636272,
+      "learning_rate": 2e-05,
+      "loss": 5.3879,
+      "step": 9117
+    },
+    {
+      "epoch": 0.6115974108729919,
+      "grad_norm": 0.15203075598926397,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 9118
+    },
+    {
+      "epoch": 0.6116644867022168,
+      "grad_norm": 0.16417960357222863,
+      "learning_rate": 2e-05,
+      "loss": 5.5013,
+      "step": 9119
+    },
+    {
+      "epoch": 0.6117315625314418,
+      "grad_norm": 0.153388498542335,
+      "learning_rate": 2e-05,
+      "loss": 5.4573,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6117986383606667,
+      "grad_norm": 0.16109174442139465,
+      "learning_rate": 2e-05,
+      "loss": 5.5615,
+      "step": 9121
+    },
+    {
+      "epoch": 0.6118657141898917,
+      "grad_norm": 0.1485843401001438,
+      "learning_rate": 2e-05,
+      "loss": 5.5611,
+      "step": 9122
+    },
+    {
+      "epoch": 0.6119327900191166,
+      "grad_norm": 0.1642143904041403,
+      "learning_rate": 2e-05,
+      "loss": 5.5858,
+      "step": 9123
+    },
+    {
+      "epoch": 0.6119998658483415,
+      "grad_norm": 0.15393671490996583,
+      "learning_rate": 2e-05,
+      "loss": 5.3904,
+      "step": 9124
+    },
+    {
+      "epoch": 0.6120669416775665,
+      "grad_norm": 0.15244460538309557,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6121340175067914,
+      "grad_norm": 0.15289091439834016,
+      "learning_rate": 2e-05,
+      "loss": 5.4565,
+      "step": 9126
+    },
+    {
+      "epoch": 0.6122010933360164,
+      "grad_norm": 0.16032301248126507,
+      "learning_rate": 2e-05,
+      "loss": 5.2736,
+      "step": 9127
+    },
+    {
+      "epoch": 0.6122681691652413,
+      "grad_norm": 0.14579745583600154,
+      "learning_rate": 2e-05,
+      "loss": 5.4408,
+      "step": 9128
+    },
+    {
+      "epoch": 0.6123352449944662,
+      "grad_norm": 0.15629664135501284,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 9129
+    },
+    {
+      "epoch": 0.6124023208236912,
+      "grad_norm": 0.15367126941106227,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6124693966529161,
+      "grad_norm": 0.14814662669089393,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 9131
+    },
+    {
+      "epoch": 0.612536472482141,
+      "grad_norm": 0.1570534498223316,
+      "learning_rate": 2e-05,
+      "loss": 5.3172,
+      "step": 9132
+    },
+    {
+      "epoch": 0.612603548311366,
+      "grad_norm": 0.1557785183646263,
+      "learning_rate": 2e-05,
+      "loss": 5.3776,
+      "step": 9133
+    },
+    {
+      "epoch": 0.6126706241405909,
+      "grad_norm": 0.15262552917103855,
+      "learning_rate": 2e-05,
+      "loss": 5.3925,
+      "step": 9134
+    },
+    {
+      "epoch": 0.6127376999698159,
+      "grad_norm": 0.1552633121071096,
+      "learning_rate": 2e-05,
+      "loss": 5.509,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6128047757990408,
+      "grad_norm": 0.15988477932644307,
+      "learning_rate": 2e-05,
+      "loss": 5.3074,
+      "step": 9136
+    },
+    {
+      "epoch": 0.6128718516282657,
+      "grad_norm": 0.14680909000207615,
+      "learning_rate": 2e-05,
+      "loss": 5.3249,
+      "step": 9137
+    },
+    {
+      "epoch": 0.6129389274574907,
+      "grad_norm": 0.1541405253784886,
+      "learning_rate": 2e-05,
+      "loss": 5.3345,
+      "step": 9138
+    },
+    {
+      "epoch": 0.6130060032867156,
+      "grad_norm": 0.15823490659934453,
+      "learning_rate": 2e-05,
+      "loss": 5.4797,
+      "step": 9139
+    },
+    {
+      "epoch": 0.6130730791159406,
+      "grad_norm": 0.1494131713669312,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6131401549451655,
+      "grad_norm": 0.15406933289099953,
+      "learning_rate": 2e-05,
+      "loss": 5.5951,
+      "step": 9141
+    },
+    {
+      "epoch": 0.6132072307743904,
+      "grad_norm": 0.15443994536494,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 9142
+    },
+    {
+      "epoch": 0.6132743066036154,
+      "grad_norm": 0.15571181708184184,
+      "learning_rate": 2e-05,
+      "loss": 5.3561,
+      "step": 9143
+    },
+    {
+      "epoch": 0.6133413824328403,
+      "grad_norm": 0.14768372128980053,
+      "learning_rate": 2e-05,
+      "loss": 5.3452,
+      "step": 9144
+    },
+    {
+      "epoch": 0.6134084582620652,
+      "grad_norm": 0.15748398950384546,
+      "learning_rate": 2e-05,
+      "loss": 5.3302,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6134755340912902,
+      "grad_norm": 0.16027516483837476,
+      "learning_rate": 2e-05,
+      "loss": 5.4361,
+      "step": 9146
+    },
+    {
+      "epoch": 0.6135426099205151,
+      "grad_norm": 0.15179697674075243,
+      "learning_rate": 2e-05,
+      "loss": 5.486,
+      "step": 9147
+    },
+    {
+      "epoch": 0.6136096857497401,
+      "grad_norm": 0.1500418413761482,
+      "learning_rate": 2e-05,
+      "loss": 5.3066,
+      "step": 9148
+    },
+    {
+      "epoch": 0.613676761578965,
+      "grad_norm": 0.15111073926904983,
+      "learning_rate": 2e-05,
+      "loss": 5.296,
+      "step": 9149
+    },
+    {
+      "epoch": 0.6137438374081899,
+      "grad_norm": 0.15382534483414995,
+      "learning_rate": 2e-05,
+      "loss": 5.3167,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6138109132374149,
+      "grad_norm": 0.15374774724261073,
+      "learning_rate": 2e-05,
+      "loss": 5.3998,
+      "step": 9151
+    },
+    {
+      "epoch": 0.6138779890666398,
+      "grad_norm": 0.15318503736190864,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 9152
+    },
+    {
+      "epoch": 0.6139450648958648,
+      "grad_norm": 0.1493165652209338,
+      "learning_rate": 2e-05,
+      "loss": 5.4034,
+      "step": 9153
+    },
+    {
+      "epoch": 0.6140121407250897,
+      "grad_norm": 0.15424032851476385,
+      "learning_rate": 2e-05,
+      "loss": 5.347,
+      "step": 9154
+    },
+    {
+      "epoch": 0.6140792165543146,
+      "grad_norm": 0.15221929390080222,
+      "learning_rate": 2e-05,
+      "loss": 5.4533,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6141462923835396,
+      "grad_norm": 0.15539052833037667,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 9156
+    },
+    {
+      "epoch": 0.6142133682127645,
+      "grad_norm": 0.157220684531554,
+      "learning_rate": 2e-05,
+      "loss": 5.4942,
+      "step": 9157
+    },
+    {
+      "epoch": 0.6142804440419894,
+      "grad_norm": 0.15402803173374535,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 9158
+    },
+    {
+      "epoch": 0.6143475198712144,
+      "grad_norm": 0.16419979677466964,
+      "learning_rate": 2e-05,
+      "loss": 5.3061,
+      "step": 9159
+    },
+    {
+      "epoch": 0.6144145957004393,
+      "grad_norm": 0.14800888828731965,
+      "learning_rate": 2e-05,
+      "loss": 5.4912,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6144816715296643,
+      "grad_norm": 0.1597973529119849,
+      "learning_rate": 2e-05,
+      "loss": 5.4298,
+      "step": 9161
+    },
+    {
+      "epoch": 0.6145487473588892,
+      "grad_norm": 0.17011893324697344,
+      "learning_rate": 2e-05,
+      "loss": 5.4764,
+      "step": 9162
+    },
+    {
+      "epoch": 0.6146158231881141,
+      "grad_norm": 0.1521485767284959,
+      "learning_rate": 2e-05,
+      "loss": 5.366,
+      "step": 9163
+    },
+    {
+      "epoch": 0.6146828990173391,
+      "grad_norm": 0.16489657413204853,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 9164
+    },
+    {
+      "epoch": 0.614749974846564,
+      "grad_norm": 0.1540293993278931,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 9165
+    },
+    {
+      "epoch": 0.614817050675789,
+      "grad_norm": 0.15210382781931017,
+      "learning_rate": 2e-05,
+      "loss": 5.4041,
+      "step": 9166
+    },
+    {
+      "epoch": 0.6148841265050139,
+      "grad_norm": 0.15293146763932716,
+      "learning_rate": 2e-05,
+      "loss": 5.3049,
+      "step": 9167
+    },
+    {
+      "epoch": 0.6149512023342388,
+      "grad_norm": 0.15593983988422613,
+      "learning_rate": 2e-05,
+      "loss": 5.441,
+      "step": 9168
+    },
+    {
+      "epoch": 0.6150182781634638,
+      "grad_norm": 0.1532360120458184,
+      "learning_rate": 2e-05,
+      "loss": 5.3625,
+      "step": 9169
+    },
+    {
+      "epoch": 0.6150853539926887,
+      "grad_norm": 0.16935776361391852,
+      "learning_rate": 2e-05,
+      "loss": 5.3193,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6151524298219136,
+      "grad_norm": 0.1581791835917573,
+      "learning_rate": 2e-05,
+      "loss": 5.5189,
+      "step": 9171
+    },
+    {
+      "epoch": 0.6152195056511386,
+      "grad_norm": 0.14738615703022015,
+      "learning_rate": 2e-05,
+      "loss": 5.5655,
+      "step": 9172
+    },
+    {
+      "epoch": 0.6152865814803635,
+      "grad_norm": 0.15568800198597363,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 9173
+    },
+    {
+      "epoch": 0.6153536573095885,
+      "grad_norm": 0.16307742081524065,
+      "learning_rate": 2e-05,
+      "loss": 5.2742,
+      "step": 9174
+    },
+    {
+      "epoch": 0.6154207331388134,
+      "grad_norm": 0.15525873674863813,
+      "learning_rate": 2e-05,
+      "loss": 5.4988,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6154878089680383,
+      "grad_norm": 0.15378017188946866,
+      "learning_rate": 2e-05,
+      "loss": 5.3985,
+      "step": 9176
+    },
+    {
+      "epoch": 0.6155548847972633,
+      "grad_norm": 0.1525424685866855,
+      "learning_rate": 2e-05,
+      "loss": 5.4116,
+      "step": 9177
+    },
+    {
+      "epoch": 0.6156219606264882,
+      "grad_norm": 0.15353082993622258,
+      "learning_rate": 2e-05,
+      "loss": 5.4252,
+      "step": 9178
+    },
+    {
+      "epoch": 0.6156890364557132,
+      "grad_norm": 0.1636444629934033,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 9179
+    },
+    {
+      "epoch": 0.6157561122849381,
+      "grad_norm": 0.16259004282009284,
+      "learning_rate": 2e-05,
+      "loss": 5.5672,
+      "step": 9180
+    },
+    {
+      "epoch": 0.615823188114163,
+      "grad_norm": 0.15356234227749602,
+      "learning_rate": 2e-05,
+      "loss": 5.4028,
+      "step": 9181
+    },
+    {
+      "epoch": 0.615890263943388,
+      "grad_norm": 0.16315072343889223,
+      "learning_rate": 2e-05,
+      "loss": 5.4428,
+      "step": 9182
+    },
+    {
+      "epoch": 0.6159573397726129,
+      "grad_norm": 0.1531209964329347,
+      "learning_rate": 2e-05,
+      "loss": 5.4918,
+      "step": 9183
+    },
+    {
+      "epoch": 0.6160244156018378,
+      "grad_norm": 0.16462808043905916,
+      "learning_rate": 2e-05,
+      "loss": 5.293,
+      "step": 9184
+    },
+    {
+      "epoch": 0.6160914914310628,
+      "grad_norm": 0.16668685250676135,
+      "learning_rate": 2e-05,
+      "loss": 5.4314,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6161585672602877,
+      "grad_norm": 0.14928969428787342,
+      "learning_rate": 2e-05,
+      "loss": 5.3635,
+      "step": 9186
+    },
+    {
+      "epoch": 0.6162256430895127,
+      "grad_norm": 0.15361459018128665,
+      "learning_rate": 2e-05,
+      "loss": 5.4541,
+      "step": 9187
+    },
+    {
+      "epoch": 0.6162927189187376,
+      "grad_norm": 0.17997472912383478,
+      "learning_rate": 2e-05,
+      "loss": 5.4929,
+      "step": 9188
+    },
+    {
+      "epoch": 0.6163597947479625,
+      "grad_norm": 0.1629036070384782,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 9189
+    },
+    {
+      "epoch": 0.6164268705771875,
+      "grad_norm": 0.15471409708792733,
+      "learning_rate": 2e-05,
+      "loss": 5.3964,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6164939464064124,
+      "grad_norm": 0.15609054834066374,
+      "learning_rate": 2e-05,
+      "loss": 5.2624,
+      "step": 9191
+    },
+    {
+      "epoch": 0.6165610222356374,
+      "grad_norm": 0.1697376811645988,
+      "learning_rate": 2e-05,
+      "loss": 5.3928,
+      "step": 9192
+    },
+    {
+      "epoch": 0.6166280980648623,
+      "grad_norm": 0.17537273470810671,
+      "learning_rate": 2e-05,
+      "loss": 5.3882,
+      "step": 9193
+    },
+    {
+      "epoch": 0.6166951738940872,
+      "grad_norm": 0.15064326072853487,
+      "learning_rate": 2e-05,
+      "loss": 5.4437,
+      "step": 9194
+    },
+    {
+      "epoch": 0.6167622497233122,
+      "grad_norm": 0.14948902466964062,
+      "learning_rate": 2e-05,
+      "loss": 5.3491,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6168293255525371,
+      "grad_norm": 0.15226022696021418,
+      "learning_rate": 2e-05,
+      "loss": 5.3219,
+      "step": 9196
+    },
+    {
+      "epoch": 0.616896401381762,
+      "grad_norm": 0.1512232658810304,
+      "learning_rate": 2e-05,
+      "loss": 5.2696,
+      "step": 9197
+    },
+    {
+      "epoch": 0.616963477210987,
+      "grad_norm": 0.1585885283358905,
+      "learning_rate": 2e-05,
+      "loss": 5.3069,
+      "step": 9198
+    },
+    {
+      "epoch": 0.6170305530402119,
+      "grad_norm": 0.15886283176081692,
+      "learning_rate": 2e-05,
+      "loss": 5.515,
+      "step": 9199
+    },
+    {
+      "epoch": 0.6170976288694369,
+      "grad_norm": 0.14672717450671494,
+      "learning_rate": 2e-05,
+      "loss": 5.4465,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6171647046986618,
+      "grad_norm": 0.15278016071435357,
+      "learning_rate": 2e-05,
+      "loss": 5.4353,
+      "step": 9201
+    },
+    {
+      "epoch": 0.6172317805278867,
+      "grad_norm": 0.15298482521249743,
+      "learning_rate": 2e-05,
+      "loss": 5.4276,
+      "step": 9202
+    },
+    {
+      "epoch": 0.6172988563571117,
+      "grad_norm": 0.16141157631793196,
+      "learning_rate": 2e-05,
+      "loss": 5.2524,
+      "step": 9203
+    },
+    {
+      "epoch": 0.6173659321863366,
+      "grad_norm": 0.15451190266329898,
+      "learning_rate": 2e-05,
+      "loss": 5.3512,
+      "step": 9204
+    },
+    {
+      "epoch": 0.6174330080155616,
+      "grad_norm": 0.14920141198483572,
+      "learning_rate": 2e-05,
+      "loss": 5.3964,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6175000838447865,
+      "grad_norm": 0.1527512013819126,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 9206
+    },
+    {
+      "epoch": 0.6175671596740114,
+      "grad_norm": 0.1498084075649274,
+      "learning_rate": 2e-05,
+      "loss": 5.3973,
+      "step": 9207
+    },
+    {
+      "epoch": 0.6176342355032364,
+      "grad_norm": 0.15392815816301336,
+      "learning_rate": 2e-05,
+      "loss": 5.5096,
+      "step": 9208
+    },
+    {
+      "epoch": 0.6177013113324613,
+      "grad_norm": 0.14676743550634994,
+      "learning_rate": 2e-05,
+      "loss": 5.3687,
+      "step": 9209
+    },
+    {
+      "epoch": 0.6177683871616862,
+      "grad_norm": 0.14948323884342046,
+      "learning_rate": 2e-05,
+      "loss": 5.2948,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6178354629909112,
+      "grad_norm": 0.15157900123979853,
+      "learning_rate": 2e-05,
+      "loss": 5.4342,
+      "step": 9211
+    },
+    {
+      "epoch": 0.6179025388201361,
+      "grad_norm": 0.16800049085795848,
+      "learning_rate": 2e-05,
+      "loss": 5.3318,
+      "step": 9212
+    },
+    {
+      "epoch": 0.6179696146493611,
+      "grad_norm": 0.15366618215390754,
+      "learning_rate": 2e-05,
+      "loss": 5.3933,
+      "step": 9213
+    },
+    {
+      "epoch": 0.618036690478586,
+      "grad_norm": 0.15572275478577646,
+      "learning_rate": 2e-05,
+      "loss": 5.3284,
+      "step": 9214
+    },
+    {
+      "epoch": 0.6181037663078109,
+      "grad_norm": 0.15142441703883947,
+      "learning_rate": 2e-05,
+      "loss": 5.4482,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6181708421370359,
+      "grad_norm": 0.1529834639326079,
+      "learning_rate": 2e-05,
+      "loss": 5.3493,
+      "step": 9216
+    },
+    {
+      "epoch": 0.6182379179662608,
+      "grad_norm": 0.15201360547245965,
+      "learning_rate": 2e-05,
+      "loss": 5.4335,
+      "step": 9217
+    },
+    {
+      "epoch": 0.6183049937954858,
+      "grad_norm": 0.15457462402115332,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 9218
+    },
+    {
+      "epoch": 0.6183720696247107,
+      "grad_norm": 0.1518124341080101,
+      "learning_rate": 2e-05,
+      "loss": 5.3917,
+      "step": 9219
+    },
+    {
+      "epoch": 0.6184391454539356,
+      "grad_norm": 0.15535182796987596,
+      "learning_rate": 2e-05,
+      "loss": 5.3658,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6185062212831606,
+      "grad_norm": 0.16444999053383863,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 9221
+    },
+    {
+      "epoch": 0.6185732971123855,
+      "grad_norm": 0.1526042716308769,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 9222
+    },
+    {
+      "epoch": 0.6186403729416105,
+      "grad_norm": 0.16340618231599566,
+      "learning_rate": 2e-05,
+      "loss": 5.4576,
+      "step": 9223
+    },
+    {
+      "epoch": 0.6187074487708354,
+      "grad_norm": 0.16153597707123132,
+      "learning_rate": 2e-05,
+      "loss": 5.4772,
+      "step": 9224
+    },
+    {
+      "epoch": 0.6187745246000603,
+      "grad_norm": 0.15153120135365902,
+      "learning_rate": 2e-05,
+      "loss": 5.4392,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6188416004292853,
+      "grad_norm": 0.15098788384477516,
+      "learning_rate": 2e-05,
+      "loss": 5.2693,
+      "step": 9226
+    },
+    {
+      "epoch": 0.6189086762585102,
+      "grad_norm": 0.1540565714494348,
+      "learning_rate": 2e-05,
+      "loss": 5.3871,
+      "step": 9227
+    },
+    {
+      "epoch": 0.6189757520877351,
+      "grad_norm": 0.15104596744240856,
+      "learning_rate": 2e-05,
+      "loss": 5.393,
+      "step": 9228
+    },
+    {
+      "epoch": 0.6190428279169601,
+      "grad_norm": 0.14921236768779056,
+      "learning_rate": 2e-05,
+      "loss": 5.5381,
+      "step": 9229
+    },
+    {
+      "epoch": 0.619109903746185,
+      "grad_norm": 0.15771590560397958,
+      "learning_rate": 2e-05,
+      "loss": 5.481,
+      "step": 9230
+    },
+    {
+      "epoch": 0.61917697957541,
+      "grad_norm": 0.1578443168078229,
+      "learning_rate": 2e-05,
+      "loss": 5.3935,
+      "step": 9231
+    },
+    {
+      "epoch": 0.6192440554046349,
+      "grad_norm": 0.14986919684751107,
+      "learning_rate": 2e-05,
+      "loss": 5.4854,
+      "step": 9232
+    },
+    {
+      "epoch": 0.6193111312338598,
+      "grad_norm": 0.15676876361848163,
+      "learning_rate": 2e-05,
+      "loss": 5.456,
+      "step": 9233
+    },
+    {
+      "epoch": 0.6193782070630848,
+      "grad_norm": 0.15455955742815858,
+      "learning_rate": 2e-05,
+      "loss": 5.5174,
+      "step": 9234
+    },
+    {
+      "epoch": 0.6194452828923097,
+      "grad_norm": 0.15043327985351507,
+      "learning_rate": 2e-05,
+      "loss": 5.3538,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6195123587215347,
+      "grad_norm": 0.15639055410765967,
+      "learning_rate": 2e-05,
+      "loss": 5.4732,
+      "step": 9236
+    },
+    {
+      "epoch": 0.6195794345507596,
+      "grad_norm": 0.15436831995127073,
+      "learning_rate": 2e-05,
+      "loss": 5.5127,
+      "step": 9237
+    },
+    {
+      "epoch": 0.6196465103799845,
+      "grad_norm": 0.14929570031421652,
+      "learning_rate": 2e-05,
+      "loss": 5.3753,
+      "step": 9238
+    },
+    {
+      "epoch": 0.6197135862092095,
+      "grad_norm": 0.15364852076901048,
+      "learning_rate": 2e-05,
+      "loss": 5.4078,
+      "step": 9239
+    },
+    {
+      "epoch": 0.6197806620384344,
+      "grad_norm": 0.15715848040501815,
+      "learning_rate": 2e-05,
+      "loss": 5.4248,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6198477378676593,
+      "grad_norm": 0.14749368864097412,
+      "learning_rate": 2e-05,
+      "loss": 5.4272,
+      "step": 9241
+    },
+    {
+      "epoch": 0.6199148136968843,
+      "grad_norm": 0.14762313422403509,
+      "learning_rate": 2e-05,
+      "loss": 5.437,
+      "step": 9242
+    },
+    {
+      "epoch": 0.6199818895261092,
+      "grad_norm": 0.1587359808442164,
+      "learning_rate": 2e-05,
+      "loss": 5.3629,
+      "step": 9243
+    },
+    {
+      "epoch": 0.6200489653553342,
+      "grad_norm": 0.16115104102351155,
+      "learning_rate": 2e-05,
+      "loss": 5.3916,
+      "step": 9244
+    },
+    {
+      "epoch": 0.6201160411845591,
+      "grad_norm": 0.15933100754668286,
+      "learning_rate": 2e-05,
+      "loss": 5.3029,
+      "step": 9245
+    },
+    {
+      "epoch": 0.620183117013784,
+      "grad_norm": 0.14782824217767637,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 9246
+    },
+    {
+      "epoch": 0.620250192843009,
+      "grad_norm": 0.16213781171198488,
+      "learning_rate": 2e-05,
+      "loss": 5.4933,
+      "step": 9247
+    },
+    {
+      "epoch": 0.6203172686722339,
+      "grad_norm": 0.15485605934888458,
+      "learning_rate": 2e-05,
+      "loss": 5.3357,
+      "step": 9248
+    },
+    {
+      "epoch": 0.6203843445014589,
+      "grad_norm": 0.15376228692849342,
+      "learning_rate": 2e-05,
+      "loss": 5.3623,
+      "step": 9249
+    },
+    {
+      "epoch": 0.6204514203306838,
+      "grad_norm": 0.15119549331915058,
+      "learning_rate": 2e-05,
+      "loss": 5.5072,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6205184961599087,
+      "grad_norm": 0.15901887348639573,
+      "learning_rate": 2e-05,
+      "loss": 5.5507,
+      "step": 9251
+    },
+    {
+      "epoch": 0.6205855719891337,
+      "grad_norm": 0.14744370770448922,
+      "learning_rate": 2e-05,
+      "loss": 5.3758,
+      "step": 9252
+    },
+    {
+      "epoch": 0.6206526478183586,
+      "grad_norm": 0.14612306184471374,
+      "learning_rate": 2e-05,
+      "loss": 5.4871,
+      "step": 9253
+    },
+    {
+      "epoch": 0.6207197236475835,
+      "grad_norm": 0.1482184048563821,
+      "learning_rate": 2e-05,
+      "loss": 5.4165,
+      "step": 9254
+    },
+    {
+      "epoch": 0.6207867994768085,
+      "grad_norm": 0.1521540898506381,
+      "learning_rate": 2e-05,
+      "loss": 5.3466,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6208538753060334,
+      "grad_norm": 0.1548622721559832,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 9256
+    },
+    {
+      "epoch": 0.6209209511352584,
+      "grad_norm": 0.17481689682593565,
+      "learning_rate": 2e-05,
+      "loss": 5.3233,
+      "step": 9257
+    },
+    {
+      "epoch": 0.6209880269644833,
+      "grad_norm": 0.15291788497956313,
+      "learning_rate": 2e-05,
+      "loss": 5.3914,
+      "step": 9258
+    },
+    {
+      "epoch": 0.6210551027937082,
+      "grad_norm": 0.15673546123262427,
+      "learning_rate": 2e-05,
+      "loss": 5.4705,
+      "step": 9259
+    },
+    {
+      "epoch": 0.6211221786229332,
+      "grad_norm": 0.16818870275985726,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6211892544521581,
+      "grad_norm": 0.15751993701303263,
+      "learning_rate": 2e-05,
+      "loss": 5.3668,
+      "step": 9261
+    },
+    {
+      "epoch": 0.621256330281383,
+      "grad_norm": 0.15375324174477098,
+      "learning_rate": 2e-05,
+      "loss": 5.453,
+      "step": 9262
+    },
+    {
+      "epoch": 0.621323406110608,
+      "grad_norm": 0.1622929007578182,
+      "learning_rate": 2e-05,
+      "loss": 5.557,
+      "step": 9263
+    },
+    {
+      "epoch": 0.6213904819398329,
+      "grad_norm": 0.15502368185657472,
+      "learning_rate": 2e-05,
+      "loss": 5.4975,
+      "step": 9264
+    },
+    {
+      "epoch": 0.6214575577690579,
+      "grad_norm": 0.15721823452445124,
+      "learning_rate": 2e-05,
+      "loss": 5.3247,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6215246335982828,
+      "grad_norm": 0.15609998897000488,
+      "learning_rate": 2e-05,
+      "loss": 5.4149,
+      "step": 9266
+    },
+    {
+      "epoch": 0.6215917094275077,
+      "grad_norm": 0.15346174518286448,
+      "learning_rate": 2e-05,
+      "loss": 5.4371,
+      "step": 9267
+    },
+    {
+      "epoch": 0.6216587852567327,
+      "grad_norm": 0.1504513946090627,
+      "learning_rate": 2e-05,
+      "loss": 5.4975,
+      "step": 9268
+    },
+    {
+      "epoch": 0.6217258610859576,
+      "grad_norm": 0.14606802071425437,
+      "learning_rate": 2e-05,
+      "loss": 5.4081,
+      "step": 9269
+    },
+    {
+      "epoch": 0.6217929369151826,
+      "grad_norm": 0.15402696060470902,
+      "learning_rate": 2e-05,
+      "loss": 5.4853,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6218600127444075,
+      "grad_norm": 0.16424940032249008,
+      "learning_rate": 2e-05,
+      "loss": 5.3883,
+      "step": 9271
+    },
+    {
+      "epoch": 0.6219270885736324,
+      "grad_norm": 0.15124918400644813,
+      "learning_rate": 2e-05,
+      "loss": 5.3673,
+      "step": 9272
+    },
+    {
+      "epoch": 0.6219941644028575,
+      "grad_norm": 0.15337658003638624,
+      "learning_rate": 2e-05,
+      "loss": 5.5385,
+      "step": 9273
+    },
+    {
+      "epoch": 0.6220612402320824,
+      "grad_norm": 0.1590592487034579,
+      "learning_rate": 2e-05,
+      "loss": 5.5259,
+      "step": 9274
+    },
+    {
+      "epoch": 0.6221283160613074,
+      "grad_norm": 0.14937191892098028,
+      "learning_rate": 2e-05,
+      "loss": 5.554,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6221953918905323,
+      "grad_norm": 0.1542872346044075,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 9276
+    },
+    {
+      "epoch": 0.6222624677197572,
+      "grad_norm": 0.15748514181131734,
+      "learning_rate": 2e-05,
+      "loss": 5.2733,
+      "step": 9277
+    },
+    {
+      "epoch": 0.6223295435489822,
+      "grad_norm": 0.1510330384999685,
+      "learning_rate": 2e-05,
+      "loss": 5.4119,
+      "step": 9278
+    },
+    {
+      "epoch": 0.6223966193782071,
+      "grad_norm": 0.14804926921760905,
+      "learning_rate": 2e-05,
+      "loss": 5.4605,
+      "step": 9279
+    },
+    {
+      "epoch": 0.622463695207432,
+      "grad_norm": 0.16910167875388965,
+      "learning_rate": 2e-05,
+      "loss": 5.4127,
+      "step": 9280
+    },
+    {
+      "epoch": 0.622530771036657,
+      "grad_norm": 0.1474651370154078,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 9281
+    },
+    {
+      "epoch": 0.6225978468658819,
+      "grad_norm": 0.15138118460717043,
+      "learning_rate": 2e-05,
+      "loss": 5.3828,
+      "step": 9282
+    },
+    {
+      "epoch": 0.6226649226951069,
+      "grad_norm": 0.1571288776006305,
+      "learning_rate": 2e-05,
+      "loss": 5.4088,
+      "step": 9283
+    },
+    {
+      "epoch": 0.6227319985243318,
+      "grad_norm": 0.15855090347045925,
+      "learning_rate": 2e-05,
+      "loss": 5.4448,
+      "step": 9284
+    },
+    {
+      "epoch": 0.6227990743535567,
+      "grad_norm": 0.1547046213668634,
+      "learning_rate": 2e-05,
+      "loss": 5.3135,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6228661501827817,
+      "grad_norm": 0.1517780795349633,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 9286
+    },
+    {
+      "epoch": 0.6229332260120066,
+      "grad_norm": 0.14368262950654156,
+      "learning_rate": 2e-05,
+      "loss": 5.3117,
+      "step": 9287
+    },
+    {
+      "epoch": 0.6230003018412316,
+      "grad_norm": 0.15305427608995553,
+      "learning_rate": 2e-05,
+      "loss": 5.5155,
+      "step": 9288
+    },
+    {
+      "epoch": 0.6230673776704565,
+      "grad_norm": 0.1584626184778909,
+      "learning_rate": 2e-05,
+      "loss": 5.5512,
+      "step": 9289
+    },
+    {
+      "epoch": 0.6231344534996814,
+      "grad_norm": 0.1508127855361275,
+      "learning_rate": 2e-05,
+      "loss": 5.4806,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6232015293289064,
+      "grad_norm": 0.1517974134630336,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 9291
+    },
+    {
+      "epoch": 0.6232686051581313,
+      "grad_norm": 0.14998233476124237,
+      "learning_rate": 2e-05,
+      "loss": 5.4461,
+      "step": 9292
+    },
+    {
+      "epoch": 0.6233356809873563,
+      "grad_norm": 0.15497538004340997,
+      "learning_rate": 2e-05,
+      "loss": 5.5588,
+      "step": 9293
+    },
+    {
+      "epoch": 0.6234027568165812,
+      "grad_norm": 0.1483266854991414,
+      "learning_rate": 2e-05,
+      "loss": 5.3464,
+      "step": 9294
+    },
+    {
+      "epoch": 0.6234698326458061,
+      "grad_norm": 0.14975399212582835,
+      "learning_rate": 2e-05,
+      "loss": 5.2911,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6235369084750311,
+      "grad_norm": 0.1564589109424068,
+      "learning_rate": 2e-05,
+      "loss": 5.5397,
+      "step": 9296
+    },
+    {
+      "epoch": 0.623603984304256,
+      "grad_norm": 0.15472077168768686,
+      "learning_rate": 2e-05,
+      "loss": 5.5971,
+      "step": 9297
+    },
+    {
+      "epoch": 0.623671060133481,
+      "grad_norm": 0.1448330043482016,
+      "learning_rate": 2e-05,
+      "loss": 5.3045,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6237381359627059,
+      "grad_norm": 0.15211393130395262,
+      "learning_rate": 2e-05,
+      "loss": 5.3359,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6238052117919308,
+      "grad_norm": 0.1567597430859764,
+      "learning_rate": 2e-05,
+      "loss": 5.499,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6238722876211558,
+      "grad_norm": 0.14699951264427674,
+      "learning_rate": 2e-05,
+      "loss": 5.3815,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6239393634503807,
+      "grad_norm": 0.15102526007471634,
+      "learning_rate": 2e-05,
+      "loss": 5.3734,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6240064392796056,
+      "grad_norm": 0.15770749910870713,
+      "learning_rate": 2e-05,
+      "loss": 5.3351,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6240735151088306,
+      "grad_norm": 0.15844481235934602,
+      "learning_rate": 2e-05,
+      "loss": 5.3105,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6241405909380555,
+      "grad_norm": 0.1565566434540911,
+      "learning_rate": 2e-05,
+      "loss": 5.3022,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6242076667672805,
+      "grad_norm": 0.16265464155376633,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6242747425965054,
+      "grad_norm": 0.15587220510525404,
+      "learning_rate": 2e-05,
+      "loss": 5.469,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6243418184257303,
+      "grad_norm": 0.15207698322516885,
+      "learning_rate": 2e-05,
+      "loss": 5.3421,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6244088942549553,
+      "grad_norm": 0.1511669358200155,
+      "learning_rate": 2e-05,
+      "loss": 5.3577,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6244759700841802,
+      "grad_norm": 0.14887184330129727,
+      "learning_rate": 2e-05,
+      "loss": 5.4598,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6245430459134051,
+      "grad_norm": 0.15150995394115577,
+      "learning_rate": 2e-05,
+      "loss": 5.5094,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6246101217426301,
+      "grad_norm": 0.14907178229915738,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 9312
+    },
+    {
+      "epoch": 0.624677197571855,
+      "grad_norm": 0.15568755004652848,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 9313
+    },
+    {
+      "epoch": 0.62474427340108,
+      "grad_norm": 0.1465566681168972,
+      "learning_rate": 2e-05,
+      "loss": 5.3034,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6248113492303049,
+      "grad_norm": 0.14822205795101306,
+      "learning_rate": 2e-05,
+      "loss": 5.426,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6248784250595298,
+      "grad_norm": 0.14981490514794482,
+      "learning_rate": 2e-05,
+      "loss": 5.5296,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6249455008887548,
+      "grad_norm": 0.15323454872852135,
+      "learning_rate": 2e-05,
+      "loss": 5.3426,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6250125767179797,
+      "grad_norm": 0.1529639777033927,
+      "learning_rate": 2e-05,
+      "loss": 5.5532,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6250796525472047,
+      "grad_norm": 0.15488385321554574,
+      "learning_rate": 2e-05,
+      "loss": 5.3381,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6251467283764296,
+      "grad_norm": 0.16030145817703134,
+      "learning_rate": 2e-05,
+      "loss": 5.4267,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6252138042056545,
+      "grad_norm": 0.1503653380274123,
+      "learning_rate": 2e-05,
+      "loss": 5.4494,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6252808800348795,
+      "grad_norm": 0.14779574602333148,
+      "learning_rate": 2e-05,
+      "loss": 5.483,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6253479558641044,
+      "grad_norm": 0.15192951045621747,
+      "learning_rate": 2e-05,
+      "loss": 5.4185,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6254150316933293,
+      "grad_norm": 0.15099471116304397,
+      "learning_rate": 2e-05,
+      "loss": 5.6855,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6254821075225543,
+      "grad_norm": 0.1563171986281022,
+      "learning_rate": 2e-05,
+      "loss": 5.3855,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6255491833517792,
+      "grad_norm": 0.15281134212851236,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6256162591810042,
+      "grad_norm": 0.15003369900742974,
+      "learning_rate": 2e-05,
+      "loss": 5.3597,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6256833350102291,
+      "grad_norm": 0.1555422667583891,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 9328
+    },
+    {
+      "epoch": 0.625750410839454,
+      "grad_norm": 0.1501022977513605,
+      "learning_rate": 2e-05,
+      "loss": 5.4525,
+      "step": 9329
+    },
+    {
+      "epoch": 0.625817486668679,
+      "grad_norm": 0.1484074411045815,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6258845624979039,
+      "grad_norm": 0.16300562517744996,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6259516383271289,
+      "grad_norm": 0.1514167359245985,
+      "learning_rate": 2e-05,
+      "loss": 5.3992,
+      "step": 9332
+    },
+    {
+      "epoch": 0.6260187141563538,
+      "grad_norm": 0.14752022984956242,
+      "learning_rate": 2e-05,
+      "loss": 5.5039,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6260857899855787,
+      "grad_norm": 0.16020043285345364,
+      "learning_rate": 2e-05,
+      "loss": 5.5195,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6261528658148037,
+      "grad_norm": 0.1519596144815404,
+      "learning_rate": 2e-05,
+      "loss": 5.3951,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6262199416440286,
+      "grad_norm": 0.1516130458819609,
+      "learning_rate": 2e-05,
+      "loss": 5.5023,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6262870174732535,
+      "grad_norm": 0.14887899339512722,
+      "learning_rate": 2e-05,
+      "loss": 5.3689,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6263540933024785,
+      "grad_norm": 0.15406302780116657,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6264211691317034,
+      "grad_norm": 0.15250107040792965,
+      "learning_rate": 2e-05,
+      "loss": 5.4585,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6264882449609284,
+      "grad_norm": 0.14720056301419343,
+      "learning_rate": 2e-05,
+      "loss": 5.3737,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6265553207901533,
+      "grad_norm": 0.15675190358722718,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6266223966193782,
+      "grad_norm": 0.1614702933134975,
+      "learning_rate": 2e-05,
+      "loss": 5.2947,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6266894724486032,
+      "grad_norm": 0.15340158128245057,
+      "learning_rate": 2e-05,
+      "loss": 5.3819,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6267565482778281,
+      "grad_norm": 0.17926965245092993,
+      "learning_rate": 2e-05,
+      "loss": 5.4628,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6268236241070531,
+      "grad_norm": 0.15145428590246568,
+      "learning_rate": 2e-05,
+      "loss": 5.3562,
+      "step": 9345
+    },
+    {
+      "epoch": 0.626890699936278,
+      "grad_norm": 0.1621878731263437,
+      "learning_rate": 2e-05,
+      "loss": 5.504,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6269577757655029,
+      "grad_norm": 0.15190778320919965,
+      "learning_rate": 2e-05,
+      "loss": 5.3607,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6270248515947279,
+      "grad_norm": 0.14957491826710362,
+      "learning_rate": 2e-05,
+      "loss": 5.4087,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6270919274239528,
+      "grad_norm": 0.15539299028143083,
+      "learning_rate": 2e-05,
+      "loss": 5.3471,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6271590032531777,
+      "grad_norm": 0.15994494937005252,
+      "learning_rate": 2e-05,
+      "loss": 5.4628,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6272260790824027,
+      "grad_norm": 0.15472053207213424,
+      "learning_rate": 2e-05,
+      "loss": 5.2943,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6272931549116276,
+      "grad_norm": 0.15798005667392523,
+      "learning_rate": 2e-05,
+      "loss": 5.4784,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6273602307408526,
+      "grad_norm": 0.15387718139426837,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6274273065700775,
+      "grad_norm": 0.15518824759680933,
+      "learning_rate": 2e-05,
+      "loss": 5.4795,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6274943823993024,
+      "grad_norm": 0.15289974895365846,
+      "learning_rate": 2e-05,
+      "loss": 5.444,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6275614582285274,
+      "grad_norm": 0.1472427886484463,
+      "learning_rate": 2e-05,
+      "loss": 5.4142,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6276285340577523,
+      "grad_norm": 0.16893801163157196,
+      "learning_rate": 2e-05,
+      "loss": 5.4421,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6276956098869773,
+      "grad_norm": 0.15474496836144508,
+      "learning_rate": 2e-05,
+      "loss": 5.3529,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6277626857162022,
+      "grad_norm": 0.15415978384005488,
+      "learning_rate": 2e-05,
+      "loss": 5.4808,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6278297615454271,
+      "grad_norm": 0.1592805460009479,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6278968373746521,
+      "grad_norm": 0.16600374041049634,
+      "learning_rate": 2e-05,
+      "loss": 5.4051,
+      "step": 9361
+    },
+    {
+      "epoch": 0.627963913203877,
+      "grad_norm": 0.15012955019120713,
+      "learning_rate": 2e-05,
+      "loss": 5.3874,
+      "step": 9362
+    },
+    {
+      "epoch": 0.628030989033102,
+      "grad_norm": 0.15497144827769838,
+      "learning_rate": 2e-05,
+      "loss": 5.4345,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6280980648623269,
+      "grad_norm": 0.1540268023660924,
+      "learning_rate": 2e-05,
+      "loss": 5.4256,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6281651406915518,
+      "grad_norm": 0.15150818998291563,
+      "learning_rate": 2e-05,
+      "loss": 5.4337,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6282322165207768,
+      "grad_norm": 0.15265461037526573,
+      "learning_rate": 2e-05,
+      "loss": 5.3603,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6282992923500017,
+      "grad_norm": 0.15127138597168155,
+      "learning_rate": 2e-05,
+      "loss": 5.4463,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6283663681792266,
+      "grad_norm": 0.1567021146851781,
+      "learning_rate": 2e-05,
+      "loss": 5.3723,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6284334440084516,
+      "grad_norm": 0.15288547138819844,
+      "learning_rate": 2e-05,
+      "loss": 5.4472,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6285005198376765,
+      "grad_norm": 0.1467130241723948,
+      "learning_rate": 2e-05,
+      "loss": 5.4234,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6285675956669015,
+      "grad_norm": 0.1576888183639012,
+      "learning_rate": 2e-05,
+      "loss": 5.4393,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6286346714961264,
+      "grad_norm": 0.15214648389781002,
+      "learning_rate": 2e-05,
+      "loss": 5.426,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6287017473253513,
+      "grad_norm": 0.15645788855709467,
+      "learning_rate": 2e-05,
+      "loss": 5.5096,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6287688231545763,
+      "grad_norm": 0.15269193901492767,
+      "learning_rate": 2e-05,
+      "loss": 5.2892,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6288358989838012,
+      "grad_norm": 0.15158076893113573,
+      "learning_rate": 2e-05,
+      "loss": 5.3594,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6289029748130261,
+      "grad_norm": 0.15525635080639794,
+      "learning_rate": 2e-05,
+      "loss": 5.5074,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6289700506422511,
+      "grad_norm": 0.1486004539384565,
+      "learning_rate": 2e-05,
+      "loss": 5.4829,
+      "step": 9377
+    },
+    {
+      "epoch": 0.629037126471476,
+      "grad_norm": 0.15083412180350117,
+      "learning_rate": 2e-05,
+      "loss": 5.332,
+      "step": 9378
+    },
+    {
+      "epoch": 0.629104202300701,
+      "grad_norm": 0.14951350870373342,
+      "learning_rate": 2e-05,
+      "loss": 5.3728,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6291712781299259,
+      "grad_norm": 0.15370659940179374,
+      "learning_rate": 2e-05,
+      "loss": 5.5014,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6292383539591508,
+      "grad_norm": 0.16019248966651564,
+      "learning_rate": 2e-05,
+      "loss": 5.3386,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6293054297883758,
+      "grad_norm": 0.15847224863165632,
+      "learning_rate": 2e-05,
+      "loss": 5.3947,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6293725056176007,
+      "grad_norm": 0.15346349043062782,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6294395814468257,
+      "grad_norm": 0.1537529103349642,
+      "learning_rate": 2e-05,
+      "loss": 5.331,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6295066572760506,
+      "grad_norm": 0.1574282192185749,
+      "learning_rate": 2e-05,
+      "loss": 5.3539,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6295737331052755,
+      "grad_norm": 0.15070565851796658,
+      "learning_rate": 2e-05,
+      "loss": 5.3835,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6296408089345005,
+      "grad_norm": 0.15006207444228298,
+      "learning_rate": 2e-05,
+      "loss": 5.4859,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6297078847637254,
+      "grad_norm": 0.15243322909477694,
+      "learning_rate": 2e-05,
+      "loss": 5.4736,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6297749605929504,
+      "grad_norm": 0.15442853966619624,
+      "learning_rate": 2e-05,
+      "loss": 5.3382,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6298420364221753,
+      "grad_norm": 0.16252676881111652,
+      "learning_rate": 2e-05,
+      "loss": 5.4422,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6299091122514002,
+      "grad_norm": 0.16616750900366756,
+      "learning_rate": 2e-05,
+      "loss": 5.439,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6299761880806252,
+      "grad_norm": 0.1527117511568586,
+      "learning_rate": 2e-05,
+      "loss": 5.4695,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6300432639098501,
+      "grad_norm": 0.15662737693165685,
+      "learning_rate": 2e-05,
+      "loss": 5.3748,
+      "step": 9393
+    },
+    {
+      "epoch": 0.630110339739075,
+      "grad_norm": 0.1506878953351132,
+      "learning_rate": 2e-05,
+      "loss": 5.4502,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6301774155683,
+      "grad_norm": 0.15252199467056696,
+      "learning_rate": 2e-05,
+      "loss": 5.4299,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6302444913975249,
+      "grad_norm": 0.15720438283149735,
+      "learning_rate": 2e-05,
+      "loss": 5.5455,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6303115672267499,
+      "grad_norm": 0.1471050117022988,
+      "learning_rate": 2e-05,
+      "loss": 5.4685,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6303786430559748,
+      "grad_norm": 0.1555049628220281,
+      "learning_rate": 2e-05,
+      "loss": 5.6692,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6304457188851997,
+      "grad_norm": 0.15744116662361604,
+      "learning_rate": 2e-05,
+      "loss": 5.466,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6305127947144247,
+      "grad_norm": 0.1486556196895531,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6305798705436496,
+      "grad_norm": 0.1568213523364497,
+      "learning_rate": 2e-05,
+      "loss": 5.4821,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6306469463728746,
+      "grad_norm": 0.15460415691534965,
+      "learning_rate": 2e-05,
+      "loss": 5.3509,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6307140222020995,
+      "grad_norm": 0.15183490097101865,
+      "learning_rate": 2e-05,
+      "loss": 5.5939,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6307810980313244,
+      "grad_norm": 0.1530798019791715,
+      "learning_rate": 2e-05,
+      "loss": 5.5448,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6308481738605494,
+      "grad_norm": 0.15789831841461469,
+      "learning_rate": 2e-05,
+      "loss": 5.5343,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6309152496897743,
+      "grad_norm": 0.15718429053894867,
+      "learning_rate": 2e-05,
+      "loss": 5.453,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6309823255189992,
+      "grad_norm": 0.1504412383842006,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6310494013482242,
+      "grad_norm": 0.1592730106985513,
+      "learning_rate": 2e-05,
+      "loss": 5.4972,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6311164771774491,
+      "grad_norm": 0.1535000130902462,
+      "learning_rate": 2e-05,
+      "loss": 5.4063,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6311835530066741,
+      "grad_norm": 0.15234881051960175,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 9410
+    },
+    {
+      "epoch": 0.631250628835899,
+      "grad_norm": 0.15367580365531822,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6313177046651239,
+      "grad_norm": 0.14942708984989578,
+      "learning_rate": 2e-05,
+      "loss": 5.3204,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6313847804943489,
+      "grad_norm": 0.15237663057447956,
+      "learning_rate": 2e-05,
+      "loss": 5.3596,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6314518563235738,
+      "grad_norm": 0.15039014801624928,
+      "learning_rate": 2e-05,
+      "loss": 5.3852,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6315189321527988,
+      "grad_norm": 0.1418609579568539,
+      "learning_rate": 2e-05,
+      "loss": 5.371,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6315860079820237,
+      "grad_norm": 0.16102950730656843,
+      "learning_rate": 2e-05,
+      "loss": 5.3,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6316530838112486,
+      "grad_norm": 0.15411501987330056,
+      "learning_rate": 2e-05,
+      "loss": 5.427,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6317201596404736,
+      "grad_norm": 0.14351749212136589,
+      "learning_rate": 2e-05,
+      "loss": 5.4073,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6317872354696985,
+      "grad_norm": 0.14243729819003836,
+      "learning_rate": 2e-05,
+      "loss": 5.554,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6318543112989234,
+      "grad_norm": 0.15725739367078473,
+      "learning_rate": 2e-05,
+      "loss": 5.4893,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6319213871281484,
+      "grad_norm": 0.15029349160834918,
+      "learning_rate": 2e-05,
+      "loss": 5.4891,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6319884629573733,
+      "grad_norm": 0.15233995251626753,
+      "learning_rate": 2e-05,
+      "loss": 5.2992,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6320555387865983,
+      "grad_norm": 0.15312730645058026,
+      "learning_rate": 2e-05,
+      "loss": 5.3665,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6321226146158232,
+      "grad_norm": 0.14902246635430225,
+      "learning_rate": 2e-05,
+      "loss": 5.3636,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6321896904450481,
+      "grad_norm": 0.14694468050861317,
+      "learning_rate": 2e-05,
+      "loss": 5.36,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6322567662742731,
+      "grad_norm": 0.15030944347252506,
+      "learning_rate": 2e-05,
+      "loss": 5.5301,
+      "step": 9426
+    },
+    {
+      "epoch": 0.632323842103498,
+      "grad_norm": 0.15706323630397048,
+      "learning_rate": 2e-05,
+      "loss": 5.3766,
+      "step": 9427
+    },
+    {
+      "epoch": 0.632390917932723,
+      "grad_norm": 0.14588139423171734,
+      "learning_rate": 2e-05,
+      "loss": 5.4507,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6324579937619479,
+      "grad_norm": 0.1482794940889182,
+      "learning_rate": 2e-05,
+      "loss": 5.4116,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6325250695911728,
+      "grad_norm": 0.14840695331562095,
+      "learning_rate": 2e-05,
+      "loss": 5.4227,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6325921454203978,
+      "grad_norm": 0.15174555783509674,
+      "learning_rate": 2e-05,
+      "loss": 5.5166,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6326592212496227,
+      "grad_norm": 0.15126502350126086,
+      "learning_rate": 2e-05,
+      "loss": 5.559,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6327262970788476,
+      "grad_norm": 0.14968732015439148,
+      "learning_rate": 2e-05,
+      "loss": 5.4261,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6327933729080726,
+      "grad_norm": 0.15205844697116036,
+      "learning_rate": 2e-05,
+      "loss": 5.5033,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6328604487372975,
+      "grad_norm": 0.15358327080389733,
+      "learning_rate": 2e-05,
+      "loss": 5.3541,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6329275245665225,
+      "grad_norm": 0.15705775961784946,
+      "learning_rate": 2e-05,
+      "loss": 5.5333,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6329946003957474,
+      "grad_norm": 0.15528950162057042,
+      "learning_rate": 2e-05,
+      "loss": 5.5916,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6330616762249723,
+      "grad_norm": 0.15708728869075786,
+      "learning_rate": 2e-05,
+      "loss": 5.4459,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6331287520541973,
+      "grad_norm": 0.14953097192377257,
+      "learning_rate": 2e-05,
+      "loss": 5.4639,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6331958278834222,
+      "grad_norm": 0.15699397437271584,
+      "learning_rate": 2e-05,
+      "loss": 5.3605,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6332629037126472,
+      "grad_norm": 0.15133602323775253,
+      "learning_rate": 2e-05,
+      "loss": 5.4004,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6333299795418721,
+      "grad_norm": 0.15769084881285195,
+      "learning_rate": 2e-05,
+      "loss": 5.3081,
+      "step": 9442
+    },
+    {
+      "epoch": 0.633397055371097,
+      "grad_norm": 0.15734582283651571,
+      "learning_rate": 2e-05,
+      "loss": 5.3784,
+      "step": 9443
+    },
+    {
+      "epoch": 0.633464131200322,
+      "grad_norm": 0.15602718562233542,
+      "learning_rate": 2e-05,
+      "loss": 5.3265,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6335312070295469,
+      "grad_norm": 0.15401118300491937,
+      "learning_rate": 2e-05,
+      "loss": 5.4874,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6335982828587718,
+      "grad_norm": 0.16075835087560125,
+      "learning_rate": 2e-05,
+      "loss": 5.5627,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6336653586879968,
+      "grad_norm": 0.15196628829201494,
+      "learning_rate": 2e-05,
+      "loss": 5.3497,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6337324345172217,
+      "grad_norm": 0.15940365948993876,
+      "learning_rate": 2e-05,
+      "loss": 5.3766,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6337995103464467,
+      "grad_norm": 0.1584771802454168,
+      "learning_rate": 2e-05,
+      "loss": 5.3804,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6338665861756716,
+      "grad_norm": 0.15979137578851854,
+      "learning_rate": 2e-05,
+      "loss": 5.4129,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6339336620048965,
+      "grad_norm": 0.16005521198963915,
+      "learning_rate": 2e-05,
+      "loss": 5.3559,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6340007378341215,
+      "grad_norm": 0.16101149541184528,
+      "learning_rate": 2e-05,
+      "loss": 5.4141,
+      "step": 9452
+    },
+    {
+      "epoch": 0.6340678136633464,
+      "grad_norm": 0.15983234360810777,
+      "learning_rate": 2e-05,
+      "loss": 5.3385,
+      "step": 9453
+    },
+    {
+      "epoch": 0.6341348894925714,
+      "grad_norm": 0.18166958649328357,
+      "learning_rate": 2e-05,
+      "loss": 5.2923,
+      "step": 9454
+    },
+    {
+      "epoch": 0.6342019653217963,
+      "grad_norm": 0.14818868662644574,
+      "learning_rate": 2e-05,
+      "loss": 5.3421,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6342690411510212,
+      "grad_norm": 0.1695367884009991,
+      "learning_rate": 2e-05,
+      "loss": 5.5122,
+      "step": 9456
+    },
+    {
+      "epoch": 0.6343361169802462,
+      "grad_norm": 0.16584590147701403,
+      "learning_rate": 2e-05,
+      "loss": 5.4734,
+      "step": 9457
+    },
+    {
+      "epoch": 0.6344031928094711,
+      "grad_norm": 0.16215749559117007,
+      "learning_rate": 2e-05,
+      "loss": 5.5513,
+      "step": 9458
+    },
+    {
+      "epoch": 0.634470268638696,
+      "grad_norm": 0.15551186833625985,
+      "learning_rate": 2e-05,
+      "loss": 5.4174,
+      "step": 9459
+    },
+    {
+      "epoch": 0.634537344467921,
+      "grad_norm": 0.17175033773750564,
+      "learning_rate": 2e-05,
+      "loss": 5.365,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6346044202971459,
+      "grad_norm": 0.1569553621318592,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6346714961263709,
+      "grad_norm": 0.15395521707328289,
+      "learning_rate": 2e-05,
+      "loss": 5.4681,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6347385719555958,
+      "grad_norm": 0.1524574915229782,
+      "learning_rate": 2e-05,
+      "loss": 5.48,
+      "step": 9463
+    },
+    {
+      "epoch": 0.6348056477848207,
+      "grad_norm": 0.15936305915168625,
+      "learning_rate": 2e-05,
+      "loss": 5.4239,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6348727236140457,
+      "grad_norm": 0.15501195137523027,
+      "learning_rate": 2e-05,
+      "loss": 5.6107,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6349397994432706,
+      "grad_norm": 0.15434577165326535,
+      "learning_rate": 2e-05,
+      "loss": 5.3584,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6350068752724956,
+      "grad_norm": 0.15862561652937449,
+      "learning_rate": 2e-05,
+      "loss": 5.4494,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6350739511017205,
+      "grad_norm": 0.15190057425290135,
+      "learning_rate": 2e-05,
+      "loss": 5.5224,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6351410269309454,
+      "grad_norm": 0.15353054523840573,
+      "learning_rate": 2e-05,
+      "loss": 5.4595,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6352081027601704,
+      "grad_norm": 0.15367267412943597,
+      "learning_rate": 2e-05,
+      "loss": 5.3996,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6352751785893953,
+      "grad_norm": 0.16337542946342778,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 9471
+    },
+    {
+      "epoch": 0.6353422544186202,
+      "grad_norm": 0.15562022561747302,
+      "learning_rate": 2e-05,
+      "loss": 5.4376,
+      "step": 9472
+    },
+    {
+      "epoch": 0.6354093302478452,
+      "grad_norm": 0.15786533453473975,
+      "learning_rate": 2e-05,
+      "loss": 5.2558,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6354764060770701,
+      "grad_norm": 0.1517233356145364,
+      "learning_rate": 2e-05,
+      "loss": 5.4219,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6355434819062951,
+      "grad_norm": 0.15267037078905313,
+      "learning_rate": 2e-05,
+      "loss": 5.3273,
+      "step": 9475
+    },
+    {
+      "epoch": 0.63561055773552,
+      "grad_norm": 0.1645452771582934,
+      "learning_rate": 2e-05,
+      "loss": 5.4951,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6356776335647449,
+      "grad_norm": 0.15058177062032424,
+      "learning_rate": 2e-05,
+      "loss": 5.387,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6357447093939699,
+      "grad_norm": 0.15680853032346234,
+      "learning_rate": 2e-05,
+      "loss": 5.4198,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6358117852231948,
+      "grad_norm": 0.1544590852365142,
+      "learning_rate": 2e-05,
+      "loss": 5.3767,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6358788610524198,
+      "grad_norm": 0.15594211265113536,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6359459368816447,
+      "grad_norm": 0.15499325831077612,
+      "learning_rate": 2e-05,
+      "loss": 5.2871,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6360130127108696,
+      "grad_norm": 0.15216954344933586,
+      "learning_rate": 2e-05,
+      "loss": 5.4468,
+      "step": 9482
+    },
+    {
+      "epoch": 0.6360800885400946,
+      "grad_norm": 0.15656687948172257,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6361471643693195,
+      "grad_norm": 0.15611668124033298,
+      "learning_rate": 2e-05,
+      "loss": 5.4402,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6362142401985444,
+      "grad_norm": 0.14803703945776656,
+      "learning_rate": 2e-05,
+      "loss": 5.3601,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6362813160277694,
+      "grad_norm": 0.1523071481143449,
+      "learning_rate": 2e-05,
+      "loss": 5.3994,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6363483918569943,
+      "grad_norm": 0.16388495064251501,
+      "learning_rate": 2e-05,
+      "loss": 5.3718,
+      "step": 9487
+    },
+    {
+      "epoch": 0.6364154676862193,
+      "grad_norm": 0.16095607556286004,
+      "learning_rate": 2e-05,
+      "loss": 5.4756,
+      "step": 9488
+    },
+    {
+      "epoch": 0.6364825435154442,
+      "grad_norm": 0.16216210804809375,
+      "learning_rate": 2e-05,
+      "loss": 5.3361,
+      "step": 9489
+    },
+    {
+      "epoch": 0.6365496193446691,
+      "grad_norm": 0.16532660689493436,
+      "learning_rate": 2e-05,
+      "loss": 5.5006,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6366166951738941,
+      "grad_norm": 0.1566142964465771,
+      "learning_rate": 2e-05,
+      "loss": 5.4268,
+      "step": 9491
+    },
+    {
+      "epoch": 0.636683771003119,
+      "grad_norm": 0.15885670825423792,
+      "learning_rate": 2e-05,
+      "loss": 5.372,
+      "step": 9492
+    },
+    {
+      "epoch": 0.636750846832344,
+      "grad_norm": 0.16549125155465008,
+      "learning_rate": 2e-05,
+      "loss": 5.3934,
+      "step": 9493
+    },
+    {
+      "epoch": 0.6368179226615689,
+      "grad_norm": 0.15679751839270195,
+      "learning_rate": 2e-05,
+      "loss": 5.4015,
+      "step": 9494
+    },
+    {
+      "epoch": 0.6368849984907938,
+      "grad_norm": 0.15565579707347904,
+      "learning_rate": 2e-05,
+      "loss": 5.3959,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6369520743200188,
+      "grad_norm": 0.16780204984453437,
+      "learning_rate": 2e-05,
+      "loss": 5.4178,
+      "step": 9496
+    },
+    {
+      "epoch": 0.6370191501492437,
+      "grad_norm": 0.15514808305443784,
+      "learning_rate": 2e-05,
+      "loss": 5.41,
+      "step": 9497
+    },
+    {
+      "epoch": 0.6370862259784686,
+      "grad_norm": 0.15799876107557367,
+      "learning_rate": 2e-05,
+      "loss": 5.3247,
+      "step": 9498
+    },
+    {
+      "epoch": 0.6371533018076936,
+      "grad_norm": 0.15842338652934157,
+      "learning_rate": 2e-05,
+      "loss": 5.5048,
+      "step": 9499
+    },
+    {
+      "epoch": 0.6372203776369185,
+      "grad_norm": 0.15856049367908429,
+      "learning_rate": 2e-05,
+      "loss": 5.5041,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6372874534661435,
+      "grad_norm": 0.15292321489799202,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 9501
+    },
+    {
+      "epoch": 0.6373545292953684,
+      "grad_norm": 0.17549276998480648,
+      "learning_rate": 2e-05,
+      "loss": 5.3782,
+      "step": 9502
+    },
+    {
+      "epoch": 0.6374216051245933,
+      "grad_norm": 0.1646050816010846,
+      "learning_rate": 2e-05,
+      "loss": 5.5341,
+      "step": 9503
+    },
+    {
+      "epoch": 0.6374886809538183,
+      "grad_norm": 0.15412238088672614,
+      "learning_rate": 2e-05,
+      "loss": 5.4122,
+      "step": 9504
+    },
+    {
+      "epoch": 0.6375557567830432,
+      "grad_norm": 0.1602054230736469,
+      "learning_rate": 2e-05,
+      "loss": 5.396,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6376228326122682,
+      "grad_norm": 0.15049906607923885,
+      "learning_rate": 2e-05,
+      "loss": 5.34,
+      "step": 9506
+    },
+    {
+      "epoch": 0.6376899084414931,
+      "grad_norm": 0.1440830202382656,
+      "learning_rate": 2e-05,
+      "loss": 5.4949,
+      "step": 9507
+    },
+    {
+      "epoch": 0.637756984270718,
+      "grad_norm": 0.16012268387977605,
+      "learning_rate": 2e-05,
+      "loss": 5.4742,
+      "step": 9508
+    },
+    {
+      "epoch": 0.637824060099943,
+      "grad_norm": 0.1486894587072813,
+      "learning_rate": 2e-05,
+      "loss": 5.4094,
+      "step": 9509
+    },
+    {
+      "epoch": 0.6378911359291679,
+      "grad_norm": 0.1551215678363046,
+      "learning_rate": 2e-05,
+      "loss": 5.3938,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6379582117583928,
+      "grad_norm": 0.16077793682598102,
+      "learning_rate": 2e-05,
+      "loss": 5.4199,
+      "step": 9511
+    },
+    {
+      "epoch": 0.6380252875876178,
+      "grad_norm": 0.153037503164381,
+      "learning_rate": 2e-05,
+      "loss": 5.4435,
+      "step": 9512
+    },
+    {
+      "epoch": 0.6380923634168427,
+      "grad_norm": 0.15139395757869561,
+      "learning_rate": 2e-05,
+      "loss": 5.4141,
+      "step": 9513
+    },
+    {
+      "epoch": 0.6381594392460677,
+      "grad_norm": 0.1579098244245811,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 9514
+    },
+    {
+      "epoch": 0.6382265150752926,
+      "grad_norm": 0.15689015879175175,
+      "learning_rate": 2e-05,
+      "loss": 5.4662,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6382935909045175,
+      "grad_norm": 0.14967695880729837,
+      "learning_rate": 2e-05,
+      "loss": 5.5311,
+      "step": 9516
+    },
+    {
+      "epoch": 0.6383606667337425,
+      "grad_norm": 0.1530207355040865,
+      "learning_rate": 2e-05,
+      "loss": 5.4214,
+      "step": 9517
+    },
+    {
+      "epoch": 0.6384277425629674,
+      "grad_norm": 0.15480594194552336,
+      "learning_rate": 2e-05,
+      "loss": 5.4134,
+      "step": 9518
+    },
+    {
+      "epoch": 0.6384948183921924,
+      "grad_norm": 0.1530039141984057,
+      "learning_rate": 2e-05,
+      "loss": 5.4399,
+      "step": 9519
+    },
+    {
+      "epoch": 0.6385618942214173,
+      "grad_norm": 0.1438741193033246,
+      "learning_rate": 2e-05,
+      "loss": 5.3394,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6386289700506422,
+      "grad_norm": 0.14766797689218333,
+      "learning_rate": 2e-05,
+      "loss": 5.3965,
+      "step": 9521
+    },
+    {
+      "epoch": 0.6386960458798672,
+      "grad_norm": 0.15047304335181527,
+      "learning_rate": 2e-05,
+      "loss": 5.3687,
+      "step": 9522
+    },
+    {
+      "epoch": 0.6387631217090921,
+      "grad_norm": 0.14809226237755085,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 9523
+    },
+    {
+      "epoch": 0.638830197538317,
+      "grad_norm": 0.1493599491928748,
+      "learning_rate": 2e-05,
+      "loss": 5.5055,
+      "step": 9524
+    },
+    {
+      "epoch": 0.638897273367542,
+      "grad_norm": 0.14506340590924596,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6389643491967669,
+      "grad_norm": 0.1495272273568717,
+      "learning_rate": 2e-05,
+      "loss": 5.409,
+      "step": 9526
+    },
+    {
+      "epoch": 0.6390314250259919,
+      "grad_norm": 0.15176155856799417,
+      "learning_rate": 2e-05,
+      "loss": 5.2607,
+      "step": 9527
+    },
+    {
+      "epoch": 0.6390985008552168,
+      "grad_norm": 0.1566061231279341,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 9528
+    },
+    {
+      "epoch": 0.6391655766844417,
+      "grad_norm": 0.14905132127029663,
+      "learning_rate": 2e-05,
+      "loss": 5.4128,
+      "step": 9529
+    },
+    {
+      "epoch": 0.6392326525136667,
+      "grad_norm": 0.15250591747960102,
+      "learning_rate": 2e-05,
+      "loss": 5.3926,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6392997283428916,
+      "grad_norm": 0.1584260215854498,
+      "learning_rate": 2e-05,
+      "loss": 5.5026,
+      "step": 9531
+    },
+    {
+      "epoch": 0.6393668041721166,
+      "grad_norm": 0.15039580862967292,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 9532
+    },
+    {
+      "epoch": 0.6394338800013415,
+      "grad_norm": 0.14373221965362223,
+      "learning_rate": 2e-05,
+      "loss": 5.2855,
+      "step": 9533
+    },
+    {
+      "epoch": 0.6395009558305664,
+      "grad_norm": 0.1560624855566293,
+      "learning_rate": 2e-05,
+      "loss": 5.3199,
+      "step": 9534
+    },
+    {
+      "epoch": 0.6395680316597914,
+      "grad_norm": 0.15318114467177987,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6396351074890163,
+      "grad_norm": 0.14731199097696887,
+      "learning_rate": 2e-05,
+      "loss": 5.5541,
+      "step": 9536
+    },
+    {
+      "epoch": 0.6397021833182412,
+      "grad_norm": 0.15836986589374136,
+      "learning_rate": 2e-05,
+      "loss": 5.3941,
+      "step": 9537
+    },
+    {
+      "epoch": 0.6397692591474662,
+      "grad_norm": 0.15254663923278802,
+      "learning_rate": 2e-05,
+      "loss": 5.3661,
+      "step": 9538
+    },
+    {
+      "epoch": 0.6398363349766911,
+      "grad_norm": 0.150000251669178,
+      "learning_rate": 2e-05,
+      "loss": 5.4193,
+      "step": 9539
+    },
+    {
+      "epoch": 0.6399034108059161,
+      "grad_norm": 0.14905334281666166,
+      "learning_rate": 2e-05,
+      "loss": 5.4141,
+      "step": 9540
+    },
+    {
+      "epoch": 0.639970486635141,
+      "grad_norm": 0.15004866555456353,
+      "learning_rate": 2e-05,
+      "loss": 5.4549,
+      "step": 9541
+    },
+    {
+      "epoch": 0.6400375624643659,
+      "grad_norm": 0.15624355406442722,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 9542
+    },
+    {
+      "epoch": 0.6401046382935909,
+      "grad_norm": 0.1539572352266018,
+      "learning_rate": 2e-05,
+      "loss": 5.3286,
+      "step": 9543
+    },
+    {
+      "epoch": 0.6401717141228158,
+      "grad_norm": 0.15034559469437175,
+      "learning_rate": 2e-05,
+      "loss": 5.35,
+      "step": 9544
+    },
+    {
+      "epoch": 0.6402387899520408,
+      "grad_norm": 0.15296272309790715,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6403058657812657,
+      "grad_norm": 0.15793724901545556,
+      "learning_rate": 2e-05,
+      "loss": 5.4591,
+      "step": 9546
+    },
+    {
+      "epoch": 0.6403729416104906,
+      "grad_norm": 0.1534399353405954,
+      "learning_rate": 2e-05,
+      "loss": 5.4501,
+      "step": 9547
+    },
+    {
+      "epoch": 0.6404400174397156,
+      "grad_norm": 0.16256117721775698,
+      "learning_rate": 2e-05,
+      "loss": 5.4045,
+      "step": 9548
+    },
+    {
+      "epoch": 0.6405070932689405,
+      "grad_norm": 0.15841655407378616,
+      "learning_rate": 2e-05,
+      "loss": 5.3571,
+      "step": 9549
+    },
+    {
+      "epoch": 0.6405741690981654,
+      "grad_norm": 0.15158164568722682,
+      "learning_rate": 2e-05,
+      "loss": 5.3399,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6406412449273904,
+      "grad_norm": 0.14944731691075602,
+      "learning_rate": 2e-05,
+      "loss": 5.2318,
+      "step": 9551
+    },
+    {
+      "epoch": 0.6407083207566153,
+      "grad_norm": 0.14974698008869625,
+      "learning_rate": 2e-05,
+      "loss": 5.2541,
+      "step": 9552
+    },
+    {
+      "epoch": 0.6407753965858403,
+      "grad_norm": 0.16173997198561788,
+      "learning_rate": 2e-05,
+      "loss": 5.4241,
+      "step": 9553
+    },
+    {
+      "epoch": 0.6408424724150652,
+      "grad_norm": 0.15834911716411407,
+      "learning_rate": 2e-05,
+      "loss": 5.4092,
+      "step": 9554
+    },
+    {
+      "epoch": 0.6409095482442901,
+      "grad_norm": 0.1486880772581723,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6409766240735151,
+      "grad_norm": 0.14635236463697232,
+      "learning_rate": 2e-05,
+      "loss": 5.253,
+      "step": 9556
+    },
+    {
+      "epoch": 0.64104369990274,
+      "grad_norm": 0.16517672803869055,
+      "learning_rate": 2e-05,
+      "loss": 5.4043,
+      "step": 9557
+    },
+    {
+      "epoch": 0.641110775731965,
+      "grad_norm": 0.15659244679530115,
+      "learning_rate": 2e-05,
+      "loss": 5.4211,
+      "step": 9558
+    },
+    {
+      "epoch": 0.6411778515611899,
+      "grad_norm": 0.15190532009580324,
+      "learning_rate": 2e-05,
+      "loss": 5.3821,
+      "step": 9559
+    },
+    {
+      "epoch": 0.6412449273904148,
+      "grad_norm": 0.15027250902769634,
+      "learning_rate": 2e-05,
+      "loss": 5.4438,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6413120032196398,
+      "grad_norm": 0.1608383529271978,
+      "learning_rate": 2e-05,
+      "loss": 5.4229,
+      "step": 9561
+    },
+    {
+      "epoch": 0.6413790790488647,
+      "grad_norm": 0.15561615468813103,
+      "learning_rate": 2e-05,
+      "loss": 5.3995,
+      "step": 9562
+    },
+    {
+      "epoch": 0.6414461548780896,
+      "grad_norm": 0.1459332140437825,
+      "learning_rate": 2e-05,
+      "loss": 5.4423,
+      "step": 9563
+    },
+    {
+      "epoch": 0.6415132307073146,
+      "grad_norm": 0.16492041435749713,
+      "learning_rate": 2e-05,
+      "loss": 5.5011,
+      "step": 9564
+    },
+    {
+      "epoch": 0.6415803065365395,
+      "grad_norm": 0.1644600226412454,
+      "learning_rate": 2e-05,
+      "loss": 5.3469,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6416473823657645,
+      "grad_norm": 0.1483640072839371,
+      "learning_rate": 2e-05,
+      "loss": 5.4591,
+      "step": 9566
+    },
+    {
+      "epoch": 0.6417144581949894,
+      "grad_norm": 0.1500838888756209,
+      "learning_rate": 2e-05,
+      "loss": 5.4474,
+      "step": 9567
+    },
+    {
+      "epoch": 0.6417815340242143,
+      "grad_norm": 0.14898070614796155,
+      "learning_rate": 2e-05,
+      "loss": 5.398,
+      "step": 9568
+    },
+    {
+      "epoch": 0.6418486098534393,
+      "grad_norm": 0.15492230138064775,
+      "learning_rate": 2e-05,
+      "loss": 5.4485,
+      "step": 9569
+    },
+    {
+      "epoch": 0.6419156856826642,
+      "grad_norm": 0.1516304257826873,
+      "learning_rate": 2e-05,
+      "loss": 5.4324,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6419827615118892,
+      "grad_norm": 0.15851503462586322,
+      "learning_rate": 2e-05,
+      "loss": 5.3148,
+      "step": 9571
+    },
+    {
+      "epoch": 0.6420498373411141,
+      "grad_norm": 0.14819808163335543,
+      "learning_rate": 2e-05,
+      "loss": 5.496,
+      "step": 9572
+    },
+    {
+      "epoch": 0.642116913170339,
+      "grad_norm": 0.15652302649547695,
+      "learning_rate": 2e-05,
+      "loss": 5.3795,
+      "step": 9573
+    },
+    {
+      "epoch": 0.642183988999564,
+      "grad_norm": 0.15050768352783667,
+      "learning_rate": 2e-05,
+      "loss": 5.3452,
+      "step": 9574
+    },
+    {
+      "epoch": 0.6422510648287889,
+      "grad_norm": 0.14859297718326564,
+      "learning_rate": 2e-05,
+      "loss": 5.4501,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6423181406580138,
+      "grad_norm": 0.15575916060066813,
+      "learning_rate": 2e-05,
+      "loss": 5.4579,
+      "step": 9576
+    },
+    {
+      "epoch": 0.6423852164872388,
+      "grad_norm": 0.15776725321324744,
+      "learning_rate": 2e-05,
+      "loss": 5.4338,
+      "step": 9577
+    },
+    {
+      "epoch": 0.6424522923164637,
+      "grad_norm": 0.14662149555613055,
+      "learning_rate": 2e-05,
+      "loss": 5.4537,
+      "step": 9578
+    },
+    {
+      "epoch": 0.6425193681456887,
+      "grad_norm": 0.15930252210172244,
+      "learning_rate": 2e-05,
+      "loss": 5.5338,
+      "step": 9579
+    },
+    {
+      "epoch": 0.6425864439749136,
+      "grad_norm": 0.1619046507808526,
+      "learning_rate": 2e-05,
+      "loss": 5.4065,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6426535198041385,
+      "grad_norm": 0.1472921510344544,
+      "learning_rate": 2e-05,
+      "loss": 5.5831,
+      "step": 9581
+    },
+    {
+      "epoch": 0.6427205956333635,
+      "grad_norm": 0.14904885700584777,
+      "learning_rate": 2e-05,
+      "loss": 5.5278,
+      "step": 9582
+    },
+    {
+      "epoch": 0.6427876714625884,
+      "grad_norm": 0.1597061010071644,
+      "learning_rate": 2e-05,
+      "loss": 5.5262,
+      "step": 9583
+    },
+    {
+      "epoch": 0.6428547472918134,
+      "grad_norm": 0.16602821829194145,
+      "learning_rate": 2e-05,
+      "loss": 5.5237,
+      "step": 9584
+    },
+    {
+      "epoch": 0.6429218231210383,
+      "grad_norm": 0.15668672272663148,
+      "learning_rate": 2e-05,
+      "loss": 5.5582,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6429888989502632,
+      "grad_norm": 0.1529725659445679,
+      "learning_rate": 2e-05,
+      "loss": 5.4347,
+      "step": 9586
+    },
+    {
+      "epoch": 0.6430559747794882,
+      "grad_norm": 0.1537540291076745,
+      "learning_rate": 2e-05,
+      "loss": 5.6293,
+      "step": 9587
+    },
+    {
+      "epoch": 0.6431230506087131,
+      "grad_norm": 0.15368245768464844,
+      "learning_rate": 2e-05,
+      "loss": 5.5426,
+      "step": 9588
+    },
+    {
+      "epoch": 0.643190126437938,
+      "grad_norm": 0.15819871933031493,
+      "learning_rate": 2e-05,
+      "loss": 5.4072,
+      "step": 9589
+    },
+    {
+      "epoch": 0.643257202267163,
+      "grad_norm": 0.16250194688436745,
+      "learning_rate": 2e-05,
+      "loss": 5.6052,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6433242780963879,
+      "grad_norm": 0.15155280490478998,
+      "learning_rate": 2e-05,
+      "loss": 5.3599,
+      "step": 9591
+    },
+    {
+      "epoch": 0.6433913539256129,
+      "grad_norm": 0.14503235620929636,
+      "learning_rate": 2e-05,
+      "loss": 5.4645,
+      "step": 9592
+    },
+    {
+      "epoch": 0.6434584297548378,
+      "grad_norm": 0.14895107121186837,
+      "learning_rate": 2e-05,
+      "loss": 5.4637,
+      "step": 9593
+    },
+    {
+      "epoch": 0.6435255055840627,
+      "grad_norm": 0.14524401637305384,
+      "learning_rate": 2e-05,
+      "loss": 5.5015,
+      "step": 9594
+    },
+    {
+      "epoch": 0.6435925814132877,
+      "grad_norm": 0.14749287350553403,
+      "learning_rate": 2e-05,
+      "loss": 5.328,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6436596572425126,
+      "grad_norm": 0.14784108059777107,
+      "learning_rate": 2e-05,
+      "loss": 5.3873,
+      "step": 9596
+    },
+    {
+      "epoch": 0.6437267330717376,
+      "grad_norm": 0.15548942435785398,
+      "learning_rate": 2e-05,
+      "loss": 5.4945,
+      "step": 9597
+    },
+    {
+      "epoch": 0.6437938089009625,
+      "grad_norm": 0.14987273044441574,
+      "learning_rate": 2e-05,
+      "loss": 5.402,
+      "step": 9598
+    },
+    {
+      "epoch": 0.6438608847301874,
+      "grad_norm": 0.1439594517853578,
+      "learning_rate": 2e-05,
+      "loss": 5.5188,
+      "step": 9599
+    },
+    {
+      "epoch": 0.6439279605594124,
+      "grad_norm": 0.15361733539856087,
+      "learning_rate": 2e-05,
+      "loss": 5.3251,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6439950363886373,
+      "grad_norm": 0.15643793221075902,
+      "learning_rate": 2e-05,
+      "loss": 5.5434,
+      "step": 9601
+    },
+    {
+      "epoch": 0.6440621122178622,
+      "grad_norm": 0.1599749182758217,
+      "learning_rate": 2e-05,
+      "loss": 5.5225,
+      "step": 9602
+    },
+    {
+      "epoch": 0.6441291880470872,
+      "grad_norm": 0.14748607964426283,
+      "learning_rate": 2e-05,
+      "loss": 5.3477,
+      "step": 9603
+    },
+    {
+      "epoch": 0.6441962638763121,
+      "grad_norm": 0.15054424171417108,
+      "learning_rate": 2e-05,
+      "loss": 5.3704,
+      "step": 9604
+    },
+    {
+      "epoch": 0.6442633397055371,
+      "grad_norm": 0.15332051110462994,
+      "learning_rate": 2e-05,
+      "loss": 5.3122,
+      "step": 9605
+    },
+    {
+      "epoch": 0.644330415534762,
+      "grad_norm": 0.15263031265372884,
+      "learning_rate": 2e-05,
+      "loss": 5.3981,
+      "step": 9606
+    },
+    {
+      "epoch": 0.6443974913639869,
+      "grad_norm": 0.15209941640546784,
+      "learning_rate": 2e-05,
+      "loss": 5.3452,
+      "step": 9607
+    },
+    {
+      "epoch": 0.6444645671932119,
+      "grad_norm": 0.14865894251510642,
+      "learning_rate": 2e-05,
+      "loss": 5.2803,
+      "step": 9608
+    },
+    {
+      "epoch": 0.6445316430224368,
+      "grad_norm": 0.15093259848735682,
+      "learning_rate": 2e-05,
+      "loss": 5.3327,
+      "step": 9609
+    },
+    {
+      "epoch": 0.6445987188516618,
+      "grad_norm": 0.15162854245403987,
+      "learning_rate": 2e-05,
+      "loss": 5.4916,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6446657946808867,
+      "grad_norm": 0.15767933702372677,
+      "learning_rate": 2e-05,
+      "loss": 5.3672,
+      "step": 9611
+    },
+    {
+      "epoch": 0.6447328705101116,
+      "grad_norm": 0.14834180438120798,
+      "learning_rate": 2e-05,
+      "loss": 5.3431,
+      "step": 9612
+    },
+    {
+      "epoch": 0.6447999463393366,
+      "grad_norm": 0.15272372418579383,
+      "learning_rate": 2e-05,
+      "loss": 5.3523,
+      "step": 9613
+    },
+    {
+      "epoch": 0.6448670221685615,
+      "grad_norm": 0.14360414554288878,
+      "learning_rate": 2e-05,
+      "loss": 5.4112,
+      "step": 9614
+    },
+    {
+      "epoch": 0.6449340979977864,
+      "grad_norm": 0.1508209731383697,
+      "learning_rate": 2e-05,
+      "loss": 5.3585,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6450011738270114,
+      "grad_norm": 0.15998607968695755,
+      "learning_rate": 2e-05,
+      "loss": 5.4022,
+      "step": 9616
+    },
+    {
+      "epoch": 0.6450682496562363,
+      "grad_norm": 0.14920330566234075,
+      "learning_rate": 2e-05,
+      "loss": 5.4747,
+      "step": 9617
+    },
+    {
+      "epoch": 0.6451353254854613,
+      "grad_norm": 0.15472019841954224,
+      "learning_rate": 2e-05,
+      "loss": 5.3639,
+      "step": 9618
+    },
+    {
+      "epoch": 0.6452024013146862,
+      "grad_norm": 0.15699400810064046,
+      "learning_rate": 2e-05,
+      "loss": 5.4612,
+      "step": 9619
+    },
+    {
+      "epoch": 0.6452694771439111,
+      "grad_norm": 0.15464956821750228,
+      "learning_rate": 2e-05,
+      "loss": 5.3826,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6453365529731361,
+      "grad_norm": 0.1490510984055217,
+      "learning_rate": 2e-05,
+      "loss": 5.2621,
+      "step": 9621
+    },
+    {
+      "epoch": 0.645403628802361,
+      "grad_norm": 0.15522416890347834,
+      "learning_rate": 2e-05,
+      "loss": 5.2824,
+      "step": 9622
+    },
+    {
+      "epoch": 0.645470704631586,
+      "grad_norm": 0.1496237029900653,
+      "learning_rate": 2e-05,
+      "loss": 5.351,
+      "step": 9623
+    },
+    {
+      "epoch": 0.6455377804608109,
+      "grad_norm": 0.1503525572270139,
+      "learning_rate": 2e-05,
+      "loss": 5.3106,
+      "step": 9624
+    },
+    {
+      "epoch": 0.6456048562900358,
+      "grad_norm": 0.15804641285287227,
+      "learning_rate": 2e-05,
+      "loss": 5.3167,
+      "step": 9625
+    },
+    {
+      "epoch": 0.6456719321192608,
+      "grad_norm": 0.15017998534092108,
+      "learning_rate": 2e-05,
+      "loss": 5.4032,
+      "step": 9626
+    },
+    {
+      "epoch": 0.6457390079484857,
+      "grad_norm": 0.14650505146820358,
+      "learning_rate": 2e-05,
+      "loss": 5.4673,
+      "step": 9627
+    },
+    {
+      "epoch": 0.6458060837777106,
+      "grad_norm": 0.16487629361197814,
+      "learning_rate": 2e-05,
+      "loss": 5.4184,
+      "step": 9628
+    },
+    {
+      "epoch": 0.6458731596069356,
+      "grad_norm": 0.153476945233987,
+      "learning_rate": 2e-05,
+      "loss": 5.4452,
+      "step": 9629
+    },
+    {
+      "epoch": 0.6459402354361605,
+      "grad_norm": 0.1509269149129582,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6460073112653855,
+      "grad_norm": 0.17117592528764228,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 9631
+    },
+    {
+      "epoch": 0.6460743870946104,
+      "grad_norm": 0.15434687163730193,
+      "learning_rate": 2e-05,
+      "loss": 5.2649,
+      "step": 9632
+    },
+    {
+      "epoch": 0.6461414629238353,
+      "grad_norm": 0.16314195587782038,
+      "learning_rate": 2e-05,
+      "loss": 5.4383,
+      "step": 9633
+    },
+    {
+      "epoch": 0.6462085387530603,
+      "grad_norm": 0.16359275633009607,
+      "learning_rate": 2e-05,
+      "loss": 5.3854,
+      "step": 9634
+    },
+    {
+      "epoch": 0.6462756145822852,
+      "grad_norm": 0.15049527329063864,
+      "learning_rate": 2e-05,
+      "loss": 5.3036,
+      "step": 9635
+    },
+    {
+      "epoch": 0.6463426904115103,
+      "grad_norm": 0.15914597557193916,
+      "learning_rate": 2e-05,
+      "loss": 5.4497,
+      "step": 9636
+    },
+    {
+      "epoch": 0.6464097662407352,
+      "grad_norm": 0.1697964179466436,
+      "learning_rate": 2e-05,
+      "loss": 5.3579,
+      "step": 9637
+    },
+    {
+      "epoch": 0.6464768420699601,
+      "grad_norm": 0.1503832042783177,
+      "learning_rate": 2e-05,
+      "loss": 5.5467,
+      "step": 9638
+    },
+    {
+      "epoch": 0.6465439178991851,
+      "grad_norm": 0.14641791135113585,
+      "learning_rate": 2e-05,
+      "loss": 5.5749,
+      "step": 9639
+    },
+    {
+      "epoch": 0.64661099372841,
+      "grad_norm": 0.17906276387635536,
+      "learning_rate": 2e-05,
+      "loss": 5.3968,
+      "step": 9640
+    },
+    {
+      "epoch": 0.646678069557635,
+      "grad_norm": 0.15723063140968163,
+      "learning_rate": 2e-05,
+      "loss": 5.3632,
+      "step": 9641
+    },
+    {
+      "epoch": 0.6467451453868599,
+      "grad_norm": 0.1483949907105092,
+      "learning_rate": 2e-05,
+      "loss": 5.392,
+      "step": 9642
+    },
+    {
+      "epoch": 0.6468122212160848,
+      "grad_norm": 0.1550574374350714,
+      "learning_rate": 2e-05,
+      "loss": 5.3785,
+      "step": 9643
+    },
+    {
+      "epoch": 0.6468792970453098,
+      "grad_norm": 0.15702300703700758,
+      "learning_rate": 2e-05,
+      "loss": 5.3813,
+      "step": 9644
+    },
+    {
+      "epoch": 0.6469463728745347,
+      "grad_norm": 0.14884135161043074,
+      "learning_rate": 2e-05,
+      "loss": 5.4426,
+      "step": 9645
+    },
+    {
+      "epoch": 0.6470134487037597,
+      "grad_norm": 0.15135356994217963,
+      "learning_rate": 2e-05,
+      "loss": 5.3643,
+      "step": 9646
+    },
+    {
+      "epoch": 0.6470805245329846,
+      "grad_norm": 0.1612984187819237,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 9647
+    },
+    {
+      "epoch": 0.6471476003622095,
+      "grad_norm": 0.15457457045945183,
+      "learning_rate": 2e-05,
+      "loss": 5.4196,
+      "step": 9648
+    },
+    {
+      "epoch": 0.6472146761914345,
+      "grad_norm": 0.14629722448328464,
+      "learning_rate": 2e-05,
+      "loss": 5.3428,
+      "step": 9649
+    },
+    {
+      "epoch": 0.6472817520206594,
+      "grad_norm": 0.14902633897594186,
+      "learning_rate": 2e-05,
+      "loss": 5.2753,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6473488278498843,
+      "grad_norm": 0.15681053314119572,
+      "learning_rate": 2e-05,
+      "loss": 5.4201,
+      "step": 9651
+    },
+    {
+      "epoch": 0.6474159036791093,
+      "grad_norm": 0.15720639401103692,
+      "learning_rate": 2e-05,
+      "loss": 5.4707,
+      "step": 9652
+    },
+    {
+      "epoch": 0.6474829795083342,
+      "grad_norm": 0.14759248500496833,
+      "learning_rate": 2e-05,
+      "loss": 5.2691,
+      "step": 9653
+    },
+    {
+      "epoch": 0.6475500553375592,
+      "grad_norm": 0.17007938770935263,
+      "learning_rate": 2e-05,
+      "loss": 5.406,
+      "step": 9654
+    },
+    {
+      "epoch": 0.6476171311667841,
+      "grad_norm": 0.1606364706128057,
+      "learning_rate": 2e-05,
+      "loss": 5.531,
+      "step": 9655
+    },
+    {
+      "epoch": 0.647684206996009,
+      "grad_norm": 0.1549971328062089,
+      "learning_rate": 2e-05,
+      "loss": 5.3729,
+      "step": 9656
+    },
+    {
+      "epoch": 0.647751282825234,
+      "grad_norm": 0.16455897125253183,
+      "learning_rate": 2e-05,
+      "loss": 5.4554,
+      "step": 9657
+    },
+    {
+      "epoch": 0.6478183586544589,
+      "grad_norm": 0.15711973876693794,
+      "learning_rate": 2e-05,
+      "loss": 5.3649,
+      "step": 9658
+    },
+    {
+      "epoch": 0.6478854344836839,
+      "grad_norm": 0.1538204425278777,
+      "learning_rate": 2e-05,
+      "loss": 5.4946,
+      "step": 9659
+    },
+    {
+      "epoch": 0.6479525103129088,
+      "grad_norm": 0.152455089406907,
+      "learning_rate": 2e-05,
+      "loss": 5.3582,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6480195861421337,
+      "grad_norm": 0.154847811106169,
+      "learning_rate": 2e-05,
+      "loss": 5.3905,
+      "step": 9661
+    },
+    {
+      "epoch": 0.6480866619713587,
+      "grad_norm": 0.16275527001741494,
+      "learning_rate": 2e-05,
+      "loss": 5.3243,
+      "step": 9662
+    },
+    {
+      "epoch": 0.6481537378005836,
+      "grad_norm": 0.15099188448131698,
+      "learning_rate": 2e-05,
+      "loss": 5.4117,
+      "step": 9663
+    },
+    {
+      "epoch": 0.6482208136298085,
+      "grad_norm": 0.15216426817689976,
+      "learning_rate": 2e-05,
+      "loss": 5.322,
+      "step": 9664
+    },
+    {
+      "epoch": 0.6482878894590335,
+      "grad_norm": 0.1493831038399376,
+      "learning_rate": 2e-05,
+      "loss": 5.4245,
+      "step": 9665
+    },
+    {
+      "epoch": 0.6483549652882584,
+      "grad_norm": 0.1463620006663565,
+      "learning_rate": 2e-05,
+      "loss": 5.446,
+      "step": 9666
+    },
+    {
+      "epoch": 0.6484220411174834,
+      "grad_norm": 0.14950009332585928,
+      "learning_rate": 2e-05,
+      "loss": 5.4908,
+      "step": 9667
+    },
+    {
+      "epoch": 0.6484891169467083,
+      "grad_norm": 0.15889861240055037,
+      "learning_rate": 2e-05,
+      "loss": 5.3659,
+      "step": 9668
+    },
+    {
+      "epoch": 0.6485561927759332,
+      "grad_norm": 0.14877195560874495,
+      "learning_rate": 2e-05,
+      "loss": 5.4918,
+      "step": 9669
+    },
+    {
+      "epoch": 0.6486232686051582,
+      "grad_norm": 0.14538447206602842,
+      "learning_rate": 2e-05,
+      "loss": 5.5437,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6486903444343831,
+      "grad_norm": 0.15554620918001608,
+      "learning_rate": 2e-05,
+      "loss": 5.3196,
+      "step": 9671
+    },
+    {
+      "epoch": 0.648757420263608,
+      "grad_norm": 0.1466811795265504,
+      "learning_rate": 2e-05,
+      "loss": 5.3822,
+      "step": 9672
+    },
+    {
+      "epoch": 0.648824496092833,
+      "grad_norm": 0.15388283618245058,
+      "learning_rate": 2e-05,
+      "loss": 5.5141,
+      "step": 9673
+    },
+    {
+      "epoch": 0.6488915719220579,
+      "grad_norm": 0.1504590181638396,
+      "learning_rate": 2e-05,
+      "loss": 5.4691,
+      "step": 9674
+    },
+    {
+      "epoch": 0.6489586477512829,
+      "grad_norm": 0.15436366922026779,
+      "learning_rate": 2e-05,
+      "loss": 5.5069,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6490257235805078,
+      "grad_norm": 0.15118546393689233,
+      "learning_rate": 2e-05,
+      "loss": 5.3641,
+      "step": 9676
+    },
+    {
+      "epoch": 0.6490927994097327,
+      "grad_norm": 0.14818090312019802,
+      "learning_rate": 2e-05,
+      "loss": 5.3921,
+      "step": 9677
+    },
+    {
+      "epoch": 0.6491598752389577,
+      "grad_norm": 0.1498858592520816,
+      "learning_rate": 2e-05,
+      "loss": 5.5255,
+      "step": 9678
+    },
+    {
+      "epoch": 0.6492269510681826,
+      "grad_norm": 0.15056689992987476,
+      "learning_rate": 2e-05,
+      "loss": 5.4798,
+      "step": 9679
+    },
+    {
+      "epoch": 0.6492940268974076,
+      "grad_norm": 0.14675061640112003,
+      "learning_rate": 2e-05,
+      "loss": 5.4805,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6493611027266325,
+      "grad_norm": 0.15584844145344556,
+      "learning_rate": 2e-05,
+      "loss": 5.3618,
+      "step": 9681
+    },
+    {
+      "epoch": 0.6494281785558574,
+      "grad_norm": 0.14720545171611701,
+      "learning_rate": 2e-05,
+      "loss": 5.5196,
+      "step": 9682
+    },
+    {
+      "epoch": 0.6494952543850824,
+      "grad_norm": 0.15527082907330772,
+      "learning_rate": 2e-05,
+      "loss": 5.476,
+      "step": 9683
+    },
+    {
+      "epoch": 0.6495623302143073,
+      "grad_norm": 0.14858509586423685,
+      "learning_rate": 2e-05,
+      "loss": 5.4638,
+      "step": 9684
+    },
+    {
+      "epoch": 0.6496294060435323,
+      "grad_norm": 0.15636126088326038,
+      "learning_rate": 2e-05,
+      "loss": 5.4385,
+      "step": 9685
+    },
+    {
+      "epoch": 0.6496964818727572,
+      "grad_norm": 0.15341723451294387,
+      "learning_rate": 2e-05,
+      "loss": 5.4711,
+      "step": 9686
+    },
+    {
+      "epoch": 0.6497635577019821,
+      "grad_norm": 0.15217583351572564,
+      "learning_rate": 2e-05,
+      "loss": 5.6358,
+      "step": 9687
+    },
+    {
+      "epoch": 0.6498306335312071,
+      "grad_norm": 0.15850348571017153,
+      "learning_rate": 2e-05,
+      "loss": 5.4725,
+      "step": 9688
+    },
+    {
+      "epoch": 0.649897709360432,
+      "grad_norm": 0.14900684363873393,
+      "learning_rate": 2e-05,
+      "loss": 5.2956,
+      "step": 9689
+    },
+    {
+      "epoch": 0.649964785189657,
+      "grad_norm": 0.15388501847347733,
+      "learning_rate": 2e-05,
+      "loss": 5.5143,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6500318610188819,
+      "grad_norm": 0.14785419298304237,
+      "learning_rate": 2e-05,
+      "loss": 5.3518,
+      "step": 9691
+    },
+    {
+      "epoch": 0.6500989368481068,
+      "grad_norm": 0.16032422588146475,
+      "learning_rate": 2e-05,
+      "loss": 5.4099,
+      "step": 9692
+    },
+    {
+      "epoch": 0.6501660126773318,
+      "grad_norm": 0.1489017434975433,
+      "learning_rate": 2e-05,
+      "loss": 5.4881,
+      "step": 9693
+    },
+    {
+      "epoch": 0.6502330885065567,
+      "grad_norm": 0.15033123086003328,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 9694
+    },
+    {
+      "epoch": 0.6503001643357816,
+      "grad_norm": 0.1490755673605477,
+      "learning_rate": 2e-05,
+      "loss": 5.4569,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6503672401650066,
+      "grad_norm": 0.1455192906923948,
+      "learning_rate": 2e-05,
+      "loss": 5.4437,
+      "step": 9696
+    },
+    {
+      "epoch": 0.6504343159942315,
+      "grad_norm": 0.15851015820476502,
+      "learning_rate": 2e-05,
+      "loss": 5.3692,
+      "step": 9697
+    },
+    {
+      "epoch": 0.6505013918234565,
+      "grad_norm": 0.15605767118599992,
+      "learning_rate": 2e-05,
+      "loss": 5.4862,
+      "step": 9698
+    },
+    {
+      "epoch": 0.6505684676526814,
+      "grad_norm": 0.150036782090852,
+      "learning_rate": 2e-05,
+      "loss": 5.4284,
+      "step": 9699
+    },
+    {
+      "epoch": 0.6506355434819063,
+      "grad_norm": 0.15306919954950293,
+      "learning_rate": 2e-05,
+      "loss": 5.3268,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6507026193111313,
+      "grad_norm": 0.16049231085586452,
+      "learning_rate": 2e-05,
+      "loss": 5.5044,
+      "step": 9701
+    },
+    {
+      "epoch": 0.6507696951403562,
+      "grad_norm": 0.15667542761614614,
+      "learning_rate": 2e-05,
+      "loss": 5.3606,
+      "step": 9702
+    },
+    {
+      "epoch": 0.6508367709695811,
+      "grad_norm": 0.14680535660701238,
+      "learning_rate": 2e-05,
+      "loss": 5.3638,
+      "step": 9703
+    },
+    {
+      "epoch": 0.6509038467988061,
+      "grad_norm": 0.14592356570410042,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 9704
+    },
+    {
+      "epoch": 0.650970922628031,
+      "grad_norm": 0.15754036697729962,
+      "learning_rate": 2e-05,
+      "loss": 5.4911,
+      "step": 9705
+    },
+    {
+      "epoch": 0.651037998457256,
+      "grad_norm": 0.150721787919293,
+      "learning_rate": 2e-05,
+      "loss": 5.332,
+      "step": 9706
+    },
+    {
+      "epoch": 0.6511050742864809,
+      "grad_norm": 0.14645099263297584,
+      "learning_rate": 2e-05,
+      "loss": 5.4232,
+      "step": 9707
+    },
+    {
+      "epoch": 0.6511721501157058,
+      "grad_norm": 0.16442496910151358,
+      "learning_rate": 2e-05,
+      "loss": 5.583,
+      "step": 9708
+    },
+    {
+      "epoch": 0.6512392259449308,
+      "grad_norm": 0.15030930710170437,
+      "learning_rate": 2e-05,
+      "loss": 5.2164,
+      "step": 9709
+    },
+    {
+      "epoch": 0.6513063017741557,
+      "grad_norm": 0.14835391404011564,
+      "learning_rate": 2e-05,
+      "loss": 5.3287,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6513733776033807,
+      "grad_norm": 0.15387801010542362,
+      "learning_rate": 2e-05,
+      "loss": 5.474,
+      "step": 9711
+    },
+    {
+      "epoch": 0.6514404534326056,
+      "grad_norm": 0.14448224246602356,
+      "learning_rate": 2e-05,
+      "loss": 5.4702,
+      "step": 9712
+    },
+    {
+      "epoch": 0.6515075292618305,
+      "grad_norm": 0.14812716260112102,
+      "learning_rate": 2e-05,
+      "loss": 5.3738,
+      "step": 9713
+    },
+    {
+      "epoch": 0.6515746050910555,
+      "grad_norm": 0.1501509166975408,
+      "learning_rate": 2e-05,
+      "loss": 5.4491,
+      "step": 9714
+    },
+    {
+      "epoch": 0.6516416809202804,
+      "grad_norm": 0.15508437064436953,
+      "learning_rate": 2e-05,
+      "loss": 5.4579,
+      "step": 9715
+    },
+    {
+      "epoch": 0.6517087567495053,
+      "grad_norm": 0.15586539713051176,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 9716
+    },
+    {
+      "epoch": 0.6517758325787303,
+      "grad_norm": 0.15499243680302235,
+      "learning_rate": 2e-05,
+      "loss": 5.4505,
+      "step": 9717
+    },
+    {
+      "epoch": 0.6518429084079552,
+      "grad_norm": 0.16033869783097102,
+      "learning_rate": 2e-05,
+      "loss": 5.3909,
+      "step": 9718
+    },
+    {
+      "epoch": 0.6519099842371802,
+      "grad_norm": 0.14826897708759246,
+      "learning_rate": 2e-05,
+      "loss": 5.4069,
+      "step": 9719
+    },
+    {
+      "epoch": 0.6519770600664051,
+      "grad_norm": 0.15827799391634742,
+      "learning_rate": 2e-05,
+      "loss": 5.5011,
+      "step": 9720
+    },
+    {
+      "epoch": 0.65204413589563,
+      "grad_norm": 0.15746106067755197,
+      "learning_rate": 2e-05,
+      "loss": 5.4692,
+      "step": 9721
+    },
+    {
+      "epoch": 0.652111211724855,
+      "grad_norm": 0.15146908103666518,
+      "learning_rate": 2e-05,
+      "loss": 5.5231,
+      "step": 9722
+    },
+    {
+      "epoch": 0.6521782875540799,
+      "grad_norm": 0.1472013456762202,
+      "learning_rate": 2e-05,
+      "loss": 5.3172,
+      "step": 9723
+    },
+    {
+      "epoch": 0.6522453633833049,
+      "grad_norm": 0.14920610153620467,
+      "learning_rate": 2e-05,
+      "loss": 5.4055,
+      "step": 9724
+    },
+    {
+      "epoch": 0.6523124392125298,
+      "grad_norm": 0.15103080591969462,
+      "learning_rate": 2e-05,
+      "loss": 5.4457,
+      "step": 9725
+    },
+    {
+      "epoch": 0.6523795150417547,
+      "grad_norm": 0.1481868326931773,
+      "learning_rate": 2e-05,
+      "loss": 5.3747,
+      "step": 9726
+    },
+    {
+      "epoch": 0.6524465908709797,
+      "grad_norm": 0.15071498672963218,
+      "learning_rate": 2e-05,
+      "loss": 5.3479,
+      "step": 9727
+    },
+    {
+      "epoch": 0.6525136667002046,
+      "grad_norm": 0.1474889049591453,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 9728
+    },
+    {
+      "epoch": 0.6525807425294295,
+      "grad_norm": 0.14706361219606814,
+      "learning_rate": 2e-05,
+      "loss": 5.3908,
+      "step": 9729
+    },
+    {
+      "epoch": 0.6526478183586545,
+      "grad_norm": 0.15193411454709,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6527148941878794,
+      "grad_norm": 0.14962464733518022,
+      "learning_rate": 2e-05,
+      "loss": 5.4915,
+      "step": 9731
+    },
+    {
+      "epoch": 0.6527819700171044,
+      "grad_norm": 0.1519044596161463,
+      "learning_rate": 2e-05,
+      "loss": 5.51,
+      "step": 9732
+    },
+    {
+      "epoch": 0.6528490458463293,
+      "grad_norm": 0.1503191436689999,
+      "learning_rate": 2e-05,
+      "loss": 5.5556,
+      "step": 9733
+    },
+    {
+      "epoch": 0.6529161216755542,
+      "grad_norm": 0.14734650392312643,
+      "learning_rate": 2e-05,
+      "loss": 5.4152,
+      "step": 9734
+    },
+    {
+      "epoch": 0.6529831975047792,
+      "grad_norm": 0.15492935870134697,
+      "learning_rate": 2e-05,
+      "loss": 5.436,
+      "step": 9735
+    },
+    {
+      "epoch": 0.6530502733340041,
+      "grad_norm": 0.15723288938345406,
+      "learning_rate": 2e-05,
+      "loss": 5.5335,
+      "step": 9736
+    },
+    {
+      "epoch": 0.653117349163229,
+      "grad_norm": 0.1499202146698686,
+      "learning_rate": 2e-05,
+      "loss": 5.3931,
+      "step": 9737
+    },
+    {
+      "epoch": 0.653184424992454,
+      "grad_norm": 0.14646712810709975,
+      "learning_rate": 2e-05,
+      "loss": 5.3749,
+      "step": 9738
+    },
+    {
+      "epoch": 0.6532515008216789,
+      "grad_norm": 0.15371108878213988,
+      "learning_rate": 2e-05,
+      "loss": 5.2358,
+      "step": 9739
+    },
+    {
+      "epoch": 0.6533185766509039,
+      "grad_norm": 0.1499715086094196,
+      "learning_rate": 2e-05,
+      "loss": 5.3972,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6533856524801288,
+      "grad_norm": 0.14396547088514178,
+      "learning_rate": 2e-05,
+      "loss": 5.3311,
+      "step": 9741
+    },
+    {
+      "epoch": 0.6534527283093537,
+      "grad_norm": 0.1507742439550711,
+      "learning_rate": 2e-05,
+      "loss": 5.4997,
+      "step": 9742
+    },
+    {
+      "epoch": 0.6535198041385787,
+      "grad_norm": 0.1514004632994984,
+      "learning_rate": 2e-05,
+      "loss": 5.4991,
+      "step": 9743
+    },
+    {
+      "epoch": 0.6535868799678036,
+      "grad_norm": 0.15060164591874142,
+      "learning_rate": 2e-05,
+      "loss": 5.347,
+      "step": 9744
+    },
+    {
+      "epoch": 0.6536539557970286,
+      "grad_norm": 0.15741586485560138,
+      "learning_rate": 2e-05,
+      "loss": 5.3845,
+      "step": 9745
+    },
+    {
+      "epoch": 0.6537210316262535,
+      "grad_norm": 0.15437227537508724,
+      "learning_rate": 2e-05,
+      "loss": 5.4934,
+      "step": 9746
+    },
+    {
+      "epoch": 0.6537881074554784,
+      "grad_norm": 0.1500921368292464,
+      "learning_rate": 2e-05,
+      "loss": 5.4643,
+      "step": 9747
+    },
+    {
+      "epoch": 0.6538551832847034,
+      "grad_norm": 0.14649586974287884,
+      "learning_rate": 2e-05,
+      "loss": 5.4684,
+      "step": 9748
+    },
+    {
+      "epoch": 0.6539222591139283,
+      "grad_norm": 0.159540979961666,
+      "learning_rate": 2e-05,
+      "loss": 5.5118,
+      "step": 9749
+    },
+    {
+      "epoch": 0.6539893349431533,
+      "grad_norm": 0.15298702290974547,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6540564107723782,
+      "grad_norm": 0.15071914022637994,
+      "learning_rate": 2e-05,
+      "loss": 5.4387,
+      "step": 9751
+    },
+    {
+      "epoch": 0.6541234866016031,
+      "grad_norm": 0.154075876512942,
+      "learning_rate": 2e-05,
+      "loss": 5.4352,
+      "step": 9752
+    },
+    {
+      "epoch": 0.6541905624308281,
+      "grad_norm": 0.1595109307758106,
+      "learning_rate": 2e-05,
+      "loss": 5.3417,
+      "step": 9753
+    },
+    {
+      "epoch": 0.654257638260053,
+      "grad_norm": 0.14371380609704684,
+      "learning_rate": 2e-05,
+      "loss": 5.4191,
+      "step": 9754
+    },
+    {
+      "epoch": 0.654324714089278,
+      "grad_norm": 0.1548107546362262,
+      "learning_rate": 2e-05,
+      "loss": 5.4776,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6543917899185029,
+      "grad_norm": 0.15275272293759626,
+      "learning_rate": 2e-05,
+      "loss": 5.4465,
+      "step": 9756
+    },
+    {
+      "epoch": 0.6544588657477278,
+      "grad_norm": 0.15048179067965053,
+      "learning_rate": 2e-05,
+      "loss": 5.4403,
+      "step": 9757
+    },
+    {
+      "epoch": 0.6545259415769528,
+      "grad_norm": 0.15544665053420284,
+      "learning_rate": 2e-05,
+      "loss": 5.3241,
+      "step": 9758
+    },
+    {
+      "epoch": 0.6545930174061777,
+      "grad_norm": 0.15418361831437774,
+      "learning_rate": 2e-05,
+      "loss": 5.5554,
+      "step": 9759
+    },
+    {
+      "epoch": 0.6546600932354026,
+      "grad_norm": 0.1563502494861759,
+      "learning_rate": 2e-05,
+      "loss": 5.3445,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6547271690646276,
+      "grad_norm": 0.1498925512154019,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 9761
+    },
+    {
+      "epoch": 0.6547942448938525,
+      "grad_norm": 0.15435867747186563,
+      "learning_rate": 2e-05,
+      "loss": 5.4559,
+      "step": 9762
+    },
+    {
+      "epoch": 0.6548613207230775,
+      "grad_norm": 0.14810807463013037,
+      "learning_rate": 2e-05,
+      "loss": 5.4719,
+      "step": 9763
+    },
+    {
+      "epoch": 0.6549283965523024,
+      "grad_norm": 0.1615476208184419,
+      "learning_rate": 2e-05,
+      "loss": 5.4436,
+      "step": 9764
+    },
+    {
+      "epoch": 0.6549954723815273,
+      "grad_norm": 0.15392980339223797,
+      "learning_rate": 2e-05,
+      "loss": 5.4179,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6550625482107523,
+      "grad_norm": 0.15413187517278365,
+      "learning_rate": 2e-05,
+      "loss": 5.3757,
+      "step": 9766
+    },
+    {
+      "epoch": 0.6551296240399772,
+      "grad_norm": 0.1562688733262742,
+      "learning_rate": 2e-05,
+      "loss": 5.394,
+      "step": 9767
+    },
+    {
+      "epoch": 0.6551966998692021,
+      "grad_norm": 0.15296545104180503,
+      "learning_rate": 2e-05,
+      "loss": 5.3995,
+      "step": 9768
+    },
+    {
+      "epoch": 0.6552637756984271,
+      "grad_norm": 0.15496639551159605,
+      "learning_rate": 2e-05,
+      "loss": 5.5259,
+      "step": 9769
+    },
+    {
+      "epoch": 0.655330851527652,
+      "grad_norm": 0.15764311720966634,
+      "learning_rate": 2e-05,
+      "loss": 5.3077,
+      "step": 9770
+    },
+    {
+      "epoch": 0.655397927356877,
+      "grad_norm": 0.15246722322270803,
+      "learning_rate": 2e-05,
+      "loss": 5.2912,
+      "step": 9771
+    },
+    {
+      "epoch": 0.6554650031861019,
+      "grad_norm": 0.1569902194975582,
+      "learning_rate": 2e-05,
+      "loss": 5.5355,
+      "step": 9772
+    },
+    {
+      "epoch": 0.6555320790153268,
+      "grad_norm": 0.15921108955847868,
+      "learning_rate": 2e-05,
+      "loss": 5.4516,
+      "step": 9773
+    },
+    {
+      "epoch": 0.6555991548445518,
+      "grad_norm": 0.15541377941918472,
+      "learning_rate": 2e-05,
+      "loss": 5.4043,
+      "step": 9774
+    },
+    {
+      "epoch": 0.6556662306737767,
+      "grad_norm": 0.15491730226992895,
+      "learning_rate": 2e-05,
+      "loss": 5.4954,
+      "step": 9775
+    },
+    {
+      "epoch": 0.6557333065030017,
+      "grad_norm": 0.16319095932142988,
+      "learning_rate": 2e-05,
+      "loss": 5.3827,
+      "step": 9776
+    },
+    {
+      "epoch": 0.6558003823322266,
+      "grad_norm": 0.1565608109547019,
+      "learning_rate": 2e-05,
+      "loss": 5.3732,
+      "step": 9777
+    },
+    {
+      "epoch": 0.6558674581614515,
+      "grad_norm": 0.16949436820630132,
+      "learning_rate": 2e-05,
+      "loss": 5.3068,
+      "step": 9778
+    },
+    {
+      "epoch": 0.6559345339906765,
+      "grad_norm": 0.1555073115364532,
+      "learning_rate": 2e-05,
+      "loss": 5.5027,
+      "step": 9779
+    },
+    {
+      "epoch": 0.6560016098199014,
+      "grad_norm": 0.14934408395956927,
+      "learning_rate": 2e-05,
+      "loss": 5.4861,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6560686856491263,
+      "grad_norm": 0.14742049285007605,
+      "learning_rate": 2e-05,
+      "loss": 5.2656,
+      "step": 9781
+    },
+    {
+      "epoch": 0.6561357614783513,
+      "grad_norm": 0.1519211461153154,
+      "learning_rate": 2e-05,
+      "loss": 5.3991,
+      "step": 9782
+    },
+    {
+      "epoch": 0.6562028373075762,
+      "grad_norm": 0.15343436761034385,
+      "learning_rate": 2e-05,
+      "loss": 5.4188,
+      "step": 9783
+    },
+    {
+      "epoch": 0.6562699131368012,
+      "grad_norm": 0.14692460151912007,
+      "learning_rate": 2e-05,
+      "loss": 5.4953,
+      "step": 9784
+    },
+    {
+      "epoch": 0.6563369889660261,
+      "grad_norm": 0.15237148397257205,
+      "learning_rate": 2e-05,
+      "loss": 5.2604,
+      "step": 9785
+    },
+    {
+      "epoch": 0.656404064795251,
+      "grad_norm": 0.1519391882914495,
+      "learning_rate": 2e-05,
+      "loss": 5.3691,
+      "step": 9786
+    },
+    {
+      "epoch": 0.656471140624476,
+      "grad_norm": 0.14787898602531654,
+      "learning_rate": 2e-05,
+      "loss": 5.3635,
+      "step": 9787
+    },
+    {
+      "epoch": 0.6565382164537009,
+      "grad_norm": 0.15425924193875706,
+      "learning_rate": 2e-05,
+      "loss": 5.4941,
+      "step": 9788
+    },
+    {
+      "epoch": 0.6566052922829259,
+      "grad_norm": 0.1587655269196565,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 9789
+    },
+    {
+      "epoch": 0.6566723681121508,
+      "grad_norm": 0.15387335242446631,
+      "learning_rate": 2e-05,
+      "loss": 5.4487,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6567394439413757,
+      "grad_norm": 0.15612251596125895,
+      "learning_rate": 2e-05,
+      "loss": 5.4104,
+      "step": 9791
+    },
+    {
+      "epoch": 0.6568065197706007,
+      "grad_norm": 0.1622354168740302,
+      "learning_rate": 2e-05,
+      "loss": 5.4466,
+      "step": 9792
+    },
+    {
+      "epoch": 0.6568735955998256,
+      "grad_norm": 0.15727627168120237,
+      "learning_rate": 2e-05,
+      "loss": 5.4056,
+      "step": 9793
+    },
+    {
+      "epoch": 0.6569406714290505,
+      "grad_norm": 0.1584620256843739,
+      "learning_rate": 2e-05,
+      "loss": 5.3286,
+      "step": 9794
+    },
+    {
+      "epoch": 0.6570077472582755,
+      "grad_norm": 0.1505885989545561,
+      "learning_rate": 2e-05,
+      "loss": 5.4442,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6570748230875004,
+      "grad_norm": 0.15204429170423917,
+      "learning_rate": 2e-05,
+      "loss": 5.348,
+      "step": 9796
+    },
+    {
+      "epoch": 0.6571418989167254,
+      "grad_norm": 0.16065073934853688,
+      "learning_rate": 2e-05,
+      "loss": 5.43,
+      "step": 9797
+    },
+    {
+      "epoch": 0.6572089747459503,
+      "grad_norm": 0.15557768627247187,
+      "learning_rate": 2e-05,
+      "loss": 5.373,
+      "step": 9798
+    },
+    {
+      "epoch": 0.6572760505751752,
+      "grad_norm": 0.1474928019301523,
+      "learning_rate": 2e-05,
+      "loss": 5.4937,
+      "step": 9799
+    },
+    {
+      "epoch": 0.6573431264044002,
+      "grad_norm": 0.1530401715609346,
+      "learning_rate": 2e-05,
+      "loss": 5.3176,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6574102022336251,
+      "grad_norm": 0.15368437891943496,
+      "learning_rate": 2e-05,
+      "loss": 5.4306,
+      "step": 9801
+    },
+    {
+      "epoch": 0.6574772780628501,
+      "grad_norm": 0.15905047542823494,
+      "learning_rate": 2e-05,
+      "loss": 5.4137,
+      "step": 9802
+    },
+    {
+      "epoch": 0.657544353892075,
+      "grad_norm": 0.1468036774153878,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 9803
+    },
+    {
+      "epoch": 0.6576114297212999,
+      "grad_norm": 0.15158080273208072,
+      "learning_rate": 2e-05,
+      "loss": 5.2619,
+      "step": 9804
+    },
+    {
+      "epoch": 0.6576785055505249,
+      "grad_norm": 0.1533306412684322,
+      "learning_rate": 2e-05,
+      "loss": 5.3967,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6577455813797498,
+      "grad_norm": 0.14764223827495807,
+      "learning_rate": 2e-05,
+      "loss": 5.4724,
+      "step": 9806
+    },
+    {
+      "epoch": 0.6578126572089747,
+      "grad_norm": 0.15535224019718089,
+      "learning_rate": 2e-05,
+      "loss": 5.4867,
+      "step": 9807
+    },
+    {
+      "epoch": 0.6578797330381997,
+      "grad_norm": 0.15394775410161804,
+      "learning_rate": 2e-05,
+      "loss": 5.4987,
+      "step": 9808
+    },
+    {
+      "epoch": 0.6579468088674246,
+      "grad_norm": 0.14843091371813202,
+      "learning_rate": 2e-05,
+      "loss": 5.4895,
+      "step": 9809
+    },
+    {
+      "epoch": 0.6580138846966496,
+      "grad_norm": 0.15262448911071694,
+      "learning_rate": 2e-05,
+      "loss": 5.4716,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6580809605258745,
+      "grad_norm": 0.1454589214530626,
+      "learning_rate": 2e-05,
+      "loss": 5.3451,
+      "step": 9811
+    },
+    {
+      "epoch": 0.6581480363550994,
+      "grad_norm": 0.1447030825401799,
+      "learning_rate": 2e-05,
+      "loss": 5.3114,
+      "step": 9812
+    },
+    {
+      "epoch": 0.6582151121843244,
+      "grad_norm": 0.15336216056136306,
+      "learning_rate": 2e-05,
+      "loss": 5.503,
+      "step": 9813
+    },
+    {
+      "epoch": 0.6582821880135493,
+      "grad_norm": 0.16095039326216143,
+      "learning_rate": 2e-05,
+      "loss": 5.3576,
+      "step": 9814
+    },
+    {
+      "epoch": 0.6583492638427743,
+      "grad_norm": 0.15099872139001738,
+      "learning_rate": 2e-05,
+      "loss": 5.3701,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6584163396719992,
+      "grad_norm": 0.14798140547894278,
+      "learning_rate": 2e-05,
+      "loss": 5.36,
+      "step": 9816
+    },
+    {
+      "epoch": 0.6584834155012241,
+      "grad_norm": 0.15316151411498008,
+      "learning_rate": 2e-05,
+      "loss": 5.3361,
+      "step": 9817
+    },
+    {
+      "epoch": 0.6585504913304491,
+      "grad_norm": 0.14843405765425724,
+      "learning_rate": 2e-05,
+      "loss": 5.228,
+      "step": 9818
+    },
+    {
+      "epoch": 0.658617567159674,
+      "grad_norm": 0.1566834544649518,
+      "learning_rate": 2e-05,
+      "loss": 5.5083,
+      "step": 9819
+    },
+    {
+      "epoch": 0.658684642988899,
+      "grad_norm": 0.1533381523250546,
+      "learning_rate": 2e-05,
+      "loss": 5.3839,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6587517188181239,
+      "grad_norm": 0.16101767186345364,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 9821
+    },
+    {
+      "epoch": 0.6588187946473488,
+      "grad_norm": 0.1513442133428081,
+      "learning_rate": 2e-05,
+      "loss": 5.6083,
+      "step": 9822
+    },
+    {
+      "epoch": 0.6588858704765738,
+      "grad_norm": 0.1465606653892973,
+      "learning_rate": 2e-05,
+      "loss": 5.3769,
+      "step": 9823
+    },
+    {
+      "epoch": 0.6589529463057987,
+      "grad_norm": 0.1523367260167032,
+      "learning_rate": 2e-05,
+      "loss": 5.4594,
+      "step": 9824
+    },
+    {
+      "epoch": 0.6590200221350236,
+      "grad_norm": 0.16342229837553954,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 9825
+    },
+    {
+      "epoch": 0.6590870979642486,
+      "grad_norm": 0.14601310614923427,
+      "learning_rate": 2e-05,
+      "loss": 5.3886,
+      "step": 9826
+    },
+    {
+      "epoch": 0.6591541737934735,
+      "grad_norm": 0.1477118313340132,
+      "learning_rate": 2e-05,
+      "loss": 5.3564,
+      "step": 9827
+    },
+    {
+      "epoch": 0.6592212496226985,
+      "grad_norm": 0.1643049003513351,
+      "learning_rate": 2e-05,
+      "loss": 5.4272,
+      "step": 9828
+    },
+    {
+      "epoch": 0.6592883254519234,
+      "grad_norm": 0.15779434032722361,
+      "learning_rate": 2e-05,
+      "loss": 5.3405,
+      "step": 9829
+    },
+    {
+      "epoch": 0.6593554012811483,
+      "grad_norm": 0.15503361925521592,
+      "learning_rate": 2e-05,
+      "loss": 5.3335,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6594224771103733,
+      "grad_norm": 0.18614289670721915,
+      "learning_rate": 2e-05,
+      "loss": 5.4074,
+      "step": 9831
+    },
+    {
+      "epoch": 0.6594895529395982,
+      "grad_norm": 0.15264048785770074,
+      "learning_rate": 2e-05,
+      "loss": 5.4631,
+      "step": 9832
+    },
+    {
+      "epoch": 0.6595566287688231,
+      "grad_norm": 0.15127216222838277,
+      "learning_rate": 2e-05,
+      "loss": 5.297,
+      "step": 9833
+    },
+    {
+      "epoch": 0.6596237045980481,
+      "grad_norm": 0.16301623055547051,
+      "learning_rate": 2e-05,
+      "loss": 5.3732,
+      "step": 9834
+    },
+    {
+      "epoch": 0.659690780427273,
+      "grad_norm": 0.1572879260840644,
+      "learning_rate": 2e-05,
+      "loss": 5.3801,
+      "step": 9835
+    },
+    {
+      "epoch": 0.659757856256498,
+      "grad_norm": 0.14589633273398284,
+      "learning_rate": 2e-05,
+      "loss": 5.3105,
+      "step": 9836
+    },
+    {
+      "epoch": 0.6598249320857229,
+      "grad_norm": 0.1516130606056624,
+      "learning_rate": 2e-05,
+      "loss": 5.3121,
+      "step": 9837
+    },
+    {
+      "epoch": 0.6598920079149478,
+      "grad_norm": 0.17385985279439833,
+      "learning_rate": 2e-05,
+      "loss": 5.3967,
+      "step": 9838
+    },
+    {
+      "epoch": 0.6599590837441728,
+      "grad_norm": 0.15380223473937704,
+      "learning_rate": 2e-05,
+      "loss": 5.3275,
+      "step": 9839
+    },
+    {
+      "epoch": 0.6600261595733977,
+      "grad_norm": 0.14911571266792814,
+      "learning_rate": 2e-05,
+      "loss": 5.4139,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6600932354026227,
+      "grad_norm": 0.15092318991705292,
+      "learning_rate": 2e-05,
+      "loss": 5.3353,
+      "step": 9841
+    },
+    {
+      "epoch": 0.6601603112318476,
+      "grad_norm": 0.16014341073568292,
+      "learning_rate": 2e-05,
+      "loss": 5.3556,
+      "step": 9842
+    },
+    {
+      "epoch": 0.6602273870610725,
+      "grad_norm": 0.15356742200288648,
+      "learning_rate": 2e-05,
+      "loss": 5.5064,
+      "step": 9843
+    },
+    {
+      "epoch": 0.6602944628902975,
+      "grad_norm": 0.14734004543635065,
+      "learning_rate": 2e-05,
+      "loss": 5.4398,
+      "step": 9844
+    },
+    {
+      "epoch": 0.6603615387195224,
+      "grad_norm": 0.15611781911789288,
+      "learning_rate": 2e-05,
+      "loss": 5.4811,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6604286145487474,
+      "grad_norm": 0.16048694007051464,
+      "learning_rate": 2e-05,
+      "loss": 5.4075,
+      "step": 9846
+    },
+    {
+      "epoch": 0.6604956903779723,
+      "grad_norm": 0.15175523800922774,
+      "learning_rate": 2e-05,
+      "loss": 5.3391,
+      "step": 9847
+    },
+    {
+      "epoch": 0.6605627662071972,
+      "grad_norm": 0.14698453372214348,
+      "learning_rate": 2e-05,
+      "loss": 5.4527,
+      "step": 9848
+    },
+    {
+      "epoch": 0.6606298420364222,
+      "grad_norm": 0.16057614934436715,
+      "learning_rate": 2e-05,
+      "loss": 5.4368,
+      "step": 9849
+    },
+    {
+      "epoch": 0.6606969178656471,
+      "grad_norm": 0.1563890243848878,
+      "learning_rate": 2e-05,
+      "loss": 5.3454,
+      "step": 9850
+    },
+    {
+      "epoch": 0.660763993694872,
+      "grad_norm": 0.15093231204648833,
+      "learning_rate": 2e-05,
+      "loss": 5.3735,
+      "step": 9851
+    },
+    {
+      "epoch": 0.660831069524097,
+      "grad_norm": 0.15749073826072976,
+      "learning_rate": 2e-05,
+      "loss": 5.4921,
+      "step": 9852
+    },
+    {
+      "epoch": 0.6608981453533219,
+      "grad_norm": 0.15572626256204597,
+      "learning_rate": 2e-05,
+      "loss": 5.3504,
+      "step": 9853
+    },
+    {
+      "epoch": 0.6609652211825469,
+      "grad_norm": 0.15271494140700653,
+      "learning_rate": 2e-05,
+      "loss": 5.5018,
+      "step": 9854
+    },
+    {
+      "epoch": 0.6610322970117718,
+      "grad_norm": 0.155349696296429,
+      "learning_rate": 2e-05,
+      "loss": 5.3971,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6610993728409967,
+      "grad_norm": 0.1542571306768787,
+      "learning_rate": 2e-05,
+      "loss": 5.4896,
+      "step": 9856
+    },
+    {
+      "epoch": 0.6611664486702217,
+      "grad_norm": 0.15392697228871702,
+      "learning_rate": 2e-05,
+      "loss": 5.4657,
+      "step": 9857
+    },
+    {
+      "epoch": 0.6612335244994466,
+      "grad_norm": 0.15250722466333888,
+      "learning_rate": 2e-05,
+      "loss": 5.4928,
+      "step": 9858
+    },
+    {
+      "epoch": 0.6613006003286716,
+      "grad_norm": 0.1610520956450239,
+      "learning_rate": 2e-05,
+      "loss": 5.4555,
+      "step": 9859
+    },
+    {
+      "epoch": 0.6613676761578965,
+      "grad_norm": 0.15182587435525102,
+      "learning_rate": 2e-05,
+      "loss": 5.5651,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6614347519871214,
+      "grad_norm": 0.14635665630882796,
+      "learning_rate": 2e-05,
+      "loss": 5.2996,
+      "step": 9861
+    },
+    {
+      "epoch": 0.6615018278163464,
+      "grad_norm": 0.14589756388168565,
+      "learning_rate": 2e-05,
+      "loss": 5.4172,
+      "step": 9862
+    },
+    {
+      "epoch": 0.6615689036455713,
+      "grad_norm": 0.15672717224072674,
+      "learning_rate": 2e-05,
+      "loss": 5.3076,
+      "step": 9863
+    },
+    {
+      "epoch": 0.6616359794747962,
+      "grad_norm": 0.15388871451956396,
+      "learning_rate": 2e-05,
+      "loss": 5.3605,
+      "step": 9864
+    },
+    {
+      "epoch": 0.6617030553040212,
+      "grad_norm": 0.16097679933253342,
+      "learning_rate": 2e-05,
+      "loss": 5.4105,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6617701311332461,
+      "grad_norm": 0.16273487752971674,
+      "learning_rate": 2e-05,
+      "loss": 5.4924,
+      "step": 9866
+    },
+    {
+      "epoch": 0.6618372069624711,
+      "grad_norm": 0.15253324654799935,
+      "learning_rate": 2e-05,
+      "loss": 5.3764,
+      "step": 9867
+    },
+    {
+      "epoch": 0.661904282791696,
+      "grad_norm": 0.15371648720707462,
+      "learning_rate": 2e-05,
+      "loss": 5.3878,
+      "step": 9868
+    },
+    {
+      "epoch": 0.6619713586209209,
+      "grad_norm": 0.15683857564391762,
+      "learning_rate": 2e-05,
+      "loss": 5.4085,
+      "step": 9869
+    },
+    {
+      "epoch": 0.6620384344501459,
+      "grad_norm": 0.14818698381802295,
+      "learning_rate": 2e-05,
+      "loss": 5.4218,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6621055102793708,
+      "grad_norm": 0.15101679860980657,
+      "learning_rate": 2e-05,
+      "loss": 5.6031,
+      "step": 9871
+    },
+    {
+      "epoch": 0.6621725861085958,
+      "grad_norm": 0.1584975764758613,
+      "learning_rate": 2e-05,
+      "loss": 5.3797,
+      "step": 9872
+    },
+    {
+      "epoch": 0.6622396619378207,
+      "grad_norm": 0.1576532911867124,
+      "learning_rate": 2e-05,
+      "loss": 5.5133,
+      "step": 9873
+    },
+    {
+      "epoch": 0.6623067377670456,
+      "grad_norm": 0.15331996867502032,
+      "learning_rate": 2e-05,
+      "loss": 5.3742,
+      "step": 9874
+    },
+    {
+      "epoch": 0.6623738135962706,
+      "grad_norm": 0.15498299025147153,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6624408894254955,
+      "grad_norm": 0.1528273612075145,
+      "learning_rate": 2e-05,
+      "loss": 5.3728,
+      "step": 9876
+    },
+    {
+      "epoch": 0.6625079652547204,
+      "grad_norm": 0.15229558109774904,
+      "learning_rate": 2e-05,
+      "loss": 5.4532,
+      "step": 9877
+    },
+    {
+      "epoch": 0.6625750410839454,
+      "grad_norm": 0.1545357144556755,
+      "learning_rate": 2e-05,
+      "loss": 5.4097,
+      "step": 9878
+    },
+    {
+      "epoch": 0.6626421169131703,
+      "grad_norm": 0.1501862177510736,
+      "learning_rate": 2e-05,
+      "loss": 5.388,
+      "step": 9879
+    },
+    {
+      "epoch": 0.6627091927423953,
+      "grad_norm": 0.15281586071981357,
+      "learning_rate": 2e-05,
+      "loss": 5.5405,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6627762685716202,
+      "grad_norm": 0.15417097413590003,
+      "learning_rate": 2e-05,
+      "loss": 5.4083,
+      "step": 9881
+    },
+    {
+      "epoch": 0.6628433444008451,
+      "grad_norm": 0.1541566899042459,
+      "learning_rate": 2e-05,
+      "loss": 5.3574,
+      "step": 9882
+    },
+    {
+      "epoch": 0.6629104202300701,
+      "grad_norm": 0.15025444610911184,
+      "learning_rate": 2e-05,
+      "loss": 5.5429,
+      "step": 9883
+    },
+    {
+      "epoch": 0.662977496059295,
+      "grad_norm": 0.16093013560778366,
+      "learning_rate": 2e-05,
+      "loss": 5.4978,
+      "step": 9884
+    },
+    {
+      "epoch": 0.66304457188852,
+      "grad_norm": 0.15206985541029774,
+      "learning_rate": 2e-05,
+      "loss": 5.4484,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6631116477177449,
+      "grad_norm": 0.1610471747088262,
+      "learning_rate": 2e-05,
+      "loss": 5.4778,
+      "step": 9886
+    },
+    {
+      "epoch": 0.6631787235469698,
+      "grad_norm": 0.1527154776704596,
+      "learning_rate": 2e-05,
+      "loss": 5.479,
+      "step": 9887
+    },
+    {
+      "epoch": 0.6632457993761948,
+      "grad_norm": 0.15070458139156798,
+      "learning_rate": 2e-05,
+      "loss": 5.3806,
+      "step": 9888
+    },
+    {
+      "epoch": 0.6633128752054197,
+      "grad_norm": 0.15647974227721417,
+      "learning_rate": 2e-05,
+      "loss": 5.407,
+      "step": 9889
+    },
+    {
+      "epoch": 0.6633799510346446,
+      "grad_norm": 0.15141137632383506,
+      "learning_rate": 2e-05,
+      "loss": 5.4799,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6634470268638696,
+      "grad_norm": 0.16239127230446004,
+      "learning_rate": 2e-05,
+      "loss": 5.4672,
+      "step": 9891
+    },
+    {
+      "epoch": 0.6635141026930945,
+      "grad_norm": 0.14692724707687202,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 9892
+    },
+    {
+      "epoch": 0.6635811785223195,
+      "grad_norm": 0.150474924918795,
+      "learning_rate": 2e-05,
+      "loss": 5.5457,
+      "step": 9893
+    },
+    {
+      "epoch": 0.6636482543515444,
+      "grad_norm": 0.15406601096746556,
+      "learning_rate": 2e-05,
+      "loss": 5.3547,
+      "step": 9894
+    },
+    {
+      "epoch": 0.6637153301807693,
+      "grad_norm": 0.14955263831455853,
+      "learning_rate": 2e-05,
+      "loss": 5.4264,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6637824060099943,
+      "grad_norm": 0.15968992689082484,
+      "learning_rate": 2e-05,
+      "loss": 5.412,
+      "step": 9896
+    },
+    {
+      "epoch": 0.6638494818392192,
+      "grad_norm": 0.15444232386469126,
+      "learning_rate": 2e-05,
+      "loss": 5.4224,
+      "step": 9897
+    },
+    {
+      "epoch": 0.6639165576684442,
+      "grad_norm": 0.14622616660921148,
+      "learning_rate": 2e-05,
+      "loss": 5.4496,
+      "step": 9898
+    },
+    {
+      "epoch": 0.6639836334976691,
+      "grad_norm": 0.1470881430441305,
+      "learning_rate": 2e-05,
+      "loss": 5.579,
+      "step": 9899
+    },
+    {
+      "epoch": 0.664050709326894,
+      "grad_norm": 0.15843972941787934,
+      "learning_rate": 2e-05,
+      "loss": 5.4303,
+      "step": 9900
+    },
+    {
+      "epoch": 0.664117785156119,
+      "grad_norm": 0.1526624399528818,
+      "learning_rate": 2e-05,
+      "loss": 5.3936,
+      "step": 9901
+    },
+    {
+      "epoch": 0.6641848609853439,
+      "grad_norm": 0.1589124417942036,
+      "learning_rate": 2e-05,
+      "loss": 5.3324,
+      "step": 9902
+    },
+    {
+      "epoch": 0.6642519368145688,
+      "grad_norm": 0.1575864283308364,
+      "learning_rate": 2e-05,
+      "loss": 5.5622,
+      "step": 9903
+    },
+    {
+      "epoch": 0.6643190126437938,
+      "grad_norm": 0.16254524274777046,
+      "learning_rate": 2e-05,
+      "loss": 5.468,
+      "step": 9904
+    },
+    {
+      "epoch": 0.6643860884730187,
+      "grad_norm": 0.15577523793666237,
+      "learning_rate": 2e-05,
+      "loss": 5.4665,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6644531643022437,
+      "grad_norm": 0.15862163309634714,
+      "learning_rate": 2e-05,
+      "loss": 5.6137,
+      "step": 9906
+    },
+    {
+      "epoch": 0.6645202401314686,
+      "grad_norm": 0.16278496648284885,
+      "learning_rate": 2e-05,
+      "loss": 5.4982,
+      "step": 9907
+    },
+    {
+      "epoch": 0.6645873159606935,
+      "grad_norm": 0.15193519080671128,
+      "learning_rate": 2e-05,
+      "loss": 5.3189,
+      "step": 9908
+    },
+    {
+      "epoch": 0.6646543917899185,
+      "grad_norm": 0.16461078722477643,
+      "learning_rate": 2e-05,
+      "loss": 5.5028,
+      "step": 9909
+    },
+    {
+      "epoch": 0.6647214676191434,
+      "grad_norm": 0.16210005332478203,
+      "learning_rate": 2e-05,
+      "loss": 5.3672,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6647885434483684,
+      "grad_norm": 0.14614757273420356,
+      "learning_rate": 2e-05,
+      "loss": 5.432,
+      "step": 9911
+    },
+    {
+      "epoch": 0.6648556192775933,
+      "grad_norm": 0.16450511683652738,
+      "learning_rate": 2e-05,
+      "loss": 5.3688,
+      "step": 9912
+    },
+    {
+      "epoch": 0.6649226951068182,
+      "grad_norm": 0.15715128515520854,
+      "learning_rate": 2e-05,
+      "loss": 5.3983,
+      "step": 9913
+    },
+    {
+      "epoch": 0.6649897709360432,
+      "grad_norm": 0.15529814782687984,
+      "learning_rate": 2e-05,
+      "loss": 5.4654,
+      "step": 9914
+    },
+    {
+      "epoch": 0.6650568467652681,
+      "grad_norm": 0.1565738437594841,
+      "learning_rate": 2e-05,
+      "loss": 5.3571,
+      "step": 9915
+    },
+    {
+      "epoch": 0.665123922594493,
+      "grad_norm": 0.15875237606167872,
+      "learning_rate": 2e-05,
+      "loss": 5.4879,
+      "step": 9916
+    },
+    {
+      "epoch": 0.665190998423718,
+      "grad_norm": 0.14861413964220985,
+      "learning_rate": 2e-05,
+      "loss": 5.4492,
+      "step": 9917
+    },
+    {
+      "epoch": 0.6652580742529429,
+      "grad_norm": 0.1497800882094297,
+      "learning_rate": 2e-05,
+      "loss": 5.6035,
+      "step": 9918
+    },
+    {
+      "epoch": 0.6653251500821679,
+      "grad_norm": 0.15951319392515279,
+      "learning_rate": 2e-05,
+      "loss": 5.4922,
+      "step": 9919
+    },
+    {
+      "epoch": 0.6653922259113928,
+      "grad_norm": 0.15573715800291496,
+      "learning_rate": 2e-05,
+      "loss": 5.5399,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6654593017406177,
+      "grad_norm": 0.1489900388193695,
+      "learning_rate": 2e-05,
+      "loss": 5.417,
+      "step": 9921
+    },
+    {
+      "epoch": 0.6655263775698427,
+      "grad_norm": 0.15571509160309432,
+      "learning_rate": 2e-05,
+      "loss": 5.4329,
+      "step": 9922
+    },
+    {
+      "epoch": 0.6655934533990676,
+      "grad_norm": 0.15123270495262256,
+      "learning_rate": 2e-05,
+      "loss": 5.6707,
+      "step": 9923
+    },
+    {
+      "epoch": 0.6656605292282926,
+      "grad_norm": 0.1473301388484114,
+      "learning_rate": 2e-05,
+      "loss": 5.3682,
+      "step": 9924
+    },
+    {
+      "epoch": 0.6657276050575175,
+      "grad_norm": 0.15433794000478224,
+      "learning_rate": 2e-05,
+      "loss": 5.4033,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6657946808867424,
+      "grad_norm": 0.15073388144085226,
+      "learning_rate": 2e-05,
+      "loss": 5.4613,
+      "step": 9926
+    },
+    {
+      "epoch": 0.6658617567159674,
+      "grad_norm": 0.15133290462201973,
+      "learning_rate": 2e-05,
+      "loss": 5.4732,
+      "step": 9927
+    },
+    {
+      "epoch": 0.6659288325451923,
+      "grad_norm": 0.15240706764072337,
+      "learning_rate": 2e-05,
+      "loss": 5.4425,
+      "step": 9928
+    },
+    {
+      "epoch": 0.6659959083744172,
+      "grad_norm": 0.15169672394459774,
+      "learning_rate": 2e-05,
+      "loss": 5.4423,
+      "step": 9929
+    },
+    {
+      "epoch": 0.6660629842036422,
+      "grad_norm": 0.1538894541078572,
+      "learning_rate": 2e-05,
+      "loss": 5.5374,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6661300600328671,
+      "grad_norm": 0.15951777153130123,
+      "learning_rate": 2e-05,
+      "loss": 5.4681,
+      "step": 9931
+    },
+    {
+      "epoch": 0.6661971358620921,
+      "grad_norm": 0.14720024892693503,
+      "learning_rate": 2e-05,
+      "loss": 5.2664,
+      "step": 9932
+    },
+    {
+      "epoch": 0.666264211691317,
+      "grad_norm": 0.15553592377236336,
+      "learning_rate": 2e-05,
+      "loss": 5.3918,
+      "step": 9933
+    },
+    {
+      "epoch": 0.6663312875205419,
+      "grad_norm": 0.1499927330002631,
+      "learning_rate": 2e-05,
+      "loss": 5.3682,
+      "step": 9934
+    },
+    {
+      "epoch": 0.6663983633497669,
+      "grad_norm": 0.1617817256565762,
+      "learning_rate": 2e-05,
+      "loss": 5.6564,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6664654391789918,
+      "grad_norm": 0.1527125226787073,
+      "learning_rate": 2e-05,
+      "loss": 5.4608,
+      "step": 9936
+    },
+    {
+      "epoch": 0.6665325150082168,
+      "grad_norm": 0.14882205672565302,
+      "learning_rate": 2e-05,
+      "loss": 5.4668,
+      "step": 9937
+    },
+    {
+      "epoch": 0.6665995908374417,
+      "grad_norm": 0.15502790986746764,
+      "learning_rate": 2e-05,
+      "loss": 5.495,
+      "step": 9938
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.15684880466398246,
+      "learning_rate": 2e-05,
+      "loss": 5.4302,
+      "step": 9939
+    },
+    {
+      "epoch": 0.6667337424958916,
+      "grad_norm": 0.16390565198762222,
+      "learning_rate": 2e-05,
+      "loss": 5.282,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6668008183251165,
+      "grad_norm": 0.14824204334654237,
+      "learning_rate": 2e-05,
+      "loss": 5.3143,
+      "step": 9941
+    },
+    {
+      "epoch": 0.6668678941543414,
+      "grad_norm": 0.16525798356668472,
+      "learning_rate": 2e-05,
+      "loss": 5.4624,
+      "step": 9942
+    },
+    {
+      "epoch": 0.6669349699835664,
+      "grad_norm": 0.14744544545898752,
+      "learning_rate": 2e-05,
+      "loss": 5.4221,
+      "step": 9943
+    },
+    {
+      "epoch": 0.6670020458127913,
+      "grad_norm": 0.15309622542448523,
+      "learning_rate": 2e-05,
+      "loss": 5.4132,
+      "step": 9944
+    },
+    {
+      "epoch": 0.6670691216420163,
+      "grad_norm": 0.1635222947711619,
+      "learning_rate": 2e-05,
+      "loss": 5.4118,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6671361974712412,
+      "grad_norm": 0.15482386482463392,
+      "learning_rate": 2e-05,
+      "loss": 5.3977,
+      "step": 9946
+    },
+    {
+      "epoch": 0.6672032733004661,
+      "grad_norm": 0.1543449351531162,
+      "learning_rate": 2e-05,
+      "loss": 5.3858,
+      "step": 9947
+    },
+    {
+      "epoch": 0.6672703491296911,
+      "grad_norm": 0.1620916310523679,
+      "learning_rate": 2e-05,
+      "loss": 5.4441,
+      "step": 9948
+    },
+    {
+      "epoch": 0.667337424958916,
+      "grad_norm": 0.15547543676521997,
+      "learning_rate": 2e-05,
+      "loss": 5.4328,
+      "step": 9949
+    },
+    {
+      "epoch": 0.667404500788141,
+      "grad_norm": 0.1482941582388271,
+      "learning_rate": 2e-05,
+      "loss": 5.5053,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6674715766173659,
+      "grad_norm": 0.15136432716871334,
+      "learning_rate": 2e-05,
+      "loss": 5.4064,
+      "step": 9951
+    },
+    {
+      "epoch": 0.6675386524465908,
+      "grad_norm": 0.1549246968050513,
+      "learning_rate": 2e-05,
+      "loss": 5.4454,
+      "step": 9952
+    },
+    {
+      "epoch": 0.6676057282758158,
+      "grad_norm": 0.15227538556117684,
+      "learning_rate": 2e-05,
+      "loss": 5.3511,
+      "step": 9953
+    },
+    {
+      "epoch": 0.6676728041050407,
+      "grad_norm": 0.15509379135906604,
+      "learning_rate": 2e-05,
+      "loss": 5.4369,
+      "step": 9954
+    },
+    {
+      "epoch": 0.6677398799342656,
+      "grad_norm": 0.1566494081923976,
+      "learning_rate": 2e-05,
+      "loss": 5.4864,
+      "step": 9955
+    },
+    {
+      "epoch": 0.6678069557634906,
+      "grad_norm": 0.16003163192605468,
+      "learning_rate": 2e-05,
+      "loss": 5.304,
+      "step": 9956
+    },
+    {
+      "epoch": 0.6678740315927155,
+      "grad_norm": 0.15450070262385776,
+      "learning_rate": 2e-05,
+      "loss": 5.523,
+      "step": 9957
+    },
+    {
+      "epoch": 0.6679411074219405,
+      "grad_norm": 0.15602559645805045,
+      "learning_rate": 2e-05,
+      "loss": 5.5701,
+      "step": 9958
+    },
+    {
+      "epoch": 0.6680081832511654,
+      "grad_norm": 0.15696806410664377,
+      "learning_rate": 2e-05,
+      "loss": 5.5698,
+      "step": 9959
+    },
+    {
+      "epoch": 0.6680752590803903,
+      "grad_norm": 0.1641461831020307,
+      "learning_rate": 2e-05,
+      "loss": 5.5103,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6681423349096153,
+      "grad_norm": 0.15707582656050184,
+      "learning_rate": 2e-05,
+      "loss": 5.3952,
+      "step": 9961
+    },
+    {
+      "epoch": 0.6682094107388402,
+      "grad_norm": 0.15254124388546,
+      "learning_rate": 2e-05,
+      "loss": 5.4323,
+      "step": 9962
+    },
+    {
+      "epoch": 0.6682764865680652,
+      "grad_norm": 0.16300698655026616,
+      "learning_rate": 2e-05,
+      "loss": 5.4246,
+      "step": 9963
+    },
+    {
+      "epoch": 0.6683435623972901,
+      "grad_norm": 0.1538032645616055,
+      "learning_rate": 2e-05,
+      "loss": 5.5084,
+      "step": 9964
+    },
+    {
+      "epoch": 0.668410638226515,
+      "grad_norm": 0.15121377959457702,
+      "learning_rate": 2e-05,
+      "loss": 5.4942,
+      "step": 9965
+    },
+    {
+      "epoch": 0.66847771405574,
+      "grad_norm": 0.15701088267031943,
+      "learning_rate": 2e-05,
+      "loss": 5.5637,
+      "step": 9966
+    },
+    {
+      "epoch": 0.6685447898849649,
+      "grad_norm": 0.1566332126977651,
+      "learning_rate": 2e-05,
+      "loss": 5.3458,
+      "step": 9967
+    },
+    {
+      "epoch": 0.6686118657141898,
+      "grad_norm": 0.16097437088348204,
+      "learning_rate": 2e-05,
+      "loss": 5.5506,
+      "step": 9968
+    },
+    {
+      "epoch": 0.6686789415434148,
+      "grad_norm": 0.14941595096094146,
+      "learning_rate": 2e-05,
+      "loss": 5.383,
+      "step": 9969
+    },
+    {
+      "epoch": 0.6687460173726397,
+      "grad_norm": 0.17137129633555515,
+      "learning_rate": 2e-05,
+      "loss": 5.3438,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6688130932018647,
+      "grad_norm": 0.15972501963282873,
+      "learning_rate": 2e-05,
+      "loss": 5.4639,
+      "step": 9971
+    },
+    {
+      "epoch": 0.6688801690310896,
+      "grad_norm": 0.1470685078787576,
+      "learning_rate": 2e-05,
+      "loss": 5.4223,
+      "step": 9972
+    },
+    {
+      "epoch": 0.6689472448603145,
+      "grad_norm": 0.14738151260212207,
+      "learning_rate": 2e-05,
+      "loss": 5.4187,
+      "step": 9973
+    },
+    {
+      "epoch": 0.6690143206895395,
+      "grad_norm": 0.15224231856721315,
+      "learning_rate": 2e-05,
+      "loss": 5.5428,
+      "step": 9974
+    },
+    {
+      "epoch": 0.6690813965187644,
+      "grad_norm": 0.15236931931780764,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6691484723479894,
+      "grad_norm": 0.1440581462480556,
+      "learning_rate": 2e-05,
+      "loss": 5.2196,
+      "step": 9976
+    },
+    {
+      "epoch": 0.6692155481772143,
+      "grad_norm": 0.15197166232604215,
+      "learning_rate": 2e-05,
+      "loss": 5.5157,
+      "step": 9977
+    },
+    {
+      "epoch": 0.6692826240064392,
+      "grad_norm": 0.15139655887583334,
+      "learning_rate": 2e-05,
+      "loss": 5.3791,
+      "step": 9978
+    },
+    {
+      "epoch": 0.6693496998356642,
+      "grad_norm": 0.14588590396242446,
+      "learning_rate": 2e-05,
+      "loss": 5.4614,
+      "step": 9979
+    },
+    {
+      "epoch": 0.6694167756648891,
+      "grad_norm": 0.15603723878436804,
+      "learning_rate": 2e-05,
+      "loss": 5.4584,
+      "step": 9980
+    },
+    {
+      "epoch": 0.669483851494114,
+      "grad_norm": 0.15509686999130207,
+      "learning_rate": 2e-05,
+      "loss": 5.4721,
+      "step": 9981
+    },
+    {
+      "epoch": 0.669550927323339,
+      "grad_norm": 0.15615746290872776,
+      "learning_rate": 2e-05,
+      "loss": 5.4512,
+      "step": 9982
+    },
+    {
+      "epoch": 0.6696180031525639,
+      "grad_norm": 0.1548041011453804,
+      "learning_rate": 2e-05,
+      "loss": 5.3855,
+      "step": 9983
+    },
+    {
+      "epoch": 0.6696850789817889,
+      "grad_norm": 0.1586709516583898,
+      "learning_rate": 2e-05,
+      "loss": 5.302,
+      "step": 9984
+    },
+    {
+      "epoch": 0.6697521548110138,
+      "grad_norm": 0.1544436378983193,
+      "learning_rate": 2e-05,
+      "loss": 5.3263,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6698192306402387,
+      "grad_norm": 0.1570757299380076,
+      "learning_rate": 2e-05,
+      "loss": 5.3323,
+      "step": 9986
+    },
+    {
+      "epoch": 0.6698863064694637,
+      "grad_norm": 0.15746434877463752,
+      "learning_rate": 2e-05,
+      "loss": 5.3288,
+      "step": 9987
+    },
+    {
+      "epoch": 0.6699533822986886,
+      "grad_norm": 0.15779685488525402,
+      "learning_rate": 2e-05,
+      "loss": 5.4011,
+      "step": 9988
+    },
+    {
+      "epoch": 0.6700204581279136,
+      "grad_norm": 0.16645831436404973,
+      "learning_rate": 2e-05,
+      "loss": 5.4708,
+      "step": 9989
+    },
+    {
+      "epoch": 0.6700875339571385,
+      "grad_norm": 0.15911966559437826,
+      "learning_rate": 2e-05,
+      "loss": 5.4077,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6701546097863634,
+      "grad_norm": 0.1553468468796053,
+      "learning_rate": 2e-05,
+      "loss": 5.4514,
+      "step": 9991
+    },
+    {
+      "epoch": 0.6702216856155884,
+      "grad_norm": 0.1544583189122214,
+      "learning_rate": 2e-05,
+      "loss": 5.4281,
+      "step": 9992
+    },
+    {
+      "epoch": 0.6702887614448133,
+      "grad_norm": 0.15193312231861617,
+      "learning_rate": 2e-05,
+      "loss": 5.3875,
+      "step": 9993
+    },
+    {
+      "epoch": 0.6703558372740382,
+      "grad_norm": 0.1563376417366738,
+      "learning_rate": 2e-05,
+      "loss": 5.4311,
+      "step": 9994
+    },
+    {
+      "epoch": 0.6704229131032632,
+      "grad_norm": 0.14631970063599006,
+      "learning_rate": 2e-05,
+      "loss": 5.3,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6704899889324881,
+      "grad_norm": 0.14797232255692694,
+      "learning_rate": 2e-05,
+      "loss": 5.3771,
+      "step": 9996
+    },
+    {
+      "epoch": 0.6705570647617131,
+      "grad_norm": 0.15259857235114627,
+      "learning_rate": 2e-05,
+      "loss": 5.4619,
+      "step": 9997
+    },
+    {
+      "epoch": 0.670624140590938,
+      "grad_norm": 0.1498548613638616,
+      "learning_rate": 2e-05,
+      "loss": 5.4547,
+      "step": 9998
+    },
+    {
+      "epoch": 0.6706912164201629,
+      "grad_norm": 0.15113487916886195,
+      "learning_rate": 2e-05,
+      "loss": 5.4338,
+      "step": 9999
+    },
+    {
+      "epoch": 0.670758292249388,
+      "grad_norm": 0.14881007214570116,
+      "learning_rate": 2e-05,
+      "loss": 5.3792,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 745400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4575457078280192.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}