diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9766 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.994606256742179,
+  "eval_steps": 500,
+  "global_step": 1389,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002157497303128371,
+      "grad_norm": 57.13587782353209,
+      "learning_rate": 0.0,
+      "loss": 11.1563,
+      "step": 1
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 57.57684447116968,
+      "learning_rate": 3.5971223021582736e-07,
+      "loss": 11.0961,
+      "step": 2
+    },
+    {
+      "epoch": 0.006472491909385114,
+      "grad_norm": 54.76562948065167,
+      "learning_rate": 7.194244604316547e-07,
+      "loss": 11.2027,
+      "step": 3
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 56.76231147192983,
+      "learning_rate": 1.0791366906474822e-06,
+      "loss": 11.1248,
+      "step": 4
+    },
+    {
+      "epoch": 0.010787486515641856,
+      "grad_norm": 59.88129471289495,
+      "learning_rate": 1.4388489208633094e-06,
+      "loss": 11.0418,
+      "step": 5
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 59.30119373658902,
+      "learning_rate": 1.7985611510791366e-06,
+      "loss": 11.0469,
+      "step": 6
+    },
+    {
+      "epoch": 0.015102481121898598,
+      "grad_norm": 63.23987032545693,
+      "learning_rate": 2.1582733812949645e-06,
+      "loss": 10.7637,
+      "step": 7
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 67.8264131296607,
+      "learning_rate": 2.5179856115107916e-06,
+      "loss": 10.5729,
+      "step": 8
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 96.87216003729526,
+      "learning_rate": 2.877697841726619e-06,
+      "loss": 9.33,
+      "step": 9
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 110.36643138785996,
+      "learning_rate": 3.237410071942446e-06,
+      "loss": 8.7194,
+      "step": 10
+    },
+    {
+      "epoch": 0.023732470334412083,
+      "grad_norm": 62.62903504880271,
+      "learning_rate": 3.5971223021582732e-06,
+      "loss": 3.5156,
+      "step": 11
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 55.87526116545333,
+      "learning_rate": 3.956834532374101e-06,
+      "loss": 3.2214,
+      "step": 12
+    },
+    {
+      "epoch": 0.028047464940668825,
+      "grad_norm": 38.69327465608191,
+      "learning_rate": 4.316546762589929e-06,
+      "loss": 2.5429,
+      "step": 13
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 29.949119954850115,
+      "learning_rate": 4.676258992805756e-06,
+      "loss": 2.2303,
+      "step": 14
+    },
+    {
+      "epoch": 0.032362459546925564,
+      "grad_norm": 6.23171849490708,
+      "learning_rate": 5.035971223021583e-06,
+      "loss": 1.3887,
+      "step": 15
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 4.727559417918715,
+      "learning_rate": 5.3956834532374105e-06,
+      "loss": 1.2918,
+      "step": 16
+    },
+    {
+      "epoch": 0.036677454153182305,
+      "grad_norm": 3.940521844725638,
+      "learning_rate": 5.755395683453238e-06,
+      "loss": 1.2401,
+      "step": 17
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 2.6425075308520816,
+      "learning_rate": 6.115107913669065e-06,
+      "loss": 1.1628,
+      "step": 18
+    },
+    {
+      "epoch": 0.040992448759439054,
+      "grad_norm": 2.1413189742525223,
+      "learning_rate": 6.474820143884892e-06,
+      "loss": 1.106,
+      "step": 19
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.4939789670719916,
+      "learning_rate": 6.83453237410072e-06,
+      "loss": 0.9834,
+      "step": 20
+    },
+    {
+      "epoch": 0.045307443365695796,
+      "grad_norm": 119.59331302017698,
+      "learning_rate": 7.1942446043165465e-06,
+      "loss": 1.0391,
+      "step": 21
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 74.35045753368888,
+      "learning_rate": 7.5539568345323745e-06,
+      "loss": 0.984,
+      "step": 22
+    },
+    {
+      "epoch": 0.04962243797195254,
+      "grad_norm": 2.1166562532348547,
+      "learning_rate": 7.913669064748202e-06,
+      "loss": 0.9028,
+      "step": 23
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.259997812455964,
+      "learning_rate": 8.273381294964029e-06,
+      "loss": 0.9009,
+      "step": 24
+    },
+    {
+      "epoch": 0.05393743257820928,
+      "grad_norm": 1.0042382956541405,
+      "learning_rate": 8.633093525179858e-06,
+      "loss": 0.8458,
+      "step": 25
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 0.8711320683532612,
+      "learning_rate": 8.992805755395683e-06,
+      "loss": 0.8222,
+      "step": 26
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 0.8070054635681398,
+      "learning_rate": 9.352517985611512e-06,
+      "loss": 0.7821,
+      "step": 27
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 0.907177791133324,
+      "learning_rate": 9.71223021582734e-06,
+      "loss": 0.7905,
+      "step": 28
+    },
+    {
+      "epoch": 0.06256742179072276,
+      "grad_norm": 0.8987984527990586,
+      "learning_rate": 1.0071942446043167e-05,
+      "loss": 0.7743,
+      "step": 29
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 0.766780715327364,
+      "learning_rate": 1.0431654676258994e-05,
+      "loss": 0.7393,
+      "step": 30
+    },
+    {
+      "epoch": 0.0668824163969795,
+      "grad_norm": 0.6221327677977063,
+      "learning_rate": 1.0791366906474821e-05,
+      "loss": 0.7169,
+      "step": 31
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 0.724452806351174,
+      "learning_rate": 1.1151079136690648e-05,
+      "loss": 0.6844,
+      "step": 32
+    },
+    {
+      "epoch": 0.07119741100323625,
+      "grad_norm": 0.726983862192927,
+      "learning_rate": 1.1510791366906475e-05,
+      "loss": 0.6782,
+      "step": 33
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 0.5784038572104493,
+      "learning_rate": 1.1870503597122303e-05,
+      "loss": 0.6674,
+      "step": 34
+    },
+    {
+      "epoch": 0.07551240560949299,
+      "grad_norm": 0.5549566422059363,
+      "learning_rate": 1.223021582733813e-05,
+      "loss": 0.6972,
+      "step": 35
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 0.6399396561875369,
+      "learning_rate": 1.2589928057553957e-05,
+      "loss": 0.6792,
+      "step": 36
+    },
+    {
+      "epoch": 0.07982740021574973,
+      "grad_norm": 0.5849368618649039,
+      "learning_rate": 1.2949640287769784e-05,
+      "loss": 0.6695,
+      "step": 37
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 0.5900434638838187,
+      "learning_rate": 1.3309352517985613e-05,
+      "loss": 0.6773,
+      "step": 38
+    },
+    {
+      "epoch": 0.08414239482200647,
+      "grad_norm": 0.4914556169563964,
+      "learning_rate": 1.366906474820144e-05,
+      "loss": 0.6521,
+      "step": 39
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 0.3975245564581552,
+      "learning_rate": 1.4028776978417266e-05,
+      "loss": 0.6457,
+      "step": 40
+    },
+    {
+      "epoch": 0.08845738942826321,
+      "grad_norm": 0.5020290377407021,
+      "learning_rate": 1.4388489208633093e-05,
+      "loss": 0.6393,
+      "step": 41
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 0.46341748213501166,
+      "learning_rate": 1.4748201438848922e-05,
+      "loss": 0.5959,
+      "step": 42
+    },
+    {
+      "epoch": 0.09277238403451996,
+      "grad_norm": 0.4138969749517581,
+      "learning_rate": 1.5107913669064749e-05,
+      "loss": 0.623,
+      "step": 43
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 0.4470998608888789,
+      "learning_rate": 1.5467625899280578e-05,
+      "loss": 0.6173,
+      "step": 44
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.4268271776986609,
+      "learning_rate": 1.5827338129496403e-05,
+      "loss": 0.6252,
+      "step": 45
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 0.3992607277762862,
+      "learning_rate": 1.618705035971223e-05,
+      "loss": 0.6104,
+      "step": 46
+    },
+    {
+      "epoch": 0.10140237324703344,
+      "grad_norm": 0.46730384996685315,
+      "learning_rate": 1.6546762589928058e-05,
+      "loss": 0.6067,
+      "step": 47
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 0.340962945547672,
+      "learning_rate": 1.6906474820143887e-05,
+      "loss": 0.6184,
+      "step": 48
+    },
+    {
+      "epoch": 0.10571736785329018,
+      "grad_norm": 0.3169350607679008,
+      "learning_rate": 1.7266187050359716e-05,
+      "loss": 0.5889,
+      "step": 49
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 0.42059973714399224,
+      "learning_rate": 1.7625899280575538e-05,
+      "loss": 0.6113,
+      "step": 50
+    },
+    {
+      "epoch": 0.11003236245954692,
+      "grad_norm": 0.33716661312326635,
+      "learning_rate": 1.7985611510791367e-05,
+      "loss": 0.5815,
+      "step": 51
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 0.2944086703558636,
+      "learning_rate": 1.8345323741007196e-05,
+      "loss": 0.6006,
+      "step": 52
+    },
+    {
+      "epoch": 0.11434735706580366,
+      "grad_norm": 0.349551156334338,
+      "learning_rate": 1.8705035971223024e-05,
+      "loss": 0.573,
+      "step": 53
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 0.3128341661906979,
+      "learning_rate": 1.906474820143885e-05,
+      "loss": 0.5897,
+      "step": 54
+    },
+    {
+      "epoch": 0.1186623516720604,
+      "grad_norm": 0.30337839597627236,
+      "learning_rate": 1.942446043165468e-05,
+      "loss": 0.585,
+      "step": 55
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 0.2977094639323748,
+      "learning_rate": 1.9784172661870504e-05,
+      "loss": 0.5783,
+      "step": 56
+    },
+    {
+      "epoch": 0.12297734627831715,
+      "grad_norm": 0.3163276610774343,
+      "learning_rate": 2.0143884892086333e-05,
+      "loss": 0.5844,
+      "step": 57
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 0.2909825753497844,
+      "learning_rate": 2.050359712230216e-05,
+      "loss": 0.5773,
+      "step": 58
+    },
+    {
+      "epoch": 0.1272923408845739,
+      "grad_norm": 0.24835717615595493,
+      "learning_rate": 2.0863309352517988e-05,
+      "loss": 0.5666,
+      "step": 59
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 0.30633255958737043,
+      "learning_rate": 2.1223021582733816e-05,
+      "loss": 0.5567,
+      "step": 60
+    },
+    {
+      "epoch": 0.13160733549083065,
+      "grad_norm": 0.3123982037388874,
+      "learning_rate": 2.1582733812949642e-05,
+      "loss": 0.5845,
+      "step": 61
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 0.3547886968310134,
+      "learning_rate": 2.1942446043165467e-05,
+      "loss": 0.5574,
+      "step": 62
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 0.3014066909176335,
+      "learning_rate": 2.2302158273381296e-05,
+      "loss": 0.5621,
+      "step": 63
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 0.24373585748942053,
+      "learning_rate": 2.2661870503597125e-05,
+      "loss": 0.5787,
+      "step": 64
+    },
+    {
+      "epoch": 0.14023732470334413,
+      "grad_norm": 0.2552538074958043,
+      "learning_rate": 2.302158273381295e-05,
+      "loss": 0.5376,
+      "step": 65
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 0.2602279217663108,
+      "learning_rate": 2.3381294964028776e-05,
+      "loss": 0.565,
+      "step": 66
+    },
+    {
+      "epoch": 0.14455231930960086,
+      "grad_norm": 0.2540749697215477,
+      "learning_rate": 2.3741007194244605e-05,
+      "loss": 0.5384,
+      "step": 67
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 0.265028926794241,
+      "learning_rate": 2.4100719424460434e-05,
+      "loss": 0.5404,
+      "step": 68
+    },
+    {
+      "epoch": 0.1488673139158576,
+      "grad_norm": 0.24987952808261527,
+      "learning_rate": 2.446043165467626e-05,
+      "loss": 0.555,
+      "step": 69
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 0.24626279967807332,
+      "learning_rate": 2.482014388489209e-05,
+      "loss": 0.5465,
+      "step": 70
+    },
+    {
+      "epoch": 0.15318230852211434,
+      "grad_norm": 0.24540860578710721,
+      "learning_rate": 2.5179856115107914e-05,
+      "loss": 0.5389,
+      "step": 71
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 0.2819028166039196,
+      "learning_rate": 2.5539568345323743e-05,
+      "loss": 0.5341,
+      "step": 72
+    },
+    {
+      "epoch": 0.1574973031283711,
+      "grad_norm": 0.2725924730541978,
+      "learning_rate": 2.589928057553957e-05,
+      "loss": 0.5593,
+      "step": 73
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 0.41273921517798356,
+      "learning_rate": 2.6258992805755394e-05,
+      "loss": 0.5627,
+      "step": 74
+    },
+    {
+      "epoch": 0.16181229773462782,
+      "grad_norm": 0.2582400805700941,
+      "learning_rate": 2.6618705035971226e-05,
+      "loss": 0.5306,
+      "step": 75
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 0.26006309950535167,
+      "learning_rate": 2.697841726618705e-05,
+      "loss": 0.5215,
+      "step": 76
+    },
+    {
+      "epoch": 0.16612729234088458,
+      "grad_norm": 0.24344187994219604,
+      "learning_rate": 2.733812949640288e-05,
+      "loss": 0.5243,
+      "step": 77
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 0.2520498275788824,
+      "learning_rate": 2.7697841726618706e-05,
+      "loss": 0.5263,
+      "step": 78
+    },
+    {
+      "epoch": 0.1704422869471413,
+      "grad_norm": 0.25245524340715986,
+      "learning_rate": 2.805755395683453e-05,
+      "loss": 0.5287,
+      "step": 79
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 0.2605418146662812,
+      "learning_rate": 2.841726618705036e-05,
+      "loss": 0.5301,
+      "step": 80
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 0.24212992631081523,
+      "learning_rate": 2.8776978417266186e-05,
+      "loss": 0.5359,
+      "step": 81
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 0.24539478768207,
+      "learning_rate": 2.9136690647482018e-05,
+      "loss": 0.5208,
+      "step": 82
+    },
+    {
+      "epoch": 0.1790722761596548,
+      "grad_norm": 0.24763507587346034,
+      "learning_rate": 2.9496402877697844e-05,
+      "loss": 0.5203,
+      "step": 83
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 0.27551677902256216,
+      "learning_rate": 2.985611510791367e-05,
+      "loss": 0.5228,
+      "step": 84
+    },
+    {
+      "epoch": 0.18338727076591155,
+      "grad_norm": 0.24453606868458008,
+      "learning_rate": 3.0215827338129498e-05,
+      "loss": 0.5235,
+      "step": 85
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 0.26688122063771413,
+      "learning_rate": 3.0575539568345324e-05,
+      "loss": 0.5343,
+      "step": 86
+    },
+    {
+      "epoch": 0.18770226537216828,
+      "grad_norm": 0.29529860209112546,
+      "learning_rate": 3.0935251798561156e-05,
+      "loss": 0.5367,
+      "step": 87
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 0.24266710111447526,
+      "learning_rate": 3.129496402877698e-05,
+      "loss": 0.5253,
+      "step": 88
+    },
+    {
+      "epoch": 0.19201725997842503,
+      "grad_norm": 0.23807746989279177,
+      "learning_rate": 3.165467625899281e-05,
+      "loss": 0.5185,
+      "step": 89
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 0.2626350221760916,
+      "learning_rate": 3.201438848920863e-05,
+      "loss": 0.5395,
+      "step": 90
+    },
+    {
+      "epoch": 0.19633225458468176,
+      "grad_norm": 0.2414024909993752,
+      "learning_rate": 3.237410071942446e-05,
+      "loss": 0.532,
+      "step": 91
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 0.25017076106275854,
+      "learning_rate": 3.273381294964029e-05,
+      "loss": 0.5276,
+      "step": 92
+    },
+    {
+      "epoch": 0.20064724919093851,
+      "grad_norm": 0.2918654609740857,
+      "learning_rate": 3.3093525179856116e-05,
+      "loss": 0.5204,
+      "step": 93
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 0.2624246785726397,
+      "learning_rate": 3.345323741007194e-05,
+      "loss": 0.5336,
+      "step": 94
+    },
+    {
+      "epoch": 0.20496224379719524,
+      "grad_norm": 0.2452405947687516,
+      "learning_rate": 3.3812949640287773e-05,
+      "loss": 0.5105,
+      "step": 95
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 0.27455141570832603,
+      "learning_rate": 3.41726618705036e-05,
+      "loss": 0.517,
+      "step": 96
+    },
+    {
+      "epoch": 0.209277238403452,
+      "grad_norm": 0.234849775079589,
+      "learning_rate": 3.453237410071943e-05,
+      "loss": 0.5155,
+      "step": 97
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 0.25742508943184506,
+      "learning_rate": 3.489208633093525e-05,
+      "loss": 0.5348,
+      "step": 98
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 0.23602027129846576,
+      "learning_rate": 3.5251798561151075e-05,
+      "loss": 0.5136,
+      "step": 99
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 0.2830916969833504,
+      "learning_rate": 3.561151079136691e-05,
+      "loss": 0.5141,
+      "step": 100
+    },
+    {
+      "epoch": 0.21790722761596548,
+      "grad_norm": 0.2661717768903485,
+      "learning_rate": 3.597122302158273e-05,
+      "loss": 0.514,
+      "step": 101
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 0.2486575067928459,
+      "learning_rate": 3.6330935251798566e-05,
+      "loss": 0.5146,
+      "step": 102
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.2953832672456234,
+      "learning_rate": 3.669064748201439e-05,
+      "loss": 0.5302,
+      "step": 103
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 0.28906909952910065,
+      "learning_rate": 3.7050359712230217e-05,
+      "loss": 0.4751,
+      "step": 104
+    },
+    {
+      "epoch": 0.22653721682847897,
+      "grad_norm": 0.2686201225334034,
+      "learning_rate": 3.741007194244605e-05,
+      "loss": 0.511,
+      "step": 105
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 0.2741508909189952,
+      "learning_rate": 3.776978417266187e-05,
+      "loss": 0.5184,
+      "step": 106
+    },
+    {
+      "epoch": 0.2308522114347357,
+      "grad_norm": 0.2802116060264731,
+      "learning_rate": 3.81294964028777e-05,
+      "loss": 0.4975,
+      "step": 107
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 0.2667693505204224,
+      "learning_rate": 3.8489208633093525e-05,
+      "loss": 0.4981,
+      "step": 108
+    },
+    {
+      "epoch": 0.23516720604099245,
+      "grad_norm": 0.2597113095109555,
+      "learning_rate": 3.884892086330936e-05,
+      "loss": 0.5118,
+      "step": 109
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 0.2697903200086048,
+      "learning_rate": 3.920863309352518e-05,
+      "loss": 0.5051,
+      "step": 110
+    },
+    {
+      "epoch": 0.23948220064724918,
+      "grad_norm": 0.2713556557014583,
+      "learning_rate": 3.956834532374101e-05,
+      "loss": 0.5212,
+      "step": 111
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 0.2728968900059994,
+      "learning_rate": 3.992805755395684e-05,
+      "loss": 0.5134,
+      "step": 112
+    },
+    {
+      "epoch": 0.24379719525350593,
+      "grad_norm": 0.3057783884890205,
+      "learning_rate": 4.0287769784172666e-05,
+      "loss": 0.5174,
+      "step": 113
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 0.2526607253149815,
+      "learning_rate": 4.064748201438849e-05,
+      "loss": 0.4979,
+      "step": 114
+    },
+    {
+      "epoch": 0.2481121898597627,
+      "grad_norm": 0.29935142309285345,
+      "learning_rate": 4.100719424460432e-05,
+      "loss": 0.4791,
+      "step": 115
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 0.274036916116646,
+      "learning_rate": 4.136690647482014e-05,
+      "loss": 0.503,
+      "step": 116
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 0.29373833817106976,
+      "learning_rate": 4.1726618705035975e-05,
+      "loss": 0.4979,
+      "step": 117
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 0.30379306843345605,
+      "learning_rate": 4.20863309352518e-05,
+      "loss": 0.5143,
+      "step": 118
+    },
+    {
+      "epoch": 0.25674217907227614,
+      "grad_norm": 0.3597081745599116,
+      "learning_rate": 4.244604316546763e-05,
+      "loss": 0.4963,
+      "step": 119
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 0.27784099569060233,
+      "learning_rate": 4.280575539568346e-05,
+      "loss": 0.4964,
+      "step": 120
+    },
+    {
+      "epoch": 0.26105717367853293,
+      "grad_norm": 0.2895994279044635,
+      "learning_rate": 4.3165467625899284e-05,
+      "loss": 0.4953,
+      "step": 121
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 0.34053962799665355,
+      "learning_rate": 4.352517985611511e-05,
+      "loss": 0.5335,
+      "step": 122
+    },
+    {
+      "epoch": 0.26537216828478966,
+      "grad_norm": 0.3141740622845576,
+      "learning_rate": 4.3884892086330935e-05,
+      "loss": 0.5071,
+      "step": 123
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 0.3046488749690831,
+      "learning_rate": 4.424460431654677e-05,
+      "loss": 0.4859,
+      "step": 124
+    },
+    {
+      "epoch": 0.2696871628910464,
+      "grad_norm": 0.30610931123203444,
+      "learning_rate": 4.460431654676259e-05,
+      "loss": 0.4708,
+      "step": 125
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 0.3193175150854491,
+      "learning_rate": 4.496402877697842e-05,
+      "loss": 0.4848,
+      "step": 126
+    },
+    {
+      "epoch": 0.2740021574973031,
+      "grad_norm": 0.3255701503756811,
+      "learning_rate": 4.532374100719425e-05,
+      "loss": 0.4766,
+      "step": 127
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 0.3599922285030049,
+      "learning_rate": 4.5683453237410076e-05,
+      "loss": 0.5077,
+      "step": 128
+    },
+    {
+      "epoch": 0.2783171521035599,
+      "grad_norm": 0.3449429958647818,
+      "learning_rate": 4.60431654676259e-05,
+      "loss": 0.4859,
+      "step": 129
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 0.30781106612363096,
+      "learning_rate": 4.640287769784173e-05,
+      "loss": 0.5086,
+      "step": 130
+    },
+    {
+      "epoch": 0.2826321467098166,
+      "grad_norm": 0.31034736893309267,
+      "learning_rate": 4.676258992805755e-05,
+      "loss": 0.4706,
+      "step": 131
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 0.3297773636655693,
+      "learning_rate": 4.7122302158273385e-05,
+      "loss": 0.4901,
+      "step": 132
+    },
+    {
+      "epoch": 0.28694714131607335,
+      "grad_norm": 0.31788365317052747,
+      "learning_rate": 4.748201438848921e-05,
+      "loss": 0.5098,
+      "step": 133
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 0.38398998891452074,
+      "learning_rate": 4.784172661870504e-05,
+      "loss": 0.493,
+      "step": 134
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.3273138176909412,
+      "learning_rate": 4.820143884892087e-05,
+      "loss": 0.4769,
+      "step": 135
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 0.33365763486992506,
+      "learning_rate": 4.8561151079136694e-05,
+      "loss": 0.49,
+      "step": 136
+    },
+    {
+      "epoch": 0.29557713052858686,
+      "grad_norm": 0.301373109121661,
+      "learning_rate": 4.892086330935252e-05,
+      "loss": 0.4869,
+      "step": 137
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 0.36827893859352856,
+      "learning_rate": 4.9280575539568345e-05,
+      "loss": 0.494,
+      "step": 138
+    },
+    {
+      "epoch": 0.2998921251348436,
+      "grad_norm": 0.4286223278223316,
+      "learning_rate": 4.964028776978418e-05,
+      "loss": 0.508,
+      "step": 139
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 0.3583525731124658,
+      "learning_rate": 5e-05,
+      "loss": 0.495,
+      "step": 140
+    },
+    {
+      "epoch": 0.3042071197411003,
+      "grad_norm": 0.3221902127168568,
+      "learning_rate": 4.996e-05,
+      "loss": 0.5045,
+      "step": 141
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 0.337708249785391,
+      "learning_rate": 4.992e-05,
+      "loss": 0.4929,
+      "step": 142
+    },
+    {
+      "epoch": 0.30852211434735705,
+      "grad_norm": 0.3394347724632982,
+      "learning_rate": 4.9880000000000004e-05,
+      "loss": 0.5047,
+      "step": 143
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 0.2813961175434312,
+      "learning_rate": 4.9840000000000004e-05,
+      "loss": 0.4949,
+      "step": 144
+    },
+    {
+      "epoch": 0.31283710895361383,
+      "grad_norm": 0.30421333717748744,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.486,
+      "step": 145
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 0.36470188308478374,
+      "learning_rate": 4.976e-05,
+      "loss": 0.4983,
+      "step": 146
+    },
+    {
+      "epoch": 0.31715210355987056,
+      "grad_norm": 0.27856852801649823,
+      "learning_rate": 4.972e-05,
+      "loss": 0.4883,
+      "step": 147
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 0.3044782327916934,
+      "learning_rate": 4.9680000000000005e-05,
+      "loss": 0.4922,
+      "step": 148
+    },
+    {
+      "epoch": 0.3214670981661273,
+      "grad_norm": 0.2804188353274235,
+      "learning_rate": 4.9640000000000006e-05,
+      "loss": 0.4866,
+      "step": 149
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 0.30594014263779995,
+      "learning_rate": 4.96e-05,
+      "loss": 0.4745,
+      "step": 150
+    },
+    {
+      "epoch": 0.325782092772384,
+      "grad_norm": 0.3175912214255066,
+      "learning_rate": 4.956e-05,
+      "loss": 0.4986,
+      "step": 151
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 0.31983013814077355,
+      "learning_rate": 4.952e-05,
+      "loss": 0.48,
+      "step": 152
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 0.3230897992697352,
+      "learning_rate": 4.948000000000001e-05,
+      "loss": 0.4825,
+      "step": 153
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 0.26971956302103783,
+      "learning_rate": 4.944e-05,
+      "loss": 0.4806,
+      "step": 154
+    },
+    {
+      "epoch": 0.3344120819848975,
+      "grad_norm": 0.2878515152482518,
+      "learning_rate": 4.94e-05,
+      "loss": 0.4868,
+      "step": 155
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 0.29671867961170256,
+      "learning_rate": 4.936e-05,
+      "loss": 0.4967,
+      "step": 156
+    },
+    {
+      "epoch": 0.33872707659115425,
+      "grad_norm": 0.28930159604607764,
+      "learning_rate": 4.932e-05,
+      "loss": 0.4877,
+      "step": 157
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 0.3015050117518478,
+      "learning_rate": 4.928e-05,
+      "loss": 0.4856,
+      "step": 158
+    },
+    {
+      "epoch": 0.343042071197411,
+      "grad_norm": 0.3286860170088506,
+      "learning_rate": 4.924e-05,
+      "loss": 0.4798,
+      "step": 159
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 0.30390765140489734,
+      "learning_rate": 4.92e-05,
+      "loss": 0.4841,
+      "step": 160
+    },
+    {
+      "epoch": 0.34735706580366776,
+      "grad_norm": 0.3599533471110528,
+      "learning_rate": 4.9160000000000004e-05,
+      "loss": 0.4798,
+      "step": 161
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 0.27961730157871734,
+      "learning_rate": 4.9120000000000004e-05,
+      "loss": 0.4622,
+      "step": 162
+    },
+    {
+      "epoch": 0.3516720604099245,
+      "grad_norm": 0.36326297857194084,
+      "learning_rate": 4.9080000000000004e-05,
+      "loss": 0.4941,
+      "step": 163
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 0.25877643675232387,
+      "learning_rate": 4.9040000000000005e-05,
+      "loss": 0.4916,
+      "step": 164
+    },
+    {
+      "epoch": 0.3559870550161812,
+      "grad_norm": 0.3556594334331794,
+      "learning_rate": 4.9e-05,
+      "loss": 0.4642,
+      "step": 165
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 0.32515033894088813,
+      "learning_rate": 4.896e-05,
+      "loss": 0.4864,
+      "step": 166
+    },
+    {
+      "epoch": 0.36030204962243795,
+      "grad_norm": 0.3576526726949483,
+      "learning_rate": 4.8920000000000006e-05,
+      "loss": 0.4737,
+      "step": 167
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 0.29454313757538503,
+      "learning_rate": 4.8880000000000006e-05,
+      "loss": 0.4579,
+      "step": 168
+    },
+    {
+      "epoch": 0.36461704422869473,
+      "grad_norm": 0.35162010767939217,
+      "learning_rate": 4.884e-05,
+      "loss": 0.4985,
+      "step": 169
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 0.28512732525892354,
+      "learning_rate": 4.88e-05,
+      "loss": 0.4902,
+      "step": 170
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 0.29759943236707304,
+      "learning_rate": 4.876e-05,
+      "loss": 0.4972,
+      "step": 171
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 0.3287877116113184,
+      "learning_rate": 4.872000000000001e-05,
+      "loss": 0.4908,
+      "step": 172
+    },
+    {
+      "epoch": 0.3732470334412082,
+      "grad_norm": 0.3133996309144812,
+      "learning_rate": 4.868e-05,
+      "loss": 0.4937,
+      "step": 173
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 0.3064667877265841,
+      "learning_rate": 4.864e-05,
+      "loss": 0.4864,
+      "step": 174
+    },
+    {
+      "epoch": 0.3775620280474649,
+      "grad_norm": 0.33485639441738874,
+      "learning_rate": 4.86e-05,
+      "loss": 0.4775,
+      "step": 175
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 0.305407925816577,
+      "learning_rate": 4.856e-05,
+      "loss": 0.4764,
+      "step": 176
+    },
+    {
+      "epoch": 0.3818770226537217,
+      "grad_norm": 0.3657289190953192,
+      "learning_rate": 4.852e-05,
+      "loss": 0.4658,
+      "step": 177
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 0.32585818474400624,
+      "learning_rate": 4.8480000000000003e-05,
+      "loss": 0.4759,
+      "step": 178
+    },
+    {
+      "epoch": 0.3861920172599784,
+      "grad_norm": 0.37557882227532163,
+      "learning_rate": 4.8440000000000004e-05,
+      "loss": 0.4747,
+      "step": 179
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.320749191525247,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 0.4821,
+      "step": 180
+    },
+    {
+      "epoch": 0.39050701186623515,
+      "grad_norm": 0.31740159569446824,
+      "learning_rate": 4.836e-05,
+      "loss": 0.4819,
+      "step": 181
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 0.34719515229206804,
+      "learning_rate": 4.8320000000000005e-05,
+      "loss": 0.4809,
+      "step": 182
+    },
+    {
+      "epoch": 0.3948220064724919,
+      "grad_norm": 0.29600056574488137,
+      "learning_rate": 4.8280000000000005e-05,
+      "loss": 0.473,
+      "step": 183
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 0.31700707281900425,
+      "learning_rate": 4.824e-05,
+      "loss": 0.4764,
+      "step": 184
+    },
+    {
+      "epoch": 0.39913700107874867,
+      "grad_norm": 0.2843323546229233,
+      "learning_rate": 4.82e-05,
+      "loss": 0.4706,
+      "step": 185
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 0.33405105406323765,
+      "learning_rate": 4.816e-05,
+      "loss": 0.4659,
+      "step": 186
+    },
+    {
+      "epoch": 0.4034519956850054,
+      "grad_norm": 0.3450445779367956,
+      "learning_rate": 4.812000000000001e-05,
+      "loss": 0.4848,
+      "step": 187
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 0.29904111020419,
+      "learning_rate": 4.808e-05,
+      "loss": 0.4692,
+      "step": 188
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 0.30156350801095055,
+      "learning_rate": 4.804e-05,
+      "loss": 0.4928,
+      "step": 189
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 0.29440492223089526,
+      "learning_rate": 4.8e-05,
+      "loss": 0.4714,
+      "step": 190
+    },
+    {
+      "epoch": 0.4120819848975189,
+      "grad_norm": 0.2941734869319751,
+      "learning_rate": 4.796e-05,
+      "loss": 0.4865,
+      "step": 191
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 0.2831205819699076,
+      "learning_rate": 4.792e-05,
+      "loss": 0.4923,
+      "step": 192
+    },
+    {
+      "epoch": 0.41639697950377563,
+      "grad_norm": 0.30976889318863166,
+      "learning_rate": 4.788e-05,
+      "loss": 0.4752,
+      "step": 193
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 0.2765632011545738,
+      "learning_rate": 4.784e-05,
+      "loss": 0.4501,
+      "step": 194
+    },
+    {
+      "epoch": 0.42071197411003236,
+      "grad_norm": 0.2955577997787949,
+      "learning_rate": 4.78e-05,
+      "loss": 0.4641,
+      "step": 195
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 0.35418034783340757,
+      "learning_rate": 4.7760000000000004e-05,
+      "loss": 0.4748,
+      "step": 196
+    },
+    {
+      "epoch": 0.4250269687162891,
+      "grad_norm": 0.290860398121857,
+      "learning_rate": 4.7720000000000004e-05,
+      "loss": 0.4666,
+      "step": 197
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 0.3202541928666888,
+      "learning_rate": 4.7680000000000004e-05,
+      "loss": 0.4812,
+      "step": 198
+    },
+    {
+      "epoch": 0.42934196332254587,
+      "grad_norm": 0.291794165366607,
+      "learning_rate": 4.7640000000000005e-05,
+      "loss": 0.4688,
+      "step": 199
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 0.32523032478470537,
+      "learning_rate": 4.76e-05,
+      "loss": 0.4604,
+      "step": 200
+    },
+    {
+      "epoch": 0.4336569579288026,
+      "grad_norm": 0.2661859076083175,
+      "learning_rate": 4.7560000000000005e-05,
+      "loss": 0.4693,
+      "step": 201
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 0.3403835298187506,
+      "learning_rate": 4.7520000000000006e-05,
+      "loss": 0.4651,
+      "step": 202
+    },
+    {
+      "epoch": 0.43797195253505933,
+      "grad_norm": 0.28598439991323976,
+      "learning_rate": 4.748e-05,
+      "loss": 0.4907,
+      "step": 203
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 0.383195732436593,
+      "learning_rate": 4.744e-05,
+      "loss": 0.4983,
+      "step": 204
+    },
+    {
+      "epoch": 0.44228694714131606,
+      "grad_norm": 0.27458653574722164,
+      "learning_rate": 4.74e-05,
+      "loss": 0.4698,
+      "step": 205
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.4194293652615916,
+      "learning_rate": 4.736000000000001e-05,
+      "loss": 0.4787,
+      "step": 206
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 0.29331891339359434,
+      "learning_rate": 4.732e-05,
+      "loss": 0.4743,
+      "step": 207
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 0.34631805619114103,
+      "learning_rate": 4.728e-05,
+      "loss": 0.4723,
+      "step": 208
+    },
+    {
+      "epoch": 0.45091693635382957,
+      "grad_norm": 0.2773817416509101,
+      "learning_rate": 4.724e-05,
+      "loss": 0.4568,
+      "step": 209
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 0.31793720505095835,
+      "learning_rate": 4.72e-05,
+      "loss": 0.4563,
+      "step": 210
+    },
+    {
+      "epoch": 0.4552319309600863,
+      "grad_norm": 0.3360952018884764,
+      "learning_rate": 4.716e-05,
+      "loss": 0.4758,
+      "step": 211
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 0.3238545895389713,
+      "learning_rate": 4.712e-05,
+      "loss": 0.4802,
+      "step": 212
+    },
+    {
+      "epoch": 0.459546925566343,
+      "grad_norm": 0.3299565529698189,
+      "learning_rate": 4.708e-05,
+      "loss": 0.4713,
+      "step": 213
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 0.3014892248399335,
+      "learning_rate": 4.7040000000000004e-05,
+      "loss": 0.4689,
+      "step": 214
+    },
+    {
+      "epoch": 0.4638619201725998,
+      "grad_norm": 0.36865008887233414,
+      "learning_rate": 4.7e-05,
+      "loss": 0.4845,
+      "step": 215
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 0.3010039204398615,
+      "learning_rate": 4.6960000000000004e-05,
+      "loss": 0.4717,
+      "step": 216
+    },
+    {
+      "epoch": 0.46817691477885653,
+      "grad_norm": 0.32021272907517795,
+      "learning_rate": 4.6920000000000005e-05,
+      "loss": 0.4635,
+      "step": 217
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 0.27683773975403986,
+      "learning_rate": 4.688e-05,
+      "loss": 0.4629,
+      "step": 218
+    },
+    {
+      "epoch": 0.47249190938511326,
+      "grad_norm": 0.2939694818374937,
+      "learning_rate": 4.684e-05,
+      "loss": 0.4616,
+      "step": 219
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 0.3146725242194594,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 0.4722,
+      "step": 220
+    },
+    {
+      "epoch": 0.47680690399137,
+      "grad_norm": 0.2764797117217664,
+      "learning_rate": 4.6760000000000006e-05,
+      "loss": 0.4673,
+      "step": 221
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 0.3043641853589388,
+      "learning_rate": 4.672e-05,
+      "loss": 0.4478,
+      "step": 222
+    },
+    {
+      "epoch": 0.4811218985976268,
+      "grad_norm": 0.28775200628381,
+      "learning_rate": 4.668e-05,
+      "loss": 0.4662,
+      "step": 223
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 0.27023475049837015,
+      "learning_rate": 4.664e-05,
+      "loss": 0.4661,
+      "step": 224
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.31669406491236995,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 0.4458,
+      "step": 225
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 0.2837164913050278,
+      "learning_rate": 4.656e-05,
+      "loss": 0.4684,
+      "step": 226
+    },
+    {
+      "epoch": 0.48975188781014023,
+      "grad_norm": 0.29868610479918123,
+      "learning_rate": 4.652e-05,
+      "loss": 0.4506,
+      "step": 227
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 0.2949287420608189,
+      "learning_rate": 4.648e-05,
+      "loss": 0.4632,
+      "step": 228
+    },
+    {
+      "epoch": 0.49406688241639696,
+      "grad_norm": 0.3119200088855303,
+      "learning_rate": 4.644e-05,
+      "loss": 0.4551,
+      "step": 229
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 0.290598992319338,
+      "learning_rate": 4.64e-05,
+      "loss": 0.458,
+      "step": 230
+    },
+    {
+      "epoch": 0.49838187702265374,
+      "grad_norm": 0.34483322575005054,
+      "learning_rate": 4.636e-05,
+      "loss": 0.4637,
+      "step": 231
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 0.27712821917993735,
+      "learning_rate": 4.6320000000000004e-05,
+      "loss": 0.4778,
+      "step": 232
+    },
+    {
+      "epoch": 0.5026968716289104,
+      "grad_norm": 0.28690313202566,
+      "learning_rate": 4.6280000000000004e-05,
+      "loss": 0.4718,
+      "step": 233
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 0.29881025273811784,
+      "learning_rate": 4.624e-05,
+      "loss": 0.4548,
+      "step": 234
+    },
+    {
+      "epoch": 0.5070118662351673,
+      "grad_norm": 0.3009843183384041,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 0.4734,
+      "step": 235
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 0.3216848912197773,
+      "learning_rate": 4.6160000000000005e-05,
+      "loss": 0.453,
+      "step": 236
+    },
+    {
+      "epoch": 0.511326860841424,
+      "grad_norm": 0.2622840122570054,
+      "learning_rate": 4.612e-05,
+      "loss": 0.4534,
+      "step": 237
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 0.31268030831842797,
+      "learning_rate": 4.608e-05,
+      "loss": 0.4541,
+      "step": 238
+    },
+    {
+      "epoch": 0.5156418554476807,
+      "grad_norm": 0.2655099701679723,
+      "learning_rate": 4.604e-05,
+      "loss": 0.4509,
+      "step": 239
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 0.27681618093808624,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.4765,
+      "step": 240
+    },
+    {
+      "epoch": 0.5199568500539374,
+      "grad_norm": 0.31086122245620157,
+      "learning_rate": 4.596e-05,
+      "loss": 0.4542,
+      "step": 241
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 0.26187143929707984,
+      "learning_rate": 4.592e-05,
+      "loss": 0.4548,
+      "step": 242
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.30612908409709966,
+      "learning_rate": 4.588e-05,
+      "loss": 0.4644,
+      "step": 243
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 0.28353354277654075,
+      "learning_rate": 4.584e-05,
+      "loss": 0.4777,
+      "step": 244
+    },
+    {
+      "epoch": 0.5285868392664509,
+      "grad_norm": 0.2707543786610627,
+      "learning_rate": 4.58e-05,
+      "loss": 0.4605,
+      "step": 245
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 0.25826035874506936,
+      "learning_rate": 4.576e-05,
+      "loss": 0.4733,
+      "step": 246
+    },
+    {
+      "epoch": 0.5329018338727076,
+      "grad_norm": 0.26731071019406205,
+      "learning_rate": 4.572e-05,
+      "loss": 0.4562,
+      "step": 247
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 0.23862376017091427,
+      "learning_rate": 4.568e-05,
+      "loss": 0.4481,
+      "step": 248
+    },
+    {
+      "epoch": 0.5372168284789643,
+      "grad_norm": 0.3832009951376676,
+      "learning_rate": 4.564e-05,
+      "loss": 0.4628,
+      "step": 249
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 0.29500842123610954,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 0.459,
+      "step": 250
+    },
+    {
+      "epoch": 0.5415318230852212,
+      "grad_norm": 0.2965682041233286,
+      "learning_rate": 4.5560000000000004e-05,
+      "loss": 0.4621,
+      "step": 251
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 0.28726731902710617,
+      "learning_rate": 4.5520000000000005e-05,
+      "loss": 0.463,
+      "step": 252
+    },
+    {
+      "epoch": 0.5458468176914779,
+      "grad_norm": 0.2690311478101724,
+      "learning_rate": 4.548e-05,
+      "loss": 0.4707,
+      "step": 253
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 0.30474236349865746,
+      "learning_rate": 4.5440000000000005e-05,
+      "loss": 0.4691,
+      "step": 254
+    },
+    {
+      "epoch": 0.5501618122977346,
+      "grad_norm": 0.24352372447443482,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 0.4413,
+      "step": 255
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 0.3042686272085747,
+      "learning_rate": 4.536e-05,
+      "loss": 0.4796,
+      "step": 256
+    },
+    {
+      "epoch": 0.5544768069039914,
+      "grad_norm": 0.24174468708449245,
+      "learning_rate": 4.532e-05,
+      "loss": 0.4436,
+      "step": 257
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 0.26229288392443767,
+      "learning_rate": 4.528e-05,
+      "loss": 0.4474,
+      "step": 258
+    },
+    {
+      "epoch": 0.5587918015102481,
+      "grad_norm": 0.3390509699538324,
+      "learning_rate": 4.524000000000001e-05,
+      "loss": 0.4584,
+      "step": 259
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 0.2525311110616745,
+      "learning_rate": 4.52e-05,
+      "loss": 0.4412,
+      "step": 260
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.3434472457371927,
+      "learning_rate": 4.516e-05,
+      "loss": 0.4698,
+      "step": 261
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 0.25391082344689797,
+      "learning_rate": 4.512e-05,
+      "loss": 0.4759,
+      "step": 262
+    },
+    {
+      "epoch": 0.5674217907227616,
+      "grad_norm": 0.34249204432261243,
+      "learning_rate": 4.508e-05,
+      "loss": 0.4552,
+      "step": 263
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 0.2472862544884362,
+      "learning_rate": 4.504e-05,
+      "loss": 0.4688,
+      "step": 264
+    },
+    {
+      "epoch": 0.5717367853290184,
+      "grad_norm": 0.27264729971113316,
+      "learning_rate": 4.5e-05,
+      "loss": 0.4488,
+      "step": 265
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 0.2977417387577173,
+      "learning_rate": 4.496e-05,
+      "loss": 0.4752,
+      "step": 266
+    },
+    {
+      "epoch": 0.5760517799352751,
+      "grad_norm": 0.2763251041657974,
+      "learning_rate": 4.4920000000000004e-05,
+      "loss": 0.4633,
+      "step": 267
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 0.29864566803635684,
+      "learning_rate": 4.488e-05,
+      "loss": 0.4877,
+      "step": 268
+    },
+    {
+      "epoch": 0.5803667745415318,
+      "grad_norm": 0.2817997182658353,
+      "learning_rate": 4.4840000000000004e-05,
+      "loss": 0.4599,
+      "step": 269
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.2691208231765093,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 0.4446,
+      "step": 270
+    },
+    {
+      "epoch": 0.5846817691477886,
+      "grad_norm": 0.26100659163381174,
+      "learning_rate": 4.4760000000000005e-05,
+      "loss": 0.4702,
+      "step": 271
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 0.30494432016510414,
+      "learning_rate": 4.472e-05,
+      "loss": 0.4434,
+      "step": 272
+    },
+    {
+      "epoch": 0.5889967637540453,
+      "grad_norm": 0.3217381889319848,
+      "learning_rate": 4.468e-05,
+      "loss": 0.4522,
+      "step": 273
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 0.2747807777821844,
+      "learning_rate": 4.4640000000000006e-05,
+      "loss": 0.4642,
+      "step": 274
+    },
+    {
+      "epoch": 0.593311758360302,
+      "grad_norm": 0.2771512136815011,
+      "learning_rate": 4.46e-05,
+      "loss": 0.458,
+      "step": 275
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 0.30667352637556383,
+      "learning_rate": 4.456e-05,
+      "loss": 0.4653,
+      "step": 276
+    },
+    {
+      "epoch": 0.5976267529665588,
+      "grad_norm": 0.27054508489934065,
+      "learning_rate": 4.452e-05,
+      "loss": 0.4457,
+      "step": 277
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 0.32885060108368724,
+      "learning_rate": 4.448e-05,
+      "loss": 0.4479,
+      "step": 278
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.25861633576048176,
+      "learning_rate": 4.444e-05,
+      "loss": 0.443,
+      "step": 279
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 0.3264488574979686,
+      "learning_rate": 4.44e-05,
+      "loss": 0.4546,
+      "step": 280
+    },
+    {
+      "epoch": 0.6062567421790723,
+      "grad_norm": 0.2661396893245308,
+      "learning_rate": 4.436e-05,
+      "loss": 0.4483,
+      "step": 281
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 0.2666725233168853,
+      "learning_rate": 4.432e-05,
+      "loss": 0.4455,
+      "step": 282
+    },
+    {
+      "epoch": 0.6105717367853291,
+      "grad_norm": 0.3213878317441801,
+      "learning_rate": 4.428e-05,
+      "loss": 0.4561,
+      "step": 283
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 0.27514627181060536,
+      "learning_rate": 4.424e-05,
+      "loss": 0.4533,
+      "step": 284
+    },
+    {
+      "epoch": 0.6148867313915858,
+      "grad_norm": 0.28529847199820635,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 0.4583,
+      "step": 285
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 0.3084015687846991,
+      "learning_rate": 4.4160000000000004e-05,
+      "loss": 0.4499,
+      "step": 286
+    },
+    {
+      "epoch": 0.6192017259978425,
+      "grad_norm": 0.2722920383928748,
+      "learning_rate": 4.412e-05,
+      "loss": 0.4699,
+      "step": 287
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 0.2790301022639745,
+      "learning_rate": 4.4080000000000005e-05,
+      "loss": 0.4571,
+      "step": 288
+    },
+    {
+      "epoch": 0.6235167206040992,
+      "grad_norm": 0.28022715939780435,
+      "learning_rate": 4.4040000000000005e-05,
+      "loss": 0.4486,
+      "step": 289
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 0.25791286221623794,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.4383,
+      "step": 290
+    },
+    {
+      "epoch": 0.627831715210356,
+      "grad_norm": 0.30038570760100414,
+      "learning_rate": 4.396e-05,
+      "loss": 0.4619,
+      "step": 291
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 0.29876170957810305,
+      "learning_rate": 4.392e-05,
+      "loss": 0.4707,
+      "step": 292
+    },
+    {
+      "epoch": 0.6321467098166127,
+      "grad_norm": 0.2932038829420852,
+      "learning_rate": 4.388000000000001e-05,
+      "loss": 0.4557,
+      "step": 293
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 0.2716815847789961,
+      "learning_rate": 4.384e-05,
+      "loss": 0.4361,
+      "step": 294
+    },
+    {
+      "epoch": 0.6364617044228694,
+      "grad_norm": 0.3099793080621506,
+      "learning_rate": 4.38e-05,
+      "loss": 0.4577,
+      "step": 295
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 0.2938986793353107,
+      "learning_rate": 4.376e-05,
+      "loss": 0.4476,
+      "step": 296
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.29035829007503894,
+      "learning_rate": 4.372e-05,
+      "loss": 0.4423,
+      "step": 297
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 0.2694452227101497,
+      "learning_rate": 4.368e-05,
+      "loss": 0.4545,
+      "step": 298
+    },
+    {
+      "epoch": 0.645091693635383,
+      "grad_norm": 0.28851646597054104,
+      "learning_rate": 4.364e-05,
+      "loss": 0.4431,
+      "step": 299
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 0.2593151761184194,
+      "learning_rate": 4.36e-05,
+      "loss": 0.455,
+      "step": 300
+    },
+    {
+      "epoch": 0.6494066882416397,
+      "grad_norm": 0.29309975049936027,
+      "learning_rate": 4.356e-05,
+      "loss": 0.4481,
+      "step": 301
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 0.25398889340605957,
+      "learning_rate": 4.352e-05,
+      "loss": 0.4476,
+      "step": 302
+    },
+    {
+      "epoch": 0.6537216828478964,
+      "grad_norm": 0.30940788747918896,
+      "learning_rate": 4.3480000000000004e-05,
+      "loss": 0.45,
+      "step": 303
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 0.2820801140095731,
+      "learning_rate": 4.3440000000000004e-05,
+      "loss": 0.4544,
+      "step": 304
+    },
+    {
+      "epoch": 0.6580366774541532,
+      "grad_norm": 0.3152437853102846,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 0.4492,
+      "step": 305
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.2749853135008326,
+      "learning_rate": 4.336e-05,
+      "loss": 0.4412,
+      "step": 306
+    },
+    {
+      "epoch": 0.6623516720604099,
+      "grad_norm": 0.3079252422579841,
+      "learning_rate": 4.332e-05,
+      "loss": 0.4369,
+      "step": 307
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 0.2946391275689202,
+      "learning_rate": 4.3280000000000006e-05,
+      "loss": 0.4573,
+      "step": 308
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.2668274385038352,
+      "learning_rate": 4.324e-05,
+      "loss": 0.4359,
+      "step": 309
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 0.28838034268652896,
+      "learning_rate": 4.32e-05,
+      "loss": 0.4508,
+      "step": 310
+    },
+    {
+      "epoch": 0.6709816612729234,
+      "grad_norm": 0.2834904014905993,
+      "learning_rate": 4.316e-05,
+      "loss": 0.4522,
+      "step": 311
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 0.27040013725015194,
+      "learning_rate": 4.312000000000001e-05,
+      "loss": 0.4456,
+      "step": 312
+    },
+    {
+      "epoch": 0.6752966558791802,
+      "grad_norm": 0.31722009198474627,
+      "learning_rate": 4.308e-05,
+      "loss": 0.4659,
+      "step": 313
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 0.27088552438961755,
+      "learning_rate": 4.304e-05,
+      "loss": 0.431,
+      "step": 314
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.26632771069672984,
+      "learning_rate": 4.3e-05,
+      "loss": 0.4501,
+      "step": 315
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 0.2972755674501818,
+      "learning_rate": 4.296e-05,
+      "loss": 0.4663,
+      "step": 316
+    },
+    {
+      "epoch": 0.6839266450916937,
+      "grad_norm": 0.25853931040596184,
+      "learning_rate": 4.292e-05,
+      "loss": 0.4572,
+      "step": 317
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 0.2714876932330396,
+      "learning_rate": 4.288e-05,
+      "loss": 0.4622,
+      "step": 318
+    },
+    {
+      "epoch": 0.6882416396979504,
+      "grad_norm": 0.2642767531993865,
+      "learning_rate": 4.284e-05,
+      "loss": 0.4714,
+      "step": 319
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 0.25954856405102256,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 0.4568,
+      "step": 320
+    },
+    {
+      "epoch": 0.6925566343042071,
+      "grad_norm": 0.3044540062878058,
+      "learning_rate": 4.276e-05,
+      "loss": 0.4415,
+      "step": 321
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 0.23066914239975445,
+      "learning_rate": 4.2720000000000004e-05,
+      "loss": 0.4638,
+      "step": 322
+    },
+    {
+      "epoch": 0.6968716289104638,
+      "grad_norm": 0.35554840668777676,
+      "learning_rate": 4.2680000000000005e-05,
+      "loss": 0.4573,
+      "step": 323
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.2913915204408648,
+      "learning_rate": 4.2640000000000005e-05,
+      "loss": 0.4473,
+      "step": 324
+    },
+    {
+      "epoch": 0.7011866235167206,
+      "grad_norm": 0.2684942718217648,
+      "learning_rate": 4.26e-05,
+      "loss": 0.4339,
+      "step": 325
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 0.3216708484980791,
+      "learning_rate": 4.256e-05,
+      "loss": 0.4468,
+      "step": 326
+    },
+    {
+      "epoch": 0.7055016181229773,
+      "grad_norm": 0.2756335067641377,
+      "learning_rate": 4.2520000000000006e-05,
+      "loss": 0.4558,
+      "step": 327
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 0.2764643077888995,
+      "learning_rate": 4.248e-05,
+      "loss": 0.4346,
+      "step": 328
+    },
+    {
+      "epoch": 0.7098166127292341,
+      "grad_norm": 0.32603235256791835,
+      "learning_rate": 4.244e-05,
+      "loss": 0.4559,
+      "step": 329
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 0.29424638904551764,
+      "learning_rate": 4.24e-05,
+      "loss": 0.4462,
+      "step": 330
+    },
+    {
+      "epoch": 0.7141316073354909,
+      "grad_norm": 0.3001554414730066,
+      "learning_rate": 4.236e-05,
+      "loss": 0.4468,
+      "step": 331
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 0.3129011037209938,
+      "learning_rate": 4.232e-05,
+      "loss": 0.4108,
+      "step": 332
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.28762494926699644,
+      "learning_rate": 4.228e-05,
+      "loss": 0.4594,
+      "step": 333
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 0.3287086931234326,
+      "learning_rate": 4.224e-05,
+      "loss": 0.4546,
+      "step": 334
+    },
+    {
+      "epoch": 0.7227615965480043,
+      "grad_norm": 0.2440171509135856,
+      "learning_rate": 4.22e-05,
+      "loss": 0.4395,
+      "step": 335
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 0.41734955962307646,
+      "learning_rate": 4.2159999999999996e-05,
+      "loss": 0.4562,
+      "step": 336
+    },
+    {
+      "epoch": 0.727076591154261,
+      "grad_norm": 0.23647128778868323,
+      "learning_rate": 4.212e-05,
+      "loss": 0.448,
+      "step": 337
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 0.3478141358669819,
+      "learning_rate": 4.2080000000000004e-05,
+      "loss": 0.4259,
+      "step": 338
+    },
+    {
+      "epoch": 0.7313915857605178,
+      "grad_norm": 0.22537690567688576,
+      "learning_rate": 4.2040000000000004e-05,
+      "loss": 0.4254,
+      "step": 339
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 0.32770466753425076,
+      "learning_rate": 4.2e-05,
+      "loss": 0.4493,
+      "step": 340
+    },
+    {
+      "epoch": 0.7357065803667745,
+      "grad_norm": 0.2348262809416127,
+      "learning_rate": 4.196e-05,
+      "loss": 0.4379,
+      "step": 341
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.2784106337499888,
+      "learning_rate": 4.1920000000000005e-05,
+      "loss": 0.4512,
+      "step": 342
+    },
+    {
+      "epoch": 0.7400215749730313,
+      "grad_norm": 0.24611169037260724,
+      "learning_rate": 4.1880000000000006e-05,
+      "loss": 0.4372,
+      "step": 343
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 0.28368805770265704,
+      "learning_rate": 4.184e-05,
+      "loss": 0.4641,
+      "step": 344
+    },
+    {
+      "epoch": 0.7443365695792881,
+      "grad_norm": 0.2538772376229055,
+      "learning_rate": 4.18e-05,
+      "loss": 0.446,
+      "step": 345
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 0.25580425464723444,
+      "learning_rate": 4.176000000000001e-05,
+      "loss": 0.4417,
+      "step": 346
+    },
+    {
+      "epoch": 0.7486515641855448,
+      "grad_norm": 0.27469783764584166,
+      "learning_rate": 4.172e-05,
+      "loss": 0.4611,
+      "step": 347
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 0.29130083464472767,
+      "learning_rate": 4.168e-05,
+      "loss": 0.444,
+      "step": 348
+    },
+    {
+      "epoch": 0.7529665587918015,
+      "grad_norm": 0.2651244371602352,
+      "learning_rate": 4.164e-05,
+      "loss": 0.4441,
+      "step": 349
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 0.26234665200685936,
+      "learning_rate": 4.16e-05,
+      "loss": 0.4461,
+      "step": 350
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 0.24125048647992992,
+      "learning_rate": 4.156e-05,
+      "loss": 0.4352,
+      "step": 351
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 0.2508832475792393,
+      "learning_rate": 4.152e-05,
+      "loss": 0.4463,
+      "step": 352
+    },
+    {
+      "epoch": 0.761596548004315,
+      "grad_norm": 0.26155428589274304,
+      "learning_rate": 4.148e-05,
+      "loss": 0.4369,
+      "step": 353
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 0.24472099152743595,
+      "learning_rate": 4.144e-05,
+      "loss": 0.4345,
+      "step": 354
+    },
+    {
+      "epoch": 0.7659115426105717,
+      "grad_norm": 0.24715070483697896,
+      "learning_rate": 4.14e-05,
+      "loss": 0.4505,
+      "step": 355
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 0.24298690156207697,
+      "learning_rate": 4.1360000000000004e-05,
+      "loss": 0.4422,
+      "step": 356
+    },
+    {
+      "epoch": 0.7702265372168284,
+      "grad_norm": 0.2349365195764117,
+      "learning_rate": 4.1320000000000004e-05,
+      "loss": 0.4526,
+      "step": 357
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 0.2331945559857268,
+      "learning_rate": 4.1280000000000005e-05,
+      "loss": 0.4383,
+      "step": 358
+    },
+    {
+      "epoch": 0.7745415318230853,
+      "grad_norm": 0.2595543035077313,
+      "learning_rate": 4.124e-05,
+      "loss": 0.4308,
+      "step": 359
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.21807377534792635,
+      "learning_rate": 4.12e-05,
+      "loss": 0.4349,
+      "step": 360
+    },
+    {
+      "epoch": 0.778856526429342,
+      "grad_norm": 0.28950608296716684,
+      "learning_rate": 4.1160000000000006e-05,
+      "loss": 0.4558,
+      "step": 361
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 0.2611339581514693,
+      "learning_rate": 4.1120000000000006e-05,
+      "loss": 0.4599,
+      "step": 362
+    },
+    {
+      "epoch": 0.7831715210355987,
+      "grad_norm": 0.25575793173451006,
+      "learning_rate": 4.108e-05,
+      "loss": 0.4406,
+      "step": 363
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 0.2450742657345304,
+      "learning_rate": 4.104e-05,
+      "loss": 0.4453,
+      "step": 364
+    },
+    {
+      "epoch": 0.7874865156418555,
+      "grad_norm": 0.28382377756829125,
+      "learning_rate": 4.1e-05,
+      "loss": 0.465,
+      "step": 365
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 0.2392683438388647,
+      "learning_rate": 4.096e-05,
+      "loss": 0.4416,
+      "step": 366
+    },
+    {
+      "epoch": 0.7918015102481122,
+      "grad_norm": 0.2539921883515288,
+      "learning_rate": 4.092e-05,
+      "loss": 0.4497,
+      "step": 367
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 0.27579412883884935,
+      "learning_rate": 4.088e-05,
+      "loss": 0.4464,
+      "step": 368
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.28727621982712537,
+      "learning_rate": 4.084e-05,
+      "loss": 0.4411,
+      "step": 369
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 0.2519428850122542,
+      "learning_rate": 4.08e-05,
+      "loss": 0.4479,
+      "step": 370
+    },
+    {
+      "epoch": 0.8004314994606256,
+      "grad_norm": 0.2674974847910012,
+      "learning_rate": 4.076e-05,
+      "loss": 0.44,
+      "step": 371
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 0.25530557484483807,
+      "learning_rate": 4.072e-05,
+      "loss": 0.4394,
+      "step": 372
+    },
+    {
+      "epoch": 0.8047464940668824,
+      "grad_norm": 0.26435664799974373,
+      "learning_rate": 4.0680000000000004e-05,
+      "loss": 0.4574,
+      "step": 373
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 0.2530022819289299,
+      "learning_rate": 4.064e-05,
+      "loss": 0.4454,
+      "step": 374
+    },
+    {
+      "epoch": 0.8090614886731392,
+      "grad_norm": 0.24054892754428694,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 0.4398,
+      "step": 375
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 0.2848289155610854,
+      "learning_rate": 4.0560000000000005e-05,
+      "loss": 0.4398,
+      "step": 376
+    },
+    {
+      "epoch": 0.8133764832793959,
+      "grad_norm": 0.2575053811702797,
+      "learning_rate": 4.0520000000000005e-05,
+      "loss": 0.4578,
+      "step": 377
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.27019672115837984,
+      "learning_rate": 4.048e-05,
+      "loss": 0.4323,
+      "step": 378
+    },
+    {
+      "epoch": 0.8176914778856527,
+      "grad_norm": 0.2398139963913645,
+      "learning_rate": 4.044e-05,
+      "loss": 0.4448,
+      "step": 379
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 0.2529963842405476,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 0.4296,
+      "step": 380
+    },
+    {
+      "epoch": 0.8220064724919094,
+      "grad_norm": 0.2511947818922217,
+      "learning_rate": 4.0360000000000007e-05,
+      "loss": 0.4396,
+      "step": 381
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 0.2584407655583235,
+      "learning_rate": 4.032e-05,
+      "loss": 0.4406,
+      "step": 382
+    },
+    {
+      "epoch": 0.8263214670981661,
+      "grad_norm": 0.2456123715798027,
+      "learning_rate": 4.028e-05,
+      "loss": 0.4677,
+      "step": 383
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 0.2701085902021937,
+      "learning_rate": 4.024e-05,
+      "loss": 0.466,
+      "step": 384
+    },
+    {
+      "epoch": 0.8306364617044228,
+      "grad_norm": 0.23978562401955322,
+      "learning_rate": 4.02e-05,
+      "loss": 0.4386,
+      "step": 385
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 0.2645169304521651,
+      "learning_rate": 4.016e-05,
+      "loss": 0.4502,
+      "step": 386
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.27897793156631207,
+      "learning_rate": 4.012e-05,
+      "loss": 0.4684,
+      "step": 387
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 0.2478443685184336,
+      "learning_rate": 4.008e-05,
+      "loss": 0.4376,
+      "step": 388
+    },
+    {
+      "epoch": 0.8392664509169363,
+      "grad_norm": 0.256589453456646,
+      "learning_rate": 4.004e-05,
+      "loss": 0.431,
+      "step": 389
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 0.28233657875363566,
+      "learning_rate": 4e-05,
+      "loss": 0.4315,
+      "step": 390
+    },
+    {
+      "epoch": 0.8435814455231931,
+      "grad_norm": 0.27009324457097755,
+      "learning_rate": 3.9960000000000004e-05,
+      "loss": 0.4481,
+      "step": 391
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 0.33499329840024655,
+      "learning_rate": 3.9920000000000004e-05,
+      "loss": 0.4471,
+      "step": 392
+    },
+    {
+      "epoch": 0.8478964401294499,
+      "grad_norm": 0.2841086526535015,
+      "learning_rate": 3.988e-05,
+      "loss": 0.4293,
+      "step": 393
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 0.2716488191030425,
+      "learning_rate": 3.984e-05,
+      "loss": 0.4397,
+      "step": 394
+    },
+    {
+      "epoch": 0.8522114347357066,
+      "grad_norm": 0.26620067701224137,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.4461,
+      "step": 395
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.2576197026648712,
+      "learning_rate": 3.9760000000000006e-05,
+      "loss": 0.4298,
+      "step": 396
+    },
+    {
+      "epoch": 0.8565264293419633,
+      "grad_norm": 0.2532441081675246,
+      "learning_rate": 3.972e-05,
+      "loss": 0.4383,
+      "step": 397
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 0.2679762403275455,
+      "learning_rate": 3.968e-05,
+      "loss": 0.4395,
+      "step": 398
+    },
+    {
+      "epoch": 0.86084142394822,
+      "grad_norm": 0.24275874935058916,
+      "learning_rate": 3.964e-05,
+      "loss": 0.4357,
+      "step": 399
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 0.24961166385653455,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.4383,
+      "step": 400
+    },
+    {
+      "epoch": 0.8651564185544768,
+      "grad_norm": 0.24374670649085867,
+      "learning_rate": 3.956e-05,
+      "loss": 0.4287,
+      "step": 401
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 0.23871690671281046,
+      "learning_rate": 3.952e-05,
+      "loss": 0.425,
+      "step": 402
+    },
+    {
+      "epoch": 0.8694714131607335,
+      "grad_norm": 0.3032682890616598,
+      "learning_rate": 3.948e-05,
+      "loss": 0.4537,
+      "step": 403
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 0.25117617642424706,
+      "learning_rate": 3.944e-05,
+      "loss": 0.4282,
+      "step": 404
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.22541166475315713,
+      "learning_rate": 3.94e-05,
+      "loss": 0.436,
+      "step": 405
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 0.2716408337060906,
+      "learning_rate": 3.936e-05,
+      "loss": 0.4295,
+      "step": 406
+    },
+    {
+      "epoch": 0.8781014023732471,
+      "grad_norm": 0.2556248222954383,
+      "learning_rate": 3.932e-05,
+      "loss": 0.4549,
+      "step": 407
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 0.23721706187791133,
+      "learning_rate": 3.9280000000000003e-05,
+      "loss": 0.4436,
+      "step": 408
+    },
+    {
+      "epoch": 0.8824163969795038,
+      "grad_norm": 0.24025254833177956,
+      "learning_rate": 3.9240000000000004e-05,
+      "loss": 0.4313,
+      "step": 409
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 0.2614266619623848,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.4453,
+      "step": 410
+    },
+    {
+      "epoch": 0.8867313915857605,
+      "grad_norm": 0.23773461747873242,
+      "learning_rate": 3.9160000000000005e-05,
+      "loss": 0.4422,
+      "step": 411
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.2555605177311676,
+      "learning_rate": 3.912e-05,
+      "loss": 0.4532,
+      "step": 412
+    },
+    {
+      "epoch": 0.8910463861920173,
+      "grad_norm": 0.21705114177765664,
+      "learning_rate": 3.908e-05,
+      "loss": 0.4325,
+      "step": 413
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.23143716203371478,
+      "learning_rate": 3.9040000000000006e-05,
+      "loss": 0.419,
+      "step": 414
+    },
+    {
+      "epoch": 0.895361380798274,
+      "grad_norm": 0.5742209849902357,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.4305,
+      "step": 415
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 0.2551103117883286,
+      "learning_rate": 3.896e-05,
+      "loss": 0.4189,
+      "step": 416
+    },
+    {
+      "epoch": 0.8996763754045307,
+      "grad_norm": 0.22402399804210352,
+      "learning_rate": 3.892e-05,
+      "loss": 0.428,
+      "step": 417
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 0.24116776940016732,
+      "learning_rate": 3.888e-05,
+      "loss": 0.4367,
+      "step": 418
+    },
+    {
+      "epoch": 0.9039913700107874,
+      "grad_norm": 0.22403592441484066,
+      "learning_rate": 3.884e-05,
+      "loss": 0.4485,
+      "step": 419
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 0.22482911614936518,
+      "learning_rate": 3.88e-05,
+      "loss": 0.457,
+      "step": 420
+    },
+    {
+      "epoch": 0.9083063646170443,
+      "grad_norm": 0.23857996790359456,
+      "learning_rate": 3.876e-05,
+      "loss": 0.4319,
+      "step": 421
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 0.22710273590374788,
+      "learning_rate": 3.872e-05,
+      "loss": 0.4375,
+      "step": 422
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 0.2259633074307126,
+      "learning_rate": 3.868e-05,
+      "loss": 0.4248,
+      "step": 423
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 0.22640228114318403,
+      "learning_rate": 3.864e-05,
+      "loss": 0.4161,
+      "step": 424
+    },
+    {
+      "epoch": 0.9169363538295577,
+      "grad_norm": 0.2350442466813375,
+      "learning_rate": 3.86e-05,
+      "loss": 0.4435,
+      "step": 425
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 0.44365613370637497,
+      "learning_rate": 3.8560000000000004e-05,
+      "loss": 0.4393,
+      "step": 426
+    },
+    {
+      "epoch": 0.9212513484358145,
+      "grad_norm": 0.2512897972825721,
+      "learning_rate": 3.8520000000000004e-05,
+      "loss": 0.4409,
+      "step": 427
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 0.24464436693649008,
+      "learning_rate": 3.848e-05,
+      "loss": 0.4324,
+      "step": 428
+    },
+    {
+      "epoch": 0.9255663430420712,
+      "grad_norm": 0.2672384976835338,
+      "learning_rate": 3.8440000000000005e-05,
+      "loss": 0.4216,
+      "step": 429
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 0.2871267008652053,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.4522,
+      "step": 430
+    },
+    {
+      "epoch": 0.9298813376483279,
+      "grad_norm": 0.24811968751313535,
+      "learning_rate": 3.836e-05,
+      "loss": 0.4339,
+      "step": 431
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 0.27159198887797176,
+      "learning_rate": 3.832e-05,
+      "loss": 0.4315,
+      "step": 432
+    },
+    {
+      "epoch": 0.9341963322545846,
+      "grad_norm": 0.26279585526150134,
+      "learning_rate": 3.828e-05,
+      "loss": 0.4335,
+      "step": 433
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 0.2463043675583301,
+      "learning_rate": 3.8240000000000007e-05,
+      "loss": 0.4426,
+      "step": 434
+    },
+    {
+      "epoch": 0.9385113268608414,
+      "grad_norm": 0.22582242130582073,
+      "learning_rate": 3.82e-05,
+      "loss": 0.4327,
+      "step": 435
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 0.22544591401205022,
+      "learning_rate": 3.816e-05,
+      "loss": 0.416,
+      "step": 436
+    },
+    {
+      "epoch": 0.9428263214670982,
+      "grad_norm": 0.25570195904870374,
+      "learning_rate": 3.812e-05,
+      "loss": 0.4413,
+      "step": 437
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 0.2165456325752943,
+      "learning_rate": 3.808e-05,
+      "loss": 0.42,
+      "step": 438
+    },
+    {
+      "epoch": 0.9471413160733549,
+      "grad_norm": 0.2660006977071334,
+      "learning_rate": 3.804e-05,
+      "loss": 0.4386,
+      "step": 439
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 0.27016187227336946,
+      "learning_rate": 3.8e-05,
+      "loss": 0.4497,
+      "step": 440
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.26490253003888736,
+      "learning_rate": 3.796e-05,
+      "loss": 0.4455,
+      "step": 441
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 0.24784119691094716,
+      "learning_rate": 3.792e-05,
+      "loss": 0.4312,
+      "step": 442
+    },
+    {
+      "epoch": 0.9557713052858684,
+      "grad_norm": 0.24577321550551764,
+      "learning_rate": 3.788e-05,
+      "loss": 0.4331,
+      "step": 443
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 0.23254334294044085,
+      "learning_rate": 3.7840000000000004e-05,
+      "loss": 0.4446,
+      "step": 444
+    },
+    {
+      "epoch": 0.9600862998921251,
+      "grad_norm": 0.24298642123366798,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.4349,
+      "step": 445
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 0.2267162236488151,
+      "learning_rate": 3.776e-05,
+      "loss": 0.4299,
+      "step": 446
+    },
+    {
+      "epoch": 0.9644012944983819,
+      "grad_norm": 0.2550173441070278,
+      "learning_rate": 3.772e-05,
+      "loss": 0.4457,
+      "step": 447
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.23923284262163064,
+      "learning_rate": 3.7680000000000005e-05,
+      "loss": 0.4228,
+      "step": 448
+    },
+    {
+      "epoch": 0.9687162891046386,
+      "grad_norm": 0.26139756709830914,
+      "learning_rate": 3.7640000000000006e-05,
+      "loss": 0.4264,
+      "step": 449
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.2504569204883789,
+      "learning_rate": 3.76e-05,
+      "loss": 0.436,
+      "step": 450
+    },
+    {
+      "epoch": 0.9730312837108953,
+      "grad_norm": 0.2638553178925634,
+      "learning_rate": 3.756e-05,
+      "loss": 0.4471,
+      "step": 451
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 0.2627847800007779,
+      "learning_rate": 3.752e-05,
+      "loss": 0.4362,
+      "step": 452
+    },
+    {
+      "epoch": 0.9773462783171522,
+      "grad_norm": 0.21271418115533056,
+      "learning_rate": 3.748000000000001e-05,
+      "loss": 0.4312,
+      "step": 453
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 0.282247303317703,
+      "learning_rate": 3.744e-05,
+      "loss": 0.4266,
+      "step": 454
+    },
+    {
+      "epoch": 0.9816612729234089,
+      "grad_norm": 0.22711280249989385,
+      "learning_rate": 3.74e-05,
+      "loss": 0.4443,
+      "step": 455
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 0.2543979274114635,
+      "learning_rate": 3.736e-05,
+      "loss": 0.4311,
+      "step": 456
+    },
+    {
+      "epoch": 0.9859762675296656,
+      "grad_norm": 0.2463425451861333,
+      "learning_rate": 3.732e-05,
+      "loss": 0.4315,
+      "step": 457
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 0.2633378786543858,
+      "learning_rate": 3.728e-05,
+      "loss": 0.4361,
+      "step": 458
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.26123423746373853,
+      "learning_rate": 3.724e-05,
+      "loss": 0.446,
+      "step": 459
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 0.2886530412464882,
+      "learning_rate": 3.72e-05,
+      "loss": 0.4471,
+      "step": 460
+    },
+    {
+      "epoch": 0.9946062567421791,
+      "grad_norm": 0.2403470838331088,
+      "learning_rate": 3.716e-05,
+      "loss": 0.4348,
+      "step": 461
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 0.3003434463680213,
+      "learning_rate": 3.712e-05,
+      "loss": 0.4434,
+      "step": 462
+    },
+    {
+      "epoch": 0.9989212513484358,
+      "grad_norm": 0.23330280858351574,
+      "learning_rate": 3.7080000000000004e-05,
+      "loss": 0.4225,
+      "step": 463
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23330280858351574,
+      "learning_rate": 3.7040000000000005e-05,
+      "loss": 0.4811,
+      "step": 464
+    },
+    {
+      "epoch": 1.0021574973031284,
+      "grad_norm": 0.4639893862561724,
+      "learning_rate": 3.7e-05,
+      "loss": 0.3817,
+      "step": 465
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 0.3321444998235641,
+      "learning_rate": 3.696e-05,
+      "loss": 0.3759,
+      "step": 466
+    },
+    {
+      "epoch": 1.006472491909385,
+      "grad_norm": 0.3177620819368746,
+      "learning_rate": 3.692e-05,
+      "loss": 0.3568,
+      "step": 467
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 0.3175305154761647,
+      "learning_rate": 3.6880000000000006e-05,
+      "loss": 0.3716,
+      "step": 468
+    },
+    {
+      "epoch": 1.0107874865156419,
+      "grad_norm": 0.3451229640661856,
+      "learning_rate": 3.684e-05,
+      "loss": 0.3509,
+      "step": 469
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 0.28804744838713436,
+      "learning_rate": 3.68e-05,
+      "loss": 0.3687,
+      "step": 470
+    },
+    {
+      "epoch": 1.0151024811218985,
+      "grad_norm": 0.33595860575734326,
+      "learning_rate": 3.676e-05,
+      "loss": 0.3779,
+      "step": 471
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 0.30104692221444024,
+      "learning_rate": 3.672000000000001e-05,
+      "loss": 0.3691,
+      "step": 472
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.35106754973912657,
+      "learning_rate": 3.668e-05,
+      "loss": 0.3654,
+      "step": 473
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 0.2533257256241151,
+      "learning_rate": 3.664e-05,
+      "loss": 0.3805,
+      "step": 474
+    },
+    {
+      "epoch": 1.0237324703344122,
+      "grad_norm": 0.3054072286372611,
+      "learning_rate": 3.66e-05,
+      "loss": 0.359,
+      "step": 475
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 0.2504478156879782,
+      "learning_rate": 3.656e-05,
+      "loss": 0.347,
+      "step": 476
+    },
+    {
+      "epoch": 1.0280474649406688,
+      "grad_norm": 0.2712883776564294,
+      "learning_rate": 3.652e-05,
+      "loss": 0.3754,
+      "step": 477
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 0.2782380806953332,
+      "learning_rate": 3.648e-05,
+      "loss": 0.3515,
+      "step": 478
+    },
+    {
+      "epoch": 1.0323624595469256,
+      "grad_norm": 0.26159335832572284,
+      "learning_rate": 3.6440000000000003e-05,
+      "loss": 0.3485,
+      "step": 479
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 0.2864328676648181,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 0.3631,
+      "step": 480
+    },
+    {
+      "epoch": 1.0366774541531822,
+      "grad_norm": 0.25027991189741466,
+      "learning_rate": 3.636e-05,
+      "loss": 0.3816,
+      "step": 481
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 0.2788892924014196,
+      "learning_rate": 3.6320000000000005e-05,
+      "loss": 0.3567,
+      "step": 482
+    },
+    {
+      "epoch": 1.040992448759439,
+      "grad_norm": 0.2659535510350543,
+      "learning_rate": 3.6280000000000005e-05,
+      "loss": 0.3713,
+      "step": 483
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 0.23515542903905737,
+      "learning_rate": 3.624e-05,
+      "loss": 0.3535,
+      "step": 484
+    },
+    {
+      "epoch": 1.0453074433656957,
+      "grad_norm": 0.279193625770593,
+      "learning_rate": 3.62e-05,
+      "loss": 0.3747,
+      "step": 485
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 0.22521695771346117,
+      "learning_rate": 3.616e-05,
+      "loss": 0.369,
+      "step": 486
+    },
+    {
+      "epoch": 1.0496224379719525,
+      "grad_norm": 0.2535094313179469,
+      "learning_rate": 3.6120000000000007e-05,
+      "loss": 0.3537,
+      "step": 487
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 0.27003969431487984,
+      "learning_rate": 3.608e-05,
+      "loss": 0.3827,
+      "step": 488
+    },
+    {
+      "epoch": 1.0539374325782094,
+      "grad_norm": 0.2514392546102155,
+      "learning_rate": 3.604e-05,
+      "loss": 0.3744,
+      "step": 489
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 0.2371342216949108,
+      "learning_rate": 3.6e-05,
+      "loss": 0.3703,
+      "step": 490
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.2416511923662919,
+      "learning_rate": 3.596e-05,
+      "loss": 0.3711,
+      "step": 491
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 0.22901813700686482,
+      "learning_rate": 3.592e-05,
+      "loss": 0.3434,
+      "step": 492
+    },
+    {
+      "epoch": 1.0625674217907228,
+      "grad_norm": 0.2525253353540884,
+      "learning_rate": 3.588e-05,
+      "loss": 0.363,
+      "step": 493
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 0.23895832490100163,
+      "learning_rate": 3.584e-05,
+      "loss": 0.3493,
+      "step": 494
+    },
+    {
+      "epoch": 1.0668824163969794,
+      "grad_norm": 0.23238858140574412,
+      "learning_rate": 3.58e-05,
+      "loss": 0.3713,
+      "step": 495
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 0.2683342199484544,
+      "learning_rate": 3.5759999999999996e-05,
+      "loss": 0.3819,
+      "step": 496
+    },
+    {
+      "epoch": 1.0711974110032363,
+      "grad_norm": 0.24170220244530477,
+      "learning_rate": 3.5720000000000004e-05,
+      "loss": 0.361,
+      "step": 497
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.22845523982957577,
+      "learning_rate": 3.5680000000000004e-05,
+      "loss": 0.363,
+      "step": 498
+    },
+    {
+      "epoch": 1.075512405609493,
+      "grad_norm": 0.25277862463525125,
+      "learning_rate": 3.5640000000000004e-05,
+      "loss": 0.355,
+      "step": 499
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 0.24700129560716771,
+      "learning_rate": 3.56e-05,
+      "loss": 0.3706,
+      "step": 500
+    },
+    {
+      "epoch": 1.0798274002157497,
+      "grad_norm": 0.22065070960656424,
+      "learning_rate": 3.5560000000000005e-05,
+      "loss": 0.3531,
+      "step": 501
+    },
+    {
+      "epoch": 1.0819848975188782,
+      "grad_norm": 0.2561454034108686,
+      "learning_rate": 3.5520000000000006e-05,
+      "loss": 0.3833,
+      "step": 502
+    },
+    {
+      "epoch": 1.0841423948220066,
+      "grad_norm": 0.22433131754774605,
+      "learning_rate": 3.548e-05,
+      "loss": 0.3691,
+      "step": 503
+    },
+    {
+      "epoch": 1.0862998921251348,
+      "grad_norm": 0.2522033412086021,
+      "learning_rate": 3.544e-05,
+      "loss": 0.3567,
+      "step": 504
+    },
+    {
+      "epoch": 1.0884573894282632,
+      "grad_norm": 0.21905495575288447,
+      "learning_rate": 3.54e-05,
+      "loss": 0.3848,
+      "step": 505
+    },
+    {
+      "epoch": 1.0906148867313916,
+      "grad_norm": 0.22081247386141037,
+      "learning_rate": 3.536000000000001e-05,
+      "loss": 0.3662,
+      "step": 506
+    },
+    {
+      "epoch": 1.09277238403452,
+      "grad_norm": 0.2713654247657179,
+      "learning_rate": 3.532e-05,
+      "loss": 0.3699,
+      "step": 507
+    },
+    {
+      "epoch": 1.0949298813376482,
+      "grad_norm": 0.22498743492459014,
+      "learning_rate": 3.528e-05,
+      "loss": 0.383,
+      "step": 508
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.2514196262074284,
+      "learning_rate": 3.524e-05,
+      "loss": 0.3499,
+      "step": 509
+    },
+    {
+      "epoch": 1.099244875943905,
+      "grad_norm": 0.23462639319558332,
+      "learning_rate": 3.52e-05,
+      "loss": 0.3607,
+      "step": 510
+    },
+    {
+      "epoch": 1.1014023732470335,
+      "grad_norm": 0.23003249047263888,
+      "learning_rate": 3.516e-05,
+      "loss": 0.3738,
+      "step": 511
+    },
+    {
+      "epoch": 1.103559870550162,
+      "grad_norm": 0.2188060350115714,
+      "learning_rate": 3.512e-05,
+      "loss": 0.3515,
+      "step": 512
+    },
+    {
+      "epoch": 1.1057173678532901,
+      "grad_norm": 0.22124097377197594,
+      "learning_rate": 3.508e-05,
+      "loss": 0.3686,
+      "step": 513
+    },
+    {
+      "epoch": 1.1078748651564185,
+      "grad_norm": 0.21022961752556638,
+      "learning_rate": 3.504e-05,
+      "loss": 0.3784,
+      "step": 514
+    },
+    {
+      "epoch": 1.110032362459547,
+      "grad_norm": 0.2210491921381139,
+      "learning_rate": 3.5e-05,
+      "loss": 0.3542,
+      "step": 515
+    },
+    {
+      "epoch": 1.1121898597626754,
+      "grad_norm": 0.23756038385005723,
+      "learning_rate": 3.4960000000000004e-05,
+      "loss": 0.3733,
+      "step": 516
+    },
+    {
+      "epoch": 1.1143473570658036,
+      "grad_norm": 0.23088533641732836,
+      "learning_rate": 3.4920000000000004e-05,
+      "loss": 0.3577,
+      "step": 517
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.23851750461757235,
+      "learning_rate": 3.4880000000000005e-05,
+      "loss": 0.3626,
+      "step": 518
+    },
+    {
+      "epoch": 1.1186623516720604,
+      "grad_norm": 0.25889253618573643,
+      "learning_rate": 3.484e-05,
+      "loss": 0.3657,
+      "step": 519
+    },
+    {
+      "epoch": 1.1208198489751888,
+      "grad_norm": 0.2318799442388334,
+      "learning_rate": 3.48e-05,
+      "loss": 0.3513,
+      "step": 520
+    },
+    {
+      "epoch": 1.1229773462783172,
+      "grad_norm": 0.2256074426059202,
+      "learning_rate": 3.4760000000000006e-05,
+      "loss": 0.3574,
+      "step": 521
+    },
+    {
+      "epoch": 1.1251348435814454,
+      "grad_norm": 0.22447577356156587,
+      "learning_rate": 3.472e-05,
+      "loss": 0.3619,
+      "step": 522
+    },
+    {
+      "epoch": 1.1272923408845739,
+      "grad_norm": 0.23975845975915916,
+      "learning_rate": 3.468e-05,
+      "loss": 0.3598,
+      "step": 523
+    },
+    {
+      "epoch": 1.1294498381877023,
+      "grad_norm": 0.19493992211953257,
+      "learning_rate": 3.464e-05,
+      "loss": 0.3648,
+      "step": 524
+    },
+    {
+      "epoch": 1.1316073354908307,
+      "grad_norm": 0.2870170542570617,
+      "learning_rate": 3.46e-05,
+      "loss": 0.3695,
+      "step": 525
+    },
+    {
+      "epoch": 1.133764832793959,
+      "grad_norm": 0.20812408093979423,
+      "learning_rate": 3.456e-05,
+      "loss": 0.3537,
+      "step": 526
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 0.279716636463304,
+      "learning_rate": 3.452e-05,
+      "loss": 0.3609,
+      "step": 527
+    },
+    {
+      "epoch": 1.1380798274002157,
+      "grad_norm": 0.2223889872165593,
+      "learning_rate": 3.448e-05,
+      "loss": 0.3663,
+      "step": 528
+    },
+    {
+      "epoch": 1.1402373247033442,
+      "grad_norm": 0.2433632262289244,
+      "learning_rate": 3.444e-05,
+      "loss": 0.3714,
+      "step": 529
+    },
+    {
+      "epoch": 1.1423948220064726,
+      "grad_norm": 0.24277614478782433,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 0.3869,
+      "step": 530
+    },
+    {
+      "epoch": 1.1445523193096008,
+      "grad_norm": 0.2586540206286391,
+      "learning_rate": 3.436e-05,
+      "loss": 0.3579,
+      "step": 531
+    },
+    {
+      "epoch": 1.1467098166127292,
+      "grad_norm": 0.23870929551871684,
+      "learning_rate": 3.4320000000000003e-05,
+      "loss": 0.3672,
+      "step": 532
+    },
+    {
+      "epoch": 1.1488673139158576,
+      "grad_norm": 0.23766584571259206,
+      "learning_rate": 3.4280000000000004e-05,
+      "loss": 0.3671,
+      "step": 533
+    },
+    {
+      "epoch": 1.151024811218986,
+      "grad_norm": 0.24568271253401686,
+      "learning_rate": 3.424e-05,
+      "loss": 0.3596,
+      "step": 534
+    },
+    {
+      "epoch": 1.1531823085221142,
+      "grad_norm": 0.2635556243671408,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 0.3834,
+      "step": 535
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.21921288320598883,
+      "learning_rate": 3.4160000000000005e-05,
+      "loss": 0.3767,
+      "step": 536
+    },
+    {
+      "epoch": 1.157497303128371,
+      "grad_norm": 0.2761715682909131,
+      "learning_rate": 3.412e-05,
+      "loss": 0.3931,
+      "step": 537
+    },
+    {
+      "epoch": 1.1596548004314995,
+      "grad_norm": 0.24124279593723555,
+      "learning_rate": 3.408e-05,
+      "loss": 0.3616,
+      "step": 538
+    },
+    {
+      "epoch": 1.161812297734628,
+      "grad_norm": 0.24495716516269506,
+      "learning_rate": 3.404e-05,
+      "loss": 0.3551,
+      "step": 539
+    },
+    {
+      "epoch": 1.1639697950377563,
+      "grad_norm": 0.26303797180952715,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.3808,
+      "step": 540
+    },
+    {
+      "epoch": 1.1661272923408845,
+      "grad_norm": 0.2320923059681381,
+      "learning_rate": 3.396e-05,
+      "loss": 0.3489,
+      "step": 541
+    },
+    {
+      "epoch": 1.168284789644013,
+      "grad_norm": 0.22912881919910247,
+      "learning_rate": 3.392e-05,
+      "loss": 0.3782,
+      "step": 542
+    },
+    {
+      "epoch": 1.1704422869471414,
+      "grad_norm": 0.25937778405892337,
+      "learning_rate": 3.388e-05,
+      "loss": 0.3657,
+      "step": 543
+    },
+    {
+      "epoch": 1.1725997842502698,
+      "grad_norm": 0.2480554773755587,
+      "learning_rate": 3.384e-05,
+      "loss": 0.3547,
+      "step": 544
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.23033283728714243,
+      "learning_rate": 3.38e-05,
+      "loss": 0.3514,
+      "step": 545
+    },
+    {
+      "epoch": 1.1769147788565264,
+      "grad_norm": 0.23578708348817395,
+      "learning_rate": 3.376e-05,
+      "loss": 0.3449,
+      "step": 546
+    },
+    {
+      "epoch": 1.1790722761596548,
+      "grad_norm": 0.28370201661918365,
+      "learning_rate": 3.372e-05,
+      "loss": 0.3938,
+      "step": 547
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 0.24579139057291602,
+      "learning_rate": 3.368e-05,
+      "loss": 0.3648,
+      "step": 548
+    },
+    {
+      "epoch": 1.1833872707659117,
+      "grad_norm": 0.2390462478430337,
+      "learning_rate": 3.3639999999999996e-05,
+      "loss": 0.3478,
+      "step": 549
+    },
+    {
+      "epoch": 1.1855447680690399,
+      "grad_norm": 0.26131993520239843,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.3653,
+      "step": 550
+    },
+    {
+      "epoch": 1.1877022653721683,
+      "grad_norm": 0.2602420831758332,
+      "learning_rate": 3.3560000000000004e-05,
+      "loss": 0.3522,
+      "step": 551
+    },
+    {
+      "epoch": 1.1898597626752967,
+      "grad_norm": 0.2694184023955293,
+      "learning_rate": 3.3520000000000004e-05,
+      "loss": 0.3651,
+      "step": 552
+    },
+    {
+      "epoch": 1.1920172599784251,
+      "grad_norm": 0.24345754285858087,
+      "learning_rate": 3.348e-05,
+      "loss": 0.3712,
+      "step": 553
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.23563745353239082,
+      "learning_rate": 3.344e-05,
+      "loss": 0.371,
+      "step": 554
+    },
+    {
+      "epoch": 1.1963322545846817,
+      "grad_norm": 0.23701901133791073,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 0.3643,
+      "step": 555
+    },
+    {
+      "epoch": 1.1984897518878102,
+      "grad_norm": 0.23925942262157326,
+      "learning_rate": 3.336e-05,
+      "loss": 0.3544,
+      "step": 556
+    },
+    {
+      "epoch": 1.2006472491909386,
+      "grad_norm": 0.2338316781910534,
+      "learning_rate": 3.332e-05,
+      "loss": 0.3562,
+      "step": 557
+    },
+    {
+      "epoch": 1.202804746494067,
+      "grad_norm": 0.2541377842977513,
+      "learning_rate": 3.328e-05,
+      "loss": 0.3658,
+      "step": 558
+    },
+    {
+      "epoch": 1.2049622437971952,
+      "grad_norm": 0.257947454208033,
+      "learning_rate": 3.324e-05,
+      "loss": 0.3862,
+      "step": 559
+    },
+    {
+      "epoch": 1.2071197411003236,
+      "grad_norm": 0.5529685709493989,
+      "learning_rate": 3.32e-05,
+      "loss": 0.3733,
+      "step": 560
+    },
+    {
+      "epoch": 1.209277238403452,
+      "grad_norm": 0.25807202026476894,
+      "learning_rate": 3.316e-05,
+      "loss": 0.3476,
+      "step": 561
+    },
+    {
+      "epoch": 1.2114347357065804,
+      "grad_norm": 0.21766878594901307,
+      "learning_rate": 3.312e-05,
+      "loss": 0.3404,
+      "step": 562
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.2136045851257333,
+      "learning_rate": 3.308e-05,
+      "loss": 0.3628,
+      "step": 563
+    },
+    {
+      "epoch": 1.215749730312837,
+      "grad_norm": 0.23905539491616293,
+      "learning_rate": 3.304e-05,
+      "loss": 0.3767,
+      "step": 564
+    },
+    {
+      "epoch": 1.2179072276159655,
+      "grad_norm": 0.2442819986159734,
+      "learning_rate": 3.3e-05,
+      "loss": 0.3673,
+      "step": 565
+    },
+    {
+      "epoch": 1.220064724919094,
+      "grad_norm": 0.26256977239020945,
+      "learning_rate": 3.296e-05,
+      "loss": 0.3693,
+      "step": 566
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.21905194245883036,
+      "learning_rate": 3.292e-05,
+      "loss": 0.3512,
+      "step": 567
+    },
+    {
+      "epoch": 1.2243797195253505,
+      "grad_norm": 0.24907388947059758,
+      "learning_rate": 3.288e-05,
+      "loss": 0.3589,
+      "step": 568
+    },
+    {
+      "epoch": 1.226537216828479,
+      "grad_norm": 0.2621089604884317,
+      "learning_rate": 3.2840000000000004e-05,
+      "loss": 0.3614,
+      "step": 569
+    },
+    {
+      "epoch": 1.2286947141316074,
+      "grad_norm": 0.21791101743115224,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 0.378,
+      "step": 570
+    },
+    {
+      "epoch": 1.2308522114347358,
+      "grad_norm": 0.2681150827216919,
+      "learning_rate": 3.2760000000000005e-05,
+      "loss": 0.356,
+      "step": 571
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.22042202632353994,
+      "learning_rate": 3.272e-05,
+      "loss": 0.3619,
+      "step": 572
+    },
+    {
+      "epoch": 1.2351672060409924,
+      "grad_norm": 0.22940316773804112,
+      "learning_rate": 3.268e-05,
+      "loss": 0.376,
+      "step": 573
+    },
+    {
+      "epoch": 1.2373247033441208,
+      "grad_norm": 0.2300745808427193,
+      "learning_rate": 3.2640000000000006e-05,
+      "loss": 0.352,
+      "step": 574
+    },
+    {
+      "epoch": 1.2394822006472492,
+      "grad_norm": 0.22246397846867583,
+      "learning_rate": 3.26e-05,
+      "loss": 0.3674,
+      "step": 575
+    },
+    {
+      "epoch": 1.2416396979503777,
+      "grad_norm": 0.2535242296437029,
+      "learning_rate": 3.256e-05,
+      "loss": 0.3676,
+      "step": 576
+    },
+    {
+      "epoch": 1.2437971952535058,
+      "grad_norm": 0.22889371782833975,
+      "learning_rate": 3.252e-05,
+      "loss": 0.3485,
+      "step": 577
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 0.2071146637004398,
+      "learning_rate": 3.248e-05,
+      "loss": 0.3771,
+      "step": 578
+    },
+    {
+      "epoch": 1.2481121898597627,
+      "grad_norm": 0.26197336938818744,
+      "learning_rate": 3.244e-05,
+      "loss": 0.3675,
+      "step": 579
+    },
+    {
+      "epoch": 1.250269687162891,
+      "grad_norm": 0.2241786710298959,
+      "learning_rate": 3.24e-05,
+      "loss": 0.3528,
+      "step": 580
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.23312122700817697,
+      "learning_rate": 3.236e-05,
+      "loss": 0.3544,
+      "step": 581
+    },
+    {
+      "epoch": 1.2545846817691477,
+      "grad_norm": 0.2182270396469802,
+      "learning_rate": 3.232e-05,
+      "loss": 0.3568,
+      "step": 582
+    },
+    {
+      "epoch": 1.2567421790722761,
+      "grad_norm": 0.26366238123425695,
+      "learning_rate": 3.2279999999999996e-05,
+      "loss": 0.3686,
+      "step": 583
+    },
+    {
+      "epoch": 1.2588996763754046,
+      "grad_norm": 0.23349496548207888,
+      "learning_rate": 3.224e-05,
+      "loss": 0.3702,
+      "step": 584
+    },
+    {
+      "epoch": 1.261057173678533,
+      "grad_norm": 0.2280868391944456,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 0.3744,
+      "step": 585
+    },
+    {
+      "epoch": 1.2632146709816614,
+      "grad_norm": 0.2532316820814779,
+      "learning_rate": 3.2160000000000004e-05,
+      "loss": 0.3551,
+      "step": 586
+    },
+    {
+      "epoch": 1.2653721682847896,
+      "grad_norm": 0.21080695604733463,
+      "learning_rate": 3.212e-05,
+      "loss": 0.3584,
+      "step": 587
+    },
+    {
+      "epoch": 1.267529665587918,
+      "grad_norm": 0.2176852597553473,
+      "learning_rate": 3.208e-05,
+      "loss": 0.3664,
+      "step": 588
+    },
+    {
+      "epoch": 1.2696871628910464,
+      "grad_norm": 0.2350058727944437,
+      "learning_rate": 3.2040000000000005e-05,
+      "loss": 0.3572,
+      "step": 589
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.2142412894660819,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.3622,
+      "step": 590
+    },
+    {
+      "epoch": 1.274002157497303,
+      "grad_norm": 0.2988582984016833,
+      "learning_rate": 3.196e-05,
+      "loss": 0.3641,
+      "step": 591
+    },
+    {
+      "epoch": 1.2761596548004315,
+      "grad_norm": 0.21391803913879617,
+      "learning_rate": 3.192e-05,
+      "loss": 0.3594,
+      "step": 592
+    },
+    {
+      "epoch": 1.27831715210356,
+      "grad_norm": 0.26842444375759716,
+      "learning_rate": 3.188e-05,
+      "loss": 0.3652,
+      "step": 593
+    },
+    {
+      "epoch": 1.2804746494066883,
+      "grad_norm": 0.24949721130385824,
+      "learning_rate": 3.184e-05,
+      "loss": 0.3734,
+      "step": 594
+    },
+    {
+      "epoch": 1.2826321467098167,
+      "grad_norm": 0.23021856207798744,
+      "learning_rate": 3.18e-05,
+      "loss": 0.3569,
+      "step": 595
+    },
+    {
+      "epoch": 1.284789644012945,
+      "grad_norm": 0.256911436435004,
+      "learning_rate": 3.176e-05,
+      "loss": 0.3658,
+      "step": 596
+    },
+    {
+      "epoch": 1.2869471413160734,
+      "grad_norm": 0.2442690046250457,
+      "learning_rate": 3.172e-05,
+      "loss": 0.3733,
+      "step": 597
+    },
+    {
+      "epoch": 1.2891046386192018,
+      "grad_norm": 0.2339516527126993,
+      "learning_rate": 3.168e-05,
+      "loss": 0.3594,
+      "step": 598
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.22497425413784988,
+      "learning_rate": 3.164e-05,
+      "loss": 0.3569,
+      "step": 599
+    },
+    {
+      "epoch": 1.2934196332254584,
+      "grad_norm": 0.20579662620446118,
+      "learning_rate": 3.16e-05,
+      "loss": 0.3683,
+      "step": 600
+    },
+    {
+      "epoch": 1.2955771305285868,
+      "grad_norm": 0.24608918583535308,
+      "learning_rate": 3.156e-05,
+      "loss": 0.3733,
+      "step": 601
+    },
+    {
+      "epoch": 1.2977346278317152,
+      "grad_norm": 0.2281105722902805,
+      "learning_rate": 3.1519999999999996e-05,
+      "loss": 0.3694,
+      "step": 602
+    },
+    {
+      "epoch": 1.2998921251348436,
+      "grad_norm": 0.2040966143041461,
+      "learning_rate": 3.1480000000000004e-05,
+      "loss": 0.3636,
+      "step": 603
+    },
+    {
+      "epoch": 1.302049622437972,
+      "grad_norm": 0.21460759759197517,
+      "learning_rate": 3.1440000000000004e-05,
+      "loss": 0.3477,
+      "step": 604
+    },
+    {
+      "epoch": 1.3042071197411003,
+      "grad_norm": 0.221481405151603,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 0.3699,
+      "step": 605
+    },
+    {
+      "epoch": 1.3063646170442287,
+      "grad_norm": 0.2501579732769482,
+      "learning_rate": 3.136e-05,
+      "loss": 0.4039,
+      "step": 606
+    },
+    {
+      "epoch": 1.308522114347357,
+      "grad_norm": 0.2077259213631551,
+      "learning_rate": 3.132e-05,
+      "loss": 0.358,
+      "step": 607
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.22593570412239217,
+      "learning_rate": 3.1280000000000005e-05,
+      "loss": 0.3546,
+      "step": 608
+    },
+    {
+      "epoch": 1.3128371089536137,
+      "grad_norm": 0.23482506691659347,
+      "learning_rate": 3.1240000000000006e-05,
+      "loss": 0.3932,
+      "step": 609
+    },
+    {
+      "epoch": 1.3149946062567421,
+      "grad_norm": 0.20192962328812056,
+      "learning_rate": 3.12e-05,
+      "loss": 0.359,
+      "step": 610
+    },
+    {
+      "epoch": 1.3171521035598706,
+      "grad_norm": 0.2309883568911319,
+      "learning_rate": 3.116e-05,
+      "loss": 0.3661,
+      "step": 611
+    },
+    {
+      "epoch": 1.319309600862999,
+      "grad_norm": 0.2258622386132955,
+      "learning_rate": 3.112e-05,
+      "loss": 0.3553,
+      "step": 612
+    },
+    {
+      "epoch": 1.3214670981661274,
+      "grad_norm": 0.1955163803876729,
+      "learning_rate": 3.108e-05,
+      "loss": 0.3502,
+      "step": 613
+    },
+    {
+      "epoch": 1.3236245954692556,
+      "grad_norm": 0.23352846688566287,
+      "learning_rate": 3.104e-05,
+      "loss": 0.3621,
+      "step": 614
+    },
+    {
+      "epoch": 1.325782092772384,
+      "grad_norm": 0.22786136344214358,
+      "learning_rate": 3.1e-05,
+      "loss": 0.3654,
+      "step": 615
+    },
+    {
+      "epoch": 1.3279395900755124,
+      "grad_norm": 0.23706574178741774,
+      "learning_rate": 3.096e-05,
+      "loss": 0.3708,
+      "step": 616
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.21678537837279216,
+      "learning_rate": 3.092e-05,
+      "loss": 0.3601,
+      "step": 617
+    },
+    {
+      "epoch": 1.332254584681769,
+      "grad_norm": 0.22426690469613242,
+      "learning_rate": 3.088e-05,
+      "loss": 0.3607,
+      "step": 618
+    },
+    {
+      "epoch": 1.3344120819848975,
+      "grad_norm": 0.22447348229662797,
+      "learning_rate": 3.084e-05,
+      "loss": 0.3677,
+      "step": 619
+    },
+    {
+      "epoch": 1.3365695792880259,
+      "grad_norm": 0.2414443482792685,
+      "learning_rate": 3.08e-05,
+      "loss": 0.3516,
+      "step": 620
+    },
+    {
+      "epoch": 1.3387270765911543,
+      "grad_norm": 0.2209324250516759,
+      "learning_rate": 3.076e-05,
+      "loss": 0.3465,
+      "step": 621
+    },
+    {
+      "epoch": 1.3408845738942827,
+      "grad_norm": 0.22931182094163088,
+      "learning_rate": 3.072e-05,
+      "loss": 0.3645,
+      "step": 622
+    },
+    {
+      "epoch": 1.343042071197411,
+      "grad_norm": 0.22884876391903108,
+      "learning_rate": 3.0680000000000004e-05,
+      "loss": 0.3732,
+      "step": 623
+    },
+    {
+      "epoch": 1.3451995685005393,
+      "grad_norm": 0.23344780330484063,
+      "learning_rate": 3.0640000000000005e-05,
+      "loss": 0.3603,
+      "step": 624
+    },
+    {
+      "epoch": 1.3473570658036678,
+      "grad_norm": 0.2286620307384633,
+      "learning_rate": 3.06e-05,
+      "loss": 0.3694,
+      "step": 625
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.2161365308453795,
+      "learning_rate": 3.056e-05,
+      "loss": 0.3654,
+      "step": 626
+    },
+    {
+      "epoch": 1.3516720604099244,
+      "grad_norm": 0.23046040136711576,
+      "learning_rate": 3.0520000000000006e-05,
+      "loss": 0.3592,
+      "step": 627
+    },
+    {
+      "epoch": 1.3538295577130528,
+      "grad_norm": 0.20568766413435982,
+      "learning_rate": 3.0480000000000003e-05,
+      "loss": 0.3878,
+      "step": 628
+    },
+    {
+      "epoch": 1.3559870550161812,
+      "grad_norm": 0.23495098967359113,
+      "learning_rate": 3.0440000000000003e-05,
+      "loss": 0.3619,
+      "step": 629
+    },
+    {
+      "epoch": 1.3581445523193096,
+      "grad_norm": 0.19926196422663414,
+      "learning_rate": 3.04e-05,
+      "loss": 0.3639,
+      "step": 630
+    },
+    {
+      "epoch": 1.360302049622438,
+      "grad_norm": 0.253526676183994,
+      "learning_rate": 3.036e-05,
+      "loss": 0.3649,
+      "step": 631
+    },
+    {
+      "epoch": 1.3624595469255665,
+      "grad_norm": 0.2247831223309294,
+      "learning_rate": 3.0320000000000004e-05,
+      "loss": 0.3836,
+      "step": 632
+    },
+    {
+      "epoch": 1.3646170442286947,
+      "grad_norm": 0.23980813430498824,
+      "learning_rate": 3.028e-05,
+      "loss": 0.3735,
+      "step": 633
+    },
+    {
+      "epoch": 1.366774541531823,
+      "grad_norm": 0.24839219342419572,
+      "learning_rate": 3.0240000000000002e-05,
+      "loss": 0.3555,
+      "step": 634
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.23203667590220284,
+      "learning_rate": 3.02e-05,
+      "loss": 0.3433,
+      "step": 635
+    },
+    {
+      "epoch": 1.3710895361380797,
+      "grad_norm": 0.25304868631907496,
+      "learning_rate": 3.016e-05,
+      "loss": 0.3635,
+      "step": 636
+    },
+    {
+      "epoch": 1.3732470334412081,
+      "grad_norm": 0.23269259251392885,
+      "learning_rate": 3.0120000000000003e-05,
+      "loss": 0.3961,
+      "step": 637
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 0.21293976180361354,
+      "learning_rate": 3.0080000000000003e-05,
+      "loss": 0.349,
+      "step": 638
+    },
+    {
+      "epoch": 1.377562028047465,
+      "grad_norm": 0.2848360522308427,
+      "learning_rate": 3.004e-05,
+      "loss": 0.3634,
+      "step": 639
+    },
+    {
+      "epoch": 1.3797195253505934,
+      "grad_norm": 0.23140180630070306,
+      "learning_rate": 3e-05,
+      "loss": 0.38,
+      "step": 640
+    },
+    {
+      "epoch": 1.3818770226537218,
+      "grad_norm": 0.2745994307355107,
+      "learning_rate": 2.9959999999999998e-05,
+      "loss": 0.3753,
+      "step": 641
+    },
+    {
+      "epoch": 1.38403451995685,
+      "grad_norm": 0.26144668714615754,
+      "learning_rate": 2.9920000000000005e-05,
+      "loss": 0.3663,
+      "step": 642
+    },
+    {
+      "epoch": 1.3861920172599784,
+      "grad_norm": 0.2421993564006052,
+      "learning_rate": 2.9880000000000002e-05,
+      "loss": 0.3568,
+      "step": 643
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 0.2370121873272197,
+      "learning_rate": 2.9840000000000002e-05,
+      "loss": 0.3596,
+      "step": 644
+    },
+    {
+      "epoch": 1.390507011866235,
+      "grad_norm": 0.24085023222334487,
+      "learning_rate": 2.98e-05,
+      "loss": 0.3762,
+      "step": 645
+    },
+    {
+      "epoch": 1.3926645091693635,
+      "grad_norm": 0.234755628545118,
+      "learning_rate": 2.976e-05,
+      "loss": 0.3672,
+      "step": 646
+    },
+    {
+      "epoch": 1.3948220064724919,
+      "grad_norm": 0.24182630900679758,
+      "learning_rate": 2.9720000000000003e-05,
+      "loss": 0.3687,
+      "step": 647
+    },
+    {
+      "epoch": 1.3969795037756203,
+      "grad_norm": 0.23986126807518696,
+      "learning_rate": 2.9680000000000004e-05,
+      "loss": 0.3591,
+      "step": 648
+    },
+    {
+      "epoch": 1.3991370010787487,
+      "grad_norm": 0.24262318684533668,
+      "learning_rate": 2.964e-05,
+      "loss": 0.3532,
+      "step": 649
+    },
+    {
+      "epoch": 1.4012944983818771,
+      "grad_norm": 0.2403573614795157,
+      "learning_rate": 2.96e-05,
+      "loss": 0.3695,
+      "step": 650
+    },
+    {
+      "epoch": 1.4034519956850053,
+      "grad_norm": 0.2231968626230008,
+      "learning_rate": 2.9559999999999998e-05,
+      "loss": 0.3687,
+      "step": 651
+    },
+    {
+      "epoch": 1.4056094929881338,
+      "grad_norm": 0.24521191692071806,
+      "learning_rate": 2.9520000000000002e-05,
+      "loss": 0.3616,
+      "step": 652
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.24730212938499072,
+      "learning_rate": 2.9480000000000002e-05,
+      "loss": 0.3465,
+      "step": 653
+    },
+    {
+      "epoch": 1.4099244875943904,
+      "grad_norm": 0.2334236281522416,
+      "learning_rate": 2.944e-05,
+      "loss": 0.3685,
+      "step": 654
+    },
+    {
+      "epoch": 1.4120819848975188,
+      "grad_norm": 0.23180879257563838,
+      "learning_rate": 2.94e-05,
+      "loss": 0.3516,
+      "step": 655
+    },
+    {
+      "epoch": 1.4142394822006472,
+      "grad_norm": 0.22429371025157882,
+      "learning_rate": 2.9360000000000003e-05,
+      "loss": 0.3539,
+      "step": 656
+    },
+    {
+      "epoch": 1.4163969795037756,
+      "grad_norm": 0.23286822854457767,
+      "learning_rate": 2.9320000000000004e-05,
+      "loss": 0.3673,
+      "step": 657
+    },
+    {
+      "epoch": 1.418554476806904,
+      "grad_norm": 0.21972421961678684,
+      "learning_rate": 2.928e-05,
+      "loss": 0.3564,
+      "step": 658
+    },
+    {
+      "epoch": 1.4207119741100325,
+      "grad_norm": 0.2317444957314592,
+      "learning_rate": 2.924e-05,
+      "loss": 0.3647,
+      "step": 659
+    },
+    {
+      "epoch": 1.4228694714131607,
+      "grad_norm": 0.7300728474445504,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 0.3521,
+      "step": 660
+    },
+    {
+      "epoch": 1.425026968716289,
+      "grad_norm": 0.22003003920883527,
+      "learning_rate": 2.9160000000000005e-05,
+      "loss": 0.3661,
+      "step": 661
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.2771166892544825,
+      "learning_rate": 2.9120000000000002e-05,
+      "loss": 0.3763,
+      "step": 662
+    },
+    {
+      "epoch": 1.429341963322546,
+      "grad_norm": 0.21961469846761253,
+      "learning_rate": 2.9080000000000003e-05,
+      "loss": 0.3637,
+      "step": 663
+    },
+    {
+      "epoch": 1.4314994606256741,
+      "grad_norm": 0.23524236389027967,
+      "learning_rate": 2.904e-05,
+      "loss": 0.382,
+      "step": 664
+    },
+    {
+      "epoch": 1.4336569579288025,
+      "grad_norm": 0.2527543079175038,
+      "learning_rate": 2.9e-05,
+      "loss": 0.3658,
+      "step": 665
+    },
+    {
+      "epoch": 1.435814455231931,
+      "grad_norm": 0.2580197589459597,
+      "learning_rate": 2.8960000000000004e-05,
+      "loss": 0.3618,
+      "step": 666
+    },
+    {
+      "epoch": 1.4379719525350594,
+      "grad_norm": 0.2556251991492771,
+      "learning_rate": 2.8920000000000004e-05,
+      "loss": 0.3742,
+      "step": 667
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 0.25698446665498725,
+      "learning_rate": 2.888e-05,
+      "loss": 0.3631,
+      "step": 668
+    },
+    {
+      "epoch": 1.442286947141316,
+      "grad_norm": 0.21986799830862488,
+      "learning_rate": 2.8840000000000002e-05,
+      "loss": 0.3697,
+      "step": 669
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.21959952885732004,
+      "learning_rate": 2.88e-05,
+      "loss": 0.3754,
+      "step": 670
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.24918494834725308,
+      "learning_rate": 2.8760000000000002e-05,
+      "loss": 0.3777,
+      "step": 671
+    },
+    {
+      "epoch": 1.4487594390507013,
+      "grad_norm": 0.2260301950438927,
+      "learning_rate": 2.8720000000000003e-05,
+      "loss": 0.3519,
+      "step": 672
+    },
+    {
+      "epoch": 1.4509169363538295,
+      "grad_norm": 0.2283050010652631,
+      "learning_rate": 2.868e-05,
+      "loss": 0.377,
+      "step": 673
+    },
+    {
+      "epoch": 1.4530744336569579,
+      "grad_norm": 0.22476927273463568,
+      "learning_rate": 2.864e-05,
+      "loss": 0.3476,
+      "step": 674
+    },
+    {
+      "epoch": 1.4552319309600863,
+      "grad_norm": 0.20324342452103078,
+      "learning_rate": 2.86e-05,
+      "loss": 0.3475,
+      "step": 675
+    },
+    {
+      "epoch": 1.4573894282632147,
+      "grad_norm": 0.22629462680575138,
+      "learning_rate": 2.8560000000000004e-05,
+      "loss": 0.3569,
+      "step": 676
+    },
+    {
+      "epoch": 1.4595469255663431,
+      "grad_norm": 0.19880472901319113,
+      "learning_rate": 2.852e-05,
+      "loss": 0.3564,
+      "step": 677
+    },
+    {
+      "epoch": 1.4617044228694713,
+      "grad_norm": 0.2143320586140639,
+      "learning_rate": 2.8480000000000002e-05,
+      "loss": 0.3655,
+      "step": 678
+    },
+    {
+      "epoch": 1.4638619201725998,
+      "grad_norm": 0.2362702869766217,
+      "learning_rate": 2.844e-05,
+      "loss": 0.3574,
+      "step": 679
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.22656216374024327,
+      "learning_rate": 2.84e-05,
+      "loss": 0.3391,
+      "step": 680
+    },
+    {
+      "epoch": 1.4681769147788566,
+      "grad_norm": 0.21809318515170673,
+      "learning_rate": 2.8360000000000003e-05,
+      "loss": 0.3639,
+      "step": 681
+    },
+    {
+      "epoch": 1.4703344120819848,
+      "grad_norm": 0.23993152342071317,
+      "learning_rate": 2.8320000000000003e-05,
+      "loss": 0.3492,
+      "step": 682
+    },
+    {
+      "epoch": 1.4724919093851132,
+      "grad_norm": 0.23685847874927074,
+      "learning_rate": 2.828e-05,
+      "loss": 0.3724,
+      "step": 683
+    },
+    {
+      "epoch": 1.4746494066882416,
+      "grad_norm": 0.22891344167344313,
+      "learning_rate": 2.824e-05,
+      "loss": 0.3542,
+      "step": 684
+    },
+    {
+      "epoch": 1.47680690399137,
+      "grad_norm": 0.23521390801990014,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.3609,
+      "step": 685
+    },
+    {
+      "epoch": 1.4789644012944985,
+      "grad_norm": 0.24288583571997324,
+      "learning_rate": 2.816e-05,
+      "loss": 0.3661,
+      "step": 686
+    },
+    {
+      "epoch": 1.4811218985976269,
+      "grad_norm": 0.2740848150612509,
+      "learning_rate": 2.8120000000000002e-05,
+      "loss": 0.3598,
+      "step": 687
+    },
+    {
+      "epoch": 1.483279395900755,
+      "grad_norm": 0.23571675791182756,
+      "learning_rate": 2.8080000000000002e-05,
+      "loss": 0.3654,
+      "step": 688
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 0.20129412441664046,
+      "learning_rate": 2.804e-05,
+      "loss": 0.3536,
+      "step": 689
+    },
+    {
+      "epoch": 1.487594390507012,
+      "grad_norm": 0.23166748621018965,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.367,
+      "step": 690
+    },
+    {
+      "epoch": 1.4897518878101401,
+      "grad_norm": 0.20871358422460617,
+      "learning_rate": 2.7960000000000003e-05,
+      "loss": 0.3518,
+      "step": 691
+    },
+    {
+      "epoch": 1.4919093851132685,
+      "grad_norm": 0.20339117924240924,
+      "learning_rate": 2.792e-05,
+      "loss": 0.3555,
+      "step": 692
+    },
+    {
+      "epoch": 1.494066882416397,
+      "grad_norm": 0.21272736583634902,
+      "learning_rate": 2.788e-05,
+      "loss": 0.3712,
+      "step": 693
+    },
+    {
+      "epoch": 1.4962243797195254,
+      "grad_norm": 0.23915492805442082,
+      "learning_rate": 2.7839999999999998e-05,
+      "loss": 0.3755,
+      "step": 694
+    },
+    {
+      "epoch": 1.4983818770226538,
+      "grad_norm": 0.20821942101166777,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.3695,
+      "step": 695
+    },
+    {
+      "epoch": 1.5005393743257822,
+      "grad_norm": 0.2232367633694246,
+      "learning_rate": 2.7760000000000002e-05,
+      "loss": 0.3688,
+      "step": 696
+    },
+    {
+      "epoch": 1.5026968716289104,
+      "grad_norm": 0.24499976864308812,
+      "learning_rate": 2.7720000000000002e-05,
+      "loss": 0.3543,
+      "step": 697
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.22940998225556647,
+      "learning_rate": 2.768e-05,
+      "loss": 0.3696,
+      "step": 698
+    },
+    {
+      "epoch": 1.5070118662351673,
+      "grad_norm": 0.2267128233697368,
+      "learning_rate": 2.764e-05,
+      "loss": 0.3579,
+      "step": 699
+    },
+    {
+      "epoch": 1.5091693635382954,
+      "grad_norm": 0.2307569329238415,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.3834,
+      "step": 700
+    },
+    {
+      "epoch": 1.5113268608414239,
+      "grad_norm": 0.25112226025346424,
+      "learning_rate": 2.7560000000000004e-05,
+      "loss": 0.3679,
+      "step": 701
+    },
+    {
+      "epoch": 1.5134843581445523,
+      "grad_norm": 0.25419142213454193,
+      "learning_rate": 2.752e-05,
+      "loss": 0.3542,
+      "step": 702
+    },
+    {
+      "epoch": 1.5156418554476807,
+      "grad_norm": 0.22613974746673926,
+      "learning_rate": 2.748e-05,
+      "loss": 0.3832,
+      "step": 703
+    },
+    {
+      "epoch": 1.5177993527508091,
+      "grad_norm": 0.22613424161758108,
+      "learning_rate": 2.7439999999999998e-05,
+      "loss": 0.3655,
+      "step": 704
+    },
+    {
+      "epoch": 1.5199568500539375,
+      "grad_norm": 0.23893379478624607,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 0.3519,
+      "step": 705
+    },
+    {
+      "epoch": 1.522114347357066,
+      "grad_norm": 0.2775995269354047,
+      "learning_rate": 2.7360000000000002e-05,
+      "loss": 0.3894,
+      "step": 706
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.23904561581196412,
+      "learning_rate": 2.7320000000000003e-05,
+      "loss": 0.3655,
+      "step": 707
+    },
+    {
+      "epoch": 1.5264293419633226,
+      "grad_norm": 0.23756759487868181,
+      "learning_rate": 2.728e-05,
+      "loss": 0.3519,
+      "step": 708
+    },
+    {
+      "epoch": 1.5285868392664508,
+      "grad_norm": 0.20601009974320345,
+      "learning_rate": 2.724e-05,
+      "loss": 0.3558,
+      "step": 709
+    },
+    {
+      "epoch": 1.5307443365695792,
+      "grad_norm": 0.2179130898471259,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.3788,
+      "step": 710
+    },
+    {
+      "epoch": 1.5329018338727076,
+      "grad_norm": 0.25261468141264204,
+      "learning_rate": 2.716e-05,
+      "loss": 0.369,
+      "step": 711
+    },
+    {
+      "epoch": 1.535059331175836,
+      "grad_norm": 0.20717129569980194,
+      "learning_rate": 2.712e-05,
+      "loss": 0.3579,
+      "step": 712
+    },
+    {
+      "epoch": 1.5372168284789645,
+      "grad_norm": 0.21298019748089506,
+      "learning_rate": 2.7079999999999998e-05,
+      "loss": 0.3586,
+      "step": 713
+    },
+    {
+      "epoch": 1.5393743257820929,
+      "grad_norm": 0.25207101142696386,
+      "learning_rate": 2.704e-05,
+      "loss": 0.3538,
+      "step": 714
+    },
+    {
+      "epoch": 1.5415318230852213,
+      "grad_norm": 0.20068229452734665,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.3573,
+      "step": 715
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.22821380199261126,
+      "learning_rate": 2.6960000000000003e-05,
+      "loss": 0.3634,
+      "step": 716
+    },
+    {
+      "epoch": 1.545846817691478,
+      "grad_norm": 0.21345032453764343,
+      "learning_rate": 2.692e-05,
+      "loss": 0.364,
+      "step": 717
+    },
+    {
+      "epoch": 1.5480043149946061,
+      "grad_norm": 0.20494875666911705,
+      "learning_rate": 2.688e-05,
+      "loss": 0.3697,
+      "step": 718
+    },
+    {
+      "epoch": 1.5501618122977345,
+      "grad_norm": 0.2154910856310443,
+      "learning_rate": 2.6840000000000004e-05,
+      "loss": 0.3574,
+      "step": 719
+    },
+    {
+      "epoch": 1.552319309600863,
+      "grad_norm": 0.20193810414215643,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 0.3531,
+      "step": 720
+    },
+    {
+      "epoch": 1.5544768069039914,
+      "grad_norm": 0.2312442332707919,
+      "learning_rate": 2.676e-05,
+      "loss": 0.3687,
+      "step": 721
+    },
+    {
+      "epoch": 1.5566343042071198,
+      "grad_norm": 0.23089370538006676,
+      "learning_rate": 2.672e-05,
+      "loss": 0.3713,
+      "step": 722
+    },
+    {
+      "epoch": 1.5587918015102482,
+      "grad_norm": 0.206819035037506,
+      "learning_rate": 2.668e-05,
+      "loss": 0.3636,
+      "step": 723
+    },
+    {
+      "epoch": 1.5609492988133766,
+      "grad_norm": 0.24663426452102197,
+      "learning_rate": 2.6640000000000002e-05,
+      "loss": 0.3757,
+      "step": 724
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 0.24688884560578506,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 0.3747,
+      "step": 725
+    },
+    {
+      "epoch": 1.5652642934196332,
+      "grad_norm": 0.2188001289885843,
+      "learning_rate": 2.6560000000000003e-05,
+      "loss": 0.3702,
+      "step": 726
+    },
+    {
+      "epoch": 1.5674217907227614,
+      "grad_norm": 0.24246127603696868,
+      "learning_rate": 2.652e-05,
+      "loss": 0.3841,
+      "step": 727
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 0.2203909069469141,
+      "learning_rate": 2.648e-05,
+      "loss": 0.3546,
+      "step": 728
+    },
+    {
+      "epoch": 1.5717367853290183,
+      "grad_norm": 0.21585230186912005,
+      "learning_rate": 2.6440000000000004e-05,
+      "loss": 0.3477,
+      "step": 729
+    },
+    {
+      "epoch": 1.5738942826321467,
+      "grad_norm": 0.22218825993341798,
+      "learning_rate": 2.64e-05,
+      "loss": 0.3632,
+      "step": 730
+    },
+    {
+      "epoch": 1.5760517799352751,
+      "grad_norm": 0.24371912936443485,
+      "learning_rate": 2.6360000000000002e-05,
+      "loss": 0.38,
+      "step": 731
+    },
+    {
+      "epoch": 1.5782092772384035,
+      "grad_norm": 0.2227192482351511,
+      "learning_rate": 2.632e-05,
+      "loss": 0.3703,
+      "step": 732
+    },
+    {
+      "epoch": 1.580366774541532,
+      "grad_norm": 0.21456759816988139,
+      "learning_rate": 2.628e-05,
+      "loss": 0.3374,
+      "step": 733
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.23342928415132325,
+      "learning_rate": 2.6240000000000003e-05,
+      "loss": 0.3563,
+      "step": 734
+    },
+    {
+      "epoch": 1.5846817691477886,
+      "grad_norm": 0.23418445062714002,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 0.3748,
+      "step": 735
+    },
+    {
+      "epoch": 1.5868392664509168,
+      "grad_norm": 0.2336619338904283,
+      "learning_rate": 2.616e-05,
+      "loss": 0.3495,
+      "step": 736
+    },
+    {
+      "epoch": 1.5889967637540452,
+      "grad_norm": 0.24381867912545538,
+      "learning_rate": 2.612e-05,
+      "loss": 0.3611,
+      "step": 737
+    },
+    {
+      "epoch": 1.5911542610571736,
+      "grad_norm": 0.3090239665378897,
+      "learning_rate": 2.6079999999999998e-05,
+      "loss": 0.3717,
+      "step": 738
+    },
+    {
+      "epoch": 1.593311758360302,
+      "grad_norm": 0.2646555898668475,
+      "learning_rate": 2.6040000000000005e-05,
+      "loss": 0.3589,
+      "step": 739
+    },
+    {
+      "epoch": 1.5954692556634305,
+      "grad_norm": 0.2296470939633854,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.3642,
+      "step": 740
+    },
+    {
+      "epoch": 1.5976267529665589,
+      "grad_norm": 0.2419732243746287,
+      "learning_rate": 2.5960000000000002e-05,
+      "loss": 0.3917,
+      "step": 741
+    },
+    {
+      "epoch": 1.5997842502696873,
+      "grad_norm": 0.2322415069928087,
+      "learning_rate": 2.592e-05,
+      "loss": 0.3584,
+      "step": 742
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.24405015607804642,
+      "learning_rate": 2.588e-05,
+      "loss": 0.3727,
+      "step": 743
+    },
+    {
+      "epoch": 1.604099244875944,
+      "grad_norm": 0.22590347323709656,
+      "learning_rate": 2.5840000000000003e-05,
+      "loss": 0.356,
+      "step": 744
+    },
+    {
+      "epoch": 1.6062567421790723,
+      "grad_norm": 0.24229824285904206,
+      "learning_rate": 2.58e-05,
+      "loss": 0.3768,
+      "step": 745
+    },
+    {
+      "epoch": 1.6084142394822005,
+      "grad_norm": 0.22910048467078928,
+      "learning_rate": 2.576e-05,
+      "loss": 0.3757,
+      "step": 746
+    },
+    {
+      "epoch": 1.610571736785329,
+      "grad_norm": 0.22282531058063634,
+      "learning_rate": 2.572e-05,
+      "loss": 0.3584,
+      "step": 747
+    },
+    {
+      "epoch": 1.6127292340884574,
+      "grad_norm": 0.23674151401492222,
+      "learning_rate": 2.5679999999999998e-05,
+      "loss": 0.3551,
+      "step": 748
+    },
+    {
+      "epoch": 1.6148867313915858,
+      "grad_norm": 0.23673152399904154,
+      "learning_rate": 2.5640000000000002e-05,
+      "loss": 0.3868,
+      "step": 749
+    },
+    {
+      "epoch": 1.6170442286947142,
+      "grad_norm": 0.23017444547110794,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.3694,
+      "step": 750
+    },
+    {
+      "epoch": 1.6192017259978426,
+      "grad_norm": 0.2464613175786792,
+      "learning_rate": 2.556e-05,
+      "loss": 0.362,
+      "step": 751
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.22557125082732474,
+      "learning_rate": 2.552e-05,
+      "loss": 0.3789,
+      "step": 752
+    },
+    {
+      "epoch": 1.6235167206040992,
+      "grad_norm": 0.23750344656685524,
+      "learning_rate": 2.5480000000000003e-05,
+      "loss": 0.373,
+      "step": 753
+    },
+    {
+      "epoch": 1.6256742179072277,
+      "grad_norm": 0.22516492800062368,
+      "learning_rate": 2.5440000000000004e-05,
+      "loss": 0.3499,
+      "step": 754
+    },
+    {
+      "epoch": 1.6278317152103559,
+      "grad_norm": 0.22543762231722306,
+      "learning_rate": 2.54e-05,
+      "loss": 0.3691,
+      "step": 755
+    },
+    {
+      "epoch": 1.6299892125134843,
+      "grad_norm": 0.21901698270170006,
+      "learning_rate": 2.536e-05,
+      "loss": 0.3541,
+      "step": 756
+    },
+    {
+      "epoch": 1.6321467098166127,
+      "grad_norm": 0.20145275001402113,
+      "learning_rate": 2.5319999999999998e-05,
+      "loss": 0.3582,
+      "step": 757
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.2375280443344114,
+      "learning_rate": 2.5280000000000005e-05,
+      "loss": 0.3659,
+      "step": 758
+    },
+    {
+      "epoch": 1.6364617044228695,
+      "grad_norm": 0.20472034873672557,
+      "learning_rate": 2.5240000000000002e-05,
+      "loss": 0.3571,
+      "step": 759
+    },
+    {
+      "epoch": 1.638619201725998,
+      "grad_norm": 0.2264353999031357,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 0.3582,
+      "step": 760
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.23609128223309025,
+      "learning_rate": 2.516e-05,
+      "loss": 0.3463,
+      "step": 761
+    },
+    {
+      "epoch": 1.6429341963322546,
+      "grad_norm": 0.20195916155379995,
+      "learning_rate": 2.512e-05,
+      "loss": 0.3736,
+      "step": 762
+    },
+    {
+      "epoch": 1.645091693635383,
+      "grad_norm": 0.22128929062222866,
+      "learning_rate": 2.5080000000000004e-05,
+      "loss": 0.3838,
+      "step": 763
+    },
+    {
+      "epoch": 1.6472491909385112,
+      "grad_norm": 0.22940217162415266,
+      "learning_rate": 2.504e-05,
+      "loss": 0.3434,
+      "step": 764
+    },
+    {
+      "epoch": 1.6494066882416396,
+      "grad_norm": 0.20765834519446938,
+      "learning_rate": 2.5e-05,
+      "loss": 0.36,
+      "step": 765
+    },
+    {
+      "epoch": 1.651564185544768,
+      "grad_norm": 0.22652122912706935,
+      "learning_rate": 2.496e-05,
+      "loss": 0.3537,
+      "step": 766
+    },
+    {
+      "epoch": 1.6537216828478964,
+      "grad_norm": 0.209486467241231,
+      "learning_rate": 2.4920000000000002e-05,
+      "loss": 0.38,
+      "step": 767
+    },
+    {
+      "epoch": 1.6558791801510249,
+      "grad_norm": 0.21329391063224873,
+      "learning_rate": 2.488e-05,
+      "loss": 0.3471,
+      "step": 768
+    },
+    {
+      "epoch": 1.6580366774541533,
+      "grad_norm": 0.204874907654904,
+      "learning_rate": 2.4840000000000003e-05,
+      "loss": 0.3874,
+      "step": 769
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 0.21401626438471516,
+      "learning_rate": 2.48e-05,
+      "loss": 0.3747,
+      "step": 770
+    },
+    {
+      "epoch": 1.66235167206041,
+      "grad_norm": 0.2259053066704498,
+      "learning_rate": 2.476e-05,
+      "loss": 0.3595,
+      "step": 771
+    },
+    {
+      "epoch": 1.6645091693635383,
+      "grad_norm": 0.2096015711688781,
+      "learning_rate": 2.472e-05,
+      "loss": 0.3635,
+      "step": 772
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.19181190032847367,
+      "learning_rate": 2.468e-05,
+      "loss": 0.3624,
+      "step": 773
+    },
+    {
+      "epoch": 1.668824163969795,
+      "grad_norm": 0.2244869637071479,
+      "learning_rate": 2.464e-05,
+      "loss": 0.3595,
+      "step": 774
+    },
+    {
+      "epoch": 1.6709816612729234,
+      "grad_norm": 0.222169368065215,
+      "learning_rate": 2.46e-05,
+      "loss": 0.3729,
+      "step": 775
+    },
+    {
+      "epoch": 1.6731391585760518,
+      "grad_norm": 0.20169251773681174,
+      "learning_rate": 2.4560000000000002e-05,
+      "loss": 0.3651,
+      "step": 776
+    },
+    {
+      "epoch": 1.6752966558791802,
+      "grad_norm": 0.25372193384583863,
+      "learning_rate": 2.4520000000000002e-05,
+      "loss": 0.3454,
+      "step": 777
+    },
+    {
+      "epoch": 1.6774541531823086,
+      "grad_norm": 0.2197114197850221,
+      "learning_rate": 2.448e-05,
+      "loss": 0.3537,
+      "step": 778
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 0.3314925595487828,
+      "learning_rate": 2.4440000000000003e-05,
+      "loss": 0.3621,
+      "step": 779
+    },
+    {
+      "epoch": 1.6817691477885652,
+      "grad_norm": 0.20939091826537268,
+      "learning_rate": 2.44e-05,
+      "loss": 0.3502,
+      "step": 780
+    },
+    {
+      "epoch": 1.6839266450916937,
+      "grad_norm": 0.21279544211352294,
+      "learning_rate": 2.4360000000000004e-05,
+      "loss": 0.35,
+      "step": 781
+    },
+    {
+      "epoch": 1.6860841423948218,
+      "grad_norm": 0.24701488636019142,
+      "learning_rate": 2.432e-05,
+      "loss": 0.3594,
+      "step": 782
+    },
+    {
+      "epoch": 1.6882416396979503,
+      "grad_norm": 0.20477887450410062,
+      "learning_rate": 2.428e-05,
+      "loss": 0.3479,
+      "step": 783
+    },
+    {
+      "epoch": 1.6903991370010787,
+      "grad_norm": 0.23481629228705686,
+      "learning_rate": 2.4240000000000002e-05,
+      "loss": 0.3456,
+      "step": 784
+    },
+    {
+      "epoch": 1.692556634304207,
+      "grad_norm": 0.192287940231301,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 0.3565,
+      "step": 785
+    },
+    {
+      "epoch": 1.6947141316073355,
+      "grad_norm": 0.2185166952318352,
+      "learning_rate": 2.4160000000000002e-05,
+      "loss": 0.3512,
+      "step": 786
+    },
+    {
+      "epoch": 1.696871628910464,
+      "grad_norm": 0.2096624106569505,
+      "learning_rate": 2.412e-05,
+      "loss": 0.3783,
+      "step": 787
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.21166468282471723,
+      "learning_rate": 2.408e-05,
+      "loss": 0.3748,
+      "step": 788
+    },
+    {
+      "epoch": 1.7011866235167206,
+      "grad_norm": 0.19873344543315882,
+      "learning_rate": 2.404e-05,
+      "loss": 0.3571,
+      "step": 789
+    },
+    {
+      "epoch": 1.703344120819849,
+      "grad_norm": 0.22110925981619345,
+      "learning_rate": 2.4e-05,
+      "loss": 0.3629,
+      "step": 790
+    },
+    {
+      "epoch": 1.7055016181229772,
+      "grad_norm": 0.2021215637434402,
+      "learning_rate": 2.396e-05,
+      "loss": 0.3716,
+      "step": 791
+    },
+    {
+      "epoch": 1.7076591154261056,
+      "grad_norm": 0.2111765564318759,
+      "learning_rate": 2.392e-05,
+      "loss": 0.3588,
+      "step": 792
+    },
+    {
+      "epoch": 1.709816612729234,
+      "grad_norm": 0.2223622417383566,
+      "learning_rate": 2.3880000000000002e-05,
+      "loss": 0.3536,
+      "step": 793
+    },
+    {
+      "epoch": 1.7119741100323624,
+      "grad_norm": 0.22406668807582533,
+      "learning_rate": 2.3840000000000002e-05,
+      "loss": 0.3596,
+      "step": 794
+    },
+    {
+      "epoch": 1.7141316073354909,
+      "grad_norm": 0.2047427267119515,
+      "learning_rate": 2.38e-05,
+      "loss": 0.362,
+      "step": 795
+    },
+    {
+      "epoch": 1.7162891046386193,
+      "grad_norm": 0.2487591943558242,
+      "learning_rate": 2.3760000000000003e-05,
+      "loss": 0.3929,
+      "step": 796
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.23436633449529568,
+      "learning_rate": 2.372e-05,
+      "loss": 0.3683,
+      "step": 797
+    },
+    {
+      "epoch": 1.720604099244876,
+      "grad_norm": 0.20926682274018532,
+      "learning_rate": 2.3680000000000004e-05,
+      "loss": 0.3553,
+      "step": 798
+    },
+    {
+      "epoch": 1.7227615965480043,
+      "grad_norm": 0.2084020707494825,
+      "learning_rate": 2.364e-05,
+      "loss": 0.3673,
+      "step": 799
+    },
+    {
+      "epoch": 1.7249190938511327,
+      "grad_norm": 0.21090087654277367,
+      "learning_rate": 2.36e-05,
+      "loss": 0.3491,
+      "step": 800
+    },
+    {
+      "epoch": 1.727076591154261,
+      "grad_norm": 0.21079538661713704,
+      "learning_rate": 2.356e-05,
+      "loss": 0.3617,
+      "step": 801
+    },
+    {
+      "epoch": 1.7292340884573894,
+      "grad_norm": 0.19785208126669376,
+      "learning_rate": 2.3520000000000002e-05,
+      "loss": 0.3734,
+      "step": 802
+    },
+    {
+      "epoch": 1.7313915857605178,
+      "grad_norm": 0.20051304335215855,
+      "learning_rate": 2.3480000000000002e-05,
+      "loss": 0.3561,
+      "step": 803
+    },
+    {
+      "epoch": 1.7335490830636462,
+      "grad_norm": 0.20991073713252967,
+      "learning_rate": 2.344e-05,
+      "loss": 0.3456,
+      "step": 804
+    },
+    {
+      "epoch": 1.7357065803667746,
+      "grad_norm": 0.19847852943956665,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.3636,
+      "step": 805
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 0.19228778135303196,
+      "learning_rate": 2.336e-05,
+      "loss": 0.3546,
+      "step": 806
+    },
+    {
+      "epoch": 1.7400215749730314,
+      "grad_norm": 0.19869096373294648,
+      "learning_rate": 2.332e-05,
+      "loss": 0.3535,
+      "step": 807
+    },
+    {
+      "epoch": 1.7421790722761596,
+      "grad_norm": 0.19598486859011863,
+      "learning_rate": 2.328e-05,
+      "loss": 0.3672,
+      "step": 808
+    },
+    {
+      "epoch": 1.744336569579288,
+      "grad_norm": 0.20021180617747214,
+      "learning_rate": 2.324e-05,
+      "loss": 0.3629,
+      "step": 809
+    },
+    {
+      "epoch": 1.7464940668824163,
+      "grad_norm": 0.19601550993847638,
+      "learning_rate": 2.32e-05,
+      "loss": 0.3661,
+      "step": 810
+    },
+    {
+      "epoch": 1.7486515641855447,
+      "grad_norm": 0.19629741169441461,
+      "learning_rate": 2.3160000000000002e-05,
+      "loss": 0.3568,
+      "step": 811
+    },
+    {
+      "epoch": 1.750809061488673,
+      "grad_norm": 0.2071662842083144,
+      "learning_rate": 2.312e-05,
+      "loss": 0.3813,
+      "step": 812
+    },
+    {
+      "epoch": 1.7529665587918015,
+      "grad_norm": 0.1966149798408275,
+      "learning_rate": 2.3080000000000003e-05,
+      "loss": 0.362,
+      "step": 813
+    },
+    {
+      "epoch": 1.75512405609493,
+      "grad_norm": 0.2279788745870227,
+      "learning_rate": 2.304e-05,
+      "loss": 0.364,
+      "step": 814
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.23247306079514818,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.3641,
+      "step": 815
+    },
+    {
+      "epoch": 1.7594390507011868,
+      "grad_norm": 0.19906860551229927,
+      "learning_rate": 2.296e-05,
+      "loss": 0.3428,
+      "step": 816
+    },
+    {
+      "epoch": 1.761596548004315,
+      "grad_norm": 0.2355677766543707,
+      "learning_rate": 2.292e-05,
+      "loss": 0.352,
+      "step": 817
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 0.2654888809339144,
+      "learning_rate": 2.288e-05,
+      "loss": 0.3766,
+      "step": 818
+    },
+    {
+      "epoch": 1.7659115426105716,
+      "grad_norm": 0.27302693742007983,
+      "learning_rate": 2.284e-05,
+      "loss": 0.3591,
+      "step": 819
+    },
+    {
+      "epoch": 1.7680690399137,
+      "grad_norm": 0.2179231966531101,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.3538,
+      "step": 820
+    },
+    {
+      "epoch": 1.7702265372168284,
+      "grad_norm": 0.22506397382520602,
+      "learning_rate": 2.2760000000000002e-05,
+      "loss": 0.3564,
+      "step": 821
+    },
+    {
+      "epoch": 1.7723840345199569,
+      "grad_norm": 0.25408157657001734,
+      "learning_rate": 2.2720000000000003e-05,
+      "loss": 0.3515,
+      "step": 822
+    },
+    {
+      "epoch": 1.7745415318230853,
+      "grad_norm": 0.21062218688731194,
+      "learning_rate": 2.268e-05,
+      "loss": 0.3651,
+      "step": 823
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 0.22725053193586056,
+      "learning_rate": 2.264e-05,
+      "loss": 0.3508,
+      "step": 824
+    },
+    {
+      "epoch": 1.778856526429342,
+      "grad_norm": 0.24341685689233888,
+      "learning_rate": 2.26e-05,
+      "loss": 0.3911,
+      "step": 825
+    },
+    {
+      "epoch": 1.7810140237324703,
+      "grad_norm": 0.28389285536430425,
+      "learning_rate": 2.256e-05,
+      "loss": 0.3767,
+      "step": 826
+    },
+    {
+      "epoch": 1.7831715210355987,
+      "grad_norm": 0.23070034042434137,
+      "learning_rate": 2.252e-05,
+      "loss": 0.3494,
+      "step": 827
+    },
+    {
+      "epoch": 1.785329018338727,
+      "grad_norm": 0.2706944129377681,
+      "learning_rate": 2.248e-05,
+      "loss": 0.354,
+      "step": 828
+    },
+    {
+      "epoch": 1.7874865156418553,
+      "grad_norm": 0.2610014419098948,
+      "learning_rate": 2.244e-05,
+      "loss": 0.3468,
+      "step": 829
+    },
+    {
+      "epoch": 1.7896440129449838,
+      "grad_norm": 0.2514731314471683,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.3629,
+      "step": 830
+    },
+    {
+      "epoch": 1.7918015102481122,
+      "grad_norm": 0.2633751369955124,
+      "learning_rate": 2.236e-05,
+      "loss": 0.3753,
+      "step": 831
+    },
+    {
+      "epoch": 1.7939590075512406,
+      "grad_norm": 0.25424651459631,
+      "learning_rate": 2.2320000000000003e-05,
+      "loss": 0.3416,
+      "step": 832
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.2250335462372855,
+      "learning_rate": 2.228e-05,
+      "loss": 0.3553,
+      "step": 833
+    },
+    {
+      "epoch": 1.7982740021574974,
+      "grad_norm": 0.2589022819689555,
+      "learning_rate": 2.224e-05,
+      "loss": 0.3748,
+      "step": 834
+    },
+    {
+      "epoch": 1.8004314994606256,
+      "grad_norm": 0.23838039943520098,
+      "learning_rate": 2.22e-05,
+      "loss": 0.3714,
+      "step": 835
+    },
+    {
+      "epoch": 1.802588996763754,
+      "grad_norm": 0.21958281718490288,
+      "learning_rate": 2.216e-05,
+      "loss": 0.3602,
+      "step": 836
+    },
+    {
+      "epoch": 1.8047464940668823,
+      "grad_norm": 0.22055042786226484,
+      "learning_rate": 2.212e-05,
+      "loss": 0.3542,
+      "step": 837
+    },
+    {
+      "epoch": 1.8069039913700107,
+      "grad_norm": 0.2178260520716557,
+      "learning_rate": 2.2080000000000002e-05,
+      "loss": 0.3643,
+      "step": 838
+    },
+    {
+      "epoch": 1.809061488673139,
+      "grad_norm": 0.22651969517407028,
+      "learning_rate": 2.2040000000000002e-05,
+      "loss": 0.367,
+      "step": 839
+    },
+    {
+      "epoch": 1.8112189859762675,
+      "grad_norm": 0.23148159521722647,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.3609,
+      "step": 840
+    },
+    {
+      "epoch": 1.813376483279396,
+      "grad_norm": 0.23027659960031496,
+      "learning_rate": 2.196e-05,
+      "loss": 0.3844,
+      "step": 841
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.21521666010080176,
+      "learning_rate": 2.192e-05,
+      "loss": 0.363,
+      "step": 842
+    },
+    {
+      "epoch": 1.8176914778856528,
+      "grad_norm": 0.23580779756655815,
+      "learning_rate": 2.188e-05,
+      "loss": 0.3719,
+      "step": 843
+    },
+    {
+      "epoch": 1.819848975188781,
+      "grad_norm": 0.24788937903653036,
+      "learning_rate": 2.184e-05,
+      "loss": 0.3586,
+      "step": 844
+    },
+    {
+      "epoch": 1.8220064724919094,
+      "grad_norm": 0.22247443155209357,
+      "learning_rate": 2.18e-05,
+      "loss": 0.3581,
+      "step": 845
+    },
+    {
+      "epoch": 1.8241639697950378,
+      "grad_norm": 0.23770870455599036,
+      "learning_rate": 2.176e-05,
+      "loss": 0.3614,
+      "step": 846
+    },
+    {
+      "epoch": 1.826321467098166,
+      "grad_norm": 0.22632423401344884,
+      "learning_rate": 2.1720000000000002e-05,
+      "loss": 0.3624,
+      "step": 847
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 0.2524920921986345,
+      "learning_rate": 2.168e-05,
+      "loss": 0.3675,
+      "step": 848
+    },
+    {
+      "epoch": 1.8306364617044228,
+      "grad_norm": 0.3215840056574261,
+      "learning_rate": 2.1640000000000003e-05,
+      "loss": 0.3538,
+      "step": 849
+    },
+    {
+      "epoch": 1.8327939590075513,
+      "grad_norm": 0.21486175758377893,
+      "learning_rate": 2.16e-05,
+      "loss": 0.3632,
+      "step": 850
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.2330894514205108,
+      "learning_rate": 2.1560000000000004e-05,
+      "loss": 0.341,
+      "step": 851
+    },
+    {
+      "epoch": 1.837108953613808,
+      "grad_norm": 0.24218781444766682,
+      "learning_rate": 2.152e-05,
+      "loss": 0.3677,
+      "step": 852
+    },
+    {
+      "epoch": 1.8392664509169363,
+      "grad_norm": 0.21963366595590614,
+      "learning_rate": 2.148e-05,
+      "loss": 0.3606,
+      "step": 853
+    },
+    {
+      "epoch": 1.8414239482200647,
+      "grad_norm": 0.21006868250745636,
+      "learning_rate": 2.144e-05,
+      "loss": 0.3489,
+      "step": 854
+    },
+    {
+      "epoch": 1.8435814455231931,
+      "grad_norm": 0.23174473879298813,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 0.3587,
+      "step": 855
+    },
+    {
+      "epoch": 1.8457389428263213,
+      "grad_norm": 0.2321069704400467,
+      "learning_rate": 2.1360000000000002e-05,
+      "loss": 0.3486,
+      "step": 856
+    },
+    {
+      "epoch": 1.8478964401294498,
+      "grad_norm": 0.20025509216353718,
+      "learning_rate": 2.1320000000000003e-05,
+      "loss": 0.3436,
+      "step": 857
+    },
+    {
+      "epoch": 1.8500539374325782,
+      "grad_norm": 0.22250953140085006,
+      "learning_rate": 2.128e-05,
+      "loss": 0.3603,
+      "step": 858
+    },
+    {
+      "epoch": 1.8522114347357066,
+      "grad_norm": 0.2173021358838602,
+      "learning_rate": 2.124e-05,
+      "loss": 0.3467,
+      "step": 859
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.1930203129733309,
+      "learning_rate": 2.12e-05,
+      "loss": 0.3615,
+      "step": 860
+    },
+    {
+      "epoch": 1.8565264293419634,
+      "grad_norm": 0.19862427723542597,
+      "learning_rate": 2.116e-05,
+      "loss": 0.3579,
+      "step": 861
+    },
+    {
+      "epoch": 1.8586839266450919,
+      "grad_norm": 0.2096060299104676,
+      "learning_rate": 2.112e-05,
+      "loss": 0.3517,
+      "step": 862
+    },
+    {
+      "epoch": 1.86084142394822,
+      "grad_norm": 0.25983349292984365,
+      "learning_rate": 2.1079999999999998e-05,
+      "loss": 0.3469,
+      "step": 863
+    },
+    {
+      "epoch": 1.8629989212513485,
+      "grad_norm": 0.19568780893613127,
+      "learning_rate": 2.1040000000000002e-05,
+      "loss": 0.3586,
+      "step": 864
+    },
+    {
+      "epoch": 1.8651564185544767,
+      "grad_norm": 0.20440754302808392,
+      "learning_rate": 2.1e-05,
+      "loss": 0.3454,
+      "step": 865
+    },
+    {
+      "epoch": 1.867313915857605,
+      "grad_norm": 0.2617960410505945,
+      "learning_rate": 2.0960000000000003e-05,
+      "loss": 0.3722,
+      "step": 866
+    },
+    {
+      "epoch": 1.8694714131607335,
+      "grad_norm": 0.21484645618030715,
+      "learning_rate": 2.092e-05,
+      "loss": 0.363,
+      "step": 867
+    },
+    {
+      "epoch": 1.871628910463862,
+      "grad_norm": 0.2089417062699798,
+      "learning_rate": 2.0880000000000003e-05,
+      "loss": 0.3575,
+      "step": 868
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.21870690017344033,
+      "learning_rate": 2.084e-05,
+      "loss": 0.3441,
+      "step": 869
+    },
+    {
+      "epoch": 1.8759439050701188,
+      "grad_norm": 0.2015916258019001,
+      "learning_rate": 2.08e-05,
+      "loss": 0.3649,
+      "step": 870
+    },
+    {
+      "epoch": 1.8781014023732472,
+      "grad_norm": 0.20445394487249433,
+      "learning_rate": 2.076e-05,
+      "loss": 0.3633,
+      "step": 871
+    },
+    {
+      "epoch": 1.8802588996763754,
+      "grad_norm": 0.21933907619881754,
+      "learning_rate": 2.072e-05,
+      "loss": 0.3565,
+      "step": 872
+    },
+    {
+      "epoch": 1.8824163969795038,
+      "grad_norm": 0.22898545335887432,
+      "learning_rate": 2.0680000000000002e-05,
+      "loss": 0.3489,
+      "step": 873
+    },
+    {
+      "epoch": 1.884573894282632,
+      "grad_norm": 0.19190043665285303,
+      "learning_rate": 2.0640000000000002e-05,
+      "loss": 0.3811,
+      "step": 874
+    },
+    {
+      "epoch": 1.8867313915857604,
+      "grad_norm": 0.22405276330574816,
+      "learning_rate": 2.06e-05,
+      "loss": 0.3514,
+      "step": 875
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.20664352560729662,
+      "learning_rate": 2.0560000000000003e-05,
+      "loss": 0.3564,
+      "step": 876
+    },
+    {
+      "epoch": 1.8910463861920173,
+      "grad_norm": 0.19849201500058444,
+      "learning_rate": 2.052e-05,
+      "loss": 0.3791,
+      "step": 877
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.21514720161140888,
+      "learning_rate": 2.048e-05,
+      "loss": 0.343,
+      "step": 878
+    },
+    {
+      "epoch": 1.895361380798274,
+      "grad_norm": 0.19591290106236087,
+      "learning_rate": 2.044e-05,
+      "loss": 0.3617,
+      "step": 879
+    },
+    {
+      "epoch": 1.8975188781014025,
+      "grad_norm": 0.20056100252045797,
+      "learning_rate": 2.04e-05,
+      "loss": 0.3692,
+      "step": 880
+    },
+    {
+      "epoch": 1.8996763754045307,
+      "grad_norm": 0.6263284386911345,
+      "learning_rate": 2.036e-05,
+      "loss": 0.3819,
+      "step": 881
+    },
+    {
+      "epoch": 1.9018338727076591,
+      "grad_norm": 0.23922694733352964,
+      "learning_rate": 2.032e-05,
+      "loss": 0.3643,
+      "step": 882
+    },
+    {
+      "epoch": 1.9039913700107873,
+      "grad_norm": 0.19781471120361346,
+      "learning_rate": 2.0280000000000002e-05,
+      "loss": 0.3611,
+      "step": 883
+    },
+    {
+      "epoch": 1.9061488673139158,
+      "grad_norm": 0.21165098306207256,
+      "learning_rate": 2.024e-05,
+      "loss": 0.3529,
+      "step": 884
+    },
+    {
+      "epoch": 1.9083063646170442,
+      "grad_norm": 0.21721904598543817,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 0.3492,
+      "step": 885
+    },
+    {
+      "epoch": 1.9104638619201726,
+      "grad_norm": 0.21422381596626322,
+      "learning_rate": 2.016e-05,
+      "loss": 0.3602,
+      "step": 886
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.20452655904024958,
+      "learning_rate": 2.012e-05,
+      "loss": 0.3564,
+      "step": 887
+    },
+    {
+      "epoch": 1.9147788565264294,
+      "grad_norm": 0.21580439136445786,
+      "learning_rate": 2.008e-05,
+      "loss": 0.3711,
+      "step": 888
+    },
+    {
+      "epoch": 1.9169363538295578,
+      "grad_norm": 0.19423056851782128,
+      "learning_rate": 2.004e-05,
+      "loss": 0.3701,
+      "step": 889
+    },
+    {
+      "epoch": 1.919093851132686,
+      "grad_norm": 0.20048302260105516,
+      "learning_rate": 2e-05,
+      "loss": 0.3672,
+      "step": 890
+    },
+    {
+      "epoch": 1.9212513484358145,
+      "grad_norm": 0.2117075068441905,
+      "learning_rate": 1.9960000000000002e-05,
+      "loss": 0.3584,
+      "step": 891
+    },
+    {
+      "epoch": 1.9234088457389427,
+      "grad_norm": 0.20275015688207176,
+      "learning_rate": 1.992e-05,
+      "loss": 0.3722,
+      "step": 892
+    },
+    {
+      "epoch": 1.925566343042071,
+      "grad_norm": 0.18554538922099112,
+      "learning_rate": 1.9880000000000003e-05,
+      "loss": 0.3688,
+      "step": 893
+    },
+    {
+      "epoch": 1.9277238403451995,
+      "grad_norm": 0.2264891336459159,
+      "learning_rate": 1.984e-05,
+      "loss": 0.3516,
+      "step": 894
+    },
+    {
+      "epoch": 1.929881337648328,
+      "grad_norm": 0.20419289874494262,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.368,
+      "step": 895
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.2253379202949314,
+      "learning_rate": 1.976e-05,
+      "loss": 0.3684,
+      "step": 896
+    },
+    {
+      "epoch": 1.9341963322545848,
+      "grad_norm": 0.21080590277290626,
+      "learning_rate": 1.972e-05,
+      "loss": 0.3559,
+      "step": 897
+    },
+    {
+      "epoch": 1.9363538295577132,
+      "grad_norm": 0.18855517288473167,
+      "learning_rate": 1.968e-05,
+      "loss": 0.3683,
+      "step": 898
+    },
+    {
+      "epoch": 1.9385113268608414,
+      "grad_norm": 0.22336015878998725,
+      "learning_rate": 1.9640000000000002e-05,
+      "loss": 0.3501,
+      "step": 899
+    },
+    {
+      "epoch": 1.9406688241639698,
+      "grad_norm": 0.20681307053495754,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.3339,
+      "step": 900
+    },
+    {
+      "epoch": 1.9428263214670982,
+      "grad_norm": 0.21956872185132326,
+      "learning_rate": 1.956e-05,
+      "loss": 0.3643,
+      "step": 901
+    },
+    {
+      "epoch": 1.9449838187702264,
+      "grad_norm": 0.22995896123381177,
+      "learning_rate": 1.9520000000000003e-05,
+      "loss": 0.3539,
+      "step": 902
+    },
+    {
+      "epoch": 1.9471413160733548,
+      "grad_norm": 0.24289096552867204,
+      "learning_rate": 1.948e-05,
+      "loss": 0.3529,
+      "step": 903
+    },
+    {
+      "epoch": 1.9492988133764833,
+      "grad_norm": 0.18977445221480282,
+      "learning_rate": 1.944e-05,
+      "loss": 0.3761,
+      "step": 904
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.19537181851448127,
+      "learning_rate": 1.94e-05,
+      "loss": 0.3776,
+      "step": 905
+    },
+    {
+      "epoch": 1.95361380798274,
+      "grad_norm": 0.21168864149685598,
+      "learning_rate": 1.936e-05,
+      "loss": 0.3564,
+      "step": 906
+    },
+    {
+      "epoch": 1.9557713052858685,
+      "grad_norm": 0.21082909120102009,
+      "learning_rate": 1.932e-05,
+      "loss": 0.3677,
+      "step": 907
+    },
+    {
+      "epoch": 1.9579288025889967,
+      "grad_norm": 0.2092667221652132,
+      "learning_rate": 1.9280000000000002e-05,
+      "loss": 0.3666,
+      "step": 908
+    },
+    {
+      "epoch": 1.9600862998921251,
+      "grad_norm": 0.28440317520339464,
+      "learning_rate": 1.924e-05,
+      "loss": 0.3616,
+      "step": 909
+    },
+    {
+      "epoch": 1.9622437971952535,
+      "grad_norm": 0.21372393833816974,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.3724,
+      "step": 910
+    },
+    {
+      "epoch": 1.9644012944983817,
+      "grad_norm": 0.21177604197905148,
+      "learning_rate": 1.916e-05,
+      "loss": 0.3547,
+      "step": 911
+    },
+    {
+      "epoch": 1.9665587918015102,
+      "grad_norm": 0.22087482780256842,
+      "learning_rate": 1.9120000000000003e-05,
+      "loss": 0.3552,
+      "step": 912
+    },
+    {
+      "epoch": 1.9687162891046386,
+      "grad_norm": 0.2293776507660548,
+      "learning_rate": 1.908e-05,
+      "loss": 0.3539,
+      "step": 913
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.20763279035660182,
+      "learning_rate": 1.904e-05,
+      "loss": 0.3609,
+      "step": 914
+    },
+    {
+      "epoch": 1.9730312837108954,
+      "grad_norm": 0.2130132604223223,
+      "learning_rate": 1.9e-05,
+      "loss": 0.387,
+      "step": 915
+    },
+    {
+      "epoch": 1.9751887810140238,
+      "grad_norm": 0.22693796782628164,
+      "learning_rate": 1.896e-05,
+      "loss": 0.3632,
+      "step": 916
+    },
+    {
+      "epoch": 1.9773462783171523,
+      "grad_norm": 0.2075508453226733,
+      "learning_rate": 1.8920000000000002e-05,
+      "loss": 0.3504,
+      "step": 917
+    },
+    {
+      "epoch": 1.9795037756202805,
+      "grad_norm": 0.21670939076491838,
+      "learning_rate": 1.888e-05,
+      "loss": 0.3573,
+      "step": 918
+    },
+    {
+      "epoch": 1.9816612729234089,
+      "grad_norm": 0.20202624393739851,
+      "learning_rate": 1.8840000000000003e-05,
+      "loss": 0.3549,
+      "step": 919
+    },
+    {
+      "epoch": 1.983818770226537,
+      "grad_norm": 0.20167979698326996,
+      "learning_rate": 1.88e-05,
+      "loss": 0.3672,
+      "step": 920
+    },
+    {
+      "epoch": 1.9859762675296655,
+      "grad_norm": 0.1948573998079774,
+      "learning_rate": 1.876e-05,
+      "loss": 0.355,
+      "step": 921
+    },
+    {
+      "epoch": 1.988133764832794,
+      "grad_norm": 0.22586589053280912,
+      "learning_rate": 1.872e-05,
+      "loss": 0.3892,
+      "step": 922
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.21193504103673555,
+      "learning_rate": 1.868e-05,
+      "loss": 0.3573,
+      "step": 923
+    },
+    {
+      "epoch": 1.9924487594390508,
+      "grad_norm": 0.19869474028310133,
+      "learning_rate": 1.864e-05,
+      "loss": 0.3524,
+      "step": 924
+    },
+    {
+      "epoch": 1.9946062567421792,
+      "grad_norm": 0.21694327316611497,
+      "learning_rate": 1.86e-05,
+      "loss": 0.3623,
+      "step": 925
+    },
+    {
+      "epoch": 1.9967637540453076,
+      "grad_norm": 0.19165371716250784,
+      "learning_rate": 1.856e-05,
+      "loss": 0.3674,
+      "step": 926
+    },
+    {
+      "epoch": 1.9989212513484358,
+      "grad_norm": 0.25911714915290096,
+      "learning_rate": 1.8520000000000002e-05,
+      "loss": 0.3711,
+      "step": 927
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3166913102001167,
+      "learning_rate": 1.848e-05,
+      "loss": 0.3554,
+      "step": 928
+    },
+    {
+      "epoch": 2.0021574973031284,
+      "grad_norm": 0.2952601206226685,
+      "learning_rate": 1.8440000000000003e-05,
+      "loss": 0.2966,
+      "step": 929
+    },
+    {
+      "epoch": 2.004314994606257,
+      "grad_norm": 0.2737583904942107,
+      "learning_rate": 1.84e-05,
+      "loss": 0.2773,
+      "step": 930
+    },
+    {
+      "epoch": 2.0064724919093853,
+      "grad_norm": 0.2883228038451365,
+      "learning_rate": 1.8360000000000004e-05,
+      "loss": 0.2972,
+      "step": 931
+    },
+    {
+      "epoch": 2.0086299892125137,
+      "grad_norm": 0.3045158995400534,
+      "learning_rate": 1.832e-05,
+      "loss": 0.2912,
+      "step": 932
+    },
+    {
+      "epoch": 2.0107874865156417,
+      "grad_norm": 0.21425516278784199,
+      "learning_rate": 1.828e-05,
+      "loss": 0.2797,
+      "step": 933
+    },
+    {
+      "epoch": 2.01294498381877,
+      "grad_norm": 0.26815968591614264,
+      "learning_rate": 1.824e-05,
+      "loss": 0.2899,
+      "step": 934
+    },
+    {
+      "epoch": 2.0151024811218985,
+      "grad_norm": 0.2525848880526318,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.2754,
+      "step": 935
+    },
+    {
+      "epoch": 2.017259978425027,
+      "grad_norm": 0.22940718108680191,
+      "learning_rate": 1.8160000000000002e-05,
+      "loss": 0.2739,
+      "step": 936
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.2930727493312503,
+      "learning_rate": 1.812e-05,
+      "loss": 0.2928,
+      "step": 937
+    },
+    {
+      "epoch": 2.0215749730312838,
+      "grad_norm": 0.2730010795221549,
+      "learning_rate": 1.808e-05,
+      "loss": 0.2799,
+      "step": 938
+    },
+    {
+      "epoch": 2.023732470334412,
+      "grad_norm": 0.21843464575067598,
+      "learning_rate": 1.804e-05,
+      "loss": 0.2759,
+      "step": 939
+    },
+    {
+      "epoch": 2.0258899676375406,
+      "grad_norm": 0.2287920671205098,
+      "learning_rate": 1.8e-05,
+      "loss": 0.2686,
+      "step": 940
+    },
+    {
+      "epoch": 2.028047464940669,
+      "grad_norm": 0.24632463641906233,
+      "learning_rate": 1.796e-05,
+      "loss": 0.2901,
+      "step": 941
+    },
+    {
+      "epoch": 2.030204962243797,
+      "grad_norm": 0.21554911391559797,
+      "learning_rate": 1.792e-05,
+      "loss": 0.2646,
+      "step": 942
+    },
+    {
+      "epoch": 2.0323624595469254,
+      "grad_norm": 0.23974674214671038,
+      "learning_rate": 1.7879999999999998e-05,
+      "loss": 0.2865,
+      "step": 943
+    },
+    {
+      "epoch": 2.034519956850054,
+      "grad_norm": 0.23389305666437907,
+      "learning_rate": 1.7840000000000002e-05,
+      "loss": 0.2659,
+      "step": 944
+    },
+    {
+      "epoch": 2.0366774541531822,
+      "grad_norm": 0.23454932706996612,
+      "learning_rate": 1.78e-05,
+      "loss": 0.2889,
+      "step": 945
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.2110811652752538,
+      "learning_rate": 1.7760000000000003e-05,
+      "loss": 0.2715,
+      "step": 946
+    },
+    {
+      "epoch": 2.040992448759439,
+      "grad_norm": 0.2086924969772193,
+      "learning_rate": 1.772e-05,
+      "loss": 0.2827,
+      "step": 947
+    },
+    {
+      "epoch": 2.0431499460625675,
+      "grad_norm": 0.2178166620120031,
+      "learning_rate": 1.7680000000000004e-05,
+      "loss": 0.2823,
+      "step": 948
+    },
+    {
+      "epoch": 2.045307443365696,
+      "grad_norm": 0.21332967807788294,
+      "learning_rate": 1.764e-05,
+      "loss": 0.2814,
+      "step": 949
+    },
+    {
+      "epoch": 2.0474649406688243,
+      "grad_norm": 0.21111966487487638,
+      "learning_rate": 1.76e-05,
+      "loss": 0.278,
+      "step": 950
+    },
+    {
+      "epoch": 2.0496224379719523,
+      "grad_norm": 0.20634749644599865,
+      "learning_rate": 1.756e-05,
+      "loss": 0.2862,
+      "step": 951
+    },
+    {
+      "epoch": 2.0517799352750807,
+      "grad_norm": 0.20256588641450451,
+      "learning_rate": 1.752e-05,
+      "loss": 0.2754,
+      "step": 952
+    },
+    {
+      "epoch": 2.053937432578209,
+      "grad_norm": 0.21546847922922335,
+      "learning_rate": 1.7480000000000002e-05,
+      "loss": 0.2859,
+      "step": 953
+    },
+    {
+      "epoch": 2.0560949298813376,
+      "grad_norm": 0.2012891240538272,
+      "learning_rate": 1.7440000000000002e-05,
+      "loss": 0.2667,
+      "step": 954
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 0.21828886379868126,
+      "learning_rate": 1.74e-05,
+      "loss": 0.2768,
+      "step": 955
+    },
+    {
+      "epoch": 2.0604099244875944,
+      "grad_norm": 0.23319314864900412,
+      "learning_rate": 1.736e-05,
+      "loss": 0.2712,
+      "step": 956
+    },
+    {
+      "epoch": 2.062567421790723,
+      "grad_norm": 0.21570652212000496,
+      "learning_rate": 1.732e-05,
+      "loss": 0.2704,
+      "step": 957
+    },
+    {
+      "epoch": 2.0647249190938513,
+      "grad_norm": 0.2050136511501782,
+      "learning_rate": 1.728e-05,
+      "loss": 0.2768,
+      "step": 958
+    },
+    {
+      "epoch": 2.0668824163969797,
+      "grad_norm": 0.19908285569355766,
+      "learning_rate": 1.724e-05,
+      "loss": 0.2699,
+      "step": 959
+    },
+    {
+      "epoch": 2.0690399137001076,
+      "grad_norm": 0.22341990371652531,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.2802,
+      "step": 960
+    },
+    {
+      "epoch": 2.071197411003236,
+      "grad_norm": 0.21245912373702183,
+      "learning_rate": 1.7160000000000002e-05,
+      "loss": 0.2732,
+      "step": 961
+    },
+    {
+      "epoch": 2.0733549083063645,
+      "grad_norm": 0.2315866528262126,
+      "learning_rate": 1.712e-05,
+      "loss": 0.2666,
+      "step": 962
+    },
+    {
+      "epoch": 2.075512405609493,
+      "grad_norm": 0.21250700843322592,
+      "learning_rate": 1.7080000000000002e-05,
+      "loss": 0.287,
+      "step": 963
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 0.20591733874775808,
+      "learning_rate": 1.704e-05,
+      "loss": 0.2805,
+      "step": 964
+    },
+    {
+      "epoch": 2.0798274002157497,
+      "grad_norm": 0.20270061570174747,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.2724,
+      "step": 965
+    },
+    {
+      "epoch": 2.081984897518878,
+      "grad_norm": 0.22011780420584517,
+      "learning_rate": 1.696e-05,
+      "loss": 0.27,
+      "step": 966
+    },
+    {
+      "epoch": 2.0841423948220066,
+      "grad_norm": 0.2086945807906332,
+      "learning_rate": 1.692e-05,
+      "loss": 0.2652,
+      "step": 967
+    },
+    {
+      "epoch": 2.086299892125135,
+      "grad_norm": 0.2017480638957166,
+      "learning_rate": 1.688e-05,
+      "loss": 0.2807,
+      "step": 968
+    },
+    {
+      "epoch": 2.0884573894282634,
+      "grad_norm": 0.23221771597231639,
+      "learning_rate": 1.684e-05,
+      "loss": 0.2908,
+      "step": 969
+    },
+    {
+      "epoch": 2.0906148867313914,
+      "grad_norm": 0.19787472936299969,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.2602,
+      "step": 970
+    },
+    {
+      "epoch": 2.09277238403452,
+      "grad_norm": 0.1911068238953241,
+      "learning_rate": 1.6760000000000002e-05,
+      "loss": 0.262,
+      "step": 971
+    },
+    {
+      "epoch": 2.0949298813376482,
+      "grad_norm": 0.21404892409121634,
+      "learning_rate": 1.672e-05,
+      "loss": 0.2883,
+      "step": 972
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.19651822000095875,
+      "learning_rate": 1.668e-05,
+      "loss": 0.2787,
+      "step": 973
+    },
+    {
+      "epoch": 2.099244875943905,
+      "grad_norm": 0.19629483862396524,
+      "learning_rate": 1.664e-05,
+      "loss": 0.2682,
+      "step": 974
+    },
+    {
+      "epoch": 2.1014023732470335,
+      "grad_norm": 0.195771507449239,
+      "learning_rate": 1.66e-05,
+      "loss": 0.2595,
+      "step": 975
+    },
+    {
+      "epoch": 2.103559870550162,
+      "grad_norm": 0.2012720434683137,
+      "learning_rate": 1.656e-05,
+      "loss": 0.2924,
+      "step": 976
+    },
+    {
+      "epoch": 2.1057173678532903,
+      "grad_norm": 0.20022994629900823,
+      "learning_rate": 1.652e-05,
+      "loss": 0.2623,
+      "step": 977
+    },
+    {
+      "epoch": 2.1078748651564188,
+      "grad_norm": 0.1988415836362952,
+      "learning_rate": 1.648e-05,
+      "loss": 0.2665,
+      "step": 978
+    },
+    {
+      "epoch": 2.1100323624595467,
+      "grad_norm": 0.20067992175054306,
+      "learning_rate": 1.644e-05,
+      "loss": 0.2708,
+      "step": 979
+    },
+    {
+      "epoch": 2.112189859762675,
+      "grad_norm": 0.1965836467645065,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.2781,
+      "step": 980
+    },
+    {
+      "epoch": 2.1143473570658036,
+      "grad_norm": 0.2106067939768535,
+      "learning_rate": 1.636e-05,
+      "loss": 0.2928,
+      "step": 981
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 0.1912874552435929,
+      "learning_rate": 1.6320000000000003e-05,
+      "loss": 0.2653,
+      "step": 982
+    },
+    {
+      "epoch": 2.1186623516720604,
+      "grad_norm": 0.21480216660403667,
+      "learning_rate": 1.628e-05,
+      "loss": 0.2795,
+      "step": 983
+    },
+    {
+      "epoch": 2.120819848975189,
+      "grad_norm": 0.19894898831802282,
+      "learning_rate": 1.624e-05,
+      "loss": 0.2723,
+      "step": 984
+    },
+    {
+      "epoch": 2.1229773462783172,
+      "grad_norm": 0.18868913838930684,
+      "learning_rate": 1.62e-05,
+      "loss": 0.2701,
+      "step": 985
+    },
+    {
+      "epoch": 2.1251348435814457,
+      "grad_norm": 0.22014773725924427,
+      "learning_rate": 1.616e-05,
+      "loss": 0.2956,
+      "step": 986
+    },
+    {
+      "epoch": 2.127292340884574,
+      "grad_norm": 0.2161914198087177,
+      "learning_rate": 1.612e-05,
+      "loss": 0.2694,
+      "step": 987
+    },
+    {
+      "epoch": 2.129449838187702,
+      "grad_norm": 0.18693419466428068,
+      "learning_rate": 1.6080000000000002e-05,
+      "loss": 0.2722,
+      "step": 988
+    },
+    {
+      "epoch": 2.1316073354908305,
+      "grad_norm": 0.19845446294819882,
+      "learning_rate": 1.604e-05,
+      "loss": 0.2699,
+      "step": 989
+    },
+    {
+      "epoch": 2.133764832793959,
+      "grad_norm": 0.2243502306601146,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2898,
+      "step": 990
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.189061438536264,
+      "learning_rate": 1.596e-05,
+      "loss": 0.2613,
+      "step": 991
+    },
+    {
+      "epoch": 2.1380798274002157,
+      "grad_norm": 0.196308108347633,
+      "learning_rate": 1.592e-05,
+      "loss": 0.2802,
+      "step": 992
+    },
+    {
+      "epoch": 2.140237324703344,
+      "grad_norm": 0.20505568263166546,
+      "learning_rate": 1.588e-05,
+      "loss": 0.2754,
+      "step": 993
+    },
+    {
+      "epoch": 2.1423948220064726,
+      "grad_norm": 0.19371375505033042,
+      "learning_rate": 1.584e-05,
+      "loss": 0.2797,
+      "step": 994
+    },
+    {
+      "epoch": 2.144552319309601,
+      "grad_norm": 0.2038159697548706,
+      "learning_rate": 1.58e-05,
+      "loss": 0.2748,
+      "step": 995
+    },
+    {
+      "epoch": 2.1467098166127294,
+      "grad_norm": 0.20104064716873768,
+      "learning_rate": 1.5759999999999998e-05,
+      "loss": 0.2517,
+      "step": 996
+    },
+    {
+      "epoch": 2.148867313915858,
+      "grad_norm": 0.1993147823205055,
+      "learning_rate": 1.5720000000000002e-05,
+      "loss": 0.2806,
+      "step": 997
+    },
+    {
+      "epoch": 2.151024811218986,
+      "grad_norm": 0.23201462938785924,
+      "learning_rate": 1.568e-05,
+      "loss": 0.2843,
+      "step": 998
+    },
+    {
+      "epoch": 2.1531823085221142,
+      "grad_norm": 0.194482970909287,
+      "learning_rate": 1.5640000000000003e-05,
+      "loss": 0.2704,
+      "step": 999
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.19741489729000775,
+      "learning_rate": 1.56e-05,
+      "loss": 0.2868,
+      "step": 1000
+    },
+    {
+      "epoch": 2.157497303128371,
+      "grad_norm": 0.19338218576211613,
+      "learning_rate": 1.556e-05,
+      "loss": 0.2807,
+      "step": 1001
+    },
+    {
+      "epoch": 2.1596548004314995,
+      "grad_norm": 0.18815117551278576,
+      "learning_rate": 1.552e-05,
+      "loss": 0.2865,
+      "step": 1002
+    },
+    {
+      "epoch": 2.161812297734628,
+      "grad_norm": 0.19320001320317376,
+      "learning_rate": 1.548e-05,
+      "loss": 0.2893,
+      "step": 1003
+    },
+    {
+      "epoch": 2.1639697950377563,
+      "grad_norm": 0.19912625340767592,
+      "learning_rate": 1.544e-05,
+      "loss": 0.2861,
+      "step": 1004
+    },
+    {
+      "epoch": 2.1661272923408847,
+      "grad_norm": 0.20055637118837155,
+      "learning_rate": 1.54e-05,
+      "loss": 0.2832,
+      "step": 1005
+    },
+    {
+      "epoch": 2.168284789644013,
+      "grad_norm": 0.18881522044079282,
+      "learning_rate": 1.536e-05,
+      "loss": 0.2735,
+      "step": 1006
+    },
+    {
+      "epoch": 2.170442286947141,
+      "grad_norm": 0.19424371584292754,
+      "learning_rate": 1.5320000000000002e-05,
+      "loss": 0.2741,
+      "step": 1007
+    },
+    {
+      "epoch": 2.1725997842502696,
+      "grad_norm": 0.19749981503516376,
+      "learning_rate": 1.528e-05,
+      "loss": 0.2814,
+      "step": 1008
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.21172397165526702,
+      "learning_rate": 1.5240000000000001e-05,
+      "loss": 0.2821,
+      "step": 1009
+    },
+    {
+      "epoch": 2.1769147788565264,
+      "grad_norm": 0.2008055699713837,
+      "learning_rate": 1.52e-05,
+      "loss": 0.2772,
+      "step": 1010
+    },
+    {
+      "epoch": 2.179072276159655,
+      "grad_norm": 0.18785347695619867,
+      "learning_rate": 1.5160000000000002e-05,
+      "loss": 0.2766,
+      "step": 1011
+    },
+    {
+      "epoch": 2.1812297734627832,
+      "grad_norm": 0.1963392396504042,
+      "learning_rate": 1.5120000000000001e-05,
+      "loss": 0.2724,
+      "step": 1012
+    },
+    {
+      "epoch": 2.1833872707659117,
+      "grad_norm": 0.20123664109137754,
+      "learning_rate": 1.508e-05,
+      "loss": 0.2756,
+      "step": 1013
+    },
+    {
+      "epoch": 2.18554476806904,
+      "grad_norm": 0.20444367434824906,
+      "learning_rate": 1.5040000000000002e-05,
+      "loss": 0.2924,
+      "step": 1014
+    },
+    {
+      "epoch": 2.1877022653721685,
+      "grad_norm": 0.1853221427084223,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2647,
+      "step": 1015
+    },
+    {
+      "epoch": 2.1898597626752965,
+      "grad_norm": 0.19532713814867492,
+      "learning_rate": 1.4960000000000002e-05,
+      "loss": 0.2885,
+      "step": 1016
+    },
+    {
+      "epoch": 2.192017259978425,
+      "grad_norm": 0.194567249405803,
+      "learning_rate": 1.4920000000000001e-05,
+      "loss": 0.2751,
+      "step": 1017
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 0.19156850642372739,
+      "learning_rate": 1.488e-05,
+      "loss": 0.2902,
+      "step": 1018
+    },
+    {
+      "epoch": 2.1963322545846817,
+      "grad_norm": 0.19068854832443224,
+      "learning_rate": 1.4840000000000002e-05,
+      "loss": 0.2785,
+      "step": 1019
+    },
+    {
+      "epoch": 2.19848975188781,
+      "grad_norm": 0.1907888424930383,
+      "learning_rate": 1.48e-05,
+      "loss": 0.2833,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2006472491909386,
+      "grad_norm": 0.2022398514747435,
+      "learning_rate": 1.4760000000000001e-05,
+      "loss": 0.2812,
+      "step": 1021
+    },
+    {
+      "epoch": 2.202804746494067,
+      "grad_norm": 0.20650152647961673,
+      "learning_rate": 1.472e-05,
+      "loss": 0.2808,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2049622437971954,
+      "grad_norm": 0.19783560549543305,
+      "learning_rate": 1.4680000000000002e-05,
+      "loss": 0.2835,
+      "step": 1023
+    },
+    {
+      "epoch": 2.207119741100324,
+      "grad_norm": 0.19791637326958472,
+      "learning_rate": 1.464e-05,
+      "loss": 0.2686,
+      "step": 1024
+    },
+    {
+      "epoch": 2.209277238403452,
+      "grad_norm": 0.19785476348240078,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.2763,
+      "step": 1025
+    },
+    {
+      "epoch": 2.2114347357065802,
+      "grad_norm": 0.2200012984407047,
+      "learning_rate": 1.4560000000000001e-05,
+      "loss": 0.2788,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.23124340363606444,
+      "learning_rate": 1.452e-05,
+      "loss": 0.278,
+      "step": 1027
+    },
+    {
+      "epoch": 2.215749730312837,
+      "grad_norm": 0.19294677227886092,
+      "learning_rate": 1.4480000000000002e-05,
+      "loss": 0.2662,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2179072276159655,
+      "grad_norm": 0.21467839292037788,
+      "learning_rate": 1.444e-05,
+      "loss": 0.293,
+      "step": 1029
+    },
+    {
+      "epoch": 2.220064724919094,
+      "grad_norm": 0.20226064234278612,
+      "learning_rate": 1.44e-05,
+      "loss": 0.2808,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.19651529866983977,
+      "learning_rate": 1.4360000000000001e-05,
+      "loss": 0.2776,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2243797195253507,
+      "grad_norm": 0.19847228391981836,
+      "learning_rate": 1.432e-05,
+      "loss": 0.286,
+      "step": 1032
+    },
+    {
+      "epoch": 2.226537216828479,
+      "grad_norm": 0.21369656352490973,
+      "learning_rate": 1.4280000000000002e-05,
+      "loss": 0.2782,
+      "step": 1033
+    },
+    {
+      "epoch": 2.228694714131607,
+      "grad_norm": 0.1919725559755518,
+      "learning_rate": 1.4240000000000001e-05,
+      "loss": 0.2775,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2308522114347356,
+      "grad_norm": 0.19131850346902102,
+      "learning_rate": 1.42e-05,
+      "loss": 0.2657,
+      "step": 1035
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.19442842880711014,
+      "learning_rate": 1.4160000000000002e-05,
+      "loss": 0.2724,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2351672060409924,
+      "grad_norm": 0.1981714390416587,
+      "learning_rate": 1.412e-05,
+      "loss": 0.285,
+      "step": 1037
+    },
+    {
+      "epoch": 2.237324703344121,
+      "grad_norm": 0.20521747405247104,
+      "learning_rate": 1.408e-05,
+      "loss": 0.266,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2394822006472492,
+      "grad_norm": 0.20356927339863315,
+      "learning_rate": 1.4040000000000001e-05,
+      "loss": 0.287,
+      "step": 1039
+    },
+    {
+      "epoch": 2.2416396979503777,
+      "grad_norm": 0.19892125410866876,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.2835,
+      "step": 1040
+    },
+    {
+      "epoch": 2.243797195253506,
+      "grad_norm": 0.19475194941115584,
+      "learning_rate": 1.396e-05,
+      "loss": 0.2804,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2459546925566345,
+      "grad_norm": 0.19481027849756127,
+      "learning_rate": 1.3919999999999999e-05,
+      "loss": 0.2867,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2481121898597625,
+      "grad_norm": 0.20053383340444886,
+      "learning_rate": 1.3880000000000001e-05,
+      "loss": 0.2796,
+      "step": 1043
+    },
+    {
+      "epoch": 2.250269687162891,
+      "grad_norm": 0.18996447992057114,
+      "learning_rate": 1.384e-05,
+      "loss": 0.2694,
+      "step": 1044
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.19590330294229358,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.2792,
+      "step": 1045
+    },
+    {
+      "epoch": 2.2545846817691477,
+      "grad_norm": 0.20196142682316945,
+      "learning_rate": 1.376e-05,
+      "loss": 0.2781,
+      "step": 1046
+    },
+    {
+      "epoch": 2.256742179072276,
+      "grad_norm": 0.18937385936130618,
+      "learning_rate": 1.3719999999999999e-05,
+      "loss": 0.2765,
+      "step": 1047
+    },
+    {
+      "epoch": 2.2588996763754046,
+      "grad_norm": 0.20180588042942285,
+      "learning_rate": 1.3680000000000001e-05,
+      "loss": 0.2884,
+      "step": 1048
+    },
+    {
+      "epoch": 2.261057173678533,
+      "grad_norm": 0.19623045696350683,
+      "learning_rate": 1.364e-05,
+      "loss": 0.2901,
+      "step": 1049
+    },
+    {
+      "epoch": 2.2632146709816614,
+      "grad_norm": 0.20787520367981113,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.2855,
+      "step": 1050
+    },
+    {
+      "epoch": 2.26537216828479,
+      "grad_norm": 0.1962467077184201,
+      "learning_rate": 1.356e-05,
+      "loss": 0.279,
+      "step": 1051
+    },
+    {
+      "epoch": 2.267529665587918,
+      "grad_norm": 0.1995464508539786,
+      "learning_rate": 1.352e-05,
+      "loss": 0.2805,
+      "step": 1052
+    },
+    {
+      "epoch": 2.269687162891046,
+      "grad_norm": 0.19046158544146177,
+      "learning_rate": 1.3480000000000001e-05,
+      "loss": 0.2684,
+      "step": 1053
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 0.19663063830525268,
+      "learning_rate": 1.344e-05,
+      "loss": 0.2815,
+      "step": 1054
+    },
+    {
+      "epoch": 2.274002157497303,
+      "grad_norm": 0.1999712561033732,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.2879,
+      "step": 1055
+    },
+    {
+      "epoch": 2.2761596548004315,
+      "grad_norm": 0.19284565246163382,
+      "learning_rate": 1.336e-05,
+      "loss": 0.2835,
+      "step": 1056
+    },
+    {
+      "epoch": 2.27831715210356,
+      "grad_norm": 0.1939460481831784,
+      "learning_rate": 1.3320000000000001e-05,
+      "loss": 0.2852,
+      "step": 1057
+    },
+    {
+      "epoch": 2.2804746494066883,
+      "grad_norm": 0.17751216415258458,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 0.2553,
+      "step": 1058
+    },
+    {
+      "epoch": 2.2826321467098167,
+      "grad_norm": 0.19432324788325034,
+      "learning_rate": 1.324e-05,
+      "loss": 0.283,
+      "step": 1059
+    },
+    {
+      "epoch": 2.284789644012945,
+      "grad_norm": 0.18964666635161925,
+      "learning_rate": 1.32e-05,
+      "loss": 0.2659,
+      "step": 1060
+    },
+    {
+      "epoch": 2.286947141316073,
+      "grad_norm": 0.19072254885569248,
+      "learning_rate": 1.316e-05,
+      "loss": 0.283,
+      "step": 1061
+    },
+    {
+      "epoch": 2.2891046386192015,
+      "grad_norm": 0.19930445238105082,
+      "learning_rate": 1.3120000000000001e-05,
+      "loss": 0.2783,
+      "step": 1062
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.19310586069082192,
+      "learning_rate": 1.308e-05,
+      "loss": 0.2633,
+      "step": 1063
+    },
+    {
+      "epoch": 2.2934196332254584,
+      "grad_norm": 0.19118685628714194,
+      "learning_rate": 1.3039999999999999e-05,
+      "loss": 0.2836,
+      "step": 1064
+    },
+    {
+      "epoch": 2.295577130528587,
+      "grad_norm": 0.19463719617986558,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.2817,
+      "step": 1065
+    },
+    {
+      "epoch": 2.2977346278317152,
+      "grad_norm": 0.19904862455983952,
+      "learning_rate": 1.296e-05,
+      "loss": 0.269,
+      "step": 1066
+    },
+    {
+      "epoch": 2.2998921251348436,
+      "grad_norm": 0.18016591793334202,
+      "learning_rate": 1.2920000000000002e-05,
+      "loss": 0.2573,
+      "step": 1067
+    },
+    {
+      "epoch": 2.302049622437972,
+      "grad_norm": 0.1870821714575586,
+      "learning_rate": 1.288e-05,
+      "loss": 0.2704,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3042071197411005,
+      "grad_norm": 0.1958000410540272,
+      "learning_rate": 1.2839999999999999e-05,
+      "loss": 0.2804,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3063646170442285,
+      "grad_norm": 0.19053367756271009,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.2793,
+      "step": 1070
+    },
+    {
+      "epoch": 2.308522114347357,
+      "grad_norm": 0.21616331965307153,
+      "learning_rate": 1.276e-05,
+      "loss": 0.2711,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 0.1889243925400608,
+      "learning_rate": 1.2720000000000002e-05,
+      "loss": 0.2861,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3128371089536137,
+      "grad_norm": 0.18365302232747097,
+      "learning_rate": 1.268e-05,
+      "loss": 0.2785,
+      "step": 1073
+    },
+    {
+      "epoch": 2.314994606256742,
+      "grad_norm": 0.21361395439929864,
+      "learning_rate": 1.2640000000000003e-05,
+      "loss": 0.273,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3171521035598706,
+      "grad_norm": 0.2076208646434828,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.2933,
+      "step": 1075
+    },
+    {
+      "epoch": 2.319309600862999,
+      "grad_norm": 0.2139822291459261,
+      "learning_rate": 1.256e-05,
+      "loss": 0.2681,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3214670981661274,
+      "grad_norm": 0.20701974653718327,
+      "learning_rate": 1.252e-05,
+      "loss": 0.2756,
+      "step": 1077
+    },
+    {
+      "epoch": 2.323624595469256,
+      "grad_norm": 0.2145284687747517,
+      "learning_rate": 1.248e-05,
+      "loss": 0.2747,
+      "step": 1078
+    },
+    {
+      "epoch": 2.325782092772384,
+      "grad_norm": 0.21677836163848008,
+      "learning_rate": 1.244e-05,
+      "loss": 0.2867,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3279395900755127,
+      "grad_norm": 0.1949952709667068,
+      "learning_rate": 1.24e-05,
+      "loss": 0.2738,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.20454609129744541,
+      "learning_rate": 1.236e-05,
+      "loss": 0.2778,
+      "step": 1081
+    },
+    {
+      "epoch": 2.332254584681769,
+      "grad_norm": 0.20493348135141942,
+      "learning_rate": 1.232e-05,
+      "loss": 0.2702,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3344120819848975,
+      "grad_norm": 0.18716804503177586,
+      "learning_rate": 1.2280000000000001e-05,
+      "loss": 0.2632,
+      "step": 1083
+    },
+    {
+      "epoch": 2.336569579288026,
+      "grad_norm": 0.19834904622075808,
+      "learning_rate": 1.224e-05,
+      "loss": 0.2707,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3387270765911543,
+      "grad_norm": 0.21807191652834473,
+      "learning_rate": 1.22e-05,
+      "loss": 0.2938,
+      "step": 1085
+    },
+    {
+      "epoch": 2.3408845738942827,
+      "grad_norm": 0.21076158567002148,
+      "learning_rate": 1.216e-05,
+      "loss": 0.2832,
+      "step": 1086
+    },
+    {
+      "epoch": 2.343042071197411,
+      "grad_norm": 0.2145838491982327,
+      "learning_rate": 1.2120000000000001e-05,
+      "loss": 0.2927,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3451995685005396,
+      "grad_norm": 0.19885442417194185,
+      "learning_rate": 1.2080000000000001e-05,
+      "loss": 0.2671,
+      "step": 1088
+    },
+    {
+      "epoch": 2.347357065803668,
+      "grad_norm": 0.19696825750002114,
+      "learning_rate": 1.204e-05,
+      "loss": 0.2737,
+      "step": 1089
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 0.18667737946503235,
+      "learning_rate": 1.2e-05,
+      "loss": 0.28,
+      "step": 1090
+    },
+    {
+      "epoch": 2.3516720604099244,
+      "grad_norm": 0.1908481063008943,
+      "learning_rate": 1.196e-05,
+      "loss": 0.2737,
+      "step": 1091
+    },
+    {
+      "epoch": 2.353829557713053,
+      "grad_norm": 0.19318351140026296,
+      "learning_rate": 1.1920000000000001e-05,
+      "loss": 0.2757,
+      "step": 1092
+    },
+    {
+      "epoch": 2.355987055016181,
+      "grad_norm": 0.1961228433938055,
+      "learning_rate": 1.1880000000000001e-05,
+      "loss": 0.2862,
+      "step": 1093
+    },
+    {
+      "epoch": 2.3581445523193096,
+      "grad_norm": 0.18057536406104824,
+      "learning_rate": 1.1840000000000002e-05,
+      "loss": 0.2596,
+      "step": 1094
+    },
+    {
+      "epoch": 2.360302049622438,
+      "grad_norm": 0.20362429416941197,
+      "learning_rate": 1.18e-05,
+      "loss": 0.2949,
+      "step": 1095
+    },
+    {
+      "epoch": 2.3624595469255665,
+      "grad_norm": 0.1991741123926773,
+      "learning_rate": 1.1760000000000001e-05,
+      "loss": 0.2759,
+      "step": 1096
+    },
+    {
+      "epoch": 2.364617044228695,
+      "grad_norm": 0.19011374313750604,
+      "learning_rate": 1.172e-05,
+      "loss": 0.2803,
+      "step": 1097
+    },
+    {
+      "epoch": 2.3667745415318233,
+      "grad_norm": 0.18495183732229703,
+      "learning_rate": 1.168e-05,
+      "loss": 0.275,
+      "step": 1098
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.183331748886235,
+      "learning_rate": 1.164e-05,
+      "loss": 0.2722,
+      "step": 1099
+    },
+    {
+      "epoch": 2.3710895361380797,
+      "grad_norm": 0.19464092804866787,
+      "learning_rate": 1.16e-05,
+      "loss": 0.2694,
+      "step": 1100
+    },
+    {
+      "epoch": 2.373247033441208,
+      "grad_norm": 0.20595615106535306,
+      "learning_rate": 1.156e-05,
+      "loss": 0.2765,
+      "step": 1101
+    },
+    {
+      "epoch": 2.3754045307443366,
+      "grad_norm": 0.20027835316245082,
+      "learning_rate": 1.152e-05,
+      "loss": 0.2916,
+      "step": 1102
+    },
+    {
+      "epoch": 2.377562028047465,
+      "grad_norm": 0.18806800378915345,
+      "learning_rate": 1.148e-05,
+      "loss": 0.2691,
+      "step": 1103
+    },
+    {
+      "epoch": 2.3797195253505934,
+      "grad_norm": 0.44925086549855475,
+      "learning_rate": 1.144e-05,
+      "loss": 0.2671,
+      "step": 1104
+    },
+    {
+      "epoch": 2.381877022653722,
+      "grad_norm": 0.1956910478238287,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.2828,
+      "step": 1105
+    },
+    {
+      "epoch": 2.3840345199568502,
+      "grad_norm": 0.19656830448675866,
+      "learning_rate": 1.1360000000000001e-05,
+      "loss": 0.2762,
+      "step": 1106
+    },
+    {
+      "epoch": 2.3861920172599786,
+      "grad_norm": 0.2012449425045171,
+      "learning_rate": 1.132e-05,
+      "loss": 0.2878,
+      "step": 1107
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 0.21769469381273382,
+      "learning_rate": 1.128e-05,
+      "loss": 0.2794,
+      "step": 1108
+    },
+    {
+      "epoch": 2.390507011866235,
+      "grad_norm": 0.19950183867784813,
+      "learning_rate": 1.124e-05,
+      "loss": 0.2888,
+      "step": 1109
+    },
+    {
+      "epoch": 2.3926645091693635,
+      "grad_norm": 0.20120005430252597,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.2835,
+      "step": 1110
+    },
+    {
+      "epoch": 2.394822006472492,
+      "grad_norm": 0.20369048867315678,
+      "learning_rate": 1.1160000000000002e-05,
+      "loss": 0.2842,
+      "step": 1111
+    },
+    {
+      "epoch": 2.3969795037756203,
+      "grad_norm": 0.1931772151237011,
+      "learning_rate": 1.112e-05,
+      "loss": 0.2857,
+      "step": 1112
+    },
+    {
+      "epoch": 2.3991370010787487,
+      "grad_norm": 0.19883071154560317,
+      "learning_rate": 1.108e-05,
+      "loss": 0.2653,
+      "step": 1113
+    },
+    {
+      "epoch": 2.401294498381877,
+      "grad_norm": 0.19563569996619778,
+      "learning_rate": 1.1040000000000001e-05,
+      "loss": 0.2799,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4034519956850056,
+      "grad_norm": 0.20341050384160864,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.2738,
+      "step": 1115
+    },
+    {
+      "epoch": 2.405609492988134,
+      "grad_norm": 0.1945190789786455,
+      "learning_rate": 1.096e-05,
+      "loss": 0.2793,
+      "step": 1116
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.21314612733031374,
+      "learning_rate": 1.092e-05,
+      "loss": 0.2841,
+      "step": 1117
+    },
+    {
+      "epoch": 2.4099244875943904,
+      "grad_norm": 0.20091792860429103,
+      "learning_rate": 1.088e-05,
+      "loss": 0.2852,
+      "step": 1118
+    },
+    {
+      "epoch": 2.412081984897519,
+      "grad_norm": 0.1912326765301699,
+      "learning_rate": 1.084e-05,
+      "loss": 0.2859,
+      "step": 1119
+    },
+    {
+      "epoch": 2.414239482200647,
+      "grad_norm": 0.19350501585802152,
+      "learning_rate": 1.08e-05,
+      "loss": 0.2819,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4163969795037756,
+      "grad_norm": 0.19011528786997062,
+      "learning_rate": 1.076e-05,
+      "loss": 0.2684,
+      "step": 1121
+    },
+    {
+      "epoch": 2.418554476806904,
+      "grad_norm": 0.19626087415761062,
+      "learning_rate": 1.072e-05,
+      "loss": 0.2863,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4207119741100325,
+      "grad_norm": 0.18700947119186567,
+      "learning_rate": 1.0680000000000001e-05,
+      "loss": 0.2771,
+      "step": 1123
+    },
+    {
+      "epoch": 2.422869471413161,
+      "grad_norm": 0.20648852488177072,
+      "learning_rate": 1.064e-05,
+      "loss": 0.2834,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4250269687162893,
+      "grad_norm": 0.2012989868600837,
+      "learning_rate": 1.06e-05,
+      "loss": 0.2716,
+      "step": 1125
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.199888643532734,
+      "learning_rate": 1.056e-05,
+      "loss": 0.2791,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4293419633225457,
+      "grad_norm": 0.18259826425673448,
+      "learning_rate": 1.0520000000000001e-05,
+      "loss": 0.2541,
+      "step": 1127
+    },
+    {
+      "epoch": 2.431499460625674,
+      "grad_norm": 0.19649477554378358,
+      "learning_rate": 1.0480000000000001e-05,
+      "loss": 0.2801,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4336569579288025,
+      "grad_norm": 0.18199564261537443,
+      "learning_rate": 1.0440000000000002e-05,
+      "loss": 0.2559,
+      "step": 1129
+    },
+    {
+      "epoch": 2.435814455231931,
+      "grad_norm": 0.1952536593316178,
+      "learning_rate": 1.04e-05,
+      "loss": 0.2745,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4379719525350594,
+      "grad_norm": 0.19876226657601323,
+      "learning_rate": 1.036e-05,
+      "loss": 0.2742,
+      "step": 1131
+    },
+    {
+      "epoch": 2.440129449838188,
+      "grad_norm": 0.199938722386871,
+      "learning_rate": 1.0320000000000001e-05,
+      "loss": 0.2877,
+      "step": 1132
+    },
+    {
+      "epoch": 2.4422869471413162,
+      "grad_norm": 0.1793115472237926,
+      "learning_rate": 1.0280000000000002e-05,
+      "loss": 0.2672,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.20872212600922832,
+      "learning_rate": 1.024e-05,
+      "loss": 0.2776,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.19129962341137918,
+      "learning_rate": 1.02e-05,
+      "loss": 0.2819,
+      "step": 1135
+    },
+    {
+      "epoch": 2.448759439050701,
+      "grad_norm": 0.19282605224929167,
+      "learning_rate": 1.016e-05,
+      "loss": 0.2864,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4509169363538295,
+      "grad_norm": 0.19659894723021457,
+      "learning_rate": 1.012e-05,
+      "loss": 0.2886,
+      "step": 1137
+    },
+    {
+      "epoch": 2.453074433656958,
+      "grad_norm": 0.19381386784264035,
+      "learning_rate": 1.008e-05,
+      "loss": 0.288,
+      "step": 1138
+    },
+    {
+      "epoch": 2.4552319309600863,
+      "grad_norm": 0.18628308033727628,
+      "learning_rate": 1.004e-05,
+      "loss": 0.2739,
+      "step": 1139
+    },
+    {
+      "epoch": 2.4573894282632147,
+      "grad_norm": 0.21441635507962467,
+      "learning_rate": 1e-05,
+      "loss": 0.2932,
+      "step": 1140
+    },
+    {
+      "epoch": 2.459546925566343,
+      "grad_norm": 0.20534818966463134,
+      "learning_rate": 9.96e-06,
+      "loss": 0.2855,
+      "step": 1141
+    },
+    {
+      "epoch": 2.4617044228694716,
+      "grad_norm": 0.1849300035193185,
+      "learning_rate": 9.92e-06,
+      "loss": 0.2632,
+      "step": 1142
+    },
+    {
+      "epoch": 2.4638619201726,
+      "grad_norm": 0.18852281346545738,
+      "learning_rate": 9.88e-06,
+      "loss": 0.278,
+      "step": 1143
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.18627944823678463,
+      "learning_rate": 9.84e-06,
+      "loss": 0.275,
+      "step": 1144
+    },
+    {
+      "epoch": 2.4681769147788564,
+      "grad_norm": 0.2045693903724477,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.2712,
+      "step": 1145
+    },
+    {
+      "epoch": 2.470334412081985,
+      "grad_norm": 0.1989352745180615,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 0.2832,
+      "step": 1146
+    },
+    {
+      "epoch": 2.472491909385113,
+      "grad_norm": 0.19608690680269333,
+      "learning_rate": 9.72e-06,
+      "loss": 0.2765,
+      "step": 1147
+    },
+    {
+      "epoch": 2.4746494066882416,
+      "grad_norm": 0.18376170966118,
+      "learning_rate": 9.68e-06,
+      "loss": 0.2877,
+      "step": 1148
+    },
+    {
+      "epoch": 2.47680690399137,
+      "grad_norm": 0.1830683736706619,
+      "learning_rate": 9.640000000000001e-06,
+      "loss": 0.273,
+      "step": 1149
+    },
+    {
+      "epoch": 2.4789644012944985,
+      "grad_norm": 0.19774471106842714,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.2768,
+      "step": 1150
+    },
+    {
+      "epoch": 2.481121898597627,
+      "grad_norm": 0.5289386953378491,
+      "learning_rate": 9.560000000000002e-06,
+      "loss": 0.2859,
+      "step": 1151
+    },
+    {
+      "epoch": 2.4832793959007553,
+      "grad_norm": 0.18054827974203572,
+      "learning_rate": 9.52e-06,
+      "loss": 0.2739,
+      "step": 1152
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.19041823794360682,
+      "learning_rate": 9.48e-06,
+      "loss": 0.2715,
+      "step": 1153
+    },
+    {
+      "epoch": 2.4875943905070117,
+      "grad_norm": 0.19696127404865826,
+      "learning_rate": 9.44e-06,
+      "loss": 0.2749,
+      "step": 1154
+    },
+    {
+      "epoch": 2.48975188781014,
+      "grad_norm": 0.20729223436965724,
+      "learning_rate": 9.4e-06,
+      "loss": 0.2665,
+      "step": 1155
+    },
+    {
+      "epoch": 2.4919093851132685,
+      "grad_norm": 0.2151672552731983,
+      "learning_rate": 9.36e-06,
+      "loss": 0.2771,
+      "step": 1156
+    },
+    {
+      "epoch": 2.494066882416397,
+      "grad_norm": 0.19970041745608494,
+      "learning_rate": 9.32e-06,
+      "loss": 0.2844,
+      "step": 1157
+    },
+    {
+      "epoch": 2.4962243797195254,
+      "grad_norm": 0.1875452071423268,
+      "learning_rate": 9.28e-06,
+      "loss": 0.269,
+      "step": 1158
+    },
+    {
+      "epoch": 2.498381877022654,
+      "grad_norm": 0.1815111230711524,
+      "learning_rate": 9.24e-06,
+      "loss": 0.2776,
+      "step": 1159
+    },
+    {
+      "epoch": 2.500539374325782,
+      "grad_norm": 0.21522418787260655,
+      "learning_rate": 9.2e-06,
+      "loss": 0.2643,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5026968716289106,
+      "grad_norm": 0.21059659615733978,
+      "learning_rate": 9.16e-06,
+      "loss": 0.2859,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 0.21572007660549064,
+      "learning_rate": 9.12e-06,
+      "loss": 0.2837,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5070118662351675,
+      "grad_norm": 0.19174617150175122,
+      "learning_rate": 9.080000000000001e-06,
+      "loss": 0.2664,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5091693635382954,
+      "grad_norm": 0.18370659572207848,
+      "learning_rate": 9.04e-06,
+      "loss": 0.2799,
+      "step": 1164
+    },
+    {
+      "epoch": 2.511326860841424,
+      "grad_norm": 0.18956236987690317,
+      "learning_rate": 9e-06,
+      "loss": 0.2736,
+      "step": 1165
+    },
+    {
+      "epoch": 2.5134843581445523,
+      "grad_norm": 0.21218350316632872,
+      "learning_rate": 8.96e-06,
+      "loss": 0.2745,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5156418554476807,
+      "grad_norm": 0.20564251196662126,
+      "learning_rate": 8.920000000000001e-06,
+      "loss": 0.2728,
+      "step": 1167
+    },
+    {
+      "epoch": 2.517799352750809,
+      "grad_norm": 0.18190663215352182,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 0.2717,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5199568500539375,
+      "grad_norm": 0.19047962603852486,
+      "learning_rate": 8.840000000000002e-06,
+      "loss": 0.274,
+      "step": 1169
+    },
+    {
+      "epoch": 2.522114347357066,
+      "grad_norm": 0.1956042804734501,
+      "learning_rate": 8.8e-06,
+      "loss": 0.2688,
+      "step": 1170
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.1960949651558405,
+      "learning_rate": 8.76e-06,
+      "loss": 0.2824,
+      "step": 1171
+    },
+    {
+      "epoch": 2.526429341963323,
+      "grad_norm": 0.1880455744302837,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 0.2731,
+      "step": 1172
+    },
+    {
+      "epoch": 2.528586839266451,
+      "grad_norm": 0.20463391348180632,
+      "learning_rate": 8.68e-06,
+      "loss": 0.2874,
+      "step": 1173
+    },
+    {
+      "epoch": 2.530744336569579,
+      "grad_norm": 0.1868158566005862,
+      "learning_rate": 8.64e-06,
+      "loss": 0.2733,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5329018338727076,
+      "grad_norm": 0.19201921577442338,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.2707,
+      "step": 1175
+    },
+    {
+      "epoch": 2.535059331175836,
+      "grad_norm": 0.1870377402294841,
+      "learning_rate": 8.56e-06,
+      "loss": 0.2705,
+      "step": 1176
+    },
+    {
+      "epoch": 2.5372168284789645,
+      "grad_norm": 0.20226457534658343,
+      "learning_rate": 8.52e-06,
+      "loss": 0.2642,
+      "step": 1177
+    },
+    {
+      "epoch": 2.539374325782093,
+      "grad_norm": 0.1957674767179278,
+      "learning_rate": 8.48e-06,
+      "loss": 0.2891,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5415318230852213,
+      "grad_norm": 0.18492027429247337,
+      "learning_rate": 8.44e-06,
+      "loss": 0.2724,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.18451597966638716,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.2881,
+      "step": 1180
+    },
+    {
+      "epoch": 2.545846817691478,
+      "grad_norm": 0.18042456868864143,
+      "learning_rate": 8.36e-06,
+      "loss": 0.2809,
+      "step": 1181
+    },
+    {
+      "epoch": 2.548004314994606,
+      "grad_norm": 0.1939726883551459,
+      "learning_rate": 8.32e-06,
+      "loss": 0.2707,
+      "step": 1182
+    },
+    {
+      "epoch": 2.5501618122977345,
+      "grad_norm": 0.19107636821986104,
+      "learning_rate": 8.28e-06,
+      "loss": 0.2797,
+      "step": 1183
+    },
+    {
+      "epoch": 2.552319309600863,
+      "grad_norm": 0.18956795249933478,
+      "learning_rate": 8.24e-06,
+      "loss": 0.2705,
+      "step": 1184
+    },
+    {
+      "epoch": 2.5544768069039914,
+      "grad_norm": 0.19186426285857633,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 0.2715,
+      "step": 1185
+    },
+    {
+      "epoch": 2.55663430420712,
+      "grad_norm": 0.19503695734156132,
+      "learning_rate": 8.160000000000001e-06,
+      "loss": 0.2757,
+      "step": 1186
+    },
+    {
+      "epoch": 2.558791801510248,
+      "grad_norm": 0.17801355493098456,
+      "learning_rate": 8.12e-06,
+      "loss": 0.2695,
+      "step": 1187
+    },
+    {
+      "epoch": 2.5609492988133766,
+      "grad_norm": 0.18569837697870945,
+      "learning_rate": 8.08e-06,
+      "loss": 0.2788,
+      "step": 1188
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.18209260370415475,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 0.2834,
+      "step": 1189
+    },
+    {
+      "epoch": 2.5652642934196335,
+      "grad_norm": 0.20862033912842837,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2787,
+      "step": 1190
+    },
+    {
+      "epoch": 2.5674217907227614,
+      "grad_norm": 0.18909747904684734,
+      "learning_rate": 7.96e-06,
+      "loss": 0.2712,
+      "step": 1191
+    },
+    {
+      "epoch": 2.56957928802589,
+      "grad_norm": 0.21129244434665395,
+      "learning_rate": 7.92e-06,
+      "loss": 0.287,
+      "step": 1192
+    },
+    {
+      "epoch": 2.5717367853290183,
+      "grad_norm": 0.18022619488696826,
+      "learning_rate": 7.879999999999999e-06,
+      "loss": 0.2813,
+      "step": 1193
+    },
+    {
+      "epoch": 2.5738942826321467,
+      "grad_norm": 0.19852098535134896,
+      "learning_rate": 7.84e-06,
+      "loss": 0.2862,
+      "step": 1194
+    },
+    {
+      "epoch": 2.576051779935275,
+      "grad_norm": 0.19617888636362643,
+      "learning_rate": 7.8e-06,
+      "loss": 0.2844,
+      "step": 1195
+    },
+    {
+      "epoch": 2.5782092772384035,
+      "grad_norm": 0.18969668381525537,
+      "learning_rate": 7.76e-06,
+      "loss": 0.2891,
+      "step": 1196
+    },
+    {
+      "epoch": 2.580366774541532,
+      "grad_norm": 0.18430145815088345,
+      "learning_rate": 7.72e-06,
+      "loss": 0.2721,
+      "step": 1197
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 0.18320017834506805,
+      "learning_rate": 7.68e-06,
+      "loss": 0.2774,
+      "step": 1198
+    },
+    {
+      "epoch": 2.584681769147789,
+      "grad_norm": 0.1892382939282894,
+      "learning_rate": 7.64e-06,
+      "loss": 0.2759,
+      "step": 1199
+    },
+    {
+      "epoch": 2.5868392664509168,
+      "grad_norm": 0.1880248772535744,
+      "learning_rate": 7.6e-06,
+      "loss": 0.2797,
+      "step": 1200
+    },
+    {
+      "epoch": 2.588996763754045,
+      "grad_norm": 0.18324546736309735,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 0.2939,
+      "step": 1201
+    },
+    {
+      "epoch": 2.5911542610571736,
+      "grad_norm": 0.1821909206802706,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 0.2797,
+      "step": 1202
+    },
+    {
+      "epoch": 2.593311758360302,
+      "grad_norm": 0.18785236372663316,
+      "learning_rate": 7.480000000000001e-06,
+      "loss": 0.2798,
+      "step": 1203
+    },
+    {
+      "epoch": 2.5954692556634305,
+      "grad_norm": 0.18862019277830552,
+      "learning_rate": 7.44e-06,
+      "loss": 0.268,
+      "step": 1204
+    },
+    {
+      "epoch": 2.597626752966559,
+      "grad_norm": 0.17802704840377132,
+      "learning_rate": 7.4e-06,
+      "loss": 0.2653,
+      "step": 1205
+    },
+    {
+      "epoch": 2.5997842502696873,
+      "grad_norm": 0.18808394376760307,
+      "learning_rate": 7.36e-06,
+      "loss": 0.2819,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 0.1736414972830234,
+      "learning_rate": 7.32e-06,
+      "loss": 0.2755,
+      "step": 1207
+    },
+    {
+      "epoch": 2.604099244875944,
+      "grad_norm": 0.19798820811833512,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 0.2766,
+      "step": 1208
+    },
+    {
+      "epoch": 2.606256742179072,
+      "grad_norm": 0.19360358509115017,
+      "learning_rate": 7.240000000000001e-06,
+      "loss": 0.278,
+      "step": 1209
+    },
+    {
+      "epoch": 2.6084142394822005,
+      "grad_norm": 0.1819274054858843,
+      "learning_rate": 7.2e-06,
+      "loss": 0.2748,
+      "step": 1210
+    },
+    {
+      "epoch": 2.610571736785329,
+      "grad_norm": 0.19153275592635163,
+      "learning_rate": 7.16e-06,
+      "loss": 0.2865,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6127292340884574,
+      "grad_norm": 0.18106937603713005,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 0.2836,
+      "step": 1212
+    },
+    {
+      "epoch": 2.614886731391586,
+      "grad_norm": 0.18316357873796607,
+      "learning_rate": 7.080000000000001e-06,
+      "loss": 0.2703,
+      "step": 1213
+    },
+    {
+      "epoch": 2.617044228694714,
+      "grad_norm": 0.19012692093079211,
+      "learning_rate": 7.04e-06,
+      "loss": 0.283,
+      "step": 1214
+    },
+    {
+      "epoch": 2.6192017259978426,
+      "grad_norm": 0.1886991400739291,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.2951,
+      "step": 1215
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.17872080834424994,
+      "learning_rate": 6.9599999999999994e-06,
+      "loss": 0.275,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6235167206040995,
+      "grad_norm": 0.19038957823975905,
+      "learning_rate": 6.92e-06,
+      "loss": 0.276,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6256742179072274,
+      "grad_norm": 0.18366617360792217,
+      "learning_rate": 6.88e-06,
+      "loss": 0.2866,
+      "step": 1218
+    },
+    {
+      "epoch": 2.627831715210356,
+      "grad_norm": 0.18194261875821086,
+      "learning_rate": 6.840000000000001e-06,
+      "loss": 0.2789,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6299892125134843,
+      "grad_norm": 0.18748953872554902,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.2833,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6321467098166127,
+      "grad_norm": 0.18281890368924675,
+      "learning_rate": 6.76e-06,
+      "loss": 0.2712,
+      "step": 1221
+    },
+    {
+      "epoch": 2.634304207119741,
+      "grad_norm": 0.18225058596967514,
+      "learning_rate": 6.72e-06,
+      "loss": 0.2784,
+      "step": 1222
+    },
+    {
+      "epoch": 2.6364617044228695,
+      "grad_norm": 0.1825800678446187,
+      "learning_rate": 6.68e-06,
+      "loss": 0.2724,
+      "step": 1223
+    },
+    {
+      "epoch": 2.638619201725998,
+      "grad_norm": 0.19363871055752166,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 0.2684,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.1944748649154887,
+      "learning_rate": 6.6e-06,
+      "loss": 0.2784,
+      "step": 1225
+    },
+    {
+      "epoch": 2.642934196332255,
+      "grad_norm": 0.1881742065564678,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 0.2756,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6450916936353828,
+      "grad_norm": 0.18392054853177556,
+      "learning_rate": 6.519999999999999e-06,
+      "loss": 0.2796,
+      "step": 1227
+    },
+    {
+      "epoch": 2.647249190938511,
+      "grad_norm": 0.18906848548661837,
+      "learning_rate": 6.48e-06,
+      "loss": 0.2761,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6494066882416396,
+      "grad_norm": 0.1770937107224099,
+      "learning_rate": 6.44e-06,
+      "loss": 0.2764,
+      "step": 1229
+    },
+    {
+      "epoch": 2.651564185544768,
+      "grad_norm": 0.1919361386776459,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.2762,
+      "step": 1230
+    },
+    {
+      "epoch": 2.6537216828478964,
+      "grad_norm": 0.1898063423812771,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 0.279,
+      "step": 1231
+    },
+    {
+      "epoch": 2.655879180151025,
+      "grad_norm": 0.18359024005891844,
+      "learning_rate": 6.320000000000001e-06,
+      "loss": 0.2895,
+      "step": 1232
+    },
+    {
+      "epoch": 2.6580366774541533,
+      "grad_norm": 0.18905016504891056,
+      "learning_rate": 6.28e-06,
+      "loss": 0.2841,
+      "step": 1233
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.18962361851782872,
+      "learning_rate": 6.24e-06,
+      "loss": 0.266,
+      "step": 1234
+    },
+    {
+      "epoch": 2.66235167206041,
+      "grad_norm": 0.18363662298141045,
+      "learning_rate": 6.2e-06,
+      "loss": 0.2874,
+      "step": 1235
+    },
+    {
+      "epoch": 2.664509169363538,
+      "grad_norm": 0.20630491074122706,
+      "learning_rate": 6.16e-06,
+      "loss": 0.2906,
+      "step": 1236
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.18135071779569467,
+      "learning_rate": 6.12e-06,
+      "loss": 0.285,
+      "step": 1237
+    },
+    {
+      "epoch": 2.668824163969795,
+      "grad_norm": 0.18372632774369446,
+      "learning_rate": 6.08e-06,
+      "loss": 0.2739,
+      "step": 1238
+    },
+    {
+      "epoch": 2.6709816612729234,
+      "grad_norm": 0.18428383879403848,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 0.2875,
+      "step": 1239
+    },
+    {
+      "epoch": 2.6731391585760518,
+      "grad_norm": 0.18460310424536336,
+      "learning_rate": 6e-06,
+      "loss": 0.2556,
+      "step": 1240
+    },
+    {
+      "epoch": 2.67529665587918,
+      "grad_norm": 0.18947805389880149,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 0.2798,
+      "step": 1241
+    },
+    {
+      "epoch": 2.6774541531823086,
+      "grad_norm": 0.19780847857755182,
+      "learning_rate": 5.920000000000001e-06,
+      "loss": 0.2675,
+      "step": 1242
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.18499409934277794,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 0.2806,
+      "step": 1243
+    },
+    {
+      "epoch": 2.6817691477885655,
+      "grad_norm": 0.18316817509358638,
+      "learning_rate": 5.84e-06,
+      "loss": 0.2765,
+      "step": 1244
+    },
+    {
+      "epoch": 2.6839266450916934,
+      "grad_norm": 0.17478366686579072,
+      "learning_rate": 5.8e-06,
+      "loss": 0.2644,
+      "step": 1245
+    },
+    {
+      "epoch": 2.686084142394822,
+      "grad_norm": 0.19358311792535488,
+      "learning_rate": 5.76e-06,
+      "loss": 0.264,
+      "step": 1246
+    },
+    {
+      "epoch": 2.6882416396979503,
+      "grad_norm": 0.1817491973984989,
+      "learning_rate": 5.72e-06,
+      "loss": 0.2776,
+      "step": 1247
+    },
+    {
+      "epoch": 2.6903991370010787,
+      "grad_norm": 0.17836599511127066,
+      "learning_rate": 5.680000000000001e-06,
+      "loss": 0.2729,
+      "step": 1248
+    },
+    {
+      "epoch": 2.692556634304207,
+      "grad_norm": 0.20855396251969097,
+      "learning_rate": 5.64e-06,
+      "loss": 0.2797,
+      "step": 1249
+    },
+    {
+      "epoch": 2.6947141316073355,
+      "grad_norm": 0.17710319375165606,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.2664,
+      "step": 1250
+    },
+    {
+      "epoch": 2.696871628910464,
+      "grad_norm": 0.18182561629022773,
+      "learning_rate": 5.56e-06,
+      "loss": 0.285,
+      "step": 1251
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.18254255786329926,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 0.2668,
+      "step": 1252
+    },
+    {
+      "epoch": 2.701186623516721,
+      "grad_norm": 0.18071712621398686,
+      "learning_rate": 5.48e-06,
+      "loss": 0.2645,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7033441208198488,
+      "grad_norm": 0.17992767145304728,
+      "learning_rate": 5.44e-06,
+      "loss": 0.2747,
+      "step": 1254
+    },
+    {
+      "epoch": 2.705501618122977,
+      "grad_norm": 0.184558078685827,
+      "learning_rate": 5.4e-06,
+      "loss": 0.2726,
+      "step": 1255
+    },
+    {
+      "epoch": 2.7076591154261056,
+      "grad_norm": 0.1848059441306179,
+      "learning_rate": 5.36e-06,
+      "loss": 0.2829,
+      "step": 1256
+    },
+    {
+      "epoch": 2.709816612729234,
+      "grad_norm": 0.17640374705147477,
+      "learning_rate": 5.32e-06,
+      "loss": 0.2701,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7119741100323624,
+      "grad_norm": 0.21805384839708805,
+      "learning_rate": 5.28e-06,
+      "loss": 0.2716,
+      "step": 1258
+    },
+    {
+      "epoch": 2.714131607335491,
+      "grad_norm": 0.20101963800895353,
+      "learning_rate": 5.240000000000001e-06,
+      "loss": 0.2886,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7162891046386193,
+      "grad_norm": 0.1804114201116908,
+      "learning_rate": 5.2e-06,
+      "loss": 0.2776,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.1774653180001705,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 0.2826,
+      "step": 1261
+    },
+    {
+      "epoch": 2.720604099244876,
+      "grad_norm": 0.17688387986830606,
+      "learning_rate": 5.12e-06,
+      "loss": 0.2689,
+      "step": 1262
+    },
+    {
+      "epoch": 2.722761596548004,
+      "grad_norm": 0.18501330626919668,
+      "learning_rate": 5.08e-06,
+      "loss": 0.2703,
+      "step": 1263
+    },
+    {
+      "epoch": 2.724919093851133,
+      "grad_norm": 0.17773374479207837,
+      "learning_rate": 5.04e-06,
+      "loss": 0.2697,
+      "step": 1264
+    },
+    {
+      "epoch": 2.727076591154261,
+      "grad_norm": 0.18714014475865487,
+      "learning_rate": 5e-06,
+      "loss": 0.2659,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7292340884573894,
+      "grad_norm": 0.1891342823049284,
+      "learning_rate": 4.96e-06,
+      "loss": 0.2689,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7313915857605178,
+      "grad_norm": 0.18120794072745916,
+      "learning_rate": 4.92e-06,
+      "loss": 0.2801,
+      "step": 1267
+    },
+    {
+      "epoch": 2.733549083063646,
+      "grad_norm": 0.18421472259058802,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 0.2742,
+      "step": 1268
+    },
+    {
+      "epoch": 2.7357065803667746,
+      "grad_norm": 0.19424882942881327,
+      "learning_rate": 4.84e-06,
+      "loss": 0.2835,
+      "step": 1269
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.19043478378237688,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.2689,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7400215749730314,
+      "grad_norm": 0.1849206683939454,
+      "learning_rate": 4.76e-06,
+      "loss": 0.2761,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7421790722761594,
+      "grad_norm": 0.18522900805651904,
+      "learning_rate": 4.72e-06,
+      "loss": 0.2767,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7443365695792883,
+      "grad_norm": 0.17248319008493151,
+      "learning_rate": 4.68e-06,
+      "loss": 0.2668,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7464940668824163,
+      "grad_norm": 0.18993609725197919,
+      "learning_rate": 4.64e-06,
+      "loss": 0.2744,
+      "step": 1274
+    },
+    {
+      "epoch": 2.7486515641855447,
+      "grad_norm": 0.17727961540010734,
+      "learning_rate": 4.6e-06,
+      "loss": 0.2708,
+      "step": 1275
+    },
+    {
+      "epoch": 2.750809061488673,
+      "grad_norm": 0.17989663911750914,
+      "learning_rate": 4.56e-06,
+      "loss": 0.2748,
+      "step": 1276
+    },
+    {
+      "epoch": 2.7529665587918015,
+      "grad_norm": 0.18993560472210555,
+      "learning_rate": 4.52e-06,
+      "loss": 0.2876,
+      "step": 1277
+    },
+    {
+      "epoch": 2.75512405609493,
+      "grad_norm": 0.18474539106255636,
+      "learning_rate": 4.48e-06,
+      "loss": 0.2767,
+      "step": 1278
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.18191823825028888,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 0.2854,
+      "step": 1279
+    },
+    {
+      "epoch": 2.759439050701187,
+      "grad_norm": 0.18478282075534863,
+      "learning_rate": 4.4e-06,
+      "loss": 0.2888,
+      "step": 1280
+    },
+    {
+      "epoch": 2.7615965480043148,
+      "grad_norm": 0.17755688039996642,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 0.2737,
+      "step": 1281
+    },
+    {
+      "epoch": 2.7637540453074436,
+      "grad_norm": 0.18419934859887713,
+      "learning_rate": 4.32e-06,
+      "loss": 0.2687,
+      "step": 1282
+    },
+    {
+      "epoch": 2.7659115426105716,
+      "grad_norm": 0.1900384282711362,
+      "learning_rate": 4.28e-06,
+      "loss": 0.2759,
+      "step": 1283
+    },
+    {
+      "epoch": 2.7680690399137,
+      "grad_norm": 0.1784962757580093,
+      "learning_rate": 4.24e-06,
+      "loss": 0.2716,
+      "step": 1284
+    },
+    {
+      "epoch": 2.7702265372168284,
+      "grad_norm": 0.1809036819922652,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.27,
+      "step": 1285
+    },
+    {
+      "epoch": 2.772384034519957,
+      "grad_norm": 0.18282579091259332,
+      "learning_rate": 4.16e-06,
+      "loss": 0.2865,
+      "step": 1286
+    },
+    {
+      "epoch": 2.7745415318230853,
+      "grad_norm": 0.18438768981642326,
+      "learning_rate": 4.12e-06,
+      "loss": 0.2774,
+      "step": 1287
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.19152300030259864,
+      "learning_rate": 4.080000000000001e-06,
+      "loss": 0.2796,
+      "step": 1288
+    },
+    {
+      "epoch": 2.778856526429342,
+      "grad_norm": 0.17285666874457525,
+      "learning_rate": 4.04e-06,
+      "loss": 0.2754,
+      "step": 1289
+    },
+    {
+      "epoch": 2.78101402373247,
+      "grad_norm": 0.18414001866694285,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.274,
+      "step": 1290
+    },
+    {
+      "epoch": 2.783171521035599,
+      "grad_norm": 0.17520099229494396,
+      "learning_rate": 3.96e-06,
+      "loss": 0.2766,
+      "step": 1291
+    },
+    {
+      "epoch": 2.785329018338727,
+      "grad_norm": 0.1806154595467803,
+      "learning_rate": 3.92e-06,
+      "loss": 0.2801,
+      "step": 1292
+    },
+    {
+      "epoch": 2.7874865156418553,
+      "grad_norm": 0.18980187456754713,
+      "learning_rate": 3.88e-06,
+      "loss": 0.2727,
+      "step": 1293
+    },
+    {
+      "epoch": 2.7896440129449838,
+      "grad_norm": 0.19597770737066833,
+      "learning_rate": 3.84e-06,
+      "loss": 0.285,
+      "step": 1294
+    },
+    {
+      "epoch": 2.791801510248112,
+      "grad_norm": 0.18063854460198395,
+      "learning_rate": 3.8e-06,
+      "loss": 0.2752,
+      "step": 1295
+    },
+    {
+      "epoch": 2.7939590075512406,
+      "grad_norm": 0.17977161179376194,
+      "learning_rate": 3.7600000000000004e-06,
+      "loss": 0.278,
+      "step": 1296
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.17338820852578288,
+      "learning_rate": 3.72e-06,
+      "loss": 0.2595,
+      "step": 1297
+    },
+    {
+      "epoch": 2.7982740021574974,
+      "grad_norm": 0.18386654166694605,
+      "learning_rate": 3.68e-06,
+      "loss": 0.2767,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8004314994606254,
+      "grad_norm": 0.17901186263657326,
+      "learning_rate": 3.6400000000000003e-06,
+      "loss": 0.2797,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8025889967637543,
+      "grad_norm": 0.17993331313633365,
+      "learning_rate": 3.6e-06,
+      "loss": 0.274,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8047464940668823,
+      "grad_norm": 1.7365736687950526,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 0.2749,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8069039913700107,
+      "grad_norm": 0.18738038459179354,
+      "learning_rate": 3.52e-06,
+      "loss": 0.2829,
+      "step": 1302
+    },
+    {
+      "epoch": 2.809061488673139,
+      "grad_norm": 0.17958359868903084,
+      "learning_rate": 3.4799999999999997e-06,
+      "loss": 0.2774,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8112189859762675,
+      "grad_norm": 0.17556904314160943,
+      "learning_rate": 3.44e-06,
+      "loss": 0.2751,
+      "step": 1304
+    },
+    {
+      "epoch": 2.813376483279396,
+      "grad_norm": 0.1788953079572772,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.2797,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.1881320873873,
+      "learning_rate": 3.36e-06,
+      "loss": 0.2873,
+      "step": 1306
+    },
+    {
+      "epoch": 2.8176914778856528,
+      "grad_norm": 0.17951722014152932,
+      "learning_rate": 3.3200000000000004e-06,
+      "loss": 0.2776,
+      "step": 1307
+    },
+    {
+      "epoch": 2.8198489751887807,
+      "grad_norm": 0.17798122918873352,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 0.2783,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8220064724919096,
+      "grad_norm": 0.18442039725945125,
+      "learning_rate": 3.24e-06,
+      "loss": 0.2878,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8241639697950376,
+      "grad_norm": 0.18259364011672352,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.2884,
+      "step": 1310
+    },
+    {
+      "epoch": 2.826321467098166,
+      "grad_norm": 0.1829848041342495,
+      "learning_rate": 3.1600000000000007e-06,
+      "loss": 0.2849,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8284789644012944,
+      "grad_norm": 0.179094433479786,
+      "learning_rate": 3.12e-06,
+      "loss": 0.2908,
+      "step": 1312
+    },
+    {
+      "epoch": 2.830636461704423,
+      "grad_norm": 0.17681991868248295,
+      "learning_rate": 3.08e-06,
+      "loss": 0.2807,
+      "step": 1313
+    },
+    {
+      "epoch": 2.8327939590075513,
+      "grad_norm": 0.17971050304421513,
+      "learning_rate": 3.04e-06,
+      "loss": 0.2848,
+      "step": 1314
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.1828969018946432,
+      "learning_rate": 3e-06,
+      "loss": 0.2912,
+      "step": 1315
+    },
+    {
+      "epoch": 2.837108953613808,
+      "grad_norm": 0.1751282473237561,
+      "learning_rate": 2.9600000000000005e-06,
+      "loss": 0.2695,
+      "step": 1316
+    },
+    {
+      "epoch": 2.839266450916936,
+      "grad_norm": 0.17477314465096877,
+      "learning_rate": 2.92e-06,
+      "loss": 0.2759,
+      "step": 1317
+    },
+    {
+      "epoch": 2.841423948220065,
+      "grad_norm": 0.17636269064952895,
+      "learning_rate": 2.88e-06,
+      "loss": 0.2864,
+      "step": 1318
+    },
+    {
+      "epoch": 2.843581445523193,
+      "grad_norm": 0.1818185234275195,
+      "learning_rate": 2.8400000000000003e-06,
+      "loss": 0.29,
+      "step": 1319
+    },
+    {
+      "epoch": 2.8457389428263213,
+      "grad_norm": 0.18916576451978248,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.2795,
+      "step": 1320
+    },
+    {
+      "epoch": 2.8478964401294498,
+      "grad_norm": 0.16973846926356626,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 0.2699,
+      "step": 1321
+    },
+    {
+      "epoch": 2.850053937432578,
+      "grad_norm": 0.17182169354600235,
+      "learning_rate": 2.72e-06,
+      "loss": 0.2791,
+      "step": 1322
+    },
+    {
+      "epoch": 2.8522114347357066,
+      "grad_norm": 0.1770412391282775,
+      "learning_rate": 2.68e-06,
+      "loss": 0.2765,
+      "step": 1323
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.17466980330096582,
+      "learning_rate": 2.64e-06,
+      "loss": 0.2623,
+      "step": 1324
+    },
+    {
+      "epoch": 2.8565264293419634,
+      "grad_norm": 0.174928731160995,
+      "learning_rate": 2.6e-06,
+      "loss": 0.2656,
+      "step": 1325
+    },
+    {
+      "epoch": 2.858683926645092,
+      "grad_norm": 0.18175215429014172,
+      "learning_rate": 2.56e-06,
+      "loss": 0.2819,
+      "step": 1326
+    },
+    {
+      "epoch": 2.8608414239482203,
+      "grad_norm": 0.1749725285284474,
+      "learning_rate": 2.52e-06,
+      "loss": 0.2725,
+      "step": 1327
+    },
+    {
+      "epoch": 2.8629989212513482,
+      "grad_norm": 0.17461652174068218,
+      "learning_rate": 2.48e-06,
+      "loss": 0.2696,
+      "step": 1328
+    },
+    {
+      "epoch": 2.8651564185544767,
+      "grad_norm": 0.1811005220241736,
+      "learning_rate": 2.4400000000000004e-06,
+      "loss": 0.2742,
+      "step": 1329
+    },
+    {
+      "epoch": 2.867313915857605,
+      "grad_norm": 0.17722407880597277,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.2773,
+      "step": 1330
+    },
+    {
+      "epoch": 2.8694714131607335,
+      "grad_norm": 0.18478937048714053,
+      "learning_rate": 2.36e-06,
+      "loss": 0.2762,
+      "step": 1331
+    },
+    {
+      "epoch": 2.871628910463862,
+      "grad_norm": 0.1864309200002423,
+      "learning_rate": 2.32e-06,
+      "loss": 0.2765,
+      "step": 1332
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.18835737178532266,
+      "learning_rate": 2.28e-06,
+      "loss": 0.3021,
+      "step": 1333
+    },
+    {
+      "epoch": 2.8759439050701188,
+      "grad_norm": 0.18323666534169097,
+      "learning_rate": 2.24e-06,
+      "loss": 0.2741,
+      "step": 1334
+    },
+    {
+      "epoch": 2.878101402373247,
+      "grad_norm": 0.17560962160831747,
+      "learning_rate": 2.2e-06,
+      "loss": 0.2828,
+      "step": 1335
+    },
+    {
+      "epoch": 2.8802588996763756,
+      "grad_norm": 0.18666287588971897,
+      "learning_rate": 2.16e-06,
+      "loss": 0.2954,
+      "step": 1336
+    },
+    {
+      "epoch": 2.8824163969795036,
+      "grad_norm": 0.17167493193172903,
+      "learning_rate": 2.12e-06,
+      "loss": 0.272,
+      "step": 1337
+    },
+    {
+      "epoch": 2.884573894282632,
+      "grad_norm": 0.17629098716034136,
+      "learning_rate": 2.08e-06,
+      "loss": 0.2889,
+      "step": 1338
+    },
+    {
+      "epoch": 2.8867313915857604,
+      "grad_norm": 0.18406413731232454,
+      "learning_rate": 2.0400000000000004e-06,
+      "loss": 0.2939,
+      "step": 1339
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.18163509827918725,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.2761,
+      "step": 1340
+    },
+    {
+      "epoch": 2.8910463861920173,
+      "grad_norm": 0.18223323103128686,
+      "learning_rate": 1.96e-06,
+      "loss": 0.2838,
+      "step": 1341
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.18297522132977273,
+      "learning_rate": 1.92e-06,
+      "loss": 0.2859,
+      "step": 1342
+    },
+    {
+      "epoch": 2.895361380798274,
+      "grad_norm": 0.1863689788725593,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 0.2662,
+      "step": 1343
+    },
+    {
+      "epoch": 2.8975188781014025,
+      "grad_norm": 0.17488508490885735,
+      "learning_rate": 1.84e-06,
+      "loss": 0.2756,
+      "step": 1344
+    },
+    {
+      "epoch": 2.899676375404531,
+      "grad_norm": 0.17992252244376258,
+      "learning_rate": 1.8e-06,
+      "loss": 0.264,
+      "step": 1345
+    },
+    {
+      "epoch": 2.901833872707659,
+      "grad_norm": 0.17566030078045597,
+      "learning_rate": 1.76e-06,
+      "loss": 0.2696,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9039913700107873,
+      "grad_norm": 0.17910981646975624,
+      "learning_rate": 1.72e-06,
+      "loss": 0.2812,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9061488673139158,
+      "grad_norm": 0.18324253604522966,
+      "learning_rate": 1.68e-06,
+      "loss": 0.2759,
+      "step": 1348
+    },
+    {
+      "epoch": 2.908306364617044,
+      "grad_norm": 0.18154686495512268,
+      "learning_rate": 1.6400000000000002e-06,
+      "loss": 0.2948,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9104638619201726,
+      "grad_norm": 0.1856157405155338,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.2842,
+      "step": 1350
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.167599852032292,
+      "learning_rate": 1.56e-06,
+      "loss": 0.2566,
+      "step": 1351
+    },
+    {
+      "epoch": 2.9147788565264294,
+      "grad_norm": 0.17436662357542593,
+      "learning_rate": 1.52e-06,
+      "loss": 0.2787,
+      "step": 1352
+    },
+    {
+      "epoch": 2.916936353829558,
+      "grad_norm": 0.1728537122922309,
+      "learning_rate": 1.4800000000000002e-06,
+      "loss": 0.2843,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9190938511326863,
+      "grad_norm": 0.17797631599058847,
+      "learning_rate": 1.44e-06,
+      "loss": 0.2651,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9212513484358142,
+      "grad_norm": 0.1770523988484935,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.2675,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9234088457389427,
+      "grad_norm": 0.19407900415997797,
+      "learning_rate": 1.36e-06,
+      "loss": 0.2845,
+      "step": 1356
+    },
+    {
+      "epoch": 2.925566343042071,
+      "grad_norm": 0.1788656382635837,
+      "learning_rate": 1.32e-06,
+      "loss": 0.2704,
+      "step": 1357
+    },
+    {
+      "epoch": 2.9277238403451995,
+      "grad_norm": 0.18340763904947952,
+      "learning_rate": 1.28e-06,
+      "loss": 0.2789,
+      "step": 1358
+    },
+    {
+      "epoch": 2.929881337648328,
+      "grad_norm": 0.17707734315166593,
+      "learning_rate": 1.24e-06,
+      "loss": 0.2822,
+      "step": 1359
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 0.17856073770387085,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.2727,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9341963322545848,
+      "grad_norm": 0.18303142619049678,
+      "learning_rate": 1.16e-06,
+      "loss": 0.2714,
+      "step": 1361
+    },
+    {
+      "epoch": 2.936353829557713,
+      "grad_norm": 0.17597510232889213,
+      "learning_rate": 1.12e-06,
+      "loss": 0.278,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9385113268608416,
+      "grad_norm": 0.18381504160428017,
+      "learning_rate": 1.08e-06,
+      "loss": 0.2793,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9406688241639696,
+      "grad_norm": 0.16872753735299875,
+      "learning_rate": 1.04e-06,
+      "loss": 0.2623,
+      "step": 1364
+    },
+    {
+      "epoch": 2.9428263214670984,
+      "grad_norm": 0.17739253303853683,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.2828,
+      "step": 1365
+    },
+    {
+      "epoch": 2.9449838187702264,
+      "grad_norm": 0.17698033237379485,
+      "learning_rate": 9.6e-07,
+      "loss": 0.2775,
+      "step": 1366
+    },
+    {
+      "epoch": 2.947141316073355,
+      "grad_norm": 0.17104830030403248,
+      "learning_rate": 9.2e-07,
+      "loss": 0.2748,
+      "step": 1367
+    },
+    {
+      "epoch": 2.9492988133764833,
+      "grad_norm": 0.17749083061324694,
+      "learning_rate": 8.8e-07,
+      "loss": 0.2794,
+      "step": 1368
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.1798254311534882,
+      "learning_rate": 8.4e-07,
+      "loss": 0.2788,
+      "step": 1369
+    },
+    {
+      "epoch": 2.95361380798274,
+      "grad_norm": 0.1784770144640976,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.2677,
+      "step": 1370
+    },
+    {
+      "epoch": 2.9557713052858685,
+      "grad_norm": 0.17694547600035745,
+      "learning_rate": 7.6e-07,
+      "loss": 0.2948,
+      "step": 1371
+    },
+    {
+      "epoch": 2.957928802588997,
+      "grad_norm": 0.1766645173046717,
+      "learning_rate": 7.2e-07,
+      "loss": 0.2752,
+      "step": 1372
+    },
+    {
+      "epoch": 2.960086299892125,
+      "grad_norm": 0.17231184142391467,
+      "learning_rate": 6.8e-07,
+      "loss": 0.2708,
+      "step": 1373
+    },
+    {
+      "epoch": 2.9622437971952538,
+      "grad_norm": 0.17699841769345795,
+      "learning_rate": 6.4e-07,
+      "loss": 0.283,
+      "step": 1374
+    },
+    {
+      "epoch": 2.9644012944983817,
+      "grad_norm": 0.17906935863892173,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.2805,
+      "step": 1375
+    },
+    {
+      "epoch": 2.96655879180151,
+      "grad_norm": 0.18194456358375216,
+      "learning_rate": 5.6e-07,
+      "loss": 0.2784,
+      "step": 1376
+    },
+    {
+      "epoch": 2.9687162891046386,
+      "grad_norm": 0.16849620824352807,
+      "learning_rate": 5.2e-07,
+      "loss": 0.2754,
+      "step": 1377
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 0.17761203976543524,
+      "learning_rate": 4.8e-07,
+      "loss": 0.2708,
+      "step": 1378
+    },
+    {
+      "epoch": 2.9730312837108954,
+      "grad_norm": 0.17640781813626127,
+      "learning_rate": 4.4e-07,
+      "loss": 0.2834,
+      "step": 1379
+    },
+    {
+      "epoch": 2.975188781014024,
+      "grad_norm": 0.17331909600308287,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.2743,
+      "step": 1380
+    },
+    {
+      "epoch": 2.9773462783171523,
+      "grad_norm": 0.17381405226314825,
+      "learning_rate": 3.6e-07,
+      "loss": 0.2782,
+      "step": 1381
+    },
+    {
+      "epoch": 2.9795037756202802,
+      "grad_norm": 0.17830336979426112,
+      "learning_rate": 3.2e-07,
+      "loss": 0.2758,
+      "step": 1382
+    },
+    {
+      "epoch": 2.981661272923409,
+      "grad_norm": 0.16766789787935527,
+      "learning_rate": 2.8e-07,
+      "loss": 0.2651,
+      "step": 1383
+    },
+    {
+      "epoch": 2.983818770226537,
+      "grad_norm": 0.18328063807292563,
+      "learning_rate": 2.4e-07,
+      "loss": 0.2894,
+      "step": 1384
+    },
+    {
+      "epoch": 2.9859762675296655,
+      "grad_norm": 0.17725873997381406,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.2812,
+      "step": 1385
+    },
+    {
+      "epoch": 2.988133764832794,
+      "grad_norm": 0.18575248175533493,
+      "learning_rate": 1.6e-07,
+      "loss": 0.2762,
+      "step": 1386
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.1738477732474402,
+      "learning_rate": 1.2e-07,
+      "loss": 0.2754,
+      "step": 1387
+    },
+    {
+      "epoch": 2.9924487594390508,
+      "grad_norm": 0.17133851982383583,
+      "learning_rate": 8e-08,
+      "loss": 0.2612,
+      "step": 1388
+    },
+    {
+      "epoch": 2.994606256742179,
+      "grad_norm": 0.18020219397411835,
+      "learning_rate": 4e-08,
+      "loss": 0.2814,
+      "step": 1389
+    },
+    {
+      "epoch": 2.994606256742179,
+      "step": 1389,
+      "total_flos": 1.5436844662748348e+19,
+      "train_loss": 0.4589164501432895,
+      "train_runtime": 43000.4811,
+      "train_samples_per_second": 0.517,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1389,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5436844662748348e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}