diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15297 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2181,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004585052728106373,
+      "grad_norm": 64.06384588549061,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 6.6334,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009170105456212746,
+      "grad_norm": 63.04628278224896,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 6.4631,
+      "step": 2
+    },
+    {
+      "epoch": 0.001375515818431912,
+      "grad_norm": 61.57004387037837,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 6.2854,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018340210912425492,
+      "grad_norm": 19.21088590921407,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 5.269,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022925263640531865,
+      "grad_norm": 12.98336262893473,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 4.7758,
+      "step": 5
+    },
+    {
+      "epoch": 0.002751031636863824,
+      "grad_norm": 16.81895680507727,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 4.6156,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032095369096744614,
+      "grad_norm": 9.568373630853339,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 4.6114,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036680421824850985,
+      "grad_norm": 9.863848244261543,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 4.4116,
+      "step": 8
+    },
+    {
+      "epoch": 0.0041265474552957355,
+      "grad_norm": 10.54119530793772,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 4.3961,
+      "step": 9
+    },
+    {
+      "epoch": 0.004585052728106373,
+      "grad_norm": 7.351081897297031,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 4.3068,
+      "step": 10
+    },
+    {
+      "epoch": 0.0050435580009170105,
+      "grad_norm": 13.714734834582426,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 4.3076,
+      "step": 11
+    },
+    {
+      "epoch": 0.005502063273727648,
+      "grad_norm": 5.840196582252703,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 4.2794,
+      "step": 12
+    },
+    {
+      "epoch": 0.005960568546538285,
+      "grad_norm": 4.313343742144532,
+      "learning_rate": 0.00019696969696969695,
+      "loss": 4.1323,
+      "step": 13
+    },
+    {
+      "epoch": 0.006419073819348923,
+      "grad_norm": 4.4650335713749865,
+      "learning_rate": 0.00021212121212121213,
+      "loss": 4.1877,
+      "step": 14
+    },
+    {
+      "epoch": 0.0068775790921595595,
+      "grad_norm": 4.863563048383204,
+      "learning_rate": 0.00022727272727272727,
+      "loss": 4.1201,
+      "step": 15
+    },
+    {
+      "epoch": 0.007336084364970197,
+      "grad_norm": 4.907474314949867,
+      "learning_rate": 0.00024242424242424245,
+      "loss": 4.0881,
+      "step": 16
+    },
+    {
+      "epoch": 0.007794589637780834,
+      "grad_norm": 5.769643980047134,
+      "learning_rate": 0.00025757575757575756,
+      "loss": 4.0822,
+      "step": 17
+    },
+    {
+      "epoch": 0.008253094910591471,
+      "grad_norm": 3.319450407868557,
+      "learning_rate": 0.00027272727272727274,
+      "loss": 3.9322,
+      "step": 18
+    },
+    {
+      "epoch": 0.00871160018340211,
+      "grad_norm": 2.7846990119050576,
+      "learning_rate": 0.0002878787878787879,
+      "loss": 3.9291,
+      "step": 19
+    },
+    {
+      "epoch": 0.009170105456212746,
+      "grad_norm": 3.0705868972096835,
+      "learning_rate": 0.00030303030303030303,
+      "loss": 4.0432,
+      "step": 20
+    },
+    {
+      "epoch": 0.009628610729023384,
+      "grad_norm": 2.8640057133863293,
+      "learning_rate": 0.0003181818181818182,
+      "loss": 4.034,
+      "step": 21
+    },
+    {
+      "epoch": 0.010087116001834021,
+      "grad_norm": 2.5290861829984874,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 3.8468,
+      "step": 22
+    },
+    {
+      "epoch": 0.01054562127464466,
+      "grad_norm": 2.3080507638741823,
+      "learning_rate": 0.0003484848484848485,
+      "loss": 3.8551,
+      "step": 23
+    },
+    {
+      "epoch": 0.011004126547455296,
+      "grad_norm": 2.148158752391558,
+      "learning_rate": 0.00036363636363636367,
+      "loss": 3.8665,
+      "step": 24
+    },
+    {
+      "epoch": 0.011462631820265932,
+      "grad_norm": 1.749905735376404,
+      "learning_rate": 0.0003787878787878788,
+      "loss": 3.8001,
+      "step": 25
+    },
+    {
+      "epoch": 0.01192113709307657,
+      "grad_norm": 1.5286615526998444,
+      "learning_rate": 0.0003939393939393939,
+      "loss": 3.7452,
+      "step": 26
+    },
+    {
+      "epoch": 0.012379642365887207,
+      "grad_norm": 1.465160435458748,
+      "learning_rate": 0.00040909090909090913,
+      "loss": 3.8447,
+      "step": 27
+    },
+    {
+      "epoch": 0.012838147638697846,
+      "grad_norm": 1.587829032632443,
+      "learning_rate": 0.00042424242424242425,
+      "loss": 3.7102,
+      "step": 28
+    },
+    {
+      "epoch": 0.013296652911508482,
+      "grad_norm": 1.245490242665303,
+      "learning_rate": 0.0004393939393939394,
+      "loss": 3.8242,
+      "step": 29
+    },
+    {
+      "epoch": 0.013755158184319119,
+      "grad_norm": 1.2578061178442352,
+      "learning_rate": 0.00045454545454545455,
+      "loss": 3.7161,
+      "step": 30
+    },
+    {
+      "epoch": 0.014213663457129757,
+      "grad_norm": 1.379469961281411,
+      "learning_rate": 0.0004696969696969697,
+      "loss": 3.745,
+      "step": 31
+    },
+    {
+      "epoch": 0.014672168729940394,
+      "grad_norm": 1.0044118794420578,
+      "learning_rate": 0.0004848484848484849,
+      "loss": 3.7487,
+      "step": 32
+    },
+    {
+      "epoch": 0.015130674002751032,
+      "grad_norm": 1.0299567648917083,
+      "learning_rate": 0.0005,
+      "loss": 3.7782,
+      "step": 33
+    },
+    {
+      "epoch": 0.015589179275561669,
+      "grad_norm": 1.007378521249317,
+      "learning_rate": 0.0005151515151515151,
+      "loss": 3.7056,
+      "step": 34
+    },
+    {
+      "epoch": 0.016047684548372305,
+      "grad_norm": 1.0863996052535372,
+      "learning_rate": 0.0005303030303030302,
+      "loss": 3.6394,
+      "step": 35
+    },
+    {
+      "epoch": 0.016506189821182942,
+      "grad_norm": 1.0238319257544128,
+      "learning_rate": 0.0005454545454545455,
+      "loss": 3.581,
+      "step": 36
+    },
+    {
+      "epoch": 0.016964695093993582,
+      "grad_norm": 0.774640101435605,
+      "learning_rate": 0.0005606060606060606,
+      "loss": 3.6271,
+      "step": 37
+    },
+    {
+      "epoch": 0.01742320036680422,
+      "grad_norm": 0.7752692069878444,
+      "learning_rate": 0.0005757575757575758,
+      "loss": 3.5815,
+      "step": 38
+    },
+    {
+      "epoch": 0.017881705639614855,
+      "grad_norm": 0.7167669752573647,
+      "learning_rate": 0.0005909090909090909,
+      "loss": 3.6287,
+      "step": 39
+    },
+    {
+      "epoch": 0.018340210912425492,
+      "grad_norm": 1.4405946372032805,
+      "learning_rate": 0.0006060606060606061,
+      "loss": 3.639,
+      "step": 40
+    },
+    {
+      "epoch": 0.01879871618523613,
+      "grad_norm": 0.8918710731638437,
+      "learning_rate": 0.0006212121212121212,
+      "loss": 3.4599,
+      "step": 41
+    },
+    {
+      "epoch": 0.01925722145804677,
+      "grad_norm": 0.7034387203202367,
+      "learning_rate": 0.0006363636363636364,
+      "loss": 3.5476,
+      "step": 42
+    },
+    {
+      "epoch": 0.019715726730857405,
+      "grad_norm": 0.7212247718112699,
+      "learning_rate": 0.0006515151515151515,
+      "loss": 3.6181,
+      "step": 43
+    },
+    {
+      "epoch": 0.020174232003668042,
+      "grad_norm": 1.0729455399061745,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 3.4984,
+      "step": 44
+    },
+    {
+      "epoch": 0.02063273727647868,
+      "grad_norm": 0.6955835270932273,
+      "learning_rate": 0.0006818181818181818,
+      "loss": 3.4588,
+      "step": 45
+    },
+    {
+      "epoch": 0.02109124254928932,
+      "grad_norm": 0.7559728179986343,
+      "learning_rate": 0.000696969696969697,
+      "loss": 3.5517,
+      "step": 46
+    },
+    {
+      "epoch": 0.021549747822099955,
+      "grad_norm": 0.9213356624036338,
+      "learning_rate": 0.0007121212121212122,
+      "loss": 3.4573,
+      "step": 47
+    },
+    {
+      "epoch": 0.02200825309491059,
+      "grad_norm": 0.7839176867365313,
+      "learning_rate": 0.0007272727272727273,
+      "loss": 3.4368,
+      "step": 48
+    },
+    {
+      "epoch": 0.02246675836772123,
+      "grad_norm": 0.7814550976949817,
+      "learning_rate": 0.0007424242424242425,
+      "loss": 3.6179,
+      "step": 49
+    },
+    {
+      "epoch": 0.022925263640531865,
+      "grad_norm": 0.9805199174713926,
+      "learning_rate": 0.0007575757575757576,
+      "loss": 3.4889,
+      "step": 50
+    },
+    {
+      "epoch": 0.023383768913342505,
+      "grad_norm": 0.903021565086016,
+      "learning_rate": 0.0007727272727272727,
+      "loss": 3.5133,
+      "step": 51
+    },
+    {
+      "epoch": 0.02384227418615314,
+      "grad_norm": 0.8165827895668291,
+      "learning_rate": 0.0007878787878787878,
+      "loss": 3.5525,
+      "step": 52
+    },
+    {
+      "epoch": 0.024300779458963778,
+      "grad_norm": 0.8354783891928327,
+      "learning_rate": 0.000803030303030303,
+      "loss": 3.548,
+      "step": 53
+    },
+    {
+      "epoch": 0.024759284731774415,
+      "grad_norm": 0.8651851006033142,
+      "learning_rate": 0.0008181818181818183,
+      "loss": 3.436,
+      "step": 54
+    },
+    {
+      "epoch": 0.02521779000458505,
+      "grad_norm": 0.9336437118201791,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 3.4644,
+      "step": 55
+    },
+    {
+      "epoch": 0.02567629527739569,
+      "grad_norm": 0.9532087291227316,
+      "learning_rate": 0.0008484848484848485,
+      "loss": 3.4784,
+      "step": 56
+    },
+    {
+      "epoch": 0.026134800550206328,
+      "grad_norm": 0.7672258744528782,
+      "learning_rate": 0.0008636363636363636,
+      "loss": 3.4309,
+      "step": 57
+    },
+    {
+      "epoch": 0.026593305823016965,
+      "grad_norm": 0.7730098445754204,
+      "learning_rate": 0.0008787878787878789,
+      "loss": 3.4592,
+      "step": 58
+    },
+    {
+      "epoch": 0.0270518110958276,
+      "grad_norm": 0.8532637157823285,
+      "learning_rate": 0.000893939393939394,
+      "loss": 3.3715,
+      "step": 59
+    },
+    {
+      "epoch": 0.027510316368638238,
+      "grad_norm": 0.7999969665834912,
+      "learning_rate": 0.0009090909090909091,
+      "loss": 3.3526,
+      "step": 60
+    },
+    {
+      "epoch": 0.027968821641448878,
+      "grad_norm": 0.7536202728972387,
+      "learning_rate": 0.0009242424242424242,
+      "loss": 3.2786,
+      "step": 61
+    },
+    {
+      "epoch": 0.028427326914259515,
+      "grad_norm": 0.8624918147101244,
+      "learning_rate": 0.0009393939393939394,
+      "loss": 3.316,
+      "step": 62
+    },
+    {
+      "epoch": 0.02888583218707015,
+      "grad_norm": 1.5405052709433964,
+      "learning_rate": 0.0009545454545454546,
+      "loss": 3.3668,
+      "step": 63
+    },
+    {
+      "epoch": 0.029344337459880788,
+      "grad_norm": 0.9095940577794858,
+      "learning_rate": 0.0009696969696969698,
+      "loss": 3.4612,
+      "step": 64
+    },
+    {
+      "epoch": 0.029802842732691424,
+      "grad_norm": 0.8319470923804443,
+      "learning_rate": 0.000984848484848485,
+      "loss": 3.2654,
+      "step": 65
+    },
+    {
+      "epoch": 0.030261348005502064,
+      "grad_norm": 0.92639889967443,
+      "learning_rate": 0.001,
+      "loss": 3.3401,
+      "step": 66
+    },
+    {
+      "epoch": 0.0307198532783127,
+      "grad_norm": 1.3688545451943757,
+      "learning_rate": 0.0009999994484067654,
+      "loss": 3.3137,
+      "step": 67
+    },
+    {
+      "epoch": 0.031178358551123338,
+      "grad_norm": 0.7828212964302271,
+      "learning_rate": 0.0009999977936282788,
+      "loss": 3.2398,
+      "step": 68
+    },
+    {
+      "epoch": 0.03163686382393398,
+      "grad_norm": 0.943192384286324,
+      "learning_rate": 0.0009999950356681913,
+      "loss": 3.3382,
+      "step": 69
+    },
+    {
+      "epoch": 0.03209536909674461,
+      "grad_norm": 0.9352832823034749,
+      "learning_rate": 0.0009999911745325876,
+      "loss": 3.2136,
+      "step": 70
+    },
+    {
+      "epoch": 0.03255387436955525,
+      "grad_norm": 0.8383791162027353,
+      "learning_rate": 0.0009999862102299873,
+      "loss": 3.2256,
+      "step": 71
+    },
+    {
+      "epoch": 0.033012379642365884,
+      "grad_norm": 0.9235471515939397,
+      "learning_rate": 0.0009999801427713433,
+      "loss": 3.3206,
+      "step": 72
+    },
+    {
+      "epoch": 0.033470884915176524,
+      "grad_norm": 0.8752497477331878,
+      "learning_rate": 0.0009999729721700424,
+      "loss": 3.1628,
+      "step": 73
+    },
+    {
+      "epoch": 0.033929390187987164,
+      "grad_norm": 0.8716057669989904,
+      "learning_rate": 0.000999964698441906,
+      "loss": 3.2271,
+      "step": 74
+    },
+    {
+      "epoch": 0.0343878954607978,
+      "grad_norm": 0.8719528526025119,
+      "learning_rate": 0.0009999553216051892,
+      "loss": 3.2758,
+      "step": 75
+    },
+    {
+      "epoch": 0.03484640073360844,
+      "grad_norm": 0.8359466395145875,
+      "learning_rate": 0.00099994484168058,
+      "loss": 3.2019,
+      "step": 76
+    },
+    {
+      "epoch": 0.03530490600641907,
+      "grad_norm": 0.9996800423031235,
+      "learning_rate": 0.0009999332586912019,
+      "loss": 3.2417,
+      "step": 77
+    },
+    {
+      "epoch": 0.03576341127922971,
+      "grad_norm": 0.8374052592635491,
+      "learning_rate": 0.0009999205726626108,
+      "loss": 3.2256,
+      "step": 78
+    },
+    {
+      "epoch": 0.03622191655204035,
+      "grad_norm": 0.8639575118638462,
+      "learning_rate": 0.000999906783622797,
+      "loss": 3.1663,
+      "step": 79
+    },
+    {
+      "epoch": 0.036680421824850984,
+      "grad_norm": 0.843886835649755,
+      "learning_rate": 0.0009998918916021842,
+      "loss": 3.2386,
+      "step": 80
+    },
+    {
+      "epoch": 0.037138927097661624,
+      "grad_norm": 0.9231042732904329,
+      "learning_rate": 0.0009998758966336297,
+      "loss": 3.1262,
+      "step": 81
+    },
+    {
+      "epoch": 0.03759743237047226,
+      "grad_norm": 0.9176475597084096,
+      "learning_rate": 0.0009998587987524242,
+      "loss": 3.188,
+      "step": 82
+    },
+    {
+      "epoch": 0.0380559376432829,
+      "grad_norm": 0.8697880010922603,
+      "learning_rate": 0.0009998405979962926,
+      "loss": 3.1521,
+      "step": 83
+    },
+    {
+      "epoch": 0.03851444291609354,
+      "grad_norm": 0.8790480753592875,
+      "learning_rate": 0.000999821294405392,
+      "loss": 3.1502,
+      "step": 84
+    },
+    {
+      "epoch": 0.03897294818890417,
+      "grad_norm": 0.9093272338362821,
+      "learning_rate": 0.0009998008880223134,
+      "loss": 3.2328,
+      "step": 85
+    },
+    {
+      "epoch": 0.03943145346171481,
+      "grad_norm": 1.010893019879814,
+      "learning_rate": 0.000999779378892081,
+      "loss": 3.2153,
+      "step": 86
+    },
+    {
+      "epoch": 0.039889958734525444,
+      "grad_norm": 0.9066171144044647,
+      "learning_rate": 0.0009997567670621522,
+      "loss": 3.0118,
+      "step": 87
+    },
+    {
+      "epoch": 0.040348464007336084,
+      "grad_norm": 0.9288231519140343,
+      "learning_rate": 0.0009997330525824165,
+      "loss": 3.1194,
+      "step": 88
+    },
+    {
+      "epoch": 0.040806969280146724,
+      "grad_norm": 0.8458693651493918,
+      "learning_rate": 0.0009997082355051976,
+      "loss": 3.0855,
+      "step": 89
+    },
+    {
+      "epoch": 0.04126547455295736,
+      "grad_norm": 0.9048386098405764,
+      "learning_rate": 0.000999682315885251,
+      "loss": 3.1434,
+      "step": 90
+    },
+    {
+      "epoch": 0.041723979825768,
+      "grad_norm": 0.8595175641107055,
+      "learning_rate": 0.0009996552937797645,
+      "loss": 3.051,
+      "step": 91
+    },
+    {
+      "epoch": 0.04218248509857864,
+      "grad_norm": 0.8988272184494203,
+      "learning_rate": 0.0009996271692483596,
+      "loss": 3.0146,
+      "step": 92
+    },
+    {
+      "epoch": 0.04264099037138927,
+      "grad_norm": 0.9752174145352055,
+      "learning_rate": 0.0009995979423530893,
+      "loss": 2.9835,
+      "step": 93
+    },
+    {
+      "epoch": 0.04309949564419991,
+      "grad_norm": 0.8756418396656522,
+      "learning_rate": 0.000999567613158439,
+      "loss": 3.1098,
+      "step": 94
+    },
+    {
+      "epoch": 0.04355800091701054,
+      "grad_norm": 0.9526672739655255,
+      "learning_rate": 0.0009995361817313263,
+      "loss": 3.0611,
+      "step": 95
+    },
+    {
+      "epoch": 0.04401650618982118,
+      "grad_norm": 0.9232976007103152,
+      "learning_rate": 0.0009995036481411004,
+      "loss": 3.1172,
+      "step": 96
+    },
+    {
+      "epoch": 0.044475011462631824,
+      "grad_norm": 0.8531943192196553,
+      "learning_rate": 0.0009994700124595429,
+      "loss": 3.0545,
+      "step": 97
+    },
+    {
+      "epoch": 0.04493351673544246,
+      "grad_norm": 0.8783200836563066,
+      "learning_rate": 0.0009994352747608663,
+      "loss": 3.0366,
+      "step": 98
+    },
+    {
+      "epoch": 0.0453920220082531,
+      "grad_norm": 0.9176735657026551,
+      "learning_rate": 0.0009993994351217151,
+      "loss": 2.9839,
+      "step": 99
+    },
+    {
+      "epoch": 0.04585052728106373,
+      "grad_norm": 0.8832045044523363,
+      "learning_rate": 0.000999362493621165,
+      "loss": 3.0298,
+      "step": 100
+    },
+    {
+      "epoch": 0.04630903255387437,
+      "grad_norm": 1.0941226171246186,
+      "learning_rate": 0.0009993244503407226,
+      "loss": 3.0902,
+      "step": 101
+    },
+    {
+      "epoch": 0.04676753782668501,
+      "grad_norm": 0.8910840868413406,
+      "learning_rate": 0.0009992853053643258,
+      "loss": 2.8847,
+      "step": 102
+    },
+    {
+      "epoch": 0.04722604309949564,
+      "grad_norm": 0.8431477649166056,
+      "learning_rate": 0.0009992450587783426,
+      "loss": 2.9794,
+      "step": 103
+    },
+    {
+      "epoch": 0.04768454837230628,
+      "grad_norm": 0.8872201548032451,
+      "learning_rate": 0.000999203710671572,
+      "loss": 3.0174,
+      "step": 104
+    },
+    {
+      "epoch": 0.048143053645116916,
+      "grad_norm": 0.8722540276367585,
+      "learning_rate": 0.0009991612611352438,
+      "loss": 2.9651,
+      "step": 105
+    },
+    {
+      "epoch": 0.048601558917927556,
+      "grad_norm": 0.928195521034802,
+      "learning_rate": 0.0009991177102630173,
+      "loss": 3.0096,
+      "step": 106
+    },
+    {
+      "epoch": 0.049060064190738197,
+      "grad_norm": 0.8960862839337044,
+      "learning_rate": 0.0009990730581509817,
+      "loss": 2.8581,
+      "step": 107
+    },
+    {
+      "epoch": 0.04951856946354883,
+      "grad_norm": 0.9293322956491185,
+      "learning_rate": 0.0009990273048976566,
+      "loss": 2.8872,
+      "step": 108
+    },
+    {
+      "epoch": 0.04997707473635947,
+      "grad_norm": 0.9221649476688292,
+      "learning_rate": 0.0009989804506039905,
+      "loss": 2.9163,
+      "step": 109
+    },
+    {
+      "epoch": 0.0504355800091701,
+      "grad_norm": 0.9711391210506485,
+      "learning_rate": 0.0009989324953733614,
+      "loss": 3.0088,
+      "step": 110
+    },
+    {
+      "epoch": 0.05089408528198074,
+      "grad_norm": 0.9613209168847097,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 2.9829,
+      "step": 111
+    },
+    {
+      "epoch": 0.05135259055479138,
+      "grad_norm": 0.9494268229532837,
+      "learning_rate": 0.000998833282526872,
+      "loss": 2.9281,
+      "step": 112
+    },
+    {
+      "epoch": 0.051811095827602016,
+      "grad_norm": 0.9406074977748234,
+      "learning_rate": 0.0009987820251299122,
+      "loss": 2.8763,
+      "step": 113
+    },
+    {
+      "epoch": 0.052269601100412656,
+      "grad_norm": 0.9006037811467329,
+      "learning_rate": 0.00099872966723379,
+      "loss": 2.955,
+      "step": 114
+    },
+    {
+      "epoch": 0.05272810637322329,
+      "grad_norm": 0.9187889112979474,
+      "learning_rate": 0.0009986762089540266,
+      "loss": 2.9328,
+      "step": 115
+    },
+    {
+      "epoch": 0.05318661164603393,
+      "grad_norm": 0.9222590880867172,
+      "learning_rate": 0.0009986216504085709,
+      "loss": 2.9632,
+      "step": 116
+    },
+    {
+      "epoch": 0.05364511691884457,
+      "grad_norm": 0.944393290343225,
+      "learning_rate": 0.0009985659917177991,
+      "loss": 3.0009,
+      "step": 117
+    },
+    {
+      "epoch": 0.0541036221916552,
+      "grad_norm": 0.8408588058353146,
+      "learning_rate": 0.0009985092330045155,
+      "loss": 2.8931,
+      "step": 118
+    },
+    {
+      "epoch": 0.05456212746446584,
+      "grad_norm": 0.9014564926608628,
+      "learning_rate": 0.0009984513743939508,
+      "loss": 2.9348,
+      "step": 119
+    },
+    {
+      "epoch": 0.055020632737276476,
+      "grad_norm": 0.8570085117815007,
+      "learning_rate": 0.0009983924160137626,
+      "loss": 2.9001,
+      "step": 120
+    },
+    {
+      "epoch": 0.055479138010087116,
+      "grad_norm": 0.8981020057426791,
+      "learning_rate": 0.000998332357994035,
+      "loss": 2.8184,
+      "step": 121
+    },
+    {
+      "epoch": 0.055937643282897756,
+      "grad_norm": 0.8748698489003712,
+      "learning_rate": 0.0009982712004672786,
+      "loss": 2.8567,
+      "step": 122
+    },
+    {
+      "epoch": 0.05639614855570839,
+      "grad_norm": 0.8755339824920007,
+      "learning_rate": 0.0009982089435684295,
+      "loss": 2.8533,
+      "step": 123
+    },
+    {
+      "epoch": 0.05685465382851903,
+      "grad_norm": 0.8809977489356686,
+      "learning_rate": 0.0009981455874348499,
+      "loss": 2.8844,
+      "step": 124
+    },
+    {
+      "epoch": 0.05731315910132966,
+      "grad_norm": 0.8811507847875845,
+      "learning_rate": 0.0009980811322063269,
+      "loss": 2.8869,
+      "step": 125
+    },
+    {
+      "epoch": 0.0577716643741403,
+      "grad_norm": 0.9291208547883063,
+      "learning_rate": 0.0009980155780250728,
+      "loss": 2.8429,
+      "step": 126
+    },
+    {
+      "epoch": 0.05823016964695094,
+      "grad_norm": 0.9249122727241481,
+      "learning_rate": 0.0009979489250357243,
+      "loss": 2.793,
+      "step": 127
+    },
+    {
+      "epoch": 0.058688674919761576,
+      "grad_norm": 0.8725765954345008,
+      "learning_rate": 0.0009978811733853431,
+      "loss": 2.8158,
+      "step": 128
+    },
+    {
+      "epoch": 0.059147180192572216,
+      "grad_norm": 0.8413455255409137,
+      "learning_rate": 0.0009978123232234147,
+      "loss": 2.7941,
+      "step": 129
+    },
+    {
+      "epoch": 0.05960568546538285,
+      "grad_norm": 0.8970432020136003,
+      "learning_rate": 0.000997742374701848,
+      "loss": 2.7949,
+      "step": 130
+    },
+    {
+      "epoch": 0.06006419073819349,
+      "grad_norm": 0.8854517528884317,
+      "learning_rate": 0.0009976713279749754,
+      "loss": 2.6579,
+      "step": 131
+    },
+    {
+      "epoch": 0.06052269601100413,
+      "grad_norm": 0.9982300294498815,
+      "learning_rate": 0.0009975991831995528,
+      "loss": 2.7973,
+      "step": 132
+    },
+    {
+      "epoch": 0.06098120128381476,
+      "grad_norm": 0.9448905507541657,
+      "learning_rate": 0.0009975259405347581,
+      "loss": 2.7945,
+      "step": 133
+    },
+    {
+      "epoch": 0.0614397065566254,
+      "grad_norm": 0.9054400687780981,
+      "learning_rate": 0.0009974516001421926,
+      "loss": 2.7864,
+      "step": 134
+    },
+    {
+      "epoch": 0.061898211829436035,
+      "grad_norm": 1.095173361320476,
+      "learning_rate": 0.000997376162185878,
+      "loss": 2.7962,
+      "step": 135
+    },
+    {
+      "epoch": 0.062356717102246675,
+      "grad_norm": 0.957145313355454,
+      "learning_rate": 0.0009972996268322594,
+      "loss": 2.7878,
+      "step": 136
+    },
+    {
+      "epoch": 0.06281522237505731,
+      "grad_norm": 0.9821870676801909,
+      "learning_rate": 0.0009972219942502017,
+      "loss": 2.8014,
+      "step": 137
+    },
+    {
+      "epoch": 0.06327372764786796,
+      "grad_norm": 0.8493905322722768,
+      "learning_rate": 0.0009971432646109918,
+      "loss": 2.7665,
+      "step": 138
+    },
+    {
+      "epoch": 0.06373223292067859,
+      "grad_norm": 0.8545195101048197,
+      "learning_rate": 0.0009970634380883365,
+      "loss": 2.6837,
+      "step": 139
+    },
+    {
+      "epoch": 0.06419073819348922,
+      "grad_norm": 0.8819727854827124,
+      "learning_rate": 0.0009969825148583627,
+      "loss": 2.72,
+      "step": 140
+    },
+    {
+      "epoch": 0.06464924346629987,
+      "grad_norm": 0.8884847957592712,
+      "learning_rate": 0.0009969004950996173,
+      "loss": 2.6808,
+      "step": 141
+    },
+    {
+      "epoch": 0.0651077487391105,
+      "grad_norm": 0.8668455827728292,
+      "learning_rate": 0.0009968173789930668,
+      "loss": 2.7916,
+      "step": 142
+    },
+    {
+      "epoch": 0.06556625401192114,
+      "grad_norm": 0.8994611216275532,
+      "learning_rate": 0.0009967331667220958,
+      "loss": 2.6876,
+      "step": 143
+    },
+    {
+      "epoch": 0.06602475928473177,
+      "grad_norm": 0.9238223051488685,
+      "learning_rate": 0.0009966478584725086,
+      "loss": 2.7362,
+      "step": 144
+    },
+    {
+      "epoch": 0.06648326455754242,
+      "grad_norm": 0.9355032513468372,
+      "learning_rate": 0.0009965614544325263,
+      "loss": 2.7288,
+      "step": 145
+    },
+    {
+      "epoch": 0.06694176983035305,
+      "grad_norm": 0.99584275991708,
+      "learning_rate": 0.000996473954792789,
+      "loss": 2.7345,
+      "step": 146
+    },
+    {
+      "epoch": 0.06740027510316368,
+      "grad_norm": 0.9001215441657587,
+      "learning_rate": 0.0009963853597463532,
+      "loss": 2.7016,
+      "step": 147
+    },
+    {
+      "epoch": 0.06785878037597433,
+      "grad_norm": 0.9887684480637634,
+      "learning_rate": 0.000996295669488693,
+      "loss": 2.7552,
+      "step": 148
+    },
+    {
+      "epoch": 0.06831728564878496,
+      "grad_norm": 0.8661609047091523,
+      "learning_rate": 0.0009962048842176979,
+      "loss": 2.7327,
+      "step": 149
+    },
+    {
+      "epoch": 0.0687757909215956,
+      "grad_norm": 0.9219073071718612,
+      "learning_rate": 0.0009961130041336748,
+      "loss": 2.6916,
+      "step": 150
+    },
+    {
+      "epoch": 0.06923429619440624,
+      "grad_norm": 0.9540262983265426,
+      "learning_rate": 0.0009960200294393449,
+      "loss": 2.7255,
+      "step": 151
+    },
+    {
+      "epoch": 0.06969280146721687,
+      "grad_norm": 0.899814636625909,
+      "learning_rate": 0.0009959259603398453,
+      "loss": 2.7198,
+      "step": 152
+    },
+    {
+      "epoch": 0.07015130674002751,
+      "grad_norm": 0.8765362298738244,
+      "learning_rate": 0.0009958307970427275,
+      "loss": 2.5785,
+      "step": 153
+    },
+    {
+      "epoch": 0.07060981201283814,
+      "grad_norm": 0.8340917946360522,
+      "learning_rate": 0.0009957345397579572,
+      "loss": 2.6711,
+      "step": 154
+    },
+    {
+      "epoch": 0.07106831728564879,
+      "grad_norm": 0.8980150037920029,
+      "learning_rate": 0.0009956371886979138,
+      "loss": 2.6712,
+      "step": 155
+    },
+    {
+      "epoch": 0.07152682255845942,
+      "grad_norm": 0.9505379192458385,
+      "learning_rate": 0.00099553874407739,
+      "loss": 2.6985,
+      "step": 156
+    },
+    {
+      "epoch": 0.07198532783127005,
+      "grad_norm": 0.9414803636035102,
+      "learning_rate": 0.0009954392061135916,
+      "loss": 2.6805,
+      "step": 157
+    },
+    {
+      "epoch": 0.0724438331040807,
+      "grad_norm": 1.0477344491044243,
+      "learning_rate": 0.0009953385750261364,
+      "loss": 2.667,
+      "step": 158
+    },
+    {
+      "epoch": 0.07290233837689133,
+      "grad_norm": 0.9941733196593143,
+      "learning_rate": 0.0009952368510370538,
+      "loss": 2.6779,
+      "step": 159
+    },
+    {
+      "epoch": 0.07336084364970197,
+      "grad_norm": 0.8597120019819413,
+      "learning_rate": 0.0009951340343707852,
+      "loss": 2.7003,
+      "step": 160
+    },
+    {
+      "epoch": 0.07381934892251261,
+      "grad_norm": 0.8237559241684576,
+      "learning_rate": 0.0009950301252541823,
+      "loss": 2.622,
+      "step": 161
+    },
+    {
+      "epoch": 0.07427785419532325,
+      "grad_norm": 0.8112250969179563,
+      "learning_rate": 0.0009949251239165075,
+      "loss": 2.5887,
+      "step": 162
+    },
+    {
+      "epoch": 0.07473635946813388,
+      "grad_norm": 0.8431994885143822,
+      "learning_rate": 0.000994819030589433,
+      "loss": 2.6998,
+      "step": 163
+    },
+    {
+      "epoch": 0.07519486474094451,
+      "grad_norm": 0.8557656708573522,
+      "learning_rate": 0.00099471184550704,
+      "loss": 2.5925,
+      "step": 164
+    },
+    {
+      "epoch": 0.07565337001375516,
+      "grad_norm": 0.9140707111397387,
+      "learning_rate": 0.0009946035689058189,
+      "loss": 2.6313,
+      "step": 165
+    },
+    {
+      "epoch": 0.0761118752865658,
+      "grad_norm": 0.9442091470118326,
+      "learning_rate": 0.0009944942010246681,
+      "loss": 2.7035,
+      "step": 166
+    },
+    {
+      "epoch": 0.07657038055937643,
+      "grad_norm": 0.9143456077675453,
+      "learning_rate": 0.0009943837421048942,
+      "loss": 2.6392,
+      "step": 167
+    },
+    {
+      "epoch": 0.07702888583218707,
+      "grad_norm": 0.8375702520030174,
+      "learning_rate": 0.0009942721923902106,
+      "loss": 2.5455,
+      "step": 168
+    },
+    {
+      "epoch": 0.07748739110499771,
+      "grad_norm": 0.8290606540297394,
+      "learning_rate": 0.0009941595521267377,
+      "loss": 2.489,
+      "step": 169
+    },
+    {
+      "epoch": 0.07794589637780834,
+      "grad_norm": 0.8034509175046176,
+      "learning_rate": 0.0009940458215630017,
+      "loss": 2.6001,
+      "step": 170
+    },
+    {
+      "epoch": 0.07840440165061899,
+      "grad_norm": 0.8223949586067094,
+      "learning_rate": 0.0009939310009499348,
+      "loss": 2.6797,
+      "step": 171
+    },
+    {
+      "epoch": 0.07886290692342962,
+      "grad_norm": 0.8501564812944725,
+      "learning_rate": 0.000993815090540874,
+      "loss": 2.6555,
+      "step": 172
+    },
+    {
+      "epoch": 0.07932141219624025,
+      "grad_norm": 0.8156375363133714,
+      "learning_rate": 0.000993698090591561,
+      "loss": 2.6819,
+      "step": 173
+    },
+    {
+      "epoch": 0.07977991746905089,
+      "grad_norm": 0.8210159237652641,
+      "learning_rate": 0.0009935800013601416,
+      "loss": 2.5994,
+      "step": 174
+    },
+    {
+      "epoch": 0.08023842274186153,
+      "grad_norm": 0.8185189974574195,
+      "learning_rate": 0.000993460823107164,
+      "loss": 2.5974,
+      "step": 175
+    },
+    {
+      "epoch": 0.08069692801467217,
+      "grad_norm": 0.7925881326183667,
+      "learning_rate": 0.0009933405560955803,
+      "loss": 2.6195,
+      "step": 176
+    },
+    {
+      "epoch": 0.0811554332874828,
+      "grad_norm": 0.7922701106928629,
+      "learning_rate": 0.0009932192005907446,
+      "loss": 2.5951,
+      "step": 177
+    },
+    {
+      "epoch": 0.08161393856029345,
+      "grad_norm": 0.8542433400881634,
+      "learning_rate": 0.0009930967568604118,
+      "loss": 2.6806,
+      "step": 178
+    },
+    {
+      "epoch": 0.08207244383310408,
+      "grad_norm": 0.810724356764697,
+      "learning_rate": 0.000992973225174739,
+      "loss": 2.6131,
+      "step": 179
+    },
+    {
+      "epoch": 0.08253094910591471,
+      "grad_norm": 0.8415543766741641,
+      "learning_rate": 0.0009928486058062827,
+      "loss": 2.6257,
+      "step": 180
+    },
+    {
+      "epoch": 0.08298945437872536,
+      "grad_norm": 0.8075843591602587,
+      "learning_rate": 0.0009927228990299999,
+      "loss": 2.6535,
+      "step": 181
+    },
+    {
+      "epoch": 0.083447959651536,
+      "grad_norm": 0.864572649793034,
+      "learning_rate": 0.0009925961051232468,
+      "loss": 2.6419,
+      "step": 182
+    },
+    {
+      "epoch": 0.08390646492434663,
+      "grad_norm": 0.8369293136386641,
+      "learning_rate": 0.000992468224365778,
+      "loss": 2.5631,
+      "step": 183
+    },
+    {
+      "epoch": 0.08436497019715727,
+      "grad_norm": 0.7780656484700785,
+      "learning_rate": 0.000992339257039746,
+      "loss": 2.6021,
+      "step": 184
+    },
+    {
+      "epoch": 0.08482347546996791,
+      "grad_norm": 0.8441926989100492,
+      "learning_rate": 0.0009922092034297006,
+      "loss": 2.5293,
+      "step": 185
+    },
+    {
+      "epoch": 0.08528198074277854,
+      "grad_norm": 0.827170714559148,
+      "learning_rate": 0.0009920780638225891,
+      "loss": 2.5356,
+      "step": 186
+    },
+    {
+      "epoch": 0.08574048601558917,
+      "grad_norm": 0.8332278395974163,
+      "learning_rate": 0.0009919458385077538,
+      "loss": 2.569,
+      "step": 187
+    },
+    {
+      "epoch": 0.08619899128839982,
+      "grad_norm": 0.7813759359542132,
+      "learning_rate": 0.0009918125277769336,
+      "loss": 2.5352,
+      "step": 188
+    },
+    {
+      "epoch": 0.08665749656121045,
+      "grad_norm": 0.7689297365660832,
+      "learning_rate": 0.0009916781319242614,
+      "loss": 2.6197,
+      "step": 189
+    },
+    {
+      "epoch": 0.08711600183402109,
+      "grad_norm": 0.7472739341092418,
+      "learning_rate": 0.0009915426512462646,
+      "loss": 2.612,
+      "step": 190
+    },
+    {
+      "epoch": 0.08757450710683173,
+      "grad_norm": 0.8303208843243266,
+      "learning_rate": 0.0009914060860418644,
+      "loss": 2.5895,
+      "step": 191
+    },
+    {
+      "epoch": 0.08803301237964237,
+      "grad_norm": 0.7778831098180473,
+      "learning_rate": 0.000991268436612374,
+      "loss": 2.6382,
+      "step": 192
+    },
+    {
+      "epoch": 0.088491517652453,
+      "grad_norm": 0.8275037520845242,
+      "learning_rate": 0.0009911297032614997,
+      "loss": 2.6637,
+      "step": 193
+    },
+    {
+      "epoch": 0.08895002292526365,
+      "grad_norm": 0.8694581587593712,
+      "learning_rate": 0.000990989886295339,
+      "loss": 2.5324,
+      "step": 194
+    },
+    {
+      "epoch": 0.08940852819807428,
+      "grad_norm": 0.9015639752602367,
+      "learning_rate": 0.0009908489860223804,
+      "loss": 2.577,
+      "step": 195
+    },
+    {
+      "epoch": 0.08986703347088491,
+      "grad_norm": 0.817099147714937,
+      "learning_rate": 0.000990707002753502,
+      "loss": 2.5553,
+      "step": 196
+    },
+    {
+      "epoch": 0.09032553874369555,
+      "grad_norm": 2.0863700402168135,
+      "learning_rate": 0.0009905639368019724,
+      "loss": 2.5859,
+      "step": 197
+    },
+    {
+      "epoch": 0.0907840440165062,
+      "grad_norm": 0.8615452828942325,
+      "learning_rate": 0.0009904197884834482,
+      "loss": 2.5652,
+      "step": 198
+    },
+    {
+      "epoch": 0.09124254928931683,
+      "grad_norm": 0.8535861296089272,
+      "learning_rate": 0.0009902745581159742,
+      "loss": 2.5812,
+      "step": 199
+    },
+    {
+      "epoch": 0.09170105456212746,
+      "grad_norm": 0.9651417127366376,
+      "learning_rate": 0.0009901282460199829,
+      "loss": 2.6255,
+      "step": 200
+    },
+    {
+      "epoch": 0.0921595598349381,
+      "grad_norm": 1.540670560536462,
+      "learning_rate": 0.0009899808525182935,
+      "loss": 2.5968,
+      "step": 201
+    },
+    {
+      "epoch": 0.09261806510774874,
+      "grad_norm": 0.8405256372507366,
+      "learning_rate": 0.0009898323779361107,
+      "loss": 2.5313,
+      "step": 202
+    },
+    {
+      "epoch": 0.09307657038055937,
+      "grad_norm": 0.9717689147313783,
+      "learning_rate": 0.000989682822601025,
+      "loss": 2.5481,
+      "step": 203
+    },
+    {
+      "epoch": 0.09353507565337002,
+      "grad_norm": 0.9172342192255246,
+      "learning_rate": 0.0009895321868430113,
+      "loss": 2.5391,
+      "step": 204
+    },
+    {
+      "epoch": 0.09399358092618065,
+      "grad_norm": 0.9202982396857077,
+      "learning_rate": 0.0009893804709944281,
+      "loss": 2.5944,
+      "step": 205
+    },
+    {
+      "epoch": 0.09445208619899129,
+      "grad_norm": 0.9149991508174432,
+      "learning_rate": 0.0009892276753900174,
+      "loss": 2.5879,
+      "step": 206
+    },
+    {
+      "epoch": 0.09491059147180192,
+      "grad_norm": 0.9501281059781731,
+      "learning_rate": 0.0009890738003669028,
+      "loss": 2.6055,
+      "step": 207
+    },
+    {
+      "epoch": 0.09536909674461257,
+      "grad_norm": 0.9633019736111224,
+      "learning_rate": 0.0009889188462645904,
+      "loss": 2.5845,
+      "step": 208
+    },
+    {
+      "epoch": 0.0958276020174232,
+      "grad_norm": 0.9332772570402804,
+      "learning_rate": 0.0009887628134249667,
+      "loss": 2.6135,
+      "step": 209
+    },
+    {
+      "epoch": 0.09628610729023383,
+      "grad_norm": 0.9571723532295567,
+      "learning_rate": 0.0009886057021922983,
+      "loss": 2.6681,
+      "step": 210
+    },
+    {
+      "epoch": 0.09674461256304448,
+      "grad_norm": 0.910703688840028,
+      "learning_rate": 0.0009884475129132311,
+      "loss": 2.5998,
+      "step": 211
+    },
+    {
+      "epoch": 0.09720311783585511,
+      "grad_norm": 0.8651501362159841,
+      "learning_rate": 0.0009882882459367897,
+      "loss": 2.4517,
+      "step": 212
+    },
+    {
+      "epoch": 0.09766162310866575,
+      "grad_norm": 0.7947046386688171,
+      "learning_rate": 0.0009881279016143766,
+      "loss": 2.3879,
+      "step": 213
+    },
+    {
+      "epoch": 0.09812012838147639,
+      "grad_norm": 0.8041599040563431,
+      "learning_rate": 0.0009879664802997707,
+      "loss": 2.5711,
+      "step": 214
+    },
+    {
+      "epoch": 0.09857863365428703,
+      "grad_norm": 0.816991660311495,
+      "learning_rate": 0.000987803982349128,
+      "loss": 2.6077,
+      "step": 215
+    },
+    {
+      "epoch": 0.09903713892709766,
+      "grad_norm": 0.8520209151446931,
+      "learning_rate": 0.0009876404081209796,
+      "loss": 2.6152,
+      "step": 216
+    },
+    {
+      "epoch": 0.09949564419990829,
+      "grad_norm": 1.122038263107235,
+      "learning_rate": 0.000987475757976231,
+      "loss": 2.5778,
+      "step": 217
+    },
+    {
+      "epoch": 0.09995414947271894,
+      "grad_norm": 0.7759472190931527,
+      "learning_rate": 0.000987310032278162,
+      "loss": 2.5612,
+      "step": 218
+    },
+    {
+      "epoch": 0.10041265474552957,
+      "grad_norm": 0.7761042044537182,
+      "learning_rate": 0.0009871432313924254,
+      "loss": 2.4998,
+      "step": 219
+    },
+    {
+      "epoch": 0.1008711600183402,
+      "grad_norm": 0.863617896662466,
+      "learning_rate": 0.000986975355687046,
+      "loss": 2.524,
+      "step": 220
+    },
+    {
+      "epoch": 0.10132966529115085,
+      "grad_norm": 0.8026596415089576,
+      "learning_rate": 0.0009868064055324204,
+      "loss": 2.5556,
+      "step": 221
+    },
+    {
+      "epoch": 0.10178817056396149,
+      "grad_norm": 0.8147040876400714,
+      "learning_rate": 0.0009866363813013153,
+      "loss": 2.5807,
+      "step": 222
+    },
+    {
+      "epoch": 0.10224667583677212,
+      "grad_norm": 1.2226261296690708,
+      "learning_rate": 0.0009864652833688676,
+      "loss": 2.5449,
+      "step": 223
+    },
+    {
+      "epoch": 0.10270518110958277,
+      "grad_norm": 0.8606322844826574,
+      "learning_rate": 0.0009862931121125836,
+      "loss": 2.5272,
+      "step": 224
+    },
+    {
+      "epoch": 0.1031636863823934,
+      "grad_norm": 0.8757067618905842,
+      "learning_rate": 0.000986119867912337,
+      "loss": 2.5704,
+      "step": 225
+    },
+    {
+      "epoch": 0.10362219165520403,
+      "grad_norm": 0.788923855677848,
+      "learning_rate": 0.000985945551150369,
+      "loss": 2.55,
+      "step": 226
+    },
+    {
+      "epoch": 0.10408069692801467,
+      "grad_norm": 0.8337646165245537,
+      "learning_rate": 0.0009857701622112876,
+      "loss": 2.4971,
+      "step": 227
+    },
+    {
+      "epoch": 0.10453920220082531,
+      "grad_norm": 0.8524152872351768,
+      "learning_rate": 0.000985593701482066,
+      "loss": 2.5966,
+      "step": 228
+    },
+    {
+      "epoch": 0.10499770747363595,
+      "grad_norm": 0.850597949769858,
+      "learning_rate": 0.0009854161693520424,
+      "loss": 2.4737,
+      "step": 229
+    },
+    {
+      "epoch": 0.10545621274644658,
+      "grad_norm": 0.790762714291422,
+      "learning_rate": 0.0009852375662129194,
+      "loss": 2.5293,
+      "step": 230
+    },
+    {
+      "epoch": 0.10591471801925723,
+      "grad_norm": 0.7999544405211173,
+      "learning_rate": 0.0009850578924587613,
+      "loss": 2.4728,
+      "step": 231
+    },
+    {
+      "epoch": 0.10637322329206786,
+      "grad_norm": 0.7614623521611457,
+      "learning_rate": 0.000984877148485996,
+      "loss": 2.5797,
+      "step": 232
+    },
+    {
+      "epoch": 0.10683172856487849,
+      "grad_norm": 0.8344695354579652,
+      "learning_rate": 0.000984695334693412,
+      "loss": 2.6222,
+      "step": 233
+    },
+    {
+      "epoch": 0.10729023383768914,
+      "grad_norm": 0.7623479005497719,
+      "learning_rate": 0.000984512451482158,
+      "loss": 2.561,
+      "step": 234
+    },
+    {
+      "epoch": 0.10774873911049977,
+      "grad_norm": 1.3402380456960339,
+      "learning_rate": 0.0009843284992557431,
+      "loss": 2.5035,
+      "step": 235
+    },
+    {
+      "epoch": 0.1082072443833104,
+      "grad_norm": 2.157205422687693,
+      "learning_rate": 0.000984143478420034,
+      "loss": 2.6011,
+      "step": 236
+    },
+    {
+      "epoch": 0.10866574965612105,
+      "grad_norm": 0.8154376015307658,
+      "learning_rate": 0.0009839573893832563,
+      "loss": 2.5013,
+      "step": 237
+    },
+    {
+      "epoch": 0.10912425492893169,
+      "grad_norm": 0.774321656107681,
+      "learning_rate": 0.000983770232555991,
+      "loss": 2.6261,
+      "step": 238
+    },
+    {
+      "epoch": 0.10958276020174232,
+      "grad_norm": 0.766211779324363,
+      "learning_rate": 0.0009835820083511765,
+      "loss": 2.4778,
+      "step": 239
+    },
+    {
+      "epoch": 0.11004126547455295,
+      "grad_norm": 0.7553635407596048,
+      "learning_rate": 0.0009833927171841055,
+      "loss": 2.6288,
+      "step": 240
+    },
+    {
+      "epoch": 0.1104997707473636,
+      "grad_norm": 0.7234450803322496,
+      "learning_rate": 0.0009832023594724246,
+      "loss": 2.615,
+      "step": 241
+    },
+    {
+      "epoch": 0.11095827602017423,
+      "grad_norm": 0.7903660995025273,
+      "learning_rate": 0.0009830109356361344,
+      "loss": 2.544,
+      "step": 242
+    },
+    {
+      "epoch": 0.11141678129298486,
+      "grad_norm": 0.753171436052527,
+      "learning_rate": 0.0009828184460975867,
+      "loss": 2.5393,
+      "step": 243
+    },
+    {
+      "epoch": 0.11187528656579551,
+      "grad_norm": 0.6808259681126628,
+      "learning_rate": 0.0009826248912814855,
+      "loss": 2.4951,
+      "step": 244
+    },
+    {
+      "epoch": 0.11233379183860615,
+      "grad_norm": 0.7488926987274572,
+      "learning_rate": 0.0009824302716148847,
+      "loss": 2.5503,
+      "step": 245
+    },
+    {
+      "epoch": 0.11279229711141678,
+      "grad_norm": 0.7626887777881333,
+      "learning_rate": 0.0009822345875271884,
+      "loss": 2.5025,
+      "step": 246
+    },
+    {
+      "epoch": 0.11325080238422743,
+      "grad_norm": 0.9204446983432812,
+      "learning_rate": 0.0009820378394501481,
+      "loss": 2.5503,
+      "step": 247
+    },
+    {
+      "epoch": 0.11370930765703806,
+      "grad_norm": 0.84159772380273,
+      "learning_rate": 0.0009818400278178636,
+      "loss": 2.4905,
+      "step": 248
+    },
+    {
+      "epoch": 0.11416781292984869,
+      "grad_norm": 0.7839792076671365,
+      "learning_rate": 0.0009816411530667814,
+      "loss": 2.4957,
+      "step": 249
+    },
+    {
+      "epoch": 0.11462631820265932,
+      "grad_norm": 0.779332676999668,
+      "learning_rate": 0.000981441215635693,
+      "loss": 2.5427,
+      "step": 250
+    },
+    {
+      "epoch": 0.11508482347546997,
+      "grad_norm": 0.80172189005947,
+      "learning_rate": 0.0009812402159657353,
+      "loss": 2.5808,
+      "step": 251
+    },
+    {
+      "epoch": 0.1155433287482806,
+      "grad_norm": 0.7773943787529896,
+      "learning_rate": 0.000981038154500388,
+      "loss": 2.5557,
+      "step": 252
+    },
+    {
+      "epoch": 0.11600183402109124,
+      "grad_norm": 0.8019319962483501,
+      "learning_rate": 0.0009808350316854746,
+      "loss": 2.5549,
+      "step": 253
+    },
+    {
+      "epoch": 0.11646033929390188,
+      "grad_norm": 0.7353254518774628,
+      "learning_rate": 0.0009806308479691594,
+      "loss": 2.5871,
+      "step": 254
+    },
+    {
+      "epoch": 0.11691884456671252,
+      "grad_norm": 0.7165668408176703,
+      "learning_rate": 0.0009804256038019482,
+      "loss": 2.4631,
+      "step": 255
+    },
+    {
+      "epoch": 0.11737734983952315,
+      "grad_norm": 0.7569316037535392,
+      "learning_rate": 0.0009802192996366857,
+      "loss": 2.5255,
+      "step": 256
+    },
+    {
+      "epoch": 0.1178358551123338,
+      "grad_norm": 0.7197882183974436,
+      "learning_rate": 0.0009800119359285563,
+      "loss": 2.5067,
+      "step": 257
+    },
+    {
+      "epoch": 0.11829436038514443,
+      "grad_norm": 0.6856715382422884,
+      "learning_rate": 0.0009798035131350813,
+      "loss": 2.5157,
+      "step": 258
+    },
+    {
+      "epoch": 0.11875286565795506,
+      "grad_norm": 0.6896203350946658,
+      "learning_rate": 0.0009795940317161194,
+      "loss": 2.461,
+      "step": 259
+    },
+    {
+      "epoch": 0.1192113709307657,
+      "grad_norm": 0.7416990862777822,
+      "learning_rate": 0.0009793834921338646,
+      "loss": 2.5544,
+      "step": 260
+    },
+    {
+      "epoch": 0.11966987620357634,
+      "grad_norm": 0.7553448231720905,
+      "learning_rate": 0.0009791718948528457,
+      "loss": 2.6162,
+      "step": 261
+    },
+    {
+      "epoch": 0.12012838147638698,
+      "grad_norm": 0.7352209336113454,
+      "learning_rate": 0.0009789592403399252,
+      "loss": 2.4887,
+      "step": 262
+    },
+    {
+      "epoch": 0.12058688674919761,
+      "grad_norm": 0.7642549154376351,
+      "learning_rate": 0.0009787455290642985,
+      "loss": 2.474,
+      "step": 263
+    },
+    {
+      "epoch": 0.12104539202200826,
+      "grad_norm": 0.7741556196356297,
+      "learning_rate": 0.000978530761497492,
+      "loss": 2.5252,
+      "step": 264
+    },
+    {
+      "epoch": 0.12150389729481889,
+      "grad_norm": 0.774196631513075,
+      "learning_rate": 0.0009783149381133633,
+      "loss": 2.4674,
+      "step": 265
+    },
+    {
+      "epoch": 0.12196240256762952,
+      "grad_norm": 0.7701613432294399,
+      "learning_rate": 0.0009780980593880992,
+      "loss": 2.4682,
+      "step": 266
+    },
+    {
+      "epoch": 0.12242090784044017,
+      "grad_norm": 0.7315386946826637,
+      "learning_rate": 0.0009778801258002153,
+      "loss": 2.5002,
+      "step": 267
+    },
+    {
+      "epoch": 0.1228794131132508,
+      "grad_norm": 0.7378607673975602,
+      "learning_rate": 0.000977661137830554,
+      "loss": 2.4919,
+      "step": 268
+    },
+    {
+      "epoch": 0.12333791838606144,
+      "grad_norm": 0.7239994829890649,
+      "learning_rate": 0.0009774410959622845,
+      "loss": 2.5025,
+      "step": 269
+    },
+    {
+      "epoch": 0.12379642365887207,
+      "grad_norm": 0.6822844606469755,
+      "learning_rate": 0.000977220000680901,
+      "loss": 2.4473,
+      "step": 270
+    },
+    {
+      "epoch": 0.12425492893168272,
+      "grad_norm": 0.6904541676294224,
+      "learning_rate": 0.000976997852474223,
+      "loss": 2.4113,
+      "step": 271
+    },
+    {
+      "epoch": 0.12471343420449335,
+      "grad_norm": 0.6550667375633697,
+      "learning_rate": 0.0009767746518323914,
+      "loss": 2.4707,
+      "step": 272
+    },
+    {
+      "epoch": 0.12517193947730398,
+      "grad_norm": 0.7097033885533001,
+      "learning_rate": 0.0009765503992478704,
+      "loss": 2.4252,
+      "step": 273
+    },
+    {
+      "epoch": 0.12563044475011462,
+      "grad_norm": 0.779139193259377,
+      "learning_rate": 0.0009763250952154449,
+      "loss": 2.4958,
+      "step": 274
+    },
+    {
+      "epoch": 0.12608895002292525,
+      "grad_norm": 0.7478268582409902,
+      "learning_rate": 0.0009760987402322195,
+      "loss": 2.4611,
+      "step": 275
+    },
+    {
+      "epoch": 0.1265474552957359,
+      "grad_norm": 0.7184300928742021,
+      "learning_rate": 0.0009758713347976178,
+      "loss": 2.5586,
+      "step": 276
+    },
+    {
+      "epoch": 0.12700596056854654,
+      "grad_norm": 0.6724211765536978,
+      "learning_rate": 0.000975642879413381,
+      "loss": 2.4989,
+      "step": 277
+    },
+    {
+      "epoch": 0.12746446584135718,
+      "grad_norm": 0.6270476991114001,
+      "learning_rate": 0.0009754133745835665,
+      "loss": 2.4836,
+      "step": 278
+    },
+    {
+      "epoch": 0.1279229711141678,
+      "grad_norm": 0.7035510408244395,
+      "learning_rate": 0.0009751828208145482,
+      "loss": 2.4852,
+      "step": 279
+    },
+    {
+      "epoch": 0.12838147638697844,
+      "grad_norm": 0.7261595829753892,
+      "learning_rate": 0.0009749512186150131,
+      "loss": 2.5271,
+      "step": 280
+    },
+    {
+      "epoch": 0.12883998165978908,
+      "grad_norm": 0.7015626361083215,
+      "learning_rate": 0.0009747185684959625,
+      "loss": 2.4948,
+      "step": 281
+    },
+    {
+      "epoch": 0.12929848693259974,
+      "grad_norm": 0.6271357066205661,
+      "learning_rate": 0.000974484870970709,
+      "loss": 2.4751,
+      "step": 282
+    },
+    {
+      "epoch": 0.12975699220541037,
+      "grad_norm": 0.624771194001488,
+      "learning_rate": 0.0009742501265548767,
+      "loss": 2.4635,
+      "step": 283
+    },
+    {
+      "epoch": 0.130215497478221,
+      "grad_norm": 0.6937683296359666,
+      "learning_rate": 0.0009740143357663993,
+      "loss": 2.5667,
+      "step": 284
+    },
+    {
+      "epoch": 0.13067400275103164,
+      "grad_norm": 0.6737419971763093,
+      "learning_rate": 0.000973777499125519,
+      "loss": 2.4439,
+      "step": 285
+    },
+    {
+      "epoch": 0.13113250802384227,
+      "grad_norm": 0.6567992678407314,
+      "learning_rate": 0.0009735396171547859,
+      "loss": 2.4533,
+      "step": 286
+    },
+    {
+      "epoch": 0.1315910132966529,
+      "grad_norm": 0.8778714438076232,
+      "learning_rate": 0.0009733006903790564,
+      "loss": 2.5233,
+      "step": 287
+    },
+    {
+      "epoch": 0.13204951856946354,
+      "grad_norm": 0.9797185482234102,
+      "learning_rate": 0.0009730607193254922,
+      "loss": 2.5284,
+      "step": 288
+    },
+    {
+      "epoch": 0.1325080238422742,
+      "grad_norm": 0.6211087818096137,
+      "learning_rate": 0.0009728197045235585,
+      "loss": 2.5092,
+      "step": 289
+    },
+    {
+      "epoch": 0.13296652911508483,
+      "grad_norm": 0.7077468183839278,
+      "learning_rate": 0.0009725776465050242,
+      "loss": 2.498,
+      "step": 290
+    },
+    {
+      "epoch": 0.13342503438789546,
+      "grad_norm": 0.6938520759559873,
+      "learning_rate": 0.0009723345458039594,
+      "loss": 2.4717,
+      "step": 291
+    },
+    {
+      "epoch": 0.1338835396607061,
+      "grad_norm": 0.784412625562576,
+      "learning_rate": 0.000972090402956735,
+      "loss": 2.4693,
+      "step": 292
+    },
+    {
+      "epoch": 0.13434204493351673,
+      "grad_norm": 0.6424996185318996,
+      "learning_rate": 0.0009718452185020212,
+      "loss": 2.4544,
+      "step": 293
+    },
+    {
+      "epoch": 0.13480055020632736,
+      "grad_norm": 0.6838017324403197,
+      "learning_rate": 0.0009715989929807862,
+      "loss": 2.4473,
+      "step": 294
+    },
+    {
+      "epoch": 0.13525905547913802,
+      "grad_norm": 0.6654010542331833,
+      "learning_rate": 0.0009713517269362955,
+      "loss": 2.4786,
+      "step": 295
+    },
+    {
+      "epoch": 0.13571756075194866,
+      "grad_norm": 0.6909124276441109,
+      "learning_rate": 0.0009711034209141101,
+      "loss": 2.4975,
+      "step": 296
+    },
+    {
+      "epoch": 0.1361760660247593,
+      "grad_norm": 0.6845480881945775,
+      "learning_rate": 0.0009708540754620856,
+      "loss": 2.548,
+      "step": 297
+    },
+    {
+      "epoch": 0.13663457129756992,
+      "grad_norm": 0.7457903855655589,
+      "learning_rate": 0.0009706036911303713,
+      "loss": 2.541,
+      "step": 298
+    },
+    {
+      "epoch": 0.13709307657038056,
+      "grad_norm": 0.7980343003573036,
+      "learning_rate": 0.0009703522684714083,
+      "loss": 2.4077,
+      "step": 299
+    },
+    {
+      "epoch": 0.1375515818431912,
+      "grad_norm": 0.9857084374236642,
+      "learning_rate": 0.0009700998080399286,
+      "loss": 2.4685,
+      "step": 300
+    },
+    {
+      "epoch": 0.13801008711600182,
+      "grad_norm": 0.6937764548695107,
+      "learning_rate": 0.0009698463103929542,
+      "loss": 2.4785,
+      "step": 301
+    },
+    {
+      "epoch": 0.13846859238881248,
+      "grad_norm": 0.6385965025753567,
+      "learning_rate": 0.0009695917760897954,
+      "loss": 2.4734,
+      "step": 302
+    },
+    {
+      "epoch": 0.13892709766162312,
+      "grad_norm": 0.6654828312794292,
+      "learning_rate": 0.0009693362056920501,
+      "loss": 2.5071,
+      "step": 303
+    },
+    {
+      "epoch": 0.13938560293443375,
+      "grad_norm": 0.642889039280102,
+      "learning_rate": 0.0009690795997636015,
+      "loss": 2.473,
+      "step": 304
+    },
+    {
+      "epoch": 0.13984410820724438,
+      "grad_norm": 0.6800950492629694,
+      "learning_rate": 0.0009688219588706179,
+      "loss": 2.4134,
+      "step": 305
+    },
+    {
+      "epoch": 0.14030261348005502,
+      "grad_norm": 0.6887959361778566,
+      "learning_rate": 0.0009685632835815518,
+      "loss": 2.4422,
+      "step": 306
+    },
+    {
+      "epoch": 0.14076111875286565,
+      "grad_norm": 0.7031544036810055,
+      "learning_rate": 0.0009683035744671367,
+      "loss": 2.491,
+      "step": 307
+    },
+    {
+      "epoch": 0.14121962402567628,
+      "grad_norm": 0.6651551517127353,
+      "learning_rate": 0.0009680428321003883,
+      "loss": 2.4704,
+      "step": 308
+    },
+    {
+      "epoch": 0.14167812929848694,
+      "grad_norm": 0.6895286029749306,
+      "learning_rate": 0.000967781057056601,
+      "loss": 2.4788,
+      "step": 309
+    },
+    {
+      "epoch": 0.14213663457129758,
+      "grad_norm": 0.6819523188191505,
+      "learning_rate": 0.0009675182499133485,
+      "loss": 2.5327,
+      "step": 310
+    },
+    {
+      "epoch": 0.1425951398441082,
+      "grad_norm": 0.6472564175744564,
+      "learning_rate": 0.0009672544112504813,
+      "loss": 2.422,
+      "step": 311
+    },
+    {
+      "epoch": 0.14305364511691884,
+      "grad_norm": 0.7446390423319355,
+      "learning_rate": 0.0009669895416501257,
+      "loss": 2.4547,
+      "step": 312
+    },
+    {
+      "epoch": 0.14351215038972948,
+      "grad_norm": 0.6718731668081336,
+      "learning_rate": 0.0009667236416966833,
+      "loss": 2.5368,
+      "step": 313
+    },
+    {
+      "epoch": 0.1439706556625401,
+      "grad_norm": 0.6743248535693379,
+      "learning_rate": 0.0009664567119768281,
+      "loss": 2.4563,
+      "step": 314
+    },
+    {
+      "epoch": 0.14442916093535077,
+      "grad_norm": 0.6941160331302662,
+      "learning_rate": 0.0009661887530795067,
+      "loss": 2.4311,
+      "step": 315
+    },
+    {
+      "epoch": 0.1448876662081614,
+      "grad_norm": 0.6626328597418961,
+      "learning_rate": 0.0009659197655959365,
+      "loss": 2.4497,
+      "step": 316
+    },
+    {
+      "epoch": 0.14534617148097204,
+      "grad_norm": 0.7032899441236751,
+      "learning_rate": 0.000965649750119604,
+      "loss": 2.5017,
+      "step": 317
+    },
+    {
+      "epoch": 0.14580467675378267,
+      "grad_norm": 0.7019473959917736,
+      "learning_rate": 0.0009653787072462643,
+      "loss": 2.4486,
+      "step": 318
+    },
+    {
+      "epoch": 0.1462631820265933,
+      "grad_norm": 0.632529486842431,
+      "learning_rate": 0.0009651066375739388,
+      "loss": 2.5125,
+      "step": 319
+    },
+    {
+      "epoch": 0.14672168729940394,
+      "grad_norm": 0.6933962113709652,
+      "learning_rate": 0.000964833541702915,
+      "loss": 2.4699,
+      "step": 320
+    },
+    {
+      "epoch": 0.14718019257221457,
+      "grad_norm": 0.6188187680176952,
+      "learning_rate": 0.0009645594202357438,
+      "loss": 2.4784,
+      "step": 321
+    },
+    {
+      "epoch": 0.14763869784502523,
+      "grad_norm": 0.6838722046751796,
+      "learning_rate": 0.0009642842737772397,
+      "loss": 2.4583,
+      "step": 322
+    },
+    {
+      "epoch": 0.14809720311783586,
+      "grad_norm": 0.6738477928636039,
+      "learning_rate": 0.0009640081029344782,
+      "loss": 2.4695,
+      "step": 323
+    },
+    {
+      "epoch": 0.1485557083906465,
+      "grad_norm": 0.6581742488345185,
+      "learning_rate": 0.0009637309083167956,
+      "loss": 2.5394,
+      "step": 324
+    },
+    {
+      "epoch": 0.14901421366345713,
+      "grad_norm": 0.7523601359368494,
+      "learning_rate": 0.0009634526905357859,
+      "loss": 2.4978,
+      "step": 325
+    },
+    {
+      "epoch": 0.14947271893626776,
+      "grad_norm": 0.7938639753398039,
+      "learning_rate": 0.000963173450205302,
+      "loss": 2.5335,
+      "step": 326
+    },
+    {
+      "epoch": 0.1499312242090784,
+      "grad_norm": 0.5928309735163766,
+      "learning_rate": 0.0009628931879414517,
+      "loss": 2.3718,
+      "step": 327
+    },
+    {
+      "epoch": 0.15038972948188903,
+      "grad_norm": 0.9634152039187717,
+      "learning_rate": 0.0009626119043625983,
+      "loss": 2.5071,
+      "step": 328
+    },
+    {
+      "epoch": 0.1508482347546997,
+      "grad_norm": 0.622342657899485,
+      "learning_rate": 0.0009623296000893582,
+      "loss": 2.402,
+      "step": 329
+    },
+    {
+      "epoch": 0.15130674002751032,
+      "grad_norm": 0.6368204890115748,
+      "learning_rate": 0.0009620462757446,
+      "loss": 2.439,
+      "step": 330
+    },
+    {
+      "epoch": 0.15176524530032096,
+      "grad_norm": 0.6760948881780416,
+      "learning_rate": 0.0009617619319534428,
+      "loss": 2.524,
+      "step": 331
+    },
+    {
+      "epoch": 0.1522237505731316,
+      "grad_norm": 0.6659288003686442,
+      "learning_rate": 0.000961476569343255,
+      "loss": 2.4672,
+      "step": 332
+    },
+    {
+      "epoch": 0.15268225584594222,
+      "grad_norm": 0.637767245863715,
+      "learning_rate": 0.0009611901885436529,
+      "loss": 2.4718,
+      "step": 333
+    },
+    {
+      "epoch": 0.15314076111875286,
+      "grad_norm": 0.6653918161887667,
+      "learning_rate": 0.0009609027901864996,
+      "loss": 2.4011,
+      "step": 334
+    },
+    {
+      "epoch": 0.15359926639156352,
+      "grad_norm": 0.710218139638482,
+      "learning_rate": 0.0009606143749059029,
+      "loss": 2.4559,
+      "step": 335
+    },
+    {
+      "epoch": 0.15405777166437415,
+      "grad_norm": 0.6853434398046934,
+      "learning_rate": 0.0009603249433382144,
+      "loss": 2.4761,
+      "step": 336
+    },
+    {
+      "epoch": 0.15451627693718478,
+      "grad_norm": 0.6556059561101525,
+      "learning_rate": 0.0009600344961220282,
+      "loss": 2.4782,
+      "step": 337
+    },
+    {
+      "epoch": 0.15497478220999542,
+      "grad_norm": 1.2426718945995174,
+      "learning_rate": 0.0009597430338981791,
+      "loss": 2.4355,
+      "step": 338
+    },
+    {
+      "epoch": 0.15543328748280605,
+      "grad_norm": 0.6159398567818948,
+      "learning_rate": 0.0009594505573097414,
+      "loss": 2.4364,
+      "step": 339
+    },
+    {
+      "epoch": 0.15589179275561668,
+      "grad_norm": 0.7057403257703008,
+      "learning_rate": 0.0009591570670020277,
+      "loss": 2.5406,
+      "step": 340
+    },
+    {
+      "epoch": 0.15635029802842731,
+      "grad_norm": 0.628052096649464,
+      "learning_rate": 0.0009588625636225871,
+      "loss": 2.3957,
+      "step": 341
+    },
+    {
+      "epoch": 0.15680880330123798,
+      "grad_norm": 0.6540186320645353,
+      "learning_rate": 0.0009585670478212036,
+      "loss": 2.2885,
+      "step": 342
+    },
+    {
+      "epoch": 0.1572673085740486,
+      "grad_norm": 0.7333731551707939,
+      "learning_rate": 0.0009582705202498956,
+      "loss": 2.4517,
+      "step": 343
+    },
+    {
+      "epoch": 0.15772581384685924,
+      "grad_norm": 0.6786641396367893,
+      "learning_rate": 0.0009579729815629133,
+      "loss": 2.4921,
+      "step": 344
+    },
+    {
+      "epoch": 0.15818431911966988,
+      "grad_norm": 0.6968162909344411,
+      "learning_rate": 0.0009576744324167379,
+      "loss": 2.5498,
+      "step": 345
+    },
+    {
+      "epoch": 0.1586428243924805,
+      "grad_norm": 0.6054584183119073,
+      "learning_rate": 0.0009573748734700804,
+      "loss": 2.3662,
+      "step": 346
+    },
+    {
+      "epoch": 0.15910132966529114,
+      "grad_norm": 0.6775445352736315,
+      "learning_rate": 0.0009570743053838796,
+      "loss": 2.5176,
+      "step": 347
+    },
+    {
+      "epoch": 0.15955983493810177,
+      "grad_norm": 0.6855904858729381,
+      "learning_rate": 0.0009567727288213005,
+      "loss": 2.431,
+      "step": 348
+    },
+    {
+      "epoch": 0.16001834021091244,
+      "grad_norm": 0.636401880145319,
+      "learning_rate": 0.0009564701444477337,
+      "loss": 2.4833,
+      "step": 349
+    },
+    {
+      "epoch": 0.16047684548372307,
+      "grad_norm": 1.1934066523506788,
+      "learning_rate": 0.000956166552930793,
+      "loss": 2.4148,
+      "step": 350
+    },
+    {
+      "epoch": 0.1609353507565337,
+      "grad_norm": 0.6946219681374941,
+      "learning_rate": 0.0009558619549403147,
+      "loss": 2.5828,
+      "step": 351
+    },
+    {
+      "epoch": 0.16139385602934433,
+      "grad_norm": 0.6022430180303605,
+      "learning_rate": 0.0009555563511483555,
+      "loss": 2.4442,
+      "step": 352
+    },
+    {
+      "epoch": 0.16185236130215497,
+      "grad_norm": 0.6922808775229436,
+      "learning_rate": 0.0009552497422291912,
+      "loss": 2.3898,
+      "step": 353
+    },
+    {
+      "epoch": 0.1623108665749656,
+      "grad_norm": 0.6461979996700966,
+      "learning_rate": 0.0009549421288593157,
+      "loss": 2.4764,
+      "step": 354
+    },
+    {
+      "epoch": 0.16276937184777626,
+      "grad_norm": 1.4887590329759381,
+      "learning_rate": 0.0009546335117174385,
+      "loss": 2.487,
+      "step": 355
+    },
+    {
+      "epoch": 0.1632278771205869,
+      "grad_norm": 0.8046411812284407,
+      "learning_rate": 0.0009543238914844843,
+      "loss": 2.5103,
+      "step": 356
+    },
+    {
+      "epoch": 0.16368638239339753,
+      "grad_norm": 0.7400497895845382,
+      "learning_rate": 0.0009540132688435907,
+      "loss": 2.4683,
+      "step": 357
+    },
+    {
+      "epoch": 0.16414488766620816,
+      "grad_norm": 0.6858839166986538,
+      "learning_rate": 0.0009537016444801074,
+      "loss": 2.3262,
+      "step": 358
+    },
+    {
+      "epoch": 0.1646033929390188,
+      "grad_norm": 0.7459273751981653,
+      "learning_rate": 0.0009533890190815935,
+      "loss": 2.5991,
+      "step": 359
+    },
+    {
+      "epoch": 0.16506189821182943,
+      "grad_norm": 0.7362788541837902,
+      "learning_rate": 0.0009530753933378173,
+      "loss": 2.4832,
+      "step": 360
+    },
+    {
+      "epoch": 0.16552040348464006,
+      "grad_norm": 0.6678543874790517,
+      "learning_rate": 0.0009527607679407545,
+      "loss": 2.4393,
+      "step": 361
+    },
+    {
+      "epoch": 0.16597890875745072,
+      "grad_norm": 0.6243641323271697,
+      "learning_rate": 0.0009524451435845857,
+      "loss": 2.4726,
+      "step": 362
+    },
+    {
+      "epoch": 0.16643741403026135,
+      "grad_norm": 0.7008745332070948,
+      "learning_rate": 0.0009521285209656963,
+      "loss": 2.4381,
+      "step": 363
+    },
+    {
+      "epoch": 0.166895919303072,
+      "grad_norm": 0.6985724228429964,
+      "learning_rate": 0.0009518109007826734,
+      "loss": 2.3702,
+      "step": 364
+    },
+    {
+      "epoch": 0.16735442457588262,
+      "grad_norm": 0.7078665802939645,
+      "learning_rate": 0.0009514922837363059,
+      "loss": 2.4865,
+      "step": 365
+    },
+    {
+      "epoch": 0.16781292984869325,
+      "grad_norm": 0.6286945160596263,
+      "learning_rate": 0.0009511726705295817,
+      "loss": 2.4837,
+      "step": 366
+    },
+    {
+      "epoch": 0.1682714351215039,
+      "grad_norm": 0.7343749235919143,
+      "learning_rate": 0.000950852061867687,
+      "loss": 2.4541,
+      "step": 367
+    },
+    {
+      "epoch": 0.16872994039431455,
+      "grad_norm": 0.6871997817148916,
+      "learning_rate": 0.0009505304584580038,
+      "loss": 2.4222,
+      "step": 368
+    },
+    {
+      "epoch": 0.16918844566712518,
+      "grad_norm": 0.6254288658581991,
+      "learning_rate": 0.0009502078610101092,
+      "loss": 2.5299,
+      "step": 369
+    },
+    {
+      "epoch": 0.16964695093993581,
+      "grad_norm": 0.6160284877823878,
+      "learning_rate": 0.0009498842702357736,
+      "loss": 2.4759,
+      "step": 370
+    },
+    {
+      "epoch": 0.17010545621274645,
+      "grad_norm": 0.6714024437883471,
+      "learning_rate": 0.0009495596868489587,
+      "loss": 2.4588,
+      "step": 371
+    },
+    {
+      "epoch": 0.17056396148555708,
+      "grad_norm": 0.6574319855699623,
+      "learning_rate": 0.0009492341115658167,
+      "loss": 2.5223,
+      "step": 372
+    },
+    {
+      "epoch": 0.1710224667583677,
+      "grad_norm": 0.6464217073126208,
+      "learning_rate": 0.0009489075451046879,
+      "loss": 2.4075,
+      "step": 373
+    },
+    {
+      "epoch": 0.17148097203117835,
+      "grad_norm": 0.5888800548457166,
+      "learning_rate": 0.0009485799881861,
+      "loss": 2.3675,
+      "step": 374
+    },
+    {
+      "epoch": 0.171939477303989,
+      "grad_norm": 0.7050705873662378,
+      "learning_rate": 0.0009482514415327654,
+      "loss": 2.5248,
+      "step": 375
+    },
+    {
+      "epoch": 0.17239798257679964,
+      "grad_norm": 0.7811604308401119,
+      "learning_rate": 0.000947921905869581,
+      "loss": 2.4406,
+      "step": 376
+    },
+    {
+      "epoch": 0.17285648784961027,
+      "grad_norm": 0.6875361198045673,
+      "learning_rate": 0.0009475913819236248,
+      "loss": 2.5738,
+      "step": 377
+    },
+    {
+      "epoch": 0.1733149931224209,
+      "grad_norm": 0.6194160819641699,
+      "learning_rate": 0.0009472598704241561,
+      "loss": 2.4878,
+      "step": 378
+    },
+    {
+      "epoch": 0.17377349839523154,
+      "grad_norm": 1.4965968387985127,
+      "learning_rate": 0.0009469273721026131,
+      "loss": 2.4295,
+      "step": 379
+    },
+    {
+      "epoch": 0.17423200366804217,
+      "grad_norm": 0.6609375959874257,
+      "learning_rate": 0.0009465938876926111,
+      "loss": 2.3786,
+      "step": 380
+    },
+    {
+      "epoch": 0.1746905089408528,
+      "grad_norm": 0.5958844897639317,
+      "learning_rate": 0.0009462594179299406,
+      "loss": 2.5537,
+      "step": 381
+    },
+    {
+      "epoch": 0.17514901421366347,
+      "grad_norm": 0.6548211847331097,
+      "learning_rate": 0.0009459239635525672,
+      "loss": 2.4642,
+      "step": 382
+    },
+    {
+      "epoch": 0.1756075194864741,
+      "grad_norm": 0.61690626006445,
+      "learning_rate": 0.0009455875253006281,
+      "loss": 2.4766,
+      "step": 383
+    },
+    {
+      "epoch": 0.17606602475928473,
+      "grad_norm": 0.6303648804314816,
+      "learning_rate": 0.0009452501039164315,
+      "loss": 2.5927,
+      "step": 384
+    },
+    {
+      "epoch": 0.17652453003209537,
+      "grad_norm": 0.6068718979367786,
+      "learning_rate": 0.0009449117001444549,
+      "loss": 2.3686,
+      "step": 385
+    },
+    {
+      "epoch": 0.176983035304906,
+      "grad_norm": 0.5789991436172282,
+      "learning_rate": 0.0009445723147313433,
+      "loss": 2.5056,
+      "step": 386
+    },
+    {
+      "epoch": 0.17744154057771663,
+      "grad_norm": 0.6291609353546281,
+      "learning_rate": 0.0009442319484259074,
+      "loss": 2.435,
+      "step": 387
+    },
+    {
+      "epoch": 0.1779000458505273,
+      "grad_norm": 0.6061101673074379,
+      "learning_rate": 0.0009438906019791222,
+      "loss": 2.5018,
+      "step": 388
+    },
+    {
+      "epoch": 0.17835855112333793,
+      "grad_norm": 0.8061266692270788,
+      "learning_rate": 0.0009435482761441251,
+      "loss": 2.5119,
+      "step": 389
+    },
+    {
+      "epoch": 0.17881705639614856,
+      "grad_norm": 0.6804001510671381,
+      "learning_rate": 0.000943204971676215,
+      "loss": 2.5111,
+      "step": 390
+    },
+    {
+      "epoch": 0.1792755616689592,
+      "grad_norm": 0.607455672576489,
+      "learning_rate": 0.0009428606893328493,
+      "loss": 2.4434,
+      "step": 391
+    },
+    {
+      "epoch": 0.17973406694176983,
+      "grad_norm": 0.5996489457421569,
+      "learning_rate": 0.0009425154298736432,
+      "loss": 2.4805,
+      "step": 392
+    },
+    {
+      "epoch": 0.18019257221458046,
+      "grad_norm": 0.580743766391417,
+      "learning_rate": 0.0009421691940603678,
+      "loss": 2.3847,
+      "step": 393
+    },
+    {
+      "epoch": 0.1806510774873911,
+      "grad_norm": 0.6093938703091121,
+      "learning_rate": 0.0009418219826569488,
+      "loss": 2.5662,
+      "step": 394
+    },
+    {
+      "epoch": 0.18110958276020175,
+      "grad_norm": 0.5935236619439335,
+      "learning_rate": 0.0009414737964294635,
+      "loss": 2.3106,
+      "step": 395
+    },
+    {
+      "epoch": 0.1815680880330124,
+      "grad_norm": 0.5869678743990359,
+      "learning_rate": 0.000941124636146141,
+      "loss": 2.4536,
+      "step": 396
+    },
+    {
+      "epoch": 0.18202659330582302,
+      "grad_norm": 0.6441633574344733,
+      "learning_rate": 0.0009407745025773589,
+      "loss": 2.3607,
+      "step": 397
+    },
+    {
+      "epoch": 0.18248509857863365,
+      "grad_norm": 0.5910548363383453,
+      "learning_rate": 0.0009404233964956423,
+      "loss": 2.3125,
+      "step": 398
+    },
+    {
+      "epoch": 0.1829436038514443,
+      "grad_norm": 0.638920788351737,
+      "learning_rate": 0.0009400713186756625,
+      "loss": 2.464,
+      "step": 399
+    },
+    {
+      "epoch": 0.18340210912425492,
+      "grad_norm": 0.6150940165419447,
+      "learning_rate": 0.0009397182698942342,
+      "loss": 2.4063,
+      "step": 400
+    },
+    {
+      "epoch": 0.18386061439706555,
+      "grad_norm": 0.610303359035554,
+      "learning_rate": 0.0009393642509303149,
+      "loss": 2.4845,
+      "step": 401
+    },
+    {
+      "epoch": 0.1843191196698762,
+      "grad_norm": 0.6456150515550119,
+      "learning_rate": 0.0009390092625650023,
+      "loss": 2.4289,
+      "step": 402
+    },
+    {
+      "epoch": 0.18477762494268685,
+      "grad_norm": 0.6243901814880781,
+      "learning_rate": 0.0009386533055815332,
+      "loss": 2.4258,
+      "step": 403
+    },
+    {
+      "epoch": 0.18523613021549748,
+      "grad_norm": 0.605247468081977,
+      "learning_rate": 0.0009382963807652813,
+      "loss": 2.4029,
+      "step": 404
+    },
+    {
+      "epoch": 0.1856946354883081,
+      "grad_norm": 0.6163361034213573,
+      "learning_rate": 0.000937938488903756,
+      "loss": 2.4426,
+      "step": 405
+    },
+    {
+      "epoch": 0.18615314076111875,
+      "grad_norm": 0.6241071623242778,
+      "learning_rate": 0.0009375796307866003,
+      "loss": 2.4616,
+      "step": 406
+    },
+    {
+      "epoch": 0.18661164603392938,
+      "grad_norm": 0.5696317990165988,
+      "learning_rate": 0.0009372198072055888,
+      "loss": 2.4472,
+      "step": 407
+    },
+    {
+      "epoch": 0.18707015130674004,
+      "grad_norm": 0.9380501129029856,
+      "learning_rate": 0.0009368590189546268,
+      "loss": 2.3894,
+      "step": 408
+    },
+    {
+      "epoch": 0.18752865657955067,
+      "grad_norm": 0.5891803376161919,
+      "learning_rate": 0.0009364972668297474,
+      "loss": 2.4628,
+      "step": 409
+    },
+    {
+      "epoch": 0.1879871618523613,
+      "grad_norm": 0.6097368408214533,
+      "learning_rate": 0.0009361345516291111,
+      "loss": 2.4296,
+      "step": 410
+    },
+    {
+      "epoch": 0.18844566712517194,
+      "grad_norm": 0.5884990239106032,
+      "learning_rate": 0.0009357708741530024,
+      "loss": 2.4416,
+      "step": 411
+    },
+    {
+      "epoch": 0.18890417239798257,
+      "grad_norm": 0.6300658414752518,
+      "learning_rate": 0.00093540623520383,
+      "loss": 2.4137,
+      "step": 412
+    },
+    {
+      "epoch": 0.1893626776707932,
+      "grad_norm": 0.5938338156647535,
+      "learning_rate": 0.000935040635586123,
+      "loss": 2.4155,
+      "step": 413
+    },
+    {
+      "epoch": 0.18982118294360384,
+      "grad_norm": 0.5953010175159519,
+      "learning_rate": 0.0009346740761065305,
+      "loss": 2.4144,
+      "step": 414
+    },
+    {
+      "epoch": 0.1902796882164145,
+      "grad_norm": 0.6814649271444966,
+      "learning_rate": 0.0009343065575738197,
+      "loss": 2.4272,
+      "step": 415
+    },
+    {
+      "epoch": 0.19073819348922513,
+      "grad_norm": 0.5740675489938089,
+      "learning_rate": 0.0009339380807988733,
+      "loss": 2.4077,
+      "step": 416
+    },
+    {
+      "epoch": 0.19119669876203577,
+      "grad_norm": 0.6009456233945604,
+      "learning_rate": 0.0009335686465946887,
+      "loss": 2.4338,
+      "step": 417
+    },
+    {
+      "epoch": 0.1916552040348464,
+      "grad_norm": 0.6196818629722376,
+      "learning_rate": 0.0009331982557763754,
+      "loss": 2.4804,
+      "step": 418
+    },
+    {
+      "epoch": 0.19211370930765703,
+      "grad_norm": 0.5723449105697626,
+      "learning_rate": 0.0009328269091611537,
+      "loss": 2.3406,
+      "step": 419
+    },
+    {
+      "epoch": 0.19257221458046767,
+      "grad_norm": 0.5826834852012828,
+      "learning_rate": 0.0009324546075683524,
+      "loss": 2.5087,
+      "step": 420
+    },
+    {
+      "epoch": 0.19303071985327833,
+      "grad_norm": 0.5931552298855082,
+      "learning_rate": 0.0009320813518194083,
+      "loss": 2.4234,
+      "step": 421
+    },
+    {
+      "epoch": 0.19348922512608896,
+      "grad_norm": 0.5820140312035208,
+      "learning_rate": 0.0009317071427378624,
+      "loss": 2.3415,
+      "step": 422
+    },
+    {
+      "epoch": 0.1939477303988996,
+      "grad_norm": 0.5802797011820913,
+      "learning_rate": 0.0009313319811493594,
+      "loss": 2.3841,
+      "step": 423
+    },
+    {
+      "epoch": 0.19440623567171023,
+      "grad_norm": 0.6251922103247147,
+      "learning_rate": 0.000930955867881646,
+      "loss": 2.5109,
+      "step": 424
+    },
+    {
+      "epoch": 0.19486474094452086,
+      "grad_norm": 0.5710787838522103,
+      "learning_rate": 0.0009305788037645681,
+      "loss": 2.3861,
+      "step": 425
+    },
+    {
+      "epoch": 0.1953232462173315,
+      "grad_norm": 0.6348485059915613,
+      "learning_rate": 0.0009302007896300697,
+      "loss": 2.4436,
+      "step": 426
+    },
+    {
+      "epoch": 0.19578175149014213,
+      "grad_norm": 0.6350008774354976,
+      "learning_rate": 0.0009298218263121911,
+      "loss": 2.4287,
+      "step": 427
+    },
+    {
+      "epoch": 0.19624025676295279,
+      "grad_norm": 0.6087427582829597,
+      "learning_rate": 0.0009294419146470668,
+      "loss": 2.4348,
+      "step": 428
+    },
+    {
+      "epoch": 0.19669876203576342,
+      "grad_norm": 0.6046789805880408,
+      "learning_rate": 0.0009290610554729234,
+      "loss": 2.3974,
+      "step": 429
+    },
+    {
+      "epoch": 0.19715726730857405,
+      "grad_norm": 0.6044875340829557,
+      "learning_rate": 0.0009286792496300784,
+      "loss": 2.4671,
+      "step": 430
+    },
+    {
+      "epoch": 0.19761577258138469,
+      "grad_norm": 0.5699483573616231,
+      "learning_rate": 0.0009282964979609379,
+      "loss": 2.4206,
+      "step": 431
+    },
+    {
+      "epoch": 0.19807427785419532,
+      "grad_norm": 0.5410803222962629,
+      "learning_rate": 0.0009279128013099947,
+      "loss": 2.408,
+      "step": 432
+    },
+    {
+      "epoch": 0.19853278312700595,
+      "grad_norm": 0.5561846260283513,
+      "learning_rate": 0.0009275281605238268,
+      "loss": 2.438,
+      "step": 433
+    },
+    {
+      "epoch": 0.19899128839981658,
+      "grad_norm": 0.5479339006074598,
+      "learning_rate": 0.0009271425764510953,
+      "loss": 2.4702,
+      "step": 434
+    },
+    {
+      "epoch": 0.19944979367262725,
+      "grad_norm": 0.6305419960203645,
+      "learning_rate": 0.0009267560499425423,
+      "loss": 2.3287,
+      "step": 435
+    },
+    {
+      "epoch": 0.19990829894543788,
+      "grad_norm": 0.5868381244435468,
+      "learning_rate": 0.0009263685818509895,
+      "loss": 2.3645,
+      "step": 436
+    },
+    {
+      "epoch": 0.2003668042182485,
+      "grad_norm": 0.5542512037456003,
+      "learning_rate": 0.000925980173031336,
+      "loss": 2.432,
+      "step": 437
+    },
+    {
+      "epoch": 0.20082530949105915,
+      "grad_norm": 0.5640438654062369,
+      "learning_rate": 0.0009255908243405567,
+      "loss": 2.3982,
+      "step": 438
+    },
+    {
+      "epoch": 0.20128381476386978,
+      "grad_norm": 0.5560976908793476,
+      "learning_rate": 0.0009252005366376996,
+      "loss": 2.4725,
+      "step": 439
+    },
+    {
+      "epoch": 0.2017423200366804,
+      "grad_norm": 0.5769401778057844,
+      "learning_rate": 0.0009248093107838852,
+      "loss": 2.397,
+      "step": 440
+    },
+    {
+      "epoch": 0.20220082530949107,
+      "grad_norm": 0.6104488800666076,
+      "learning_rate": 0.0009244171476423036,
+      "loss": 2.4321,
+      "step": 441
+    },
+    {
+      "epoch": 0.2026593305823017,
+      "grad_norm": 0.5914691357456412,
+      "learning_rate": 0.0009240240480782129,
+      "loss": 2.3211,
+      "step": 442
+    },
+    {
+      "epoch": 0.20311783585511234,
+      "grad_norm": 0.5492281609755851,
+      "learning_rate": 0.0009236300129589376,
+      "loss": 2.536,
+      "step": 443
+    },
+    {
+      "epoch": 0.20357634112792297,
+      "grad_norm": 0.598189998377817,
+      "learning_rate": 0.0009232350431538657,
+      "loss": 2.5165,
+      "step": 444
+    },
+    {
+      "epoch": 0.2040348464007336,
+      "grad_norm": 0.5999921631487511,
+      "learning_rate": 0.0009228391395344482,
+      "loss": 2.4248,
+      "step": 445
+    },
+    {
+      "epoch": 0.20449335167354424,
+      "grad_norm": 0.5378597114114536,
+      "learning_rate": 0.000922442302974196,
+      "loss": 2.4859,
+      "step": 446
+    },
+    {
+      "epoch": 0.20495185694635487,
+      "grad_norm": 0.5311390104740188,
+      "learning_rate": 0.0009220445343486785,
+      "loss": 2.4308,
+      "step": 447
+    },
+    {
+      "epoch": 0.20541036221916553,
+      "grad_norm": 0.5865194887415135,
+      "learning_rate": 0.0009216458345355217,
+      "loss": 2.4495,
+      "step": 448
+    },
+    {
+      "epoch": 0.20586886749197617,
+      "grad_norm": 0.5777914353155509,
+      "learning_rate": 0.0009212462044144061,
+      "loss": 2.4909,
+      "step": 449
+    },
+    {
+      "epoch": 0.2063273727647868,
+      "grad_norm": 0.6098820757508707,
+      "learning_rate": 0.0009208456448670648,
+      "loss": 2.4149,
+      "step": 450
+    },
+    {
+      "epoch": 0.20678587803759743,
+      "grad_norm": 0.5489691868237454,
+      "learning_rate": 0.0009204441567772816,
+      "loss": 2.4708,
+      "step": 451
+    },
+    {
+      "epoch": 0.20724438331040806,
+      "grad_norm": 0.5480364627301055,
+      "learning_rate": 0.0009200417410308888,
+      "loss": 2.4061,
+      "step": 452
+    },
+    {
+      "epoch": 0.2077028885832187,
+      "grad_norm": 0.6291468873331679,
+      "learning_rate": 0.0009196383985157656,
+      "loss": 2.4217,
+      "step": 453
+    },
+    {
+      "epoch": 0.20816139385602933,
+      "grad_norm": 0.6202459918690598,
+      "learning_rate": 0.000919234130121836,
+      "loss": 2.428,
+      "step": 454
+    },
+    {
+      "epoch": 0.20861989912884,
+      "grad_norm": 0.5891111386152992,
+      "learning_rate": 0.0009188289367410672,
+      "loss": 2.4402,
+      "step": 455
+    },
+    {
+      "epoch": 0.20907840440165062,
+      "grad_norm": 0.5801315503772662,
+      "learning_rate": 0.0009184228192674666,
+      "loss": 2.3302,
+      "step": 456
+    },
+    {
+      "epoch": 0.20953690967446126,
+      "grad_norm": 0.6373415147339335,
+      "learning_rate": 0.0009180157785970808,
+      "loss": 2.4785,
+      "step": 457
+    },
+    {
+      "epoch": 0.2099954149472719,
+      "grad_norm": 0.5468496255011686,
+      "learning_rate": 0.0009176078156279932,
+      "loss": 2.369,
+      "step": 458
+    },
+    {
+      "epoch": 0.21045392022008252,
+      "grad_norm": 0.5554206248687655,
+      "learning_rate": 0.0009171989312603226,
+      "loss": 2.4199,
+      "step": 459
+    },
+    {
+      "epoch": 0.21091242549289316,
+      "grad_norm": 0.8291224554549931,
+      "learning_rate": 0.0009167891263962202,
+      "loss": 2.3861,
+      "step": 460
+    },
+    {
+      "epoch": 0.21137093076570382,
+      "grad_norm": 0.6317787975537253,
+      "learning_rate": 0.0009163784019398685,
+      "loss": 2.3886,
+      "step": 461
+    },
+    {
+      "epoch": 0.21182943603851445,
+      "grad_norm": 0.594718640608244,
+      "learning_rate": 0.0009159667587974785,
+      "loss": 2.5409,
+      "step": 462
+    },
+    {
+      "epoch": 0.21228794131132508,
+      "grad_norm": 0.7909507089044695,
+      "learning_rate": 0.0009155541978772887,
+      "loss": 2.3893,
+      "step": 463
+    },
+    {
+      "epoch": 0.21274644658413572,
+      "grad_norm": 0.6001400637715768,
+      "learning_rate": 0.0009151407200895625,
+      "loss": 2.3208,
+      "step": 464
+    },
+    {
+      "epoch": 0.21320495185694635,
+      "grad_norm": 0.6296171774122176,
+      "learning_rate": 0.000914726326346586,
+      "loss": 2.3313,
+      "step": 465
+    },
+    {
+      "epoch": 0.21366345712975698,
+      "grad_norm": 0.617242381365939,
+      "learning_rate": 0.0009143110175626661,
+      "loss": 2.4775,
+      "step": 466
+    },
+    {
+      "epoch": 0.21412196240256762,
+      "grad_norm": 0.5978157023199778,
+      "learning_rate": 0.0009138947946541291,
+      "loss": 2.5134,
+      "step": 467
+    },
+    {
+      "epoch": 0.21458046767537828,
+      "grad_norm": 0.5821367124643003,
+      "learning_rate": 0.0009134776585393181,
+      "loss": 2.404,
+      "step": 468
+    },
+    {
+      "epoch": 0.2150389729481889,
+      "grad_norm": 0.6555550999620402,
+      "learning_rate": 0.0009130596101385906,
+      "loss": 2.5042,
+      "step": 469
+    },
+    {
+      "epoch": 0.21549747822099954,
+      "grad_norm": 0.5785198868653141,
+      "learning_rate": 0.0009126406503743174,
+      "loss": 2.432,
+      "step": 470
+    },
+    {
+      "epoch": 0.21595598349381018,
+      "grad_norm": 1.109994968811881,
+      "learning_rate": 0.0009122207801708802,
+      "loss": 2.372,
+      "step": 471
+    },
+    {
+      "epoch": 0.2164144887666208,
+      "grad_norm": 0.6328906501091458,
+      "learning_rate": 0.0009118000004546689,
+      "loss": 2.4421,
+      "step": 472
+    },
+    {
+      "epoch": 0.21687299403943144,
+      "grad_norm": 1.2104883537721463,
+      "learning_rate": 0.0009113783121540807,
+      "loss": 2.3763,
+      "step": 473
+    },
+    {
+      "epoch": 0.2173314993122421,
+      "grad_norm": 0.631691749664706,
+      "learning_rate": 0.0009109557161995172,
+      "loss": 2.471,
+      "step": 474
+    },
+    {
+      "epoch": 0.21779000458505274,
+      "grad_norm": 0.5867073177733104,
+      "learning_rate": 0.0009105322135233828,
+      "loss": 2.4193,
+      "step": 475
+    },
+    {
+      "epoch": 0.21824850985786337,
+      "grad_norm": 0.5588017764906784,
+      "learning_rate": 0.0009101078050600821,
+      "loss": 2.4279,
+      "step": 476
+    },
+    {
+      "epoch": 0.218707015130674,
+      "grad_norm": 0.6435331484268266,
+      "learning_rate": 0.0009096824917460186,
+      "loss": 2.4535,
+      "step": 477
+    },
+    {
+      "epoch": 0.21916552040348464,
+      "grad_norm": 0.619035983117679,
+      "learning_rate": 0.0009092562745195921,
+      "loss": 2.4703,
+      "step": 478
+    },
+    {
+      "epoch": 0.21962402567629527,
+      "grad_norm": 0.6523598241396069,
+      "learning_rate": 0.0009088291543211967,
+      "loss": 2.4101,
+      "step": 479
+    },
+    {
+      "epoch": 0.2200825309491059,
+      "grad_norm": 0.6097791413660891,
+      "learning_rate": 0.0009084011320932188,
+      "loss": 2.4184,
+      "step": 480
+    },
+    {
+      "epoch": 0.22054103622191656,
+      "grad_norm": 0.6879145133957434,
+      "learning_rate": 0.0009079722087800352,
+      "loss": 2.3555,
+      "step": 481
+    },
+    {
+      "epoch": 0.2209995414947272,
+      "grad_norm": 0.6489364154567205,
+      "learning_rate": 0.0009075423853280106,
+      "loss": 2.4421,
+      "step": 482
+    },
+    {
+      "epoch": 0.22145804676753783,
+      "grad_norm": 0.5768209965103789,
+      "learning_rate": 0.0009071116626854958,
+      "loss": 2.3902,
+      "step": 483
+    },
+    {
+      "epoch": 0.22191655204034846,
+      "grad_norm": 0.634467611414643,
+      "learning_rate": 0.0009066800418028256,
+      "loss": 2.4188,
+      "step": 484
+    },
+    {
+      "epoch": 0.2223750573131591,
+      "grad_norm": 0.6738650244478047,
+      "learning_rate": 0.0009062475236323168,
+      "loss": 2.4139,
+      "step": 485
+    },
+    {
+      "epoch": 0.22283356258596973,
+      "grad_norm": 0.5736394063228365,
+      "learning_rate": 0.0009058141091282656,
+      "loss": 2.37,
+      "step": 486
+    },
+    {
+      "epoch": 0.22329206785878036,
+      "grad_norm": 0.5779066168714002,
+      "learning_rate": 0.0009053797992469461,
+      "loss": 2.419,
+      "step": 487
+    },
+    {
+      "epoch": 0.22375057313159102,
+      "grad_norm": 0.6403152548198145,
+      "learning_rate": 0.0009049445949466078,
+      "loss": 2.4385,
+      "step": 488
+    },
+    {
+      "epoch": 0.22420907840440166,
+      "grad_norm": 0.5455186084183692,
+      "learning_rate": 0.0009045084971874737,
+      "loss": 2.4222,
+      "step": 489
+    },
+    {
+      "epoch": 0.2246675836772123,
+      "grad_norm": 0.5915668082836045,
+      "learning_rate": 0.0009040715069317382,
+      "loss": 2.3368,
+      "step": 490
+    },
+    {
+      "epoch": 0.22512608895002292,
+      "grad_norm": 0.5997298326243515,
+      "learning_rate": 0.0009036336251435648,
+      "loss": 2.4507,
+      "step": 491
+    },
+    {
+      "epoch": 0.22558459422283356,
+      "grad_norm": 1.1085493871436514,
+      "learning_rate": 0.0009031948527890839,
+      "loss": 2.3747,
+      "step": 492
+    },
+    {
+      "epoch": 0.2260430994956442,
+      "grad_norm": 0.5365620863547145,
+      "learning_rate": 0.000902755190836391,
+      "loss": 2.4182,
+      "step": 493
+    },
+    {
+      "epoch": 0.22650160476845485,
+      "grad_norm": 0.6116705902993067,
+      "learning_rate": 0.0009023146402555442,
+      "loss": 2.4551,
+      "step": 494
+    },
+    {
+      "epoch": 0.22696011004126548,
+      "grad_norm": 0.5937787105103242,
+      "learning_rate": 0.0009018732020185624,
+      "loss": 2.3837,
+      "step": 495
+    },
+    {
+      "epoch": 0.22741861531407612,
+      "grad_norm": 0.5492610201993933,
+      "learning_rate": 0.0009014308770994235,
+      "loss": 2.3784,
+      "step": 496
+    },
+    {
+      "epoch": 0.22787712058688675,
+      "grad_norm": 0.6074675312086674,
+      "learning_rate": 0.0009009876664740605,
+      "loss": 2.5215,
+      "step": 497
+    },
+    {
+      "epoch": 0.22833562585969738,
+      "grad_norm": 0.6066718212152176,
+      "learning_rate": 0.0009005435711203618,
+      "loss": 2.4552,
+      "step": 498
+    },
+    {
+      "epoch": 0.22879413113250802,
+      "grad_norm": 0.5991487160527572,
+      "learning_rate": 0.000900098592018167,
+      "loss": 2.4422,
+      "step": 499
+    },
+    {
+      "epoch": 0.22925263640531865,
+      "grad_norm": 0.5875332604001245,
+      "learning_rate": 0.0008996527301492663,
+      "loss": 2.3749,
+      "step": 500
+    },
+    {
+      "epoch": 0.2297111416781293,
+      "grad_norm": 0.5893656876728157,
+      "learning_rate": 0.0008992059864973972,
+      "loss": 2.3696,
+      "step": 501
+    },
+    {
+      "epoch": 0.23016964695093994,
+      "grad_norm": 0.6070441306878955,
+      "learning_rate": 0.0008987583620482427,
+      "loss": 2.3016,
+      "step": 502
+    },
+    {
+      "epoch": 0.23062815222375058,
+      "grad_norm": 0.5618427910950657,
+      "learning_rate": 0.0008983098577894292,
+      "loss": 2.4049,
+      "step": 503
+    },
+    {
+      "epoch": 0.2310866574965612,
+      "grad_norm": 0.6109859257951824,
+      "learning_rate": 0.0008978604747105246,
+      "loss": 2.3441,
+      "step": 504
+    },
+    {
+      "epoch": 0.23154516276937184,
+      "grad_norm": 0.6342283026947244,
+      "learning_rate": 0.0008974102138030354,
+      "loss": 2.533,
+      "step": 505
+    },
+    {
+      "epoch": 0.23200366804218248,
+      "grad_norm": 0.5804953547698418,
+      "learning_rate": 0.000896959076060405,
+      "loss": 2.3899,
+      "step": 506
+    },
+    {
+      "epoch": 0.2324621733149931,
+      "grad_norm": 0.5424054400374302,
+      "learning_rate": 0.0008965070624780116,
+      "loss": 2.3835,
+      "step": 507
+    },
+    {
+      "epoch": 0.23292067858780377,
+      "grad_norm": 0.4965987891911406,
+      "learning_rate": 0.0008960541740531658,
+      "loss": 2.3743,
+      "step": 508
+    },
+    {
+      "epoch": 0.2333791838606144,
+      "grad_norm": 0.558241212745367,
+      "learning_rate": 0.0008956004117851083,
+      "loss": 2.3427,
+      "step": 509
+    },
+    {
+      "epoch": 0.23383768913342504,
+      "grad_norm": 0.5417846615311314,
+      "learning_rate": 0.0008951457766750079,
+      "loss": 2.4892,
+      "step": 510
+    },
+    {
+      "epoch": 0.23429619440623567,
+      "grad_norm": 0.5402854119129425,
+      "learning_rate": 0.0008946902697259593,
+      "loss": 2.3431,
+      "step": 511
+    },
+    {
+      "epoch": 0.2347546996790463,
+      "grad_norm": 0.564377177385817,
+      "learning_rate": 0.0008942338919429805,
+      "loss": 2.4254,
+      "step": 512
+    },
+    {
+      "epoch": 0.23521320495185694,
+      "grad_norm": 0.559607310053975,
+      "learning_rate": 0.0008937766443330113,
+      "loss": 2.3725,
+      "step": 513
+    },
+    {
+      "epoch": 0.2356717102246676,
+      "grad_norm": 0.625570921608999,
+      "learning_rate": 0.0008933185279049103,
+      "loss": 2.4478,
+      "step": 514
+    },
+    {
+      "epoch": 0.23613021549747823,
+      "grad_norm": 0.6208243055508642,
+      "learning_rate": 0.0008928595436694532,
+      "loss": 2.378,
+      "step": 515
+    },
+    {
+      "epoch": 0.23658872077028886,
+      "grad_norm": 0.566898283095845,
+      "learning_rate": 0.0008923996926393305,
+      "loss": 2.4214,
+      "step": 516
+    },
+    {
+      "epoch": 0.2370472260430995,
+      "grad_norm": 0.5873414084902094,
+      "learning_rate": 0.0008919389758291449,
+      "loss": 2.3211,
+      "step": 517
+    },
+    {
+      "epoch": 0.23750573131591013,
+      "grad_norm": 0.5770312761582556,
+      "learning_rate": 0.0008914773942554098,
+      "loss": 2.4037,
+      "step": 518
+    },
+    {
+      "epoch": 0.23796423658872076,
+      "grad_norm": 0.5689473023954376,
+      "learning_rate": 0.000891014948936546,
+      "loss": 2.3996,
+      "step": 519
+    },
+    {
+      "epoch": 0.2384227418615314,
+      "grad_norm": 0.8418011261418884,
+      "learning_rate": 0.0008905516408928804,
+      "loss": 2.4138,
+      "step": 520
+    },
+    {
+      "epoch": 0.23888124713434206,
+      "grad_norm": 0.5753671637531373,
+      "learning_rate": 0.0008900874711466434,
+      "loss": 2.5025,
+      "step": 521
+    },
+    {
+      "epoch": 0.2393397524071527,
+      "grad_norm": 0.557717973499604,
+      "learning_rate": 0.0008896224407219666,
+      "loss": 2.3406,
+      "step": 522
+    },
+    {
+      "epoch": 0.23979825767996332,
+      "grad_norm": 0.8140279302652,
+      "learning_rate": 0.0008891565506448804,
+      "loss": 2.3717,
+      "step": 523
+    },
+    {
+      "epoch": 0.24025676295277396,
+      "grad_norm": 0.5801251682157392,
+      "learning_rate": 0.0008886898019433122,
+      "loss": 2.3226,
+      "step": 524
+    },
+    {
+      "epoch": 0.2407152682255846,
+      "grad_norm": 0.6352479558368405,
+      "learning_rate": 0.0008882221956470836,
+      "loss": 2.4495,
+      "step": 525
+    },
+    {
+      "epoch": 0.24117377349839522,
+      "grad_norm": 0.615036704650059,
+      "learning_rate": 0.0008877537327879086,
+      "loss": 2.4352,
+      "step": 526
+    },
+    {
+      "epoch": 0.24163227877120586,
+      "grad_norm": 0.5997925489834574,
+      "learning_rate": 0.0008872844143993908,
+      "loss": 2.3368,
+      "step": 527
+    },
+    {
+      "epoch": 0.24209078404401652,
+      "grad_norm": 0.5578116728722426,
+      "learning_rate": 0.0008868142415170218,
+      "loss": 2.3497,
+      "step": 528
+    },
+    {
+      "epoch": 0.24254928931682715,
+      "grad_norm": 0.6257998290277491,
+      "learning_rate": 0.0008863432151781781,
+      "loss": 2.4183,
+      "step": 529
+    },
+    {
+      "epoch": 0.24300779458963778,
+      "grad_norm": 0.5236638579650542,
+      "learning_rate": 0.0008858713364221195,
+      "loss": 2.3682,
+      "step": 530
+    },
+    {
+      "epoch": 0.24346629986244842,
+      "grad_norm": 0.5597434046628229,
+      "learning_rate": 0.0008853986062899868,
+      "loss": 2.3783,
+      "step": 531
+    },
+    {
+      "epoch": 0.24392480513525905,
+      "grad_norm": 0.6068466491370786,
+      "learning_rate": 0.0008849250258247986,
+      "loss": 2.4146,
+      "step": 532
+    },
+    {
+      "epoch": 0.24438331040806968,
+      "grad_norm": 0.5989416311447335,
+      "learning_rate": 0.0008844505960714503,
+      "loss": 2.3043,
+      "step": 533
+    },
+    {
+      "epoch": 0.24484181568088034,
+      "grad_norm": 0.705168221488245,
+      "learning_rate": 0.0008839753180767108,
+      "loss": 2.4185,
+      "step": 534
+    },
+    {
+      "epoch": 0.24530032095369098,
+      "grad_norm": 0.6521536523696724,
+      "learning_rate": 0.0008834991928892204,
+      "loss": 2.4348,
+      "step": 535
+    },
+    {
+      "epoch": 0.2457588262265016,
+      "grad_norm": 0.5821957045836091,
+      "learning_rate": 0.000883022221559489,
+      "loss": 2.4254,
+      "step": 536
+    },
+    {
+      "epoch": 0.24621733149931224,
+      "grad_norm": 0.59762097430686,
+      "learning_rate": 0.0008825444051398934,
+      "loss": 2.4602,
+      "step": 537
+    },
+    {
+      "epoch": 0.24667583677212288,
+      "grad_norm": 0.5540284677279287,
+      "learning_rate": 0.0008820657446846745,
+      "loss": 2.4374,
+      "step": 538
+    },
+    {
+      "epoch": 0.2471343420449335,
+      "grad_norm": 0.5314643745953714,
+      "learning_rate": 0.000881586241249936,
+      "loss": 2.2842,
+      "step": 539
+    },
+    {
+      "epoch": 0.24759284731774414,
+      "grad_norm": 0.5441588228024599,
+      "learning_rate": 0.0008811058958936411,
+      "loss": 2.3821,
+      "step": 540
+    },
+    {
+      "epoch": 0.2480513525905548,
+      "grad_norm": 0.6130527348503864,
+      "learning_rate": 0.000880624709675611,
+      "loss": 2.3148,
+      "step": 541
+    },
+    {
+      "epoch": 0.24850985786336544,
+      "grad_norm": 0.630110431175199,
+      "learning_rate": 0.000880142683657522,
+      "loss": 2.4582,
+      "step": 542
+    },
+    {
+      "epoch": 0.24896836313617607,
+      "grad_norm": 0.5254708380871063,
+      "learning_rate": 0.0008796598189029029,
+      "loss": 2.3448,
+      "step": 543
+    },
+    {
+      "epoch": 0.2494268684089867,
+      "grad_norm": 0.5372711913235257,
+      "learning_rate": 0.0008791761164771338,
+      "loss": 2.2926,
+      "step": 544
+    },
+    {
+      "epoch": 0.24988537368179733,
+      "grad_norm": 0.5629437477367555,
+      "learning_rate": 0.0008786915774474424,
+      "loss": 2.4596,
+      "step": 545
+    },
+    {
+      "epoch": 0.25034387895460797,
+      "grad_norm": 0.5728650706757089,
+      "learning_rate": 0.0008782062028829027,
+      "loss": 2.4756,
+      "step": 546
+    },
+    {
+      "epoch": 0.25080238422741863,
+      "grad_norm": 0.5800210677047338,
+      "learning_rate": 0.0008777199938544318,
+      "loss": 2.375,
+      "step": 547
+    },
+    {
+      "epoch": 0.25126088950022923,
+      "grad_norm": 0.5802964889649675,
+      "learning_rate": 0.0008772329514347883,
+      "loss": 2.3496,
+      "step": 548
+    },
+    {
+      "epoch": 0.2517193947730399,
+      "grad_norm": 0.5703823744087412,
+      "learning_rate": 0.0008767450766985694,
+      "loss": 2.4683,
+      "step": 549
+    },
+    {
+      "epoch": 0.2521779000458505,
+      "grad_norm": 0.510679745595259,
+      "learning_rate": 0.0008762563707222086,
+      "loss": 2.327,
+      "step": 550
+    },
+    {
+      "epoch": 0.25263640531866116,
+      "grad_norm": 0.572772334103997,
+      "learning_rate": 0.0008757668345839738,
+      "loss": 2.399,
+      "step": 551
+    },
+    {
+      "epoch": 0.2530949105914718,
+      "grad_norm": 0.5132576587740991,
+      "learning_rate": 0.0008752764693639638,
+      "loss": 2.3563,
+      "step": 552
+    },
+    {
+      "epoch": 0.2535534158642824,
+      "grad_norm": 0.5159009606819669,
+      "learning_rate": 0.0008747852761441078,
+      "loss": 2.4996,
+      "step": 553
+    },
+    {
+      "epoch": 0.2540119211370931,
+      "grad_norm": 0.555804579495914,
+      "learning_rate": 0.0008742932560081607,
+      "loss": 2.4069,
+      "step": 554
+    },
+    {
+      "epoch": 0.2544704264099037,
+      "grad_norm": 0.5059899724004829,
+      "learning_rate": 0.0008738004100417025,
+      "loss": 2.4034,
+      "step": 555
+    },
+    {
+      "epoch": 0.25492893168271435,
+      "grad_norm": 0.5485498386310822,
+      "learning_rate": 0.0008733067393321355,
+      "loss": 2.4002,
+      "step": 556
+    },
+    {
+      "epoch": 0.255387436955525,
+      "grad_norm": 0.5235339754607539,
+      "learning_rate": 0.000872812244968681,
+      "loss": 2.4273,
+      "step": 557
+    },
+    {
+      "epoch": 0.2558459422283356,
+      "grad_norm": 0.5211491737513065,
+      "learning_rate": 0.0008723169280423783,
+      "loss": 2.4386,
+      "step": 558
+    },
+    {
+      "epoch": 0.2563044475011463,
+      "grad_norm": 0.5365874581039461,
+      "learning_rate": 0.0008718207896460811,
+      "loss": 2.4125,
+      "step": 559
+    },
+    {
+      "epoch": 0.2567629527739569,
+      "grad_norm": 0.6480005214506801,
+      "learning_rate": 0.0008713238308744557,
+      "loss": 2.3689,
+      "step": 560
+    },
+    {
+      "epoch": 0.25722145804676755,
+      "grad_norm": 0.6379202065336462,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 2.3477,
+      "step": 561
+    },
+    {
+      "epoch": 0.25767996331957815,
+      "grad_norm": 0.5230972893394507,
+      "learning_rate": 0.000870327456592934,
+      "loss": 2.3421,
+      "step": 562
+    },
+    {
+      "epoch": 0.2581384685923888,
+      "grad_norm": 0.6725852652844913,
+      "learning_rate": 0.0008698280432814107,
+      "loss": 2.4165,
+      "step": 563
+    },
+    {
+      "epoch": 0.2585969738651995,
+      "grad_norm": 0.6339004375393678,
+      "learning_rate": 0.000869327813991301,
+      "loss": 2.4691,
+      "step": 564
+    },
+    {
+      "epoch": 0.2590554791380101,
+      "grad_norm": 0.5874152711268649,
+      "learning_rate": 0.0008688267698262971,
+      "loss": 2.4168,
+      "step": 565
+    },
+    {
+      "epoch": 0.25951398441082074,
+      "grad_norm": 0.5546504223631067,
+      "learning_rate": 0.0008683249118918894,
+      "loss": 2.3884,
+      "step": 566
+    },
+    {
+      "epoch": 0.25997248968363135,
+      "grad_norm": 0.6287367232033735,
+      "learning_rate": 0.0008678222412953637,
+      "loss": 2.3875,
+      "step": 567
+    },
+    {
+      "epoch": 0.260430994956442,
+      "grad_norm": 0.5870381006424814,
+      "learning_rate": 0.0008673187591457987,
+      "loss": 2.4796,
+      "step": 568
+    },
+    {
+      "epoch": 0.2608895002292526,
+      "grad_norm": 0.5430510003240951,
+      "learning_rate": 0.0008668144665540639,
+      "loss": 2.4178,
+      "step": 569
+    },
+    {
+      "epoch": 0.2613480055020633,
+      "grad_norm": 0.6292235828142542,
+      "learning_rate": 0.0008663093646328167,
+      "loss": 2.4317,
+      "step": 570
+    },
+    {
+      "epoch": 0.26180651077487394,
+      "grad_norm": 0.5671496133656164,
+      "learning_rate": 0.0008658034544965003,
+      "loss": 2.2877,
+      "step": 571
+    },
+    {
+      "epoch": 0.26226501604768454,
+      "grad_norm": 0.5420353717611501,
+      "learning_rate": 0.0008652967372613412,
+      "loss": 2.3556,
+      "step": 572
+    },
+    {
+      "epoch": 0.2627235213204952,
+      "grad_norm": 0.5116366889817214,
+      "learning_rate": 0.0008647892140453466,
+      "loss": 2.2892,
+      "step": 573
+    },
+    {
+      "epoch": 0.2631820265933058,
+      "grad_norm": 0.5316154050766173,
+      "learning_rate": 0.0008642808859683021,
+      "loss": 2.3063,
+      "step": 574
+    },
+    {
+      "epoch": 0.26364053186611647,
+      "grad_norm": 0.5746544379993335,
+      "learning_rate": 0.0008637717541517689,
+      "loss": 2.4118,
+      "step": 575
+    },
+    {
+      "epoch": 0.2640990371389271,
+      "grad_norm": 0.5538772696064993,
+      "learning_rate": 0.0008632618197190816,
+      "loss": 2.3938,
+      "step": 576
+    },
+    {
+      "epoch": 0.26455754241173773,
+      "grad_norm": 0.5637789157616306,
+      "learning_rate": 0.0008627510837953458,
+      "loss": 2.4185,
+      "step": 577
+    },
+    {
+      "epoch": 0.2650160476845484,
+      "grad_norm": 0.5326380958481975,
+      "learning_rate": 0.0008622395475074355,
+      "loss": 2.4511,
+      "step": 578
+    },
+    {
+      "epoch": 0.265474552957359,
+      "grad_norm": 0.5255018647392606,
+      "learning_rate": 0.0008617272119839903,
+      "loss": 2.3686,
+      "step": 579
+    },
+    {
+      "epoch": 0.26593305823016966,
+      "grad_norm": 0.5662000079216941,
+      "learning_rate": 0.0008612140783554136,
+      "loss": 2.4767,
+      "step": 580
+    },
+    {
+      "epoch": 0.26639156350298027,
+      "grad_norm": 0.5518577086481349,
+      "learning_rate": 0.0008607001477538696,
+      "loss": 2.3724,
+      "step": 581
+    },
+    {
+      "epoch": 0.2668500687757909,
+      "grad_norm": 0.5188242504398544,
+      "learning_rate": 0.0008601854213132807,
+      "loss": 2.4514,
+      "step": 582
+    },
+    {
+      "epoch": 0.26730857404860153,
+      "grad_norm": 0.5641747639263673,
+      "learning_rate": 0.0008596699001693256,
+      "loss": 2.3708,
+      "step": 583
+    },
+    {
+      "epoch": 0.2677670793214122,
+      "grad_norm": 0.5983943736607411,
+      "learning_rate": 0.000859153585459436,
+      "loss": 2.3701,
+      "step": 584
+    },
+    {
+      "epoch": 0.26822558459422285,
+      "grad_norm": 0.5531608838713388,
+      "learning_rate": 0.0008586364783227949,
+      "loss": 2.4272,
+      "step": 585
+    },
+    {
+      "epoch": 0.26868408986703346,
+      "grad_norm": 0.5211153629019489,
+      "learning_rate": 0.0008581185799003332,
+      "loss": 2.3988,
+      "step": 586
+    },
+    {
+      "epoch": 0.2691425951398441,
+      "grad_norm": 0.5283544305954021,
+      "learning_rate": 0.0008575998913347283,
+      "loss": 2.4427,
+      "step": 587
+    },
+    {
+      "epoch": 0.2696011004126547,
+      "grad_norm": 0.5539562276607203,
+      "learning_rate": 0.0008570804137704004,
+      "loss": 2.4061,
+      "step": 588
+    },
+    {
+      "epoch": 0.2700596056854654,
+      "grad_norm": 0.7800466672846708,
+      "learning_rate": 0.0008565601483535108,
+      "loss": 2.3228,
+      "step": 589
+    },
+    {
+      "epoch": 0.27051811095827605,
+      "grad_norm": 0.5071224234912427,
+      "learning_rate": 0.0008560390962319591,
+      "loss": 2.3758,
+      "step": 590
+    },
+    {
+      "epoch": 0.27097661623108665,
+      "grad_norm": 0.5472034670653824,
+      "learning_rate": 0.0008555172585553804,
+      "loss": 2.3751,
+      "step": 591
+    },
+    {
+      "epoch": 0.2714351215038973,
+      "grad_norm": 0.5328957681389516,
+      "learning_rate": 0.0008549946364751435,
+      "loss": 2.2769,
+      "step": 592
+    },
+    {
+      "epoch": 0.2718936267767079,
+      "grad_norm": 1.27304884674719,
+      "learning_rate": 0.0008544712311443475,
+      "loss": 2.4189,
+      "step": 593
+    },
+    {
+      "epoch": 0.2723521320495186,
+      "grad_norm": 0.5725683759770754,
+      "learning_rate": 0.0008539470437178196,
+      "loss": 2.3743,
+      "step": 594
+    },
+    {
+      "epoch": 0.2728106373223292,
+      "grad_norm": 0.5599878668920902,
+      "learning_rate": 0.000853422075352113,
+      "loss": 2.3886,
+      "step": 595
+    },
+    {
+      "epoch": 0.27326914259513985,
+      "grad_norm": 0.5434023098245148,
+      "learning_rate": 0.0008528963272055035,
+      "loss": 2.4708,
+      "step": 596
+    },
+    {
+      "epoch": 0.2737276478679505,
+      "grad_norm": 0.5269678104784802,
+      "learning_rate": 0.0008523698004379877,
+      "loss": 2.3704,
+      "step": 597
+    },
+    {
+      "epoch": 0.2741861531407611,
+      "grad_norm": 0.5749366588340156,
+      "learning_rate": 0.00085184249621128,
+      "loss": 2.3504,
+      "step": 598
+    },
+    {
+      "epoch": 0.2746446584135718,
+      "grad_norm": 0.5370315686136987,
+      "learning_rate": 0.0008513144156888101,
+      "loss": 2.3429,
+      "step": 599
+    },
+    {
+      "epoch": 0.2751031636863824,
+      "grad_norm": 0.546271222037716,
+      "learning_rate": 0.0008507855600357207,
+      "loss": 2.4531,
+      "step": 600
+    },
+    {
+      "epoch": 0.27556166895919304,
+      "grad_norm": 0.511697775560146,
+      "learning_rate": 0.0008502559304188644,
+      "loss": 2.3158,
+      "step": 601
+    },
+    {
+      "epoch": 0.27602017423200365,
+      "grad_norm": 0.5521963132485257,
+      "learning_rate": 0.0008497255280068019,
+      "loss": 2.4404,
+      "step": 602
+    },
+    {
+      "epoch": 0.2764786795048143,
+      "grad_norm": 0.5162369014782571,
+      "learning_rate": 0.0008491943539697986,
+      "loss": 2.329,
+      "step": 603
+    },
+    {
+      "epoch": 0.27693718477762497,
+      "grad_norm": 0.5340018338088954,
+      "learning_rate": 0.0008486624094798226,
+      "loss": 2.3421,
+      "step": 604
+    },
+    {
+      "epoch": 0.2773956900504356,
+      "grad_norm": 0.5144786384953455,
+      "learning_rate": 0.0008481296957105417,
+      "loss": 2.3575,
+      "step": 605
+    },
+    {
+      "epoch": 0.27785419532324623,
+      "grad_norm": 0.5342099474961054,
+      "learning_rate": 0.0008475962138373213,
+      "loss": 2.3842,
+      "step": 606
+    },
+    {
+      "epoch": 0.27831270059605684,
+      "grad_norm": 0.5489868580623048,
+      "learning_rate": 0.0008470619650372211,
+      "loss": 2.3489,
+      "step": 607
+    },
+    {
+      "epoch": 0.2787712058688675,
+      "grad_norm": 0.5471580700895529,
+      "learning_rate": 0.0008465269504889934,
+      "loss": 2.3518,
+      "step": 608
+    },
+    {
+      "epoch": 0.2792297111416781,
+      "grad_norm": 0.5206795985107469,
+      "learning_rate": 0.0008459911713730799,
+      "loss": 2.4016,
+      "step": 609
+    },
+    {
+      "epoch": 0.27968821641448877,
+      "grad_norm": 0.5330570718704636,
+      "learning_rate": 0.0008454546288716089,
+      "loss": 2.3605,
+      "step": 610
+    },
+    {
+      "epoch": 0.2801467216872994,
+      "grad_norm": 0.522541288854971,
+      "learning_rate": 0.0008449173241683935,
+      "loss": 2.3444,
+      "step": 611
+    },
+    {
+      "epoch": 0.28060522696011003,
+      "grad_norm": 0.6019656619270888,
+      "learning_rate": 0.0008443792584489281,
+      "loss": 2.4332,
+      "step": 612
+    },
+    {
+      "epoch": 0.2810637322329207,
+      "grad_norm": 0.5515838715996126,
+      "learning_rate": 0.0008438404329003863,
+      "loss": 2.3841,
+      "step": 613
+    },
+    {
+      "epoch": 0.2815222375057313,
+      "grad_norm": 0.6375095285201424,
+      "learning_rate": 0.0008433008487116183,
+      "loss": 2.3423,
+      "step": 614
+    },
+    {
+      "epoch": 0.28198074277854196,
+      "grad_norm": 0.5406706875362183,
+      "learning_rate": 0.0008427605070731481,
+      "loss": 2.3185,
+      "step": 615
+    },
+    {
+      "epoch": 0.28243924805135256,
+      "grad_norm": 0.539151759851803,
+      "learning_rate": 0.0008422194091771708,
+      "loss": 2.3714,
+      "step": 616
+    },
+    {
+      "epoch": 0.2828977533241632,
+      "grad_norm": 0.5426484267716822,
+      "learning_rate": 0.0008416775562175503,
+      "loss": 2.3839,
+      "step": 617
+    },
+    {
+      "epoch": 0.2833562585969739,
+      "grad_norm": 0.565178271508405,
+      "learning_rate": 0.000841134949389816,
+      "loss": 2.3579,
+      "step": 618
+    },
+    {
+      "epoch": 0.2838147638697845,
+      "grad_norm": 0.5623681939277834,
+      "learning_rate": 0.0008405915898911611,
+      "loss": 2.3951,
+      "step": 619
+    },
+    {
+      "epoch": 0.28427326914259515,
+      "grad_norm": 0.557100648648169,
+      "learning_rate": 0.0008400474789204396,
+      "loss": 2.3678,
+      "step": 620
+    },
+    {
+      "epoch": 0.28473177441540576,
+      "grad_norm": 0.6393645018552272,
+      "learning_rate": 0.0008395026176781626,
+      "loss": 2.3282,
+      "step": 621
+    },
+    {
+      "epoch": 0.2851902796882164,
+      "grad_norm": 0.5791991263813852,
+      "learning_rate": 0.0008389570073664976,
+      "loss": 2.4132,
+      "step": 622
+    },
+    {
+      "epoch": 0.285648784961027,
+      "grad_norm": 0.5366598181250238,
+      "learning_rate": 0.0008384106491892642,
+      "loss": 2.4364,
+      "step": 623
+    },
+    {
+      "epoch": 0.2861072902338377,
+      "grad_norm": 0.5650556169871017,
+      "learning_rate": 0.0008378635443519327,
+      "loss": 2.3619,
+      "step": 624
+    },
+    {
+      "epoch": 0.28656579550664835,
+      "grad_norm": 0.5349525858992139,
+      "learning_rate": 0.0008373156940616199,
+      "loss": 2.3597,
+      "step": 625
+    },
+    {
+      "epoch": 0.28702430077945895,
+      "grad_norm": 0.5339108013528998,
+      "learning_rate": 0.0008367670995270882,
+      "loss": 2.4172,
+      "step": 626
+    },
+    {
+      "epoch": 0.2874828060522696,
+      "grad_norm": 0.5315198916170973,
+      "learning_rate": 0.0008362177619587416,
+      "loss": 2.312,
+      "step": 627
+    },
+    {
+      "epoch": 0.2879413113250802,
+      "grad_norm": 0.503944486777112,
+      "learning_rate": 0.0008356676825686238,
+      "loss": 2.3309,
+      "step": 628
+    },
+    {
+      "epoch": 0.2883998165978909,
+      "grad_norm": 0.5232927245385984,
+      "learning_rate": 0.0008351168625704147,
+      "loss": 2.4937,
+      "step": 629
+    },
+    {
+      "epoch": 0.28885832187070154,
+      "grad_norm": 0.5259540501074851,
+      "learning_rate": 0.0008345653031794292,
+      "loss": 2.3996,
+      "step": 630
+    },
+    {
+      "epoch": 0.28931682714351215,
+      "grad_norm": 0.6843989304994758,
+      "learning_rate": 0.0008340130056126125,
+      "loss": 2.2843,
+      "step": 631
+    },
+    {
+      "epoch": 0.2897753324163228,
+      "grad_norm": 0.5897752914743187,
+      "learning_rate": 0.0008334599710885394,
+      "loss": 2.3156,
+      "step": 632
+    },
+    {
+      "epoch": 0.2902338376891334,
+      "grad_norm": 0.5560216856962238,
+      "learning_rate": 0.0008329062008274098,
+      "loss": 2.3174,
+      "step": 633
+    },
+    {
+      "epoch": 0.2906923429619441,
+      "grad_norm": 0.5521939874515143,
+      "learning_rate": 0.000832351696051048,
+      "loss": 2.3257,
+      "step": 634
+    },
+    {
+      "epoch": 0.2911508482347547,
+      "grad_norm": 0.5376565904822016,
+      "learning_rate": 0.000831796457982898,
+      "loss": 2.4059,
+      "step": 635
+    },
+    {
+      "epoch": 0.29160935350756534,
+      "grad_norm": 0.5464280925014233,
+      "learning_rate": 0.0008312404878480222,
+      "loss": 2.4136,
+      "step": 636
+    },
+    {
+      "epoch": 0.292067858780376,
+      "grad_norm": 1.4952538267147286,
+      "learning_rate": 0.0008306837868730979,
+      "loss": 2.3672,
+      "step": 637
+    },
+    {
+      "epoch": 0.2925263640531866,
+      "grad_norm": 0.5225466515294473,
+      "learning_rate": 0.0008301263562864152,
+      "loss": 2.3927,
+      "step": 638
+    },
+    {
+      "epoch": 0.29298486932599727,
+      "grad_norm": 0.5353561679686526,
+      "learning_rate": 0.0008295681973178737,
+      "loss": 2.4364,
+      "step": 639
+    },
+    {
+      "epoch": 0.29344337459880787,
+      "grad_norm": 0.6331515450696173,
+      "learning_rate": 0.0008290093111989804,
+      "loss": 2.3822,
+      "step": 640
+    },
+    {
+      "epoch": 0.29390187987161853,
+      "grad_norm": 0.5413982631853191,
+      "learning_rate": 0.0008284496991628465,
+      "loss": 2.3404,
+      "step": 641
+    },
+    {
+      "epoch": 0.29436038514442914,
+      "grad_norm": 0.5983762393501888,
+      "learning_rate": 0.0008278893624441847,
+      "loss": 2.4362,
+      "step": 642
+    },
+    {
+      "epoch": 0.2948188904172398,
+      "grad_norm": 0.6062348187834143,
+      "learning_rate": 0.000827328302279307,
+      "loss": 2.3028,
+      "step": 643
+    },
+    {
+      "epoch": 0.29527739569005046,
+      "grad_norm": 0.5128035074693732,
+      "learning_rate": 0.0008267665199061211,
+      "loss": 2.3066,
+      "step": 644
+    },
+    {
+      "epoch": 0.29573590096286106,
+      "grad_norm": 0.5736225757198515,
+      "learning_rate": 0.0008262040165641288,
+      "loss": 2.3859,
+      "step": 645
+    },
+    {
+      "epoch": 0.2961944062356717,
+      "grad_norm": 0.5484989747211276,
+      "learning_rate": 0.0008256407934944219,
+      "loss": 2.3725,
+      "step": 646
+    },
+    {
+      "epoch": 0.29665291150848233,
+      "grad_norm": 0.5227253405275298,
+      "learning_rate": 0.0008250768519396807,
+      "loss": 2.3025,
+      "step": 647
+    },
+    {
+      "epoch": 0.297111416781293,
+      "grad_norm": 0.5348623096749071,
+      "learning_rate": 0.0008245121931441706,
+      "loss": 2.3231,
+      "step": 648
+    },
+    {
+      "epoch": 0.2975699220541036,
+      "grad_norm": 0.5651193559417393,
+      "learning_rate": 0.0008239468183537393,
+      "loss": 2.4434,
+      "step": 649
+    },
+    {
+      "epoch": 0.29802842732691426,
+      "grad_norm": 0.5551124356700429,
+      "learning_rate": 0.0008233807288158146,
+      "loss": 2.3647,
+      "step": 650
+    },
+    {
+      "epoch": 0.2984869325997249,
+      "grad_norm": 0.5217070461313092,
+      "learning_rate": 0.0008228139257794012,
+      "loss": 2.3709,
+      "step": 651
+    },
+    {
+      "epoch": 0.2989454378725355,
+      "grad_norm": 0.5660066847756605,
+      "learning_rate": 0.0008222464104950778,
+      "loss": 2.3952,
+      "step": 652
+    },
+    {
+      "epoch": 0.2994039431453462,
+      "grad_norm": 0.6175548418105947,
+      "learning_rate": 0.000821678184214995,
+      "loss": 2.3966,
+      "step": 653
+    },
+    {
+      "epoch": 0.2998624484181568,
+      "grad_norm": 0.5204324560385964,
+      "learning_rate": 0.0008211092481928716,
+      "loss": 2.327,
+      "step": 654
+    },
+    {
+      "epoch": 0.30032095369096745,
+      "grad_norm": 0.5442201268590497,
+      "learning_rate": 0.0008205396036839927,
+      "loss": 2.3973,
+      "step": 655
+    },
+    {
+      "epoch": 0.30077945896377806,
+      "grad_norm": 0.540403697405995,
+      "learning_rate": 0.0008199692519452069,
+      "loss": 2.4953,
+      "step": 656
+    },
+    {
+      "epoch": 0.3012379642365887,
+      "grad_norm": 0.5736949974620128,
+      "learning_rate": 0.0008193981942349224,
+      "loss": 2.3321,
+      "step": 657
+    },
+    {
+      "epoch": 0.3016964695093994,
+      "grad_norm": 0.4968016822271614,
+      "learning_rate": 0.0008188264318131056,
+      "loss": 2.3132,
+      "step": 658
+    },
+    {
+      "epoch": 0.30215497478221,
+      "grad_norm": 0.5394754454289029,
+      "learning_rate": 0.0008182539659412776,
+      "loss": 2.3335,
+      "step": 659
+    },
+    {
+      "epoch": 0.30261348005502064,
+      "grad_norm": 0.6204128966427828,
+      "learning_rate": 0.0008176807978825118,
+      "loss": 2.3367,
+      "step": 660
+    },
+    {
+      "epoch": 0.30307198532783125,
+      "grad_norm": 0.5499057452456196,
+      "learning_rate": 0.0008171069289014306,
+      "loss": 2.3025,
+      "step": 661
+    },
+    {
+      "epoch": 0.3035304906006419,
+      "grad_norm": 0.5105799618844776,
+      "learning_rate": 0.0008165323602642028,
+      "loss": 2.4236,
+      "step": 662
+    },
+    {
+      "epoch": 0.30398899587345257,
+      "grad_norm": 0.5186799887136597,
+      "learning_rate": 0.0008159570932385414,
+      "loss": 2.3772,
+      "step": 663
+    },
+    {
+      "epoch": 0.3044475011462632,
+      "grad_norm": 0.5183290088813661,
+      "learning_rate": 0.0008153811290936999,
+      "loss": 2.4176,
+      "step": 664
+    },
+    {
+      "epoch": 0.30490600641907384,
+      "grad_norm": 0.49152954246169356,
+      "learning_rate": 0.0008148044691004698,
+      "loss": 2.4007,
+      "step": 665
+    },
+    {
+      "epoch": 0.30536451169188444,
+      "grad_norm": 0.6469966531297832,
+      "learning_rate": 0.0008142271145311783,
+      "loss": 2.3731,
+      "step": 666
+    },
+    {
+      "epoch": 0.3058230169646951,
+      "grad_norm": 0.5473175484866435,
+      "learning_rate": 0.000813649066659685,
+      "loss": 2.3749,
+      "step": 667
+    },
+    {
+      "epoch": 0.3062815222375057,
+      "grad_norm": 0.49986289331274064,
+      "learning_rate": 0.0008130703267613787,
+      "loss": 2.3574,
+      "step": 668
+    },
+    {
+      "epoch": 0.30674002751031637,
+      "grad_norm": 0.5326700161115692,
+      "learning_rate": 0.0008124908961131759,
+      "loss": 2.3594,
+      "step": 669
+    },
+    {
+      "epoch": 0.30719853278312703,
+      "grad_norm": 0.5209271225327553,
+      "learning_rate": 0.0008119107759935163,
+      "loss": 2.3386,
+      "step": 670
+    },
+    {
+      "epoch": 0.30765703805593764,
+      "grad_norm": 0.4856303801029115,
+      "learning_rate": 0.0008113299676823615,
+      "loss": 2.4237,
+      "step": 671
+    },
+    {
+      "epoch": 0.3081155433287483,
+      "grad_norm": 0.5447155072884559,
+      "learning_rate": 0.0008107484724611911,
+      "loss": 2.3465,
+      "step": 672
+    },
+    {
+      "epoch": 0.3085740486015589,
+      "grad_norm": 0.47611356841593255,
+      "learning_rate": 0.0008101662916130006,
+      "loss": 2.3938,
+      "step": 673
+    },
+    {
+      "epoch": 0.30903255387436956,
+      "grad_norm": 0.5248158518924337,
+      "learning_rate": 0.0008095834264222979,
+      "loss": 2.3507,
+      "step": 674
+    },
+    {
+      "epoch": 0.30949105914718017,
+      "grad_norm": 0.4990830752071052,
+      "learning_rate": 0.0008089998781751009,
+      "loss": 2.4834,
+      "step": 675
+    },
+    {
+      "epoch": 0.30994956441999083,
+      "grad_norm": 0.471552794232097,
+      "learning_rate": 0.0008084156481589349,
+      "loss": 2.4311,
+      "step": 676
+    },
+    {
+      "epoch": 0.3104080696928015,
+      "grad_norm": 0.5156580524254721,
+      "learning_rate": 0.0008078307376628291,
+      "loss": 2.3664,
+      "step": 677
+    },
+    {
+      "epoch": 0.3108665749656121,
+      "grad_norm": 0.5214407133279948,
+      "learning_rate": 0.0008072451479773143,
+      "loss": 2.4562,
+      "step": 678
+    },
+    {
+      "epoch": 0.31132508023842276,
+      "grad_norm": 0.4995679530039473,
+      "learning_rate": 0.0008066588803944195,
+      "loss": 2.3274,
+      "step": 679
+    },
+    {
+      "epoch": 0.31178358551123336,
+      "grad_norm": 0.5068913226929482,
+      "learning_rate": 0.0008060719362076697,
+      "loss": 2.415,
+      "step": 680
+    },
+    {
+      "epoch": 0.312242090784044,
+      "grad_norm": 0.5536946268219113,
+      "learning_rate": 0.0008054843167120826,
+      "loss": 2.3181,
+      "step": 681
+    },
+    {
+      "epoch": 0.31270059605685463,
+      "grad_norm": 0.5215604095802077,
+      "learning_rate": 0.0008048960232041663,
+      "loss": 2.405,
+      "step": 682
+    },
+    {
+      "epoch": 0.3131591013296653,
+      "grad_norm": 0.577427099095177,
+      "learning_rate": 0.0008043070569819153,
+      "loss": 2.3172,
+      "step": 683
+    },
+    {
+      "epoch": 0.31361760660247595,
+      "grad_norm": 0.5577497180751821,
+      "learning_rate": 0.0008037174193448089,
+      "loss": 2.4177,
+      "step": 684
+    },
+    {
+      "epoch": 0.31407611187528656,
+      "grad_norm": 0.5942194595976527,
+      "learning_rate": 0.0008031271115938077,
+      "loss": 2.3484,
+      "step": 685
+    },
+    {
+      "epoch": 0.3145346171480972,
+      "grad_norm": 0.5223535656894671,
+      "learning_rate": 0.0008025361350313505,
+      "loss": 2.3808,
+      "step": 686
+    },
+    {
+      "epoch": 0.3149931224209078,
+      "grad_norm": 0.5613833829118856,
+      "learning_rate": 0.0008019444909613523,
+      "loss": 2.3892,
+      "step": 687
+    },
+    {
+      "epoch": 0.3154516276937185,
+      "grad_norm": 0.5956185132268366,
+      "learning_rate": 0.0008013521806892003,
+      "loss": 2.2431,
+      "step": 688
+    },
+    {
+      "epoch": 0.3159101329665291,
+      "grad_norm": 0.5310352529962502,
+      "learning_rate": 0.000800759205521752,
+      "loss": 2.3831,
+      "step": 689
+    },
+    {
+      "epoch": 0.31636863823933975,
+      "grad_norm": 0.48903349247271477,
+      "learning_rate": 0.0008001655667673318,
+      "loss": 2.4249,
+      "step": 690
+    },
+    {
+      "epoch": 0.3168271435121504,
+      "grad_norm": 0.5235027567909062,
+      "learning_rate": 0.0007995712657357279,
+      "loss": 2.4149,
+      "step": 691
+    },
+    {
+      "epoch": 0.317285648784961,
+      "grad_norm": 0.49403546108559393,
+      "learning_rate": 0.0007989763037381904,
+      "loss": 2.4967,
+      "step": 692
+    },
+    {
+      "epoch": 0.3177441540577717,
+      "grad_norm": 0.5164271152078467,
+      "learning_rate": 0.0007983806820874271,
+      "loss": 2.3284,
+      "step": 693
+    },
+    {
+      "epoch": 0.3182026593305823,
+      "grad_norm": 0.5516599452311578,
+      "learning_rate": 0.0007977844020976016,
+      "loss": 2.3246,
+      "step": 694
+    },
+    {
+      "epoch": 0.31866116460339294,
+      "grad_norm": 0.5168584870229157,
+      "learning_rate": 0.00079718746508433,
+      "loss": 2.4086,
+      "step": 695
+    },
+    {
+      "epoch": 0.31911966987620355,
+      "grad_norm": 0.48882005884806584,
+      "learning_rate": 0.0007965898723646776,
+      "loss": 2.3756,
+      "step": 696
+    },
+    {
+      "epoch": 0.3195781751490142,
+      "grad_norm": 0.521620354309612,
+      "learning_rate": 0.0007959916252571573,
+      "loss": 2.31,
+      "step": 697
+    },
+    {
+      "epoch": 0.32003668042182487,
+      "grad_norm": 0.5222024650354833,
+      "learning_rate": 0.000795392725081725,
+      "loss": 2.2943,
+      "step": 698
+    },
+    {
+      "epoch": 0.3204951856946355,
+      "grad_norm": 0.5438518991851105,
+      "learning_rate": 0.000794793173159778,
+      "loss": 2.4359,
+      "step": 699
+    },
+    {
+      "epoch": 0.32095369096744614,
+      "grad_norm": 0.5207238990087364,
+      "learning_rate": 0.0007941929708141513,
+      "loss": 2.4471,
+      "step": 700
+    },
+    {
+      "epoch": 0.32141219624025674,
+      "grad_norm": 0.6053526405553308,
+      "learning_rate": 0.0007935921193691153,
+      "loss": 2.3667,
+      "step": 701
+    },
+    {
+      "epoch": 0.3218707015130674,
+      "grad_norm": 0.5318734307702329,
+      "learning_rate": 0.0007929906201503722,
+      "loss": 2.3805,
+      "step": 702
+    },
+    {
+      "epoch": 0.32232920678587806,
+      "grad_norm": 0.5070868704262409,
+      "learning_rate": 0.0007923884744850536,
+      "loss": 2.4145,
+      "step": 703
+    },
+    {
+      "epoch": 0.32278771205868867,
+      "grad_norm": 0.5243578837226105,
+      "learning_rate": 0.0007917856837017176,
+      "loss": 2.3674,
+      "step": 704
+    },
+    {
+      "epoch": 0.32324621733149933,
+      "grad_norm": 0.6642367772408904,
+      "learning_rate": 0.0007911822491303452,
+      "loss": 2.4122,
+      "step": 705
+    },
+    {
+      "epoch": 0.32370472260430994,
+      "grad_norm": 0.4942320722089315,
+      "learning_rate": 0.0007905781721023382,
+      "loss": 2.3567,
+      "step": 706
+    },
+    {
+      "epoch": 0.3241632278771206,
+      "grad_norm": 0.5326057773919948,
+      "learning_rate": 0.000789973453950516,
+      "loss": 2.3317,
+      "step": 707
+    },
+    {
+      "epoch": 0.3246217331499312,
+      "grad_norm": 0.5014530547761141,
+      "learning_rate": 0.000789368096009112,
+      "loss": 2.3852,
+      "step": 708
+    },
+    {
+      "epoch": 0.32508023842274186,
+      "grad_norm": 0.5411070568240991,
+      "learning_rate": 0.0007887620996137721,
+      "loss": 2.3357,
+      "step": 709
+    },
+    {
+      "epoch": 0.3255387436955525,
+      "grad_norm": 0.5336977830908979,
+      "learning_rate": 0.0007881554661015497,
+      "loss": 2.3351,
+      "step": 710
+    },
+    {
+      "epoch": 0.32599724896836313,
+      "grad_norm": 0.5817593054824945,
+      "learning_rate": 0.0007875481968109051,
+      "loss": 2.3664,
+      "step": 711
+    },
+    {
+      "epoch": 0.3264557542411738,
+      "grad_norm": 0.5389121885233782,
+      "learning_rate": 0.0007869402930817007,
+      "loss": 2.4287,
+      "step": 712
+    },
+    {
+      "epoch": 0.3269142595139844,
+      "grad_norm": 0.5623204144096784,
+      "learning_rate": 0.0007863317562551987,
+      "loss": 2.2681,
+      "step": 713
+    },
+    {
+      "epoch": 0.32737276478679506,
+      "grad_norm": 0.5024852782188429,
+      "learning_rate": 0.0007857225876740584,
+      "loss": 2.3373,
+      "step": 714
+    },
+    {
+      "epoch": 0.32783127005960566,
+      "grad_norm": 0.5405830933475162,
+      "learning_rate": 0.0007851127886823327,
+      "loss": 2.2987,
+      "step": 715
+    },
+    {
+      "epoch": 0.3282897753324163,
+      "grad_norm": 0.4691087971343938,
+      "learning_rate": 0.0007845023606254658,
+      "loss": 2.349,
+      "step": 716
+    },
+    {
+      "epoch": 0.328748280605227,
+      "grad_norm": 0.5173694574580822,
+      "learning_rate": 0.0007838913048502894,
+      "loss": 2.316,
+      "step": 717
+    },
+    {
+      "epoch": 0.3292067858780376,
+      "grad_norm": 0.5349400779397456,
+      "learning_rate": 0.0007832796227050208,
+      "loss": 2.4886,
+      "step": 718
+    },
+    {
+      "epoch": 0.32966529115084825,
+      "grad_norm": 0.5223044582247615,
+      "learning_rate": 0.0007826673155392587,
+      "loss": 2.3924,
+      "step": 719
+    },
+    {
+      "epoch": 0.33012379642365886,
+      "grad_norm": 0.5055248871894552,
+      "learning_rate": 0.000782054384703981,
+      "loss": 2.2775,
+      "step": 720
+    },
+    {
+      "epoch": 0.3305823016964695,
+      "grad_norm": 0.5356572642156102,
+      "learning_rate": 0.0007814408315515418,
+      "loss": 2.3611,
+      "step": 721
+    },
+    {
+      "epoch": 0.3310408069692801,
+      "grad_norm": 0.5157299106609934,
+      "learning_rate": 0.0007808266574356683,
+      "loss": 2.4231,
+      "step": 722
+    },
+    {
+      "epoch": 0.3314993122420908,
+      "grad_norm": 0.5327205312374573,
+      "learning_rate": 0.0007802118637114573,
+      "loss": 2.369,
+      "step": 723
+    },
+    {
+      "epoch": 0.33195781751490144,
+      "grad_norm": 0.5200177069529023,
+      "learning_rate": 0.0007795964517353734,
+      "loss": 2.4242,
+      "step": 724
+    },
+    {
+      "epoch": 0.33241632278771205,
+      "grad_norm": 0.6596414478118274,
+      "learning_rate": 0.0007789804228652449,
+      "loss": 2.3564,
+      "step": 725
+    },
+    {
+      "epoch": 0.3328748280605227,
+      "grad_norm": 0.49521167383061443,
+      "learning_rate": 0.0007783637784602609,
+      "loss": 2.3026,
+      "step": 726
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5488021284866547,
+      "learning_rate": 0.0007777465198809692,
+      "loss": 2.3652,
+      "step": 727
+    },
+    {
+      "epoch": 0.333791838606144,
+      "grad_norm": 0.5690240555781033,
+      "learning_rate": 0.0007771286484892722,
+      "loss": 2.4628,
+      "step": 728
+    },
+    {
+      "epoch": 0.3342503438789546,
+      "grad_norm": 0.5138682726069629,
+      "learning_rate": 0.000776510165648425,
+      "loss": 2.3931,
+      "step": 729
+    },
+    {
+      "epoch": 0.33470884915176524,
+      "grad_norm": 0.5276507987082376,
+      "learning_rate": 0.0007758910727230311,
+      "loss": 2.2642,
+      "step": 730
+    },
+    {
+      "epoch": 0.3351673544245759,
+      "grad_norm": 0.5743208444913899,
+      "learning_rate": 0.0007752713710790404,
+      "loss": 2.3424,
+      "step": 731
+    },
+    {
+      "epoch": 0.3356258596973865,
+      "grad_norm": 0.5234944495848588,
+      "learning_rate": 0.0007746510620837459,
+      "loss": 2.3921,
+      "step": 732
+    },
+    {
+      "epoch": 0.33608436497019717,
+      "grad_norm": 0.5259547032850798,
+      "learning_rate": 0.0007740301471057807,
+      "loss": 2.3232,
+      "step": 733
+    },
+    {
+      "epoch": 0.3365428702430078,
+      "grad_norm": 0.5331122928845768,
+      "learning_rate": 0.0007734086275151146,
+      "loss": 2.35,
+      "step": 734
+    },
+    {
+      "epoch": 0.33700137551581844,
+      "grad_norm": 0.5042601768347503,
+      "learning_rate": 0.0007727865046830517,
+      "loss": 2.2445,
+      "step": 735
+    },
+    {
+      "epoch": 0.3374598807886291,
+      "grad_norm": 0.5214289826002181,
+      "learning_rate": 0.0007721637799822269,
+      "loss": 2.3619,
+      "step": 736
+    },
+    {
+      "epoch": 0.3379183860614397,
+      "grad_norm": 0.511225688514841,
+      "learning_rate": 0.0007715404547866032,
+      "loss": 2.3206,
+      "step": 737
+    },
+    {
+      "epoch": 0.33837689133425036,
+      "grad_norm": 0.5023039445801867,
+      "learning_rate": 0.0007709165304714685,
+      "loss": 2.4163,
+      "step": 738
+    },
+    {
+      "epoch": 0.33883539660706097,
+      "grad_norm": 0.5128564086498849,
+      "learning_rate": 0.0007702920084134324,
+      "loss": 2.3806,
+      "step": 739
+    },
+    {
+      "epoch": 0.33929390187987163,
+      "grad_norm": 0.5573381633594853,
+      "learning_rate": 0.0007696668899904236,
+      "loss": 2.3486,
+      "step": 740
+    },
+    {
+      "epoch": 0.33975240715268223,
+      "grad_norm": 0.5730646996319768,
+      "learning_rate": 0.0007690411765816864,
+      "loss": 2.3863,
+      "step": 741
+    },
+    {
+      "epoch": 0.3402109124254929,
+      "grad_norm": 0.5539612734008126,
+      "learning_rate": 0.0007684148695677778,
+      "loss": 2.2979,
+      "step": 742
+    },
+    {
+      "epoch": 0.34066941769830356,
+      "grad_norm": 0.5502700731089045,
+      "learning_rate": 0.000767787970330565,
+      "loss": 2.2711,
+      "step": 743
+    },
+    {
+      "epoch": 0.34112792297111416,
+      "grad_norm": 0.5503075893054108,
+      "learning_rate": 0.000767160480253221,
+      "loss": 2.4202,
+      "step": 744
+    },
+    {
+      "epoch": 0.3415864282439248,
+      "grad_norm": 0.5317727912742012,
+      "learning_rate": 0.0007665324007202235,
+      "loss": 2.2948,
+      "step": 745
+    },
+    {
+      "epoch": 0.3420449335167354,
+      "grad_norm": 0.5040287258953108,
+      "learning_rate": 0.0007659037331173498,
+      "loss": 2.3327,
+      "step": 746
+    },
+    {
+      "epoch": 0.3425034387895461,
+      "grad_norm": 0.47998424171524884,
+      "learning_rate": 0.0007652744788316752,
+      "loss": 2.2674,
+      "step": 747
+    },
+    {
+      "epoch": 0.3429619440623567,
+      "grad_norm": 0.5864017221182993,
+      "learning_rate": 0.0007646446392515692,
+      "loss": 2.3456,
+      "step": 748
+    },
+    {
+      "epoch": 0.34342044933516735,
+      "grad_norm": 0.5079829402226119,
+      "learning_rate": 0.000764014215766693,
+      "loss": 2.3316,
+      "step": 749
+    },
+    {
+      "epoch": 0.343878954607978,
+      "grad_norm": 0.4994793876316227,
+      "learning_rate": 0.0007633832097679958,
+      "loss": 2.3271,
+      "step": 750
+    },
+    {
+      "epoch": 0.3443374598807886,
+      "grad_norm": 0.5100290786479313,
+      "learning_rate": 0.0007627516226477122,
+      "loss": 2.3518,
+      "step": 751
+    },
+    {
+      "epoch": 0.3447959651535993,
+      "grad_norm": 0.4926163032118616,
+      "learning_rate": 0.0007621194557993589,
+      "loss": 2.4618,
+      "step": 752
+    },
+    {
+      "epoch": 0.3452544704264099,
+      "grad_norm": 0.5064721807336271,
+      "learning_rate": 0.0007614867106177319,
+      "loss": 2.3643,
+      "step": 753
+    },
+    {
+      "epoch": 0.34571297569922055,
+      "grad_norm": 0.4970866795020557,
+      "learning_rate": 0.0007608533884989029,
+      "loss": 2.3926,
+      "step": 754
+    },
+    {
+      "epoch": 0.34617148097203115,
+      "grad_norm": 0.4831166038247794,
+      "learning_rate": 0.0007602194908402166,
+      "loss": 2.2746,
+      "step": 755
+    },
+    {
+      "epoch": 0.3466299862448418,
+      "grad_norm": 0.5108659530028151,
+      "learning_rate": 0.0007595850190402877,
+      "loss": 2.3692,
+      "step": 756
+    },
+    {
+      "epoch": 0.3470884915176525,
+      "grad_norm": 0.5056909058316879,
+      "learning_rate": 0.0007589499744989976,
+      "loss": 2.4599,
+      "step": 757
+    },
+    {
+      "epoch": 0.3475469967904631,
+      "grad_norm": 0.5321640157257008,
+      "learning_rate": 0.0007583143586174916,
+      "loss": 2.3352,
+      "step": 758
+    },
+    {
+      "epoch": 0.34800550206327374,
+      "grad_norm": 0.5109504032310261,
+      "learning_rate": 0.000757678172798175,
+      "loss": 2.3597,
+      "step": 759
+    },
+    {
+      "epoch": 0.34846400733608435,
+      "grad_norm": 0.5055833937281741,
+      "learning_rate": 0.0007570414184447112,
+      "loss": 2.35,
+      "step": 760
+    },
+    {
+      "epoch": 0.348922512608895,
+      "grad_norm": 0.4950924195634894,
+      "learning_rate": 0.0007564040969620179,
+      "loss": 2.3555,
+      "step": 761
+    },
+    {
+      "epoch": 0.3493810178817056,
+      "grad_norm": 0.4876402571962275,
+      "learning_rate": 0.0007557662097562636,
+      "loss": 2.2559,
+      "step": 762
+    },
+    {
+      "epoch": 0.3498395231545163,
+      "grad_norm": 0.5188598540821798,
+      "learning_rate": 0.0007551277582348658,
+      "loss": 2.48,
+      "step": 763
+    },
+    {
+      "epoch": 0.35029802842732694,
+      "grad_norm": 0.5040855446830914,
+      "learning_rate": 0.0007544887438064862,
+      "loss": 2.3339,
+      "step": 764
+    },
+    {
+      "epoch": 0.35075653370013754,
+      "grad_norm": 0.4906688174324912,
+      "learning_rate": 0.0007538491678810294,
+      "loss": 2.2564,
+      "step": 765
+    },
+    {
+      "epoch": 0.3512150389729482,
+      "grad_norm": 0.5410756564574768,
+      "learning_rate": 0.0007532090318696381,
+      "loss": 2.3854,
+      "step": 766
+    },
+    {
+      "epoch": 0.3516735442457588,
+      "grad_norm": 0.6058224219985191,
+      "learning_rate": 0.0007525683371846913,
+      "loss": 2.3787,
+      "step": 767
+    },
+    {
+      "epoch": 0.35213204951856947,
+      "grad_norm": 0.5293859496157319,
+      "learning_rate": 0.0007519270852398001,
+      "loss": 2.3568,
+      "step": 768
+    },
+    {
+      "epoch": 0.35259055479138013,
+      "grad_norm": 0.5620338094232531,
+      "learning_rate": 0.000751285277449806,
+      "loss": 2.4028,
+      "step": 769
+    },
+    {
+      "epoch": 0.35304906006419073,
+      "grad_norm": 0.5173657368397887,
+      "learning_rate": 0.0007506429152307756,
+      "loss": 2.339,
+      "step": 770
+    },
+    {
+      "epoch": 0.3535075653370014,
+      "grad_norm": 0.549929319919795,
+      "learning_rate": 0.00075,
+      "loss": 2.3048,
+      "step": 771
+    },
+    {
+      "epoch": 0.353966070609812,
+      "grad_norm": 0.48002944562527666,
+      "learning_rate": 0.00074935653317599,
+      "loss": 2.323,
+      "step": 772
+    },
+    {
+      "epoch": 0.35442457588262266,
+      "grad_norm": 0.4842699371016886,
+      "learning_rate": 0.000748712516178473,
+      "loss": 2.3403,
+      "step": 773
+    },
+    {
+      "epoch": 0.35488308115543327,
+      "grad_norm": 0.5259971113897327,
+      "learning_rate": 0.0007480679504283911,
+      "loss": 2.3626,
+      "step": 774
+    },
+    {
+      "epoch": 0.3553415864282439,
+      "grad_norm": 0.48558145461891233,
+      "learning_rate": 0.0007474228373478964,
+      "loss": 2.3549,
+      "step": 775
+    },
+    {
+      "epoch": 0.3558000917010546,
+      "grad_norm": 0.6004714881970159,
+      "learning_rate": 0.0007467771783603492,
+      "loss": 2.2954,
+      "step": 776
+    },
+    {
+      "epoch": 0.3562585969738652,
+      "grad_norm": 0.4824522948519849,
+      "learning_rate": 0.0007461309748903138,
+      "loss": 2.3609,
+      "step": 777
+    },
+    {
+      "epoch": 0.35671710224667585,
+      "grad_norm": 0.47881723698006406,
+      "learning_rate": 0.0007454842283635562,
+      "loss": 2.3312,
+      "step": 778
+    },
+    {
+      "epoch": 0.35717560751948646,
+      "grad_norm": 0.5004570962913181,
+      "learning_rate": 0.0007448369402070404,
+      "loss": 2.2403,
+      "step": 779
+    },
+    {
+      "epoch": 0.3576341127922971,
+      "grad_norm": 0.5302758640511284,
+      "learning_rate": 0.0007441891118489254,
+      "loss": 2.3623,
+      "step": 780
+    },
+    {
+      "epoch": 0.3580926180651077,
+      "grad_norm": 0.4951247340723839,
+      "learning_rate": 0.0007435407447185622,
+      "loss": 2.3282,
+      "step": 781
+    },
+    {
+      "epoch": 0.3585511233379184,
+      "grad_norm": 0.5098353263824458,
+      "learning_rate": 0.0007428918402464908,
+      "loss": 2.3191,
+      "step": 782
+    },
+    {
+      "epoch": 0.35900962861072905,
+      "grad_norm": 0.5059419660270986,
+      "learning_rate": 0.0007422423998644359,
+      "loss": 2.2288,
+      "step": 783
+    },
+    {
+      "epoch": 0.35946813388353965,
+      "grad_norm": 0.4873217372357919,
+      "learning_rate": 0.0007415924250053055,
+      "loss": 2.359,
+      "step": 784
+    },
+    {
+      "epoch": 0.3599266391563503,
+      "grad_norm": 0.5203206887187901,
+      "learning_rate": 0.0007409419171031865,
+      "loss": 2.3401,
+      "step": 785
+    },
+    {
+      "epoch": 0.3603851444291609,
+      "grad_norm": 0.49636219113405705,
+      "learning_rate": 0.0007402908775933419,
+      "loss": 2.2547,
+      "step": 786
+    },
+    {
+      "epoch": 0.3608436497019716,
+      "grad_norm": 0.505084982336452,
+      "learning_rate": 0.0007396393079122077,
+      "loss": 2.2564,
+      "step": 787
+    },
+    {
+      "epoch": 0.3613021549747822,
+      "grad_norm": 0.5373035166737021,
+      "learning_rate": 0.0007389872094973896,
+      "loss": 2.277,
+      "step": 788
+    },
+    {
+      "epoch": 0.36176066024759285,
+      "grad_norm": 0.4828929133851646,
+      "learning_rate": 0.00073833458378766,
+      "loss": 2.2981,
+      "step": 789
+    },
+    {
+      "epoch": 0.3622191655204035,
+      "grad_norm": 0.5878539905666519,
+      "learning_rate": 0.0007376814322229544,
+      "loss": 2.4617,
+      "step": 790
+    },
+    {
+      "epoch": 0.3626776707932141,
+      "grad_norm": 0.520910001665015,
+      "learning_rate": 0.0007370277562443688,
+      "loss": 2.3314,
+      "step": 791
+    },
+    {
+      "epoch": 0.3631361760660248,
+      "grad_norm": 0.5062607060123825,
+      "learning_rate": 0.0007363735572941564,
+      "loss": 2.4008,
+      "step": 792
+    },
+    {
+      "epoch": 0.3635946813388354,
+      "grad_norm": 0.5315024147914118,
+      "learning_rate": 0.0007357188368157236,
+      "loss": 2.4207,
+      "step": 793
+    },
+    {
+      "epoch": 0.36405318661164604,
+      "grad_norm": 0.5101132128187713,
+      "learning_rate": 0.0007350635962536284,
+      "loss": 2.2956,
+      "step": 794
+    },
+    {
+      "epoch": 0.36451169188445665,
+      "grad_norm": 0.49950766328711005,
+      "learning_rate": 0.0007344078370535756,
+      "loss": 2.3124,
+      "step": 795
+    },
+    {
+      "epoch": 0.3649701971572673,
+      "grad_norm": 0.4636792385656284,
+      "learning_rate": 0.0007337515606624148,
+      "loss": 2.3103,
+      "step": 796
+    },
+    {
+      "epoch": 0.36542870243007797,
+      "grad_norm": 0.5138281697289474,
+      "learning_rate": 0.0007330947685281362,
+      "loss": 2.2683,
+      "step": 797
+    },
+    {
+      "epoch": 0.3658872077028886,
+      "grad_norm": 0.5297462341408407,
+      "learning_rate": 0.0007324374620998682,
+      "loss": 2.3439,
+      "step": 798
+    },
+    {
+      "epoch": 0.36634571297569923,
+      "grad_norm": 0.47320557280262093,
+      "learning_rate": 0.000731779642827874,
+      "loss": 2.255,
+      "step": 799
+    },
+    {
+      "epoch": 0.36680421824850984,
+      "grad_norm": 0.5531955695107711,
+      "learning_rate": 0.0007311213121635483,
+      "loss": 2.358,
+      "step": 800
+    },
+    {
+      "epoch": 0.3672627235213205,
+      "grad_norm": 0.5575395073204374,
+      "learning_rate": 0.0007304624715594139,
+      "loss": 2.3307,
+      "step": 801
+    },
+    {
+      "epoch": 0.3677212287941311,
+      "grad_norm": 0.5196796528849716,
+      "learning_rate": 0.0007298031224691193,
+      "loss": 2.3653,
+      "step": 802
+    },
+    {
+      "epoch": 0.36817973406694177,
+      "grad_norm": 0.5415326232106318,
+      "learning_rate": 0.0007291432663474339,
+      "loss": 2.2836,
+      "step": 803
+    },
+    {
+      "epoch": 0.3686382393397524,
+      "grad_norm": 0.49083706698718116,
+      "learning_rate": 0.0007284829046502467,
+      "loss": 2.3106,
+      "step": 804
+    },
+    {
+      "epoch": 0.36909674461256303,
+      "grad_norm": 0.4696076863782398,
+      "learning_rate": 0.0007278220388345619,
+      "loss": 2.336,
+      "step": 805
+    },
+    {
+      "epoch": 0.3695552498853737,
+      "grad_norm": 0.4598637229342556,
+      "learning_rate": 0.0007271606703584958,
+      "loss": 2.2542,
+      "step": 806
+    },
+    {
+      "epoch": 0.3700137551581843,
+      "grad_norm": 0.4933919597218924,
+      "learning_rate": 0.000726498800681274,
+      "loss": 2.2781,
+      "step": 807
+    },
+    {
+      "epoch": 0.37047226043099496,
+      "grad_norm": 0.5077903449287385,
+      "learning_rate": 0.0007258364312632279,
+      "loss": 2.3157,
+      "step": 808
+    },
+    {
+      "epoch": 0.3709307657038056,
+      "grad_norm": 0.5034586532018497,
+      "learning_rate": 0.0007251735635657915,
+      "loss": 2.3029,
+      "step": 809
+    },
+    {
+      "epoch": 0.3713892709766162,
+      "grad_norm": 0.5003264111922728,
+      "learning_rate": 0.000724510199051498,
+      "loss": 2.3296,
+      "step": 810
+    },
+    {
+      "epoch": 0.3718477762494269,
+      "grad_norm": 0.5107589613164051,
+      "learning_rate": 0.0007238463391839769,
+      "loss": 2.3346,
+      "step": 811
+    },
+    {
+      "epoch": 0.3723062815222375,
+      "grad_norm": 0.5063688410334187,
+      "learning_rate": 0.0007231819854279508,
+      "loss": 2.4067,
+      "step": 812
+    },
+    {
+      "epoch": 0.37276478679504815,
+      "grad_norm": 0.6076857242553011,
+      "learning_rate": 0.0007225171392492316,
+      "loss": 2.3056,
+      "step": 813
+    },
+    {
+      "epoch": 0.37322329206785876,
+      "grad_norm": 0.49332470048070076,
+      "learning_rate": 0.0007218518021147182,
+      "loss": 2.3795,
+      "step": 814
+    },
+    {
+      "epoch": 0.3736817973406694,
+      "grad_norm": 0.5123086970249099,
+      "learning_rate": 0.0007211859754923923,
+      "loss": 2.3422,
+      "step": 815
+    },
+    {
+      "epoch": 0.3741403026134801,
+      "grad_norm": 0.5958529782201946,
+      "learning_rate": 0.0007205196608513158,
+      "loss": 2.2727,
+      "step": 816
+    },
+    {
+      "epoch": 0.3745988078862907,
+      "grad_norm": 0.49983207063767116,
+      "learning_rate": 0.0007198528596616272,
+      "loss": 2.4435,
+      "step": 817
+    },
+    {
+      "epoch": 0.37505731315910135,
+      "grad_norm": 0.4779226930836834,
+      "learning_rate": 0.0007191855733945387,
+      "loss": 2.2447,
+      "step": 818
+    },
+    {
+      "epoch": 0.37551581843191195,
+      "grad_norm": 0.4503218233608779,
+      "learning_rate": 0.0007185178035223327,
+      "loss": 2.2606,
+      "step": 819
+    },
+    {
+      "epoch": 0.3759743237047226,
+      "grad_norm": 0.5168320884226817,
+      "learning_rate": 0.0007178495515183583,
+      "loss": 2.3248,
+      "step": 820
+    },
+    {
+      "epoch": 0.3764328289775332,
+      "grad_norm": 0.5340502667602426,
+      "learning_rate": 0.000717180818857029,
+      "loss": 2.3186,
+      "step": 821
+    },
+    {
+      "epoch": 0.3768913342503439,
+      "grad_norm": 0.5616671581444059,
+      "learning_rate": 0.0007165116070138182,
+      "loss": 2.2337,
+      "step": 822
+    },
+    {
+      "epoch": 0.37734983952315454,
+      "grad_norm": 0.47823693178388105,
+      "learning_rate": 0.0007158419174652569,
+      "loss": 2.2082,
+      "step": 823
+    },
+    {
+      "epoch": 0.37780834479596515,
+      "grad_norm": 0.49828191404987715,
+      "learning_rate": 0.00071517175168893,
+      "loss": 2.3108,
+      "step": 824
+    },
+    {
+      "epoch": 0.3782668500687758,
+      "grad_norm": 0.5335194536681442,
+      "learning_rate": 0.0007145011111634732,
+      "loss": 2.4071,
+      "step": 825
+    },
+    {
+      "epoch": 0.3787253553415864,
+      "grad_norm": 0.5317595451786741,
+      "learning_rate": 0.0007138299973685694,
+      "loss": 2.386,
+      "step": 826
+    },
+    {
+      "epoch": 0.3791838606143971,
+      "grad_norm": 0.5203396877246573,
+      "learning_rate": 0.0007131584117849459,
+      "loss": 2.2945,
+      "step": 827
+    },
+    {
+      "epoch": 0.3796423658872077,
+      "grad_norm": 0.5293887885988519,
+      "learning_rate": 0.0007124863558943713,
+      "loss": 2.3499,
+      "step": 828
+    },
+    {
+      "epoch": 0.38010087116001834,
+      "grad_norm": 0.558538011040775,
+      "learning_rate": 0.0007118138311796514,
+      "loss": 2.4076,
+      "step": 829
+    },
+    {
+      "epoch": 0.380559376432829,
+      "grad_norm": 0.5194227667352685,
+      "learning_rate": 0.0007111408391246262,
+      "loss": 2.2966,
+      "step": 830
+    },
+    {
+      "epoch": 0.3810178817056396,
+      "grad_norm": 0.48355088094009163,
+      "learning_rate": 0.0007104673812141675,
+      "loss": 2.3135,
+      "step": 831
+    },
+    {
+      "epoch": 0.38147638697845027,
+      "grad_norm": 0.515029347561018,
+      "learning_rate": 0.0007097934589341745,
+      "loss": 2.4014,
+      "step": 832
+    },
+    {
+      "epoch": 0.38193489225126087,
+      "grad_norm": 0.5214767622927807,
+      "learning_rate": 0.0007091190737715711,
+      "loss": 2.3752,
+      "step": 833
+    },
+    {
+      "epoch": 0.38239339752407153,
+      "grad_norm": 0.515849588351304,
+      "learning_rate": 0.0007084442272143026,
+      "loss": 2.3706,
+      "step": 834
+    },
+    {
+      "epoch": 0.38285190279688214,
+      "grad_norm": 0.47438754666343674,
+      "learning_rate": 0.000707768920751332,
+      "loss": 2.3584,
+      "step": 835
+    },
+    {
+      "epoch": 0.3833104080696928,
+      "grad_norm": 0.49370094243096074,
+      "learning_rate": 0.0007070931558726373,
+      "loss": 2.3784,
+      "step": 836
+    },
+    {
+      "epoch": 0.38376891334250346,
+      "grad_norm": 0.45177894576663824,
+      "learning_rate": 0.0007064169340692076,
+      "loss": 2.3062,
+      "step": 837
+    },
+    {
+      "epoch": 0.38422741861531406,
+      "grad_norm": 0.5292922546301186,
+      "learning_rate": 0.0007057402568330407,
+      "loss": 2.206,
+      "step": 838
+    },
+    {
+      "epoch": 0.3846859238881247,
+      "grad_norm": 0.5732383222222189,
+      "learning_rate": 0.0007050631256571389,
+      "loss": 2.2552,
+      "step": 839
+    },
+    {
+      "epoch": 0.38514442916093533,
+      "grad_norm": 0.5430606247165943,
+      "learning_rate": 0.000704385542035506,
+      "loss": 2.302,
+      "step": 840
+    },
+    {
+      "epoch": 0.385602934433746,
+      "grad_norm": 0.49587513699692237,
+      "learning_rate": 0.000703707507463144,
+      "loss": 2.2661,
+      "step": 841
+    },
+    {
+      "epoch": 0.38606143970655665,
+      "grad_norm": 0.5614630612795471,
+      "learning_rate": 0.0007030290234360505,
+      "loss": 2.4366,
+      "step": 842
+    },
+    {
+      "epoch": 0.38651994497936726,
+      "grad_norm": 0.5365337002584697,
+      "learning_rate": 0.0007023500914512139,
+      "loss": 2.3422,
+      "step": 843
+    },
+    {
+      "epoch": 0.3869784502521779,
+      "grad_norm": 0.5072774147256215,
+      "learning_rate": 0.0007016707130066116,
+      "loss": 2.2878,
+      "step": 844
+    },
+    {
+      "epoch": 0.3874369555249885,
+      "grad_norm": 0.5300534630286812,
+      "learning_rate": 0.0007009908896012055,
+      "loss": 2.3215,
+      "step": 845
+    },
+    {
+      "epoch": 0.3878954607977992,
+      "grad_norm": 0.6318139250383273,
+      "learning_rate": 0.0007003106227349399,
+      "loss": 2.388,
+      "step": 846
+    },
+    {
+      "epoch": 0.3883539660706098,
+      "grad_norm": 0.6140585450674173,
+      "learning_rate": 0.000699629913908737,
+      "loss": 2.3622,
+      "step": 847
+    },
+    {
+      "epoch": 0.38881247134342045,
+      "grad_norm": 0.5246325937484992,
+      "learning_rate": 0.0006989487646244943,
+      "loss": 2.3176,
+      "step": 848
+    },
+    {
+      "epoch": 0.3892709766162311,
+      "grad_norm": 0.5119757887143743,
+      "learning_rate": 0.0006982671763850814,
+      "loss": 2.3044,
+      "step": 849
+    },
+    {
+      "epoch": 0.3897294818890417,
+      "grad_norm": 0.48100067947536984,
+      "learning_rate": 0.0006975851506943359,
+      "loss": 2.2846,
+      "step": 850
+    },
+    {
+      "epoch": 0.3901879871618524,
+      "grad_norm": 0.5542296367649585,
+      "learning_rate": 0.0006969026890570611,
+      "loss": 2.4135,
+      "step": 851
+    },
+    {
+      "epoch": 0.390646492434663,
+      "grad_norm": 0.5098985050289393,
+      "learning_rate": 0.0006962197929790216,
+      "loss": 2.2549,
+      "step": 852
+    },
+    {
+      "epoch": 0.39110499770747365,
+      "grad_norm": 0.5018123831090496,
+      "learning_rate": 0.0006955364639669409,
+      "loss": 2.3829,
+      "step": 853
+    },
+    {
+      "epoch": 0.39156350298028425,
+      "grad_norm": 0.5083898345832703,
+      "learning_rate": 0.0006948527035284978,
+      "loss": 2.3626,
+      "step": 854
+    },
+    {
+      "epoch": 0.3920220082530949,
+      "grad_norm": 0.535731837061056,
+      "learning_rate": 0.0006941685131723225,
+      "loss": 2.4153,
+      "step": 855
+    },
+    {
+      "epoch": 0.39248051352590557,
+      "grad_norm": 0.46276564434910383,
+      "learning_rate": 0.0006934838944079943,
+      "loss": 2.3182,
+      "step": 856
+    },
+    {
+      "epoch": 0.3929390187987162,
+      "grad_norm": 0.4813534436678019,
+      "learning_rate": 0.0006927988487460378,
+      "loss": 2.3067,
+      "step": 857
+    },
+    {
+      "epoch": 0.39339752407152684,
+      "grad_norm": 0.5421055312249927,
+      "learning_rate": 0.0006921133776979186,
+      "loss": 2.3034,
+      "step": 858
+    },
+    {
+      "epoch": 0.39385602934433744,
+      "grad_norm": 0.5132706057520444,
+      "learning_rate": 0.0006914274827760418,
+      "loss": 2.2901,
+      "step": 859
+    },
+    {
+      "epoch": 0.3943145346171481,
+      "grad_norm": 0.49991165601630766,
+      "learning_rate": 0.0006907411654937475,
+      "loss": 2.3619,
+      "step": 860
+    },
+    {
+      "epoch": 0.3947730398899587,
+      "grad_norm": 0.47989588855232634,
+      "learning_rate": 0.0006900544273653075,
+      "loss": 2.4495,
+      "step": 861
+    },
+    {
+      "epoch": 0.39523154516276937,
+      "grad_norm": 0.49756492899903604,
+      "learning_rate": 0.000689367269905922,
+      "loss": 2.3907,
+      "step": 862
+    },
+    {
+      "epoch": 0.39569005043558003,
+      "grad_norm": 0.491099424104108,
+      "learning_rate": 0.0006886796946317168,
+      "loss": 2.3468,
+      "step": 863
+    },
+    {
+      "epoch": 0.39614855570839064,
+      "grad_norm": 0.4819225307110801,
+      "learning_rate": 0.0006879917030597397,
+      "loss": 2.3252,
+      "step": 864
+    },
+    {
+      "epoch": 0.3966070609812013,
+      "grad_norm": 0.49371641452632786,
+      "learning_rate": 0.0006873032967079561,
+      "loss": 2.3026,
+      "step": 865
+    },
+    {
+      "epoch": 0.3970655662540119,
+      "grad_norm": 0.48724925871706615,
+      "learning_rate": 0.0006866144770952474,
+      "loss": 2.335,
+      "step": 866
+    },
+    {
+      "epoch": 0.39752407152682256,
+      "grad_norm": 0.475348612389198,
+      "learning_rate": 0.0006859252457414067,
+      "loss": 2.3914,
+      "step": 867
+    },
+    {
+      "epoch": 0.39798257679963317,
+      "grad_norm": 0.5164663928053809,
+      "learning_rate": 0.0006852356041671351,
+      "loss": 2.3013,
+      "step": 868
+    },
+    {
+      "epoch": 0.39844108207244383,
+      "grad_norm": 0.4785817093370589,
+      "learning_rate": 0.0006845455538940394,
+      "loss": 2.3201,
+      "step": 869
+    },
+    {
+      "epoch": 0.3988995873452545,
+      "grad_norm": 0.8537329261025568,
+      "learning_rate": 0.0006838550964446276,
+      "loss": 2.3324,
+      "step": 870
+    },
+    {
+      "epoch": 0.3993580926180651,
+      "grad_norm": 1.0393380748859995,
+      "learning_rate": 0.0006831642333423067,
+      "loss": 2.3989,
+      "step": 871
+    },
+    {
+      "epoch": 0.39981659789087576,
+      "grad_norm": 0.5111786208964053,
+      "learning_rate": 0.000682472966111378,
+      "loss": 2.31,
+      "step": 872
+    },
+    {
+      "epoch": 0.40027510316368636,
+      "grad_norm": 0.49291536911009215,
+      "learning_rate": 0.0006817812962770348,
+      "loss": 2.3276,
+      "step": 873
+    },
+    {
+      "epoch": 0.400733608436497,
+      "grad_norm": 0.5050132588020874,
+      "learning_rate": 0.0006810892253653589,
+      "loss": 2.293,
+      "step": 874
+    },
+    {
+      "epoch": 0.40119211370930763,
+      "grad_norm": 0.49344733949856345,
+      "learning_rate": 0.0006803967549033167,
+      "loss": 2.3188,
+      "step": 875
+    },
+    {
+      "epoch": 0.4016506189821183,
+      "grad_norm": 0.5226190468155382,
+      "learning_rate": 0.0006797038864187564,
+      "loss": 2.3906,
+      "step": 876
+    },
+    {
+      "epoch": 0.40210912425492895,
+      "grad_norm": 0.49019607502620155,
+      "learning_rate": 0.0006790106214404043,
+      "loss": 2.399,
+      "step": 877
+    },
+    {
+      "epoch": 0.40256762952773956,
+      "grad_norm": 0.5911254301844122,
+      "learning_rate": 0.0006783169614978614,
+      "loss": 2.3334,
+      "step": 878
+    },
+    {
+      "epoch": 0.4030261348005502,
+      "grad_norm": 0.5046952240846649,
+      "learning_rate": 0.0006776229081216001,
+      "loss": 2.3513,
+      "step": 879
+    },
+    {
+      "epoch": 0.4034846400733608,
+      "grad_norm": 0.4572457434019569,
+      "learning_rate": 0.0006769284628429611,
+      "loss": 2.3002,
+      "step": 880
+    },
+    {
+      "epoch": 0.4039431453461715,
+      "grad_norm": 0.5005960004303357,
+      "learning_rate": 0.0006762336271941498,
+      "loss": 2.3383,
+      "step": 881
+    },
+    {
+      "epoch": 0.40440165061898214,
+      "grad_norm": 2.8214063557307223,
+      "learning_rate": 0.0006755384027082326,
+      "loss": 2.3165,
+      "step": 882
+    },
+    {
+      "epoch": 0.40486015589179275,
+      "grad_norm": 0.509506678674573,
+      "learning_rate": 0.0006748427909191342,
+      "loss": 2.2705,
+      "step": 883
+    },
+    {
+      "epoch": 0.4053186611646034,
+      "grad_norm": 0.491279329009541,
+      "learning_rate": 0.0006741467933616335,
+      "loss": 2.3341,
+      "step": 884
+    },
+    {
+      "epoch": 0.405777166437414,
+      "grad_norm": 0.5067600549918637,
+      "learning_rate": 0.0006734504115713604,
+      "loss": 2.2896,
+      "step": 885
+    },
+    {
+      "epoch": 0.4062356717102247,
+      "grad_norm": 0.4968550362040195,
+      "learning_rate": 0.0006727536470847932,
+      "loss": 2.3295,
+      "step": 886
+    },
+    {
+      "epoch": 0.4066941769830353,
+      "grad_norm": 0.5003530961733217,
+      "learning_rate": 0.000672056501439254,
+      "loss": 2.3175,
+      "step": 887
+    },
+    {
+      "epoch": 0.40715268225584594,
+      "grad_norm": 0.5309836604421927,
+      "learning_rate": 0.0006713589761729063,
+      "loss": 2.3199,
+      "step": 888
+    },
+    {
+      "epoch": 0.4076111875286566,
+      "grad_norm": 0.5116088713729465,
+      "learning_rate": 0.0006706610728247508,
+      "loss": 2.3589,
+      "step": 889
+    },
+    {
+      "epoch": 0.4080696928014672,
+      "grad_norm": 0.4815047089148338,
+      "learning_rate": 0.0006699627929346227,
+      "loss": 2.3842,
+      "step": 890
+    },
+    {
+      "epoch": 0.40852819807427787,
+      "grad_norm": 0.503596743309734,
+      "learning_rate": 0.0006692641380431879,
+      "loss": 2.3177,
+      "step": 891
+    },
+    {
+      "epoch": 0.4089867033470885,
+      "grad_norm": 0.4747390866614862,
+      "learning_rate": 0.0006685651096919393,
+      "loss": 2.324,
+      "step": 892
+    },
+    {
+      "epoch": 0.40944520861989914,
+      "grad_norm": 0.537421435388401,
+      "learning_rate": 0.0006678657094231944,
+      "loss": 2.1521,
+      "step": 893
+    },
+    {
+      "epoch": 0.40990371389270974,
+      "grad_norm": 0.49873459167972334,
+      "learning_rate": 0.0006671659387800909,
+      "loss": 2.4866,
+      "step": 894
+    },
+    {
+      "epoch": 0.4103622191655204,
+      "grad_norm": 0.4745868093224122,
+      "learning_rate": 0.000666465799306584,
+      "loss": 2.3241,
+      "step": 895
+    },
+    {
+      "epoch": 0.41082072443833106,
+      "grad_norm": 0.5008468885353613,
+      "learning_rate": 0.0006657652925474423,
+      "loss": 2.3452,
+      "step": 896
+    },
+    {
+      "epoch": 0.41127922971114167,
+      "grad_norm": 0.5038389403092872,
+      "learning_rate": 0.000665064420048245,
+      "loss": 2.359,
+      "step": 897
+    },
+    {
+      "epoch": 0.41173773498395233,
+      "grad_norm": 0.485461593873843,
+      "learning_rate": 0.0006643631833553785,
+      "loss": 2.2804,
+      "step": 898
+    },
+    {
+      "epoch": 0.41219624025676294,
+      "grad_norm": 0.4802270866430993,
+      "learning_rate": 0.000663661584016032,
+      "loss": 2.2434,
+      "step": 899
+    },
+    {
+      "epoch": 0.4126547455295736,
+      "grad_norm": 0.49353088885544216,
+      "learning_rate": 0.0006629596235781957,
+      "loss": 2.3648,
+      "step": 900
+    },
+    {
+      "epoch": 0.4131132508023842,
+      "grad_norm": 0.48502576425830324,
+      "learning_rate": 0.0006622573035906556,
+      "loss": 2.3578,
+      "step": 901
+    },
+    {
+      "epoch": 0.41357175607519486,
+      "grad_norm": 0.5072988137996871,
+      "learning_rate": 0.0006615546256029921,
+      "loss": 2.3631,
+      "step": 902
+    },
+    {
+      "epoch": 0.4140302613480055,
+      "grad_norm": 0.5462239532656361,
+      "learning_rate": 0.0006608515911655743,
+      "loss": 2.5042,
+      "step": 903
+    },
+    {
+      "epoch": 0.41448876662081613,
+      "grad_norm": 0.49373676029166835,
+      "learning_rate": 0.0006601482018295591,
+      "loss": 2.3484,
+      "step": 904
+    },
+    {
+      "epoch": 0.4149472718936268,
+      "grad_norm": 0.500831106747745,
+      "learning_rate": 0.0006594444591468851,
+      "loss": 2.378,
+      "step": 905
+    },
+    {
+      "epoch": 0.4154057771664374,
+      "grad_norm": 0.4324445528042812,
+      "learning_rate": 0.0006587403646702713,
+      "loss": 2.3619,
+      "step": 906
+    },
+    {
+      "epoch": 0.41586428243924806,
+      "grad_norm": 0.4788234445609536,
+      "learning_rate": 0.0006580359199532126,
+      "loss": 2.3112,
+      "step": 907
+    },
+    {
+      "epoch": 0.41632278771205866,
+      "grad_norm": 0.4855897338734282,
+      "learning_rate": 0.000657331126549977,
+      "loss": 2.2753,
+      "step": 908
+    },
+    {
+      "epoch": 0.4167812929848693,
+      "grad_norm": 0.4482218057542594,
+      "learning_rate": 0.0006566259860156014,
+      "loss": 2.3048,
+      "step": 909
+    },
+    {
+      "epoch": 0.41723979825768,
+      "grad_norm": 0.4830424336860545,
+      "learning_rate": 0.0006559204999058888,
+      "loss": 2.3546,
+      "step": 910
+    },
+    {
+      "epoch": 0.4176983035304906,
+      "grad_norm": 0.4782785567987006,
+      "learning_rate": 0.0006552146697774049,
+      "loss": 2.3745,
+      "step": 911
+    },
+    {
+      "epoch": 0.41815680880330125,
+      "grad_norm": 0.5095239889634433,
+      "learning_rate": 0.0006545084971874737,
+      "loss": 2.3372,
+      "step": 912
+    },
+    {
+      "epoch": 0.41861531407611186,
+      "grad_norm": 0.4884241331376324,
+      "learning_rate": 0.0006538019836941758,
+      "loss": 2.351,
+      "step": 913
+    },
+    {
+      "epoch": 0.4190738193489225,
+      "grad_norm": 0.4848620675466508,
+      "learning_rate": 0.0006530951308563431,
+      "loss": 2.3925,
+      "step": 914
+    },
+    {
+      "epoch": 0.4195323246217332,
+      "grad_norm": 0.46120046735697584,
+      "learning_rate": 0.0006523879402335567,
+      "loss": 2.3365,
+      "step": 915
+    },
+    {
+      "epoch": 0.4199908298945438,
+      "grad_norm": 0.4940612419637646,
+      "learning_rate": 0.0006516804133861429,
+      "loss": 2.3602,
+      "step": 916
+    },
+    {
+      "epoch": 0.42044933516735444,
+      "grad_norm": 0.46522652244466806,
+      "learning_rate": 0.0006509725518751698,
+      "loss": 2.2658,
+      "step": 917
+    },
+    {
+      "epoch": 0.42090784044016505,
+      "grad_norm": 0.4954687449056919,
+      "learning_rate": 0.0006502643572624438,
+      "loss": 2.3434,
+      "step": 918
+    },
+    {
+      "epoch": 0.4213663457129757,
+      "grad_norm": 0.4859172560489846,
+      "learning_rate": 0.0006495558311105064,
+      "loss": 2.267,
+      "step": 919
+    },
+    {
+      "epoch": 0.4218248509857863,
+      "grad_norm": 0.4586674038353237,
+      "learning_rate": 0.0006488469749826305,
+      "loss": 2.3184,
+      "step": 920
+    },
+    {
+      "epoch": 0.422283356258597,
+      "grad_norm": 0.48614271957226973,
+      "learning_rate": 0.000648137790442817,
+      "loss": 2.3533,
+      "step": 921
+    },
+    {
+      "epoch": 0.42274186153140764,
+      "grad_norm": 0.46771626618352563,
+      "learning_rate": 0.0006474282790557916,
+      "loss": 2.2455,
+      "step": 922
+    },
+    {
+      "epoch": 0.42320036680421824,
+      "grad_norm": 0.5961543172411146,
+      "learning_rate": 0.000646718442387001,
+      "loss": 2.2879,
+      "step": 923
+    },
+    {
+      "epoch": 0.4236588720770289,
+      "grad_norm": 0.4611278419759929,
+      "learning_rate": 0.0006460082820026094,
+      "loss": 2.3021,
+      "step": 924
+    },
+    {
+      "epoch": 0.4241173773498395,
+      "grad_norm": 0.4722489019626595,
+      "learning_rate": 0.0006452977994694959,
+      "loss": 2.2233,
+      "step": 925
+    },
+    {
+      "epoch": 0.42457588262265017,
+      "grad_norm": 0.466201110006326,
+      "learning_rate": 0.0006445869963552496,
+      "loss": 2.2749,
+      "step": 926
+    },
+    {
+      "epoch": 0.4250343878954608,
+      "grad_norm": 0.45798441854099264,
+      "learning_rate": 0.0006438758742281672,
+      "loss": 2.3263,
+      "step": 927
+    },
+    {
+      "epoch": 0.42549289316827144,
+      "grad_norm": 0.4872282062305679,
+      "learning_rate": 0.0006431644346572495,
+      "loss": 2.2801,
+      "step": 928
+    },
+    {
+      "epoch": 0.4259513984410821,
+      "grad_norm": 0.48415165316449826,
+      "learning_rate": 0.0006424526792121974,
+      "loss": 2.2466,
+      "step": 929
+    },
+    {
+      "epoch": 0.4264099037138927,
+      "grad_norm": 0.462843417474602,
+      "learning_rate": 0.0006417406094634089,
+      "loss": 2.3398,
+      "step": 930
+    },
+    {
+      "epoch": 0.42686840898670336,
+      "grad_norm": 0.48574803601537647,
+      "learning_rate": 0.0006410282269819756,
+      "loss": 2.2943,
+      "step": 931
+    },
+    {
+      "epoch": 0.42732691425951397,
+      "grad_norm": 0.5042989980566513,
+      "learning_rate": 0.0006403155333396787,
+      "loss": 2.3835,
+      "step": 932
+    },
+    {
+      "epoch": 0.42778541953232463,
+      "grad_norm": 0.4811331517265269,
+      "learning_rate": 0.0006396025301089863,
+      "loss": 2.3891,
+      "step": 933
+    },
+    {
+      "epoch": 0.42824392480513523,
+      "grad_norm": 0.4621071216718968,
+      "learning_rate": 0.0006388892188630493,
+      "loss": 2.2592,
+      "step": 934
+    },
+    {
+      "epoch": 0.4287024300779459,
+      "grad_norm": 0.493620770908743,
+      "learning_rate": 0.0006381756011756982,
+      "loss": 2.3253,
+      "step": 935
+    },
+    {
+      "epoch": 0.42916093535075656,
+      "grad_norm": 0.4595917357135942,
+      "learning_rate": 0.0006374616786214403,
+      "loss": 2.3742,
+      "step": 936
+    },
+    {
+      "epoch": 0.42961944062356716,
+      "grad_norm": 0.4239861902530995,
+      "learning_rate": 0.0006367474527754544,
+      "loss": 2.3108,
+      "step": 937
+    },
+    {
+      "epoch": 0.4300779458963778,
+      "grad_norm": 0.4540402154348363,
+      "learning_rate": 0.0006360329252135894,
+      "loss": 2.3788,
+      "step": 938
+    },
+    {
+      "epoch": 0.4305364511691884,
+      "grad_norm": 0.448882525891929,
+      "learning_rate": 0.0006353180975123595,
+      "loss": 2.3681,
+      "step": 939
+    },
+    {
+      "epoch": 0.4309949564419991,
+      "grad_norm": 0.490079035170157,
+      "learning_rate": 0.0006346029712489413,
+      "loss": 2.3343,
+      "step": 940
+    },
+    {
+      "epoch": 0.4314534617148097,
+      "grad_norm": 0.4728909837676357,
+      "learning_rate": 0.0006338875480011698,
+      "loss": 2.4174,
+      "step": 941
+    },
+    {
+      "epoch": 0.43191196698762035,
+      "grad_norm": 0.45345192131924794,
+      "learning_rate": 0.0006331718293475357,
+      "loss": 2.2331,
+      "step": 942
+    },
+    {
+      "epoch": 0.432370472260431,
+      "grad_norm": 0.4617551806059198,
+      "learning_rate": 0.0006324558168671811,
+      "loss": 2.3143,
+      "step": 943
+    },
+    {
+      "epoch": 0.4328289775332416,
+      "grad_norm": 0.4653566814019945,
+      "learning_rate": 0.0006317395121398968,
+      "loss": 2.3628,
+      "step": 944
+    },
+    {
+      "epoch": 0.4332874828060523,
+      "grad_norm": 0.4793788332592386,
+      "learning_rate": 0.0006310229167461179,
+      "loss": 2.308,
+      "step": 945
+    },
+    {
+      "epoch": 0.4337459880788629,
+      "grad_norm": 0.5396371567317124,
+      "learning_rate": 0.0006303060322669214,
+      "loss": 2.2065,
+      "step": 946
+    },
+    {
+      "epoch": 0.43420449335167355,
+      "grad_norm": 0.5157898899789563,
+      "learning_rate": 0.0006295888602840214,
+      "loss": 2.4563,
+      "step": 947
+    },
+    {
+      "epoch": 0.4346629986244842,
+      "grad_norm": 0.517868225426727,
+      "learning_rate": 0.0006288714023797671,
+      "loss": 2.3305,
+      "step": 948
+    },
+    {
+      "epoch": 0.4351215038972948,
+      "grad_norm": 0.5178879449351766,
+      "learning_rate": 0.000628153660137138,
+      "loss": 2.4685,
+      "step": 949
+    },
+    {
+      "epoch": 0.4355800091701055,
+      "grad_norm": 0.4843856174416667,
+      "learning_rate": 0.0006274356351397413,
+      "loss": 2.3532,
+      "step": 950
+    },
+    {
+      "epoch": 0.4360385144429161,
+      "grad_norm": 0.45866043968695985,
+      "learning_rate": 0.0006267173289718079,
+      "loss": 2.2673,
+      "step": 951
+    },
+    {
+      "epoch": 0.43649701971572674,
+      "grad_norm": 0.6503733814877684,
+      "learning_rate": 0.000625998743218189,
+      "loss": 2.3281,
+      "step": 952
+    },
+    {
+      "epoch": 0.43695552498853735,
+      "grad_norm": 0.48812781617358725,
+      "learning_rate": 0.000625279879464353,
+      "loss": 2.2717,
+      "step": 953
+    },
+    {
+      "epoch": 0.437414030261348,
+      "grad_norm": 0.48781449826014234,
+      "learning_rate": 0.000624560739296381,
+      "loss": 2.3315,
+      "step": 954
+    },
+    {
+      "epoch": 0.43787253553415867,
+      "grad_norm": 0.48836540402971934,
+      "learning_rate": 0.0006238413243009648,
+      "loss": 2.2527,
+      "step": 955
+    },
+    {
+      "epoch": 0.4383310408069693,
+      "grad_norm": 0.49298361673283486,
+      "learning_rate": 0.000623121636065402,
+      "loss": 2.3316,
+      "step": 956
+    },
+    {
+      "epoch": 0.43878954607977994,
+      "grad_norm": 0.46976442826309894,
+      "learning_rate": 0.0006224016761775933,
+      "loss": 2.2871,
+      "step": 957
+    },
+    {
+      "epoch": 0.43924805135259054,
+      "grad_norm": 0.5330759183174556,
+      "learning_rate": 0.0006216814462260386,
+      "loss": 2.3943,
+      "step": 958
+    },
+    {
+      "epoch": 0.4397065566254012,
+      "grad_norm": 0.4880964702020806,
+      "learning_rate": 0.0006209609477998338,
+      "loss": 2.2553,
+      "step": 959
+    },
+    {
+      "epoch": 0.4401650618982118,
+      "grad_norm": 0.47725495319874045,
+      "learning_rate": 0.0006202401824886674,
+      "loss": 2.3407,
+      "step": 960
+    },
+    {
+      "epoch": 0.44062356717102247,
+      "grad_norm": 0.504217311952684,
+      "learning_rate": 0.0006195191518828162,
+      "loss": 2.3257,
+      "step": 961
+    },
+    {
+      "epoch": 0.44108207244383313,
+      "grad_norm": 0.8035057981942154,
+      "learning_rate": 0.0006187978575731427,
+      "loss": 2.346,
+      "step": 962
+    },
+    {
+      "epoch": 0.44154057771664373,
+      "grad_norm": 0.5048724845350064,
+      "learning_rate": 0.0006180763011510911,
+      "loss": 2.3342,
+      "step": 963
+    },
+    {
+      "epoch": 0.4419990829894544,
+      "grad_norm": 0.4846749208763762,
+      "learning_rate": 0.000617354484208684,
+      "loss": 2.2684,
+      "step": 964
+    },
+    {
+      "epoch": 0.442457588262265,
+      "grad_norm": 0.46203836997333364,
+      "learning_rate": 0.0006166324083385189,
+      "loss": 2.2906,
+      "step": 965
+    },
+    {
+      "epoch": 0.44291609353507566,
+      "grad_norm": 0.4642997915400839,
+      "learning_rate": 0.0006159100751337642,
+      "loss": 2.3409,
+      "step": 966
+    },
+    {
+      "epoch": 0.44337459880788627,
+      "grad_norm": 0.43495375976220013,
+      "learning_rate": 0.0006151874861881565,
+      "loss": 2.3447,
+      "step": 967
+    },
+    {
+      "epoch": 0.4438331040806969,
+      "grad_norm": 0.48198001429300724,
+      "learning_rate": 0.0006144646430959964,
+      "loss": 2.2948,
+      "step": 968
+    },
+    {
+      "epoch": 0.4442916093535076,
+      "grad_norm": 0.45919446233342803,
+      "learning_rate": 0.0006137415474521454,
+      "loss": 2.2758,
+      "step": 969
+    },
+    {
+      "epoch": 0.4447501146263182,
+      "grad_norm": 0.48009814287777647,
+      "learning_rate": 0.0006130182008520222,
+      "loss": 2.3153,
+      "step": 970
+    },
+    {
+      "epoch": 0.44520861989912885,
+      "grad_norm": 0.45227878681384015,
+      "learning_rate": 0.000612294604891599,
+      "loss": 2.2919,
+      "step": 971
+    },
+    {
+      "epoch": 0.44566712517193946,
+      "grad_norm": 0.48016336353026323,
+      "learning_rate": 0.0006115707611673986,
+      "loss": 2.3289,
+      "step": 972
+    },
+    {
+      "epoch": 0.4461256304447501,
+      "grad_norm": 0.49030014063346844,
+      "learning_rate": 0.0006108466712764902,
+      "loss": 2.4351,
+      "step": 973
+    },
+    {
+      "epoch": 0.4465841357175607,
+      "grad_norm": 0.4760132899787433,
+      "learning_rate": 0.0006101223368164858,
+      "loss": 2.3335,
+      "step": 974
+    },
+    {
+      "epoch": 0.4470426409903714,
+      "grad_norm": 0.4675874312888077,
+      "learning_rate": 0.0006093977593855375,
+      "loss": 2.3935,
+      "step": 975
+    },
+    {
+      "epoch": 0.44750114626318205,
+      "grad_norm": 0.47215885328115414,
+      "learning_rate": 0.0006086729405823335,
+      "loss": 2.2959,
+      "step": 976
+    },
+    {
+      "epoch": 0.44795965153599265,
+      "grad_norm": 0.4766015956233957,
+      "learning_rate": 0.0006079478820060943,
+      "loss": 2.3029,
+      "step": 977
+    },
+    {
+      "epoch": 0.4484181568088033,
+      "grad_norm": 0.44502446949031027,
+      "learning_rate": 0.0006072225852565695,
+      "loss": 2.3438,
+      "step": 978
+    },
+    {
+      "epoch": 0.4488766620816139,
+      "grad_norm": 0.47657664709902287,
+      "learning_rate": 0.0006064970519340341,
+      "loss": 2.3012,
+      "step": 979
+    },
+    {
+      "epoch": 0.4493351673544246,
+      "grad_norm": 0.4498268333129423,
+      "learning_rate": 0.0006057712836392856,
+      "loss": 2.3072,
+      "step": 980
+    },
+    {
+      "epoch": 0.4497936726272352,
+      "grad_norm": 0.46486967629570913,
+      "learning_rate": 0.0006050452819736389,
+      "loss": 2.3287,
+      "step": 981
+    },
+    {
+      "epoch": 0.45025217790004585,
+      "grad_norm": 0.49029150046773906,
+      "learning_rate": 0.000604319048538925,
+      "loss": 2.3189,
+      "step": 982
+    },
+    {
+      "epoch": 0.4507106831728565,
+      "grad_norm": 0.46032078525390874,
+      "learning_rate": 0.0006035925849374855,
+      "loss": 2.2364,
+      "step": 983
+    },
+    {
+      "epoch": 0.4511691884456671,
+      "grad_norm": 0.4716644484563857,
+      "learning_rate": 0.0006028658927721697,
+      "loss": 2.3504,
+      "step": 984
+    },
+    {
+      "epoch": 0.4516276937184778,
+      "grad_norm": 0.47331403864026816,
+      "learning_rate": 0.0006021389736463321,
+      "loss": 2.3092,
+      "step": 985
+    },
+    {
+      "epoch": 0.4520861989912884,
+      "grad_norm": 0.5034313473001398,
+      "learning_rate": 0.0006014118291638271,
+      "loss": 2.2938,
+      "step": 986
+    },
+    {
+      "epoch": 0.45254470426409904,
+      "grad_norm": 0.45314174307214544,
+      "learning_rate": 0.0006006844609290065,
+      "loss": 2.3728,
+      "step": 987
+    },
+    {
+      "epoch": 0.4530032095369097,
+      "grad_norm": 0.4722107239935127,
+      "learning_rate": 0.0005999568705467161,
+      "loss": 2.3346,
+      "step": 988
+    },
+    {
+      "epoch": 0.4534617148097203,
+      "grad_norm": 0.42698378071685333,
+      "learning_rate": 0.0005992290596222915,
+      "loss": 2.2585,
+      "step": 989
+    },
+    {
+      "epoch": 0.45392022008253097,
+      "grad_norm": 0.4555669449274572,
+      "learning_rate": 0.0005985010297615551,
+      "loss": 2.2427,
+      "step": 990
+    },
+    {
+      "epoch": 0.4543787253553416,
+      "grad_norm": 0.4572173132372357,
+      "learning_rate": 0.0005977727825708123,
+      "loss": 2.31,
+      "step": 991
+    },
+    {
+      "epoch": 0.45483723062815223,
+      "grad_norm": 0.4490337432711153,
+      "learning_rate": 0.0005970443196568478,
+      "loss": 2.3417,
+      "step": 992
+    },
+    {
+      "epoch": 0.45529573590096284,
+      "grad_norm": 0.5007637277783273,
+      "learning_rate": 0.0005963156426269227,
+      "loss": 2.3818,
+      "step": 993
+    },
+    {
+      "epoch": 0.4557542411737735,
+      "grad_norm": 0.4788548038679883,
+      "learning_rate": 0.0005955867530887702,
+      "loss": 2.2479,
+      "step": 994
+    },
+    {
+      "epoch": 0.45621274644658416,
+      "grad_norm": 0.5839914296355632,
+      "learning_rate": 0.0005948576526505923,
+      "loss": 2.3383,
+      "step": 995
+    },
+    {
+      "epoch": 0.45667125171939477,
+      "grad_norm": 0.48097170066730827,
+      "learning_rate": 0.0005941283429210568,
+      "loss": 2.2403,
+      "step": 996
+    },
+    {
+      "epoch": 0.4571297569922054,
+      "grad_norm": 0.4927783579038282,
+      "learning_rate": 0.0005933988255092926,
+      "loss": 2.3507,
+      "step": 997
+    },
+    {
+      "epoch": 0.45758826226501603,
+      "grad_norm": 0.4626115068585194,
+      "learning_rate": 0.0005926691020248874,
+      "loss": 2.3266,
+      "step": 998
+    },
+    {
+      "epoch": 0.4580467675378267,
+      "grad_norm": 0.4999449474231508,
+      "learning_rate": 0.0005919391740778833,
+      "loss": 2.3026,
+      "step": 999
+    },
+    {
+      "epoch": 0.4585052728106373,
+      "grad_norm": 0.4968538434760327,
+      "learning_rate": 0.0005912090432787736,
+      "loss": 2.3307,
+      "step": 1000
+    },
+    {
+      "epoch": 0.45896377808344796,
+      "grad_norm": 0.47110398919569285,
+      "learning_rate": 0.000590478711238499,
+      "loss": 2.3052,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4594222833562586,
+      "grad_norm": 0.4588417966795537,
+      "learning_rate": 0.0005897481795684446,
+      "loss": 2.2711,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4598807886290692,
+      "grad_norm": 0.5426352674485234,
+      "learning_rate": 0.0005890174498804355,
+      "loss": 2.3582,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4603392939018799,
+      "grad_norm": 0.47638488182314126,
+      "learning_rate": 0.0005882865237867339,
+      "loss": 2.2862,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4607977991746905,
+      "grad_norm": 0.4516194816265361,
+      "learning_rate": 0.0005875554029000353,
+      "loss": 2.3185,
+      "step": 1005
+    },
+    {
+      "epoch": 0.46125630444750115,
+      "grad_norm": 0.44290601911751915,
+      "learning_rate": 0.0005868240888334653,
+      "loss": 2.2731,
+      "step": 1006
+    },
+    {
+      "epoch": 0.46171480972031176,
+      "grad_norm": 0.4663205237635063,
+      "learning_rate": 0.0005860925832005753,
+      "loss": 2.3265,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4621733149931224,
+      "grad_norm": 0.4823643363151113,
+      "learning_rate": 0.0005853608876153395,
+      "loss": 2.3367,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4626318202659331,
+      "grad_norm": 0.4584588863735623,
+      "learning_rate": 0.0005846290036921512,
+      "loss": 2.2465,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4630903255387437,
+      "grad_norm": 0.4539599811918593,
+      "learning_rate": 0.0005838969330458195,
+      "loss": 2.2997,
+      "step": 1010
+    },
+    {
+      "epoch": 0.46354883081155435,
+      "grad_norm": 0.5350953915024513,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 2.2773,
+      "step": 1011
+    },
+    {
+      "epoch": 0.46400733608436495,
+      "grad_norm": 0.503290351922234,
+      "learning_rate": 0.0005824322380450173,
+      "loss": 2.3221,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4644658413571756,
+      "grad_norm": 0.4901413423195104,
+      "learning_rate": 0.0005816996169222102,
+      "loss": 2.323,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4649243466299862,
+      "grad_norm": 0.47517089656572487,
+      "learning_rate": 0.0005809668155395793,
+      "loss": 2.3456,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4653828519027969,
+      "grad_norm": 0.4975689723370862,
+      "learning_rate": 0.0005802338355139578,
+      "loss": 2.2526,
+      "step": 1015
+    },
+    {
+      "epoch": 0.46584135717560754,
+      "grad_norm": 0.4831223636327179,
+      "learning_rate": 0.0005795006784625728,
+      "loss": 2.3521,
+      "step": 1016
+    },
+    {
+      "epoch": 0.46629986244841815,
+      "grad_norm": 0.5151720534314618,
+      "learning_rate": 0.0005787673460030423,
+      "loss": 2.3789,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4667583677212288,
+      "grad_norm": 0.4635345691948349,
+      "learning_rate": 0.000578033839753371,
+      "loss": 2.3306,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4672168729940394,
+      "grad_norm": 0.5079579071793286,
+      "learning_rate": 0.0005773001613319476,
+      "loss": 2.2325,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4676753782668501,
+      "grad_norm": 0.4725130202577579,
+      "learning_rate": 0.00057656631235754,
+      "loss": 2.3773,
+      "step": 1020
+    },
+    {
+      "epoch": 0.46813388353966073,
+      "grad_norm": 0.47392571845448217,
+      "learning_rate": 0.0005758322944492929,
+      "loss": 2.316,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46859238881247134,
+      "grad_norm": 0.4700602910496621,
+      "learning_rate": 0.0005750981092267237,
+      "loss": 2.3171,
+      "step": 1022
+    },
+    {
+      "epoch": 0.469050894085282,
+      "grad_norm": 0.6211755041981487,
+      "learning_rate": 0.0005743637583097183,
+      "loss": 2.2164,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4695093993580926,
+      "grad_norm": 0.504059436527753,
+      "learning_rate": 0.0005736292433185291,
+      "loss": 2.229,
+      "step": 1024
+    },
+    {
+      "epoch": 0.46996790463090327,
+      "grad_norm": 0.5072019268315424,
+      "learning_rate": 0.0005728945658737699,
+      "loss": 2.365,
+      "step": 1025
+    },
+    {
+      "epoch": 0.47042640990371387,
+      "grad_norm": 0.5031376843749016,
+      "learning_rate": 0.0005721597275964133,
+      "loss": 2.2943,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47088491517652453,
+      "grad_norm": 0.4864317025339457,
+      "learning_rate": 0.0005714247301077865,
+      "loss": 2.3171,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4713434204493352,
+      "grad_norm": 0.4843302504668407,
+      "learning_rate": 0.0005706895750295682,
+      "loss": 2.344,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4718019257221458,
+      "grad_norm": 0.4785454891578169,
+      "learning_rate": 0.0005699542639837844,
+      "loss": 2.3006,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47226043099495646,
+      "grad_norm": 0.44953709587595225,
+      "learning_rate": 0.0005692187985928055,
+      "loss": 2.3853,
+      "step": 1030
+    },
+    {
+      "epoch": 0.47271893626776706,
+      "grad_norm": 0.5166133465080921,
+      "learning_rate": 0.0005684831804793427,
+      "loss": 2.5453,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4731774415405777,
+      "grad_norm": 0.5401668475005018,
+      "learning_rate": 0.0005677474112664438,
+      "loss": 2.2639,
+      "step": 1032
+    },
+    {
+      "epoch": 0.47363594681338833,
+      "grad_norm": 0.5054412779240842,
+      "learning_rate": 0.0005670114925774899,
+      "loss": 2.3611,
+      "step": 1033
+    },
+    {
+      "epoch": 0.474094452086199,
+      "grad_norm": 0.4477254434449625,
+      "learning_rate": 0.0005662754260361924,
+      "loss": 2.2835,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47455295735900965,
+      "grad_norm": 0.49557885723358797,
+      "learning_rate": 0.0005655392132665884,
+      "loss": 2.3149,
+      "step": 1035
+    },
+    {
+      "epoch": 0.47501146263182026,
+      "grad_norm": 0.499986328463355,
+      "learning_rate": 0.000564802855893038,
+      "loss": 2.2859,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4754699679046309,
+      "grad_norm": 0.4990473305754683,
+      "learning_rate": 0.0005640663555402198,
+      "loss": 2.3857,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4759284731774415,
+      "grad_norm": 0.484290941985585,
+      "learning_rate": 0.0005633297138331285,
+      "loss": 2.3707,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4763869784502522,
+      "grad_norm": 0.5182792855842383,
+      "learning_rate": 0.0005625929323970705,
+      "loss": 2.51,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4768454837230628,
+      "grad_norm": 0.47936103448891715,
+      "learning_rate": 0.0005618560128576603,
+      "loss": 2.2998,
+      "step": 1040
+    },
+    {
+      "epoch": 0.47730398899587345,
+      "grad_norm": 0.5021429542507719,
+      "learning_rate": 0.0005611189568408173,
+      "loss": 2.3285,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4777624942686841,
+      "grad_norm": 0.5069534366768079,
+      "learning_rate": 0.0005603817659727619,
+      "loss": 2.2564,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4782209995414947,
+      "grad_norm": 0.45677293172351224,
+      "learning_rate": 0.0005596444418800121,
+      "loss": 2.2716,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4786795048143054,
+      "grad_norm": 0.5001989188728971,
+      "learning_rate": 0.0005589069861893798,
+      "loss": 2.2774,
+      "step": 1044
+    },
+    {
+      "epoch": 0.479138010087116,
+      "grad_norm": 0.4586752856475308,
+      "learning_rate": 0.0005581694005279673,
+      "loss": 2.2897,
+      "step": 1045
+    },
+    {
+      "epoch": 0.47959651535992665,
+      "grad_norm": 0.46291290264301566,
+      "learning_rate": 0.0005574316865231637,
+      "loss": 2.2419,
+      "step": 1046
+    },
+    {
+      "epoch": 0.48005502063273725,
+      "grad_norm": 0.49421940899213135,
+      "learning_rate": 0.0005566938458026411,
+      "loss": 2.3222,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4805135259055479,
+      "grad_norm": 0.46879476256525243,
+      "learning_rate": 0.0005559558799943514,
+      "loss": 2.2658,
+      "step": 1048
+    },
+    {
+      "epoch": 0.48097203117835857,
+      "grad_norm": 0.5719447951494459,
+      "learning_rate": 0.0005552177907265223,
+      "loss": 2.1607,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4814305364511692,
+      "grad_norm": 0.5006487658853649,
+      "learning_rate": 0.000554479579627654,
+      "loss": 2.2984,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48188904172397984,
+      "grad_norm": 0.51442128462548,
+      "learning_rate": 0.0005537412483265157,
+      "loss": 2.3481,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48234754699679044,
+      "grad_norm": 0.4676104319566605,
+      "learning_rate": 0.0005530027984521413,
+      "loss": 2.3418,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4828060522696011,
+      "grad_norm": 0.47541486159471047,
+      "learning_rate": 0.0005522642316338268,
+      "loss": 2.2408,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4832645575424117,
+      "grad_norm": 0.4707367293196365,
+      "learning_rate": 0.0005515255495011259,
+      "loss": 2.2664,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48372306281522237,
+      "grad_norm": 0.4581220565430314,
+      "learning_rate": 0.0005507867536838472,
+      "loss": 2.2975,
+      "step": 1055
+    },
+    {
+      "epoch": 0.48418156808803303,
+      "grad_norm": 0.4956980301716918,
+      "learning_rate": 0.0005500478458120492,
+      "loss": 2.2593,
+      "step": 1056
+    },
+    {
+      "epoch": 0.48464007336084364,
+      "grad_norm": 0.4656752896933445,
+      "learning_rate": 0.0005493088275160387,
+      "loss": 2.2895,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4850985786336543,
+      "grad_norm": 0.4784940174634947,
+      "learning_rate": 0.0005485697004263657,
+      "loss": 2.314,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4855570839064649,
+      "grad_norm": 0.45343980626194924,
+      "learning_rate": 0.0005478304661738199,
+      "loss": 2.3428,
+      "step": 1059
+    },
+    {
+      "epoch": 0.48601558917927556,
+      "grad_norm": 0.5290233454707374,
+      "learning_rate": 0.0005470911263894279,
+      "loss": 2.2349,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4864740944520862,
+      "grad_norm": 0.48617079798760576,
+      "learning_rate": 0.0005463516827044491,
+      "loss": 2.3361,
+      "step": 1061
+    },
+    {
+      "epoch": 0.48693259972489683,
+      "grad_norm": 0.47277509164729614,
+      "learning_rate": 0.000545612136750372,
+      "loss": 2.2729,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4873911049977075,
+      "grad_norm": 0.481649015612465,
+      "learning_rate": 0.0005448724901589107,
+      "loss": 2.4311,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4878496102705181,
+      "grad_norm": 0.46252456306622247,
+      "learning_rate": 0.0005441327445620014,
+      "loss": 2.291,
+      "step": 1064
+    },
+    {
+      "epoch": 0.48830811554332876,
+      "grad_norm": 0.5216151385628517,
+      "learning_rate": 0.0005433929015917988,
+      "loss": 2.2526,
+      "step": 1065
+    },
+    {
+      "epoch": 0.48876662081613936,
+      "grad_norm": 0.46347916649477944,
+      "learning_rate": 0.0005426529628806724,
+      "loss": 2.2298,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48922512608895,
+      "grad_norm": 0.508834246763414,
+      "learning_rate": 0.0005419129300612029,
+      "loss": 2.3075,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4896836313617607,
+      "grad_norm": 0.4821042345258515,
+      "learning_rate": 0.000541172804766179,
+      "loss": 2.4074,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4901421366345713,
+      "grad_norm": 0.4571782660458305,
+      "learning_rate": 0.0005404325886285927,
+      "loss": 2.2367,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49060064190738195,
+      "grad_norm": 0.5396808646538432,
+      "learning_rate": 0.000539692283281637,
+      "loss": 2.3261,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49105914718019256,
+      "grad_norm": 0.5043741697077228,
+      "learning_rate": 0.0005389518903587017,
+      "loss": 2.3335,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4915176524530032,
+      "grad_norm": 0.4620883147480308,
+      "learning_rate": 0.0005382114114933695,
+      "loss": 2.2572,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4919761577258138,
+      "grad_norm": 0.46790557863428006,
+      "learning_rate": 0.0005374708483194132,
+      "loss": 2.3255,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4924346629986245,
+      "grad_norm": 0.49622865779923275,
+      "learning_rate": 0.000536730202470791,
+      "loss": 2.2734,
+      "step": 1074
+    },
+    {
+      "epoch": 0.49289316827143514,
+      "grad_norm": 0.5655220852681334,
+      "learning_rate": 0.0005359894755816443,
+      "loss": 2.294,
+      "step": 1075
+    },
+    {
+      "epoch": 0.49335167354424575,
+      "grad_norm": 0.44769365015462687,
+      "learning_rate": 0.0005352486692862926,
+      "loss": 2.3339,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4938101788170564,
+      "grad_norm": 0.460320765401225,
+      "learning_rate": 0.0005345077852192307,
+      "loss": 2.3772,
+      "step": 1077
+    },
+    {
+      "epoch": 0.494268684089867,
+      "grad_norm": 0.7593846625505101,
+      "learning_rate": 0.0005337668250151254,
+      "loss": 2.276,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4947271893626777,
+      "grad_norm": 0.47506782664995806,
+      "learning_rate": 0.0005330257903088111,
+      "loss": 2.2926,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4951856946354883,
+      "grad_norm": 0.5648110733103908,
+      "learning_rate": 0.000532284682735287,
+      "loss": 2.2475,
+      "step": 1080
+    },
+    {
+      "epoch": 0.49564419990829894,
+      "grad_norm": 0.4852364396851478,
+      "learning_rate": 0.0005315435039297124,
+      "loss": 2.2907,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4961027051811096,
+      "grad_norm": 0.4804307610769782,
+      "learning_rate": 0.0005308022555274046,
+      "loss": 2.2541,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4965612104539202,
+      "grad_norm": 0.5055534358642296,
+      "learning_rate": 0.0005300609391638336,
+      "loss": 2.2705,
+      "step": 1083
+    },
+    {
+      "epoch": 0.49701971572673087,
+      "grad_norm": 0.49346026269113974,
+      "learning_rate": 0.0005293195564746201,
+      "loss": 2.2726,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4974782209995415,
+      "grad_norm": 0.48194020154687983,
+      "learning_rate": 0.0005285781090955304,
+      "loss": 2.2316,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49793672627235214,
+      "grad_norm": 0.5266666129109214,
+      "learning_rate": 0.0005278365986624743,
+      "loss": 2.2621,
+      "step": 1086
+    },
+    {
+      "epoch": 0.49839523154516274,
+      "grad_norm": 0.47321976613684635,
+      "learning_rate": 0.0005270950268115001,
+      "loss": 2.2383,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4988537368179734,
+      "grad_norm": 0.448687382965505,
+      "learning_rate": 0.0005263533951787919,
+      "loss": 2.3198,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49931224209078406,
+      "grad_norm": 0.49257649719966096,
+      "learning_rate": 0.000525611705400666,
+      "loss": 2.3425,
+      "step": 1089
+    },
+    {
+      "epoch": 0.49977074736359467,
+      "grad_norm": 0.444594945323832,
+      "learning_rate": 0.0005248699591135664,
+      "loss": 2.3216,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5002292526364053,
+      "grad_norm": 0.4868624341645274,
+      "learning_rate": 0.0005241281579540618,
+      "loss": 2.2743,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5006877579092159,
+      "grad_norm": 0.47156523960795316,
+      "learning_rate": 0.0005233863035588427,
+      "loss": 2.3672,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5011462631820266,
+      "grad_norm": 0.4793811398067745,
+      "learning_rate": 0.0005226443975647161,
+      "loss": 2.3589,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5016047684548373,
+      "grad_norm": 0.4928844726346011,
+      "learning_rate": 0.0005219024416086036,
+      "loss": 2.2745,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5020632737276479,
+      "grad_norm": 0.47737072552446524,
+      "learning_rate": 0.0005211604373275366,
+      "loss": 2.3416,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5025217790004585,
+      "grad_norm": 0.5170980750915561,
+      "learning_rate": 0.0005204183863586533,
+      "loss": 2.2296,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5029802842732691,
+      "grad_norm": 0.5253155381672523,
+      "learning_rate": 0.0005196762903391951,
+      "loss": 2.3047,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5034387895460798,
+      "grad_norm": 0.4744911002323076,
+      "learning_rate": 0.0005189341509065023,
+      "loss": 2.2952,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5038972948188905,
+      "grad_norm": 0.47375073378273813,
+      "learning_rate": 0.0005181919696980112,
+      "loss": 2.3541,
+      "step": 1099
+    },
+    {
+      "epoch": 0.504355800091701,
+      "grad_norm": 0.4484503457228679,
+      "learning_rate": 0.0005174497483512506,
+      "loss": 2.3387,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5048143053645117,
+      "grad_norm": 0.5021340898843981,
+      "learning_rate": 0.0005167074885038374,
+      "loss": 2.2192,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5052728106373223,
+      "grad_norm": 0.4792320877465066,
+      "learning_rate": 0.0005159651917934735,
+      "loss": 2.2631,
+      "step": 1102
+    },
+    {
+      "epoch": 0.505731315910133,
+      "grad_norm": 0.5007509751697835,
+      "learning_rate": 0.0005152228598579428,
+      "loss": 2.3968,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5061898211829436,
+      "grad_norm": 0.4710449875754153,
+      "learning_rate": 0.000514480494335106,
+      "loss": 2.2833,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5066483264557542,
+      "grad_norm": 0.45370965987020284,
+      "learning_rate": 0.0005137380968628983,
+      "loss": 2.312,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5071068317285649,
+      "grad_norm": 0.46783451296811757,
+      "learning_rate": 0.0005129956690793255,
+      "loss": 2.3063,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5075653370013755,
+      "grad_norm": 0.4743889416450896,
+      "learning_rate": 0.0005122532126224601,
+      "loss": 2.3611,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5080238422741862,
+      "grad_norm": 0.4670874538406349,
+      "learning_rate": 0.0005115107291304378,
+      "loss": 2.309,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5084823475469968,
+      "grad_norm": 0.476505452986287,
+      "learning_rate": 0.0005107682202414544,
+      "loss": 2.3005,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5089408528198074,
+      "grad_norm": 0.46721642000116864,
+      "learning_rate": 0.0005100256875937613,
+      "loss": 2.2525,
+      "step": 1110
+    },
+    {
+      "epoch": 0.509399358092618,
+      "grad_norm": 0.5162795050882883,
+      "learning_rate": 0.0005092831328256625,
+      "loss": 2.3561,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5098578633654287,
+      "grad_norm": 0.4781076034640987,
+      "learning_rate": 0.0005085405575755105,
+      "loss": 2.345,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5103163686382394,
+      "grad_norm": 0.45977215489313794,
+      "learning_rate": 0.0005077979634817034,
+      "loss": 2.2238,
+      "step": 1113
+    },
+    {
+      "epoch": 0.51077487391105,
+      "grad_norm": 0.4953342719888701,
+      "learning_rate": 0.0005070553521826808,
+      "loss": 2.2558,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5112333791838606,
+      "grad_norm": 0.47380752253830977,
+      "learning_rate": 0.00050631272531692,
+      "loss": 2.3426,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5116918844566712,
+      "grad_norm": 0.5917450237692711,
+      "learning_rate": 0.0005055700845229327,
+      "loss": 2.3293,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5121503897294819,
+      "grad_norm": 0.492189379434933,
+      "learning_rate": 0.000504827431439262,
+      "loss": 2.2849,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5126088950022926,
+      "grad_norm": 0.47239165902538754,
+      "learning_rate": 0.000504084767704477,
+      "loss": 2.3262,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5130674002751031,
+      "grad_norm": 0.506670576427528,
+      "learning_rate": 0.0005033420949571712,
+      "loss": 2.2042,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5135259055479138,
+      "grad_norm": 0.46657347158959656,
+      "learning_rate": 0.0005025994148359574,
+      "loss": 2.3255,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5139844108207244,
+      "grad_norm": 0.5231608457693565,
+      "learning_rate": 0.0005018567289794651,
+      "loss": 2.1685,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5144429160935351,
+      "grad_norm": 0.4800691205015213,
+      "learning_rate": 0.0005011140390263362,
+      "loss": 2.3809,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5149014213663458,
+      "grad_norm": 0.47800977205082296,
+      "learning_rate": 0.0005003713466152218,
+      "loss": 2.3094,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5153599266391563,
+      "grad_norm": 0.44560440930916273,
+      "learning_rate": 0.0004996286533847783,
+      "loss": 2.3031,
+      "step": 1124
+    },
+    {
+      "epoch": 0.515818431911967,
+      "grad_norm": 0.4675088130955647,
+      "learning_rate": 0.000498885960973664,
+      "loss": 2.3043,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5162769371847776,
+      "grad_norm": 0.5027796512243792,
+      "learning_rate": 0.000498143271020535,
+      "loss": 2.2817,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5167354424575883,
+      "grad_norm": 0.456596682316068,
+      "learning_rate": 0.0004974005851640428,
+      "loss": 2.3096,
+      "step": 1127
+    },
+    {
+      "epoch": 0.517193947730399,
+      "grad_norm": 0.473293224242195,
+      "learning_rate": 0.000496657905042829,
+      "loss": 2.3084,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5176524530032095,
+      "grad_norm": 0.7572552903545632,
+      "learning_rate": 0.0004959152322955232,
+      "loss": 2.2919,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5181109582760202,
+      "grad_norm": 0.4671212998265469,
+      "learning_rate": 0.0004951725685607382,
+      "loss": 2.4188,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5185694635488308,
+      "grad_norm": 0.48495757767319647,
+      "learning_rate": 0.0004944299154770673,
+      "loss": 2.2322,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5190279688216415,
+      "grad_norm": 0.4778586619460251,
+      "learning_rate": 0.0004936872746830802,
+      "loss": 2.2938,
+      "step": 1132
+    },
+    {
+      "epoch": 0.519486474094452,
+      "grad_norm": 0.4614275896955302,
+      "learning_rate": 0.0004929446478173195,
+      "loss": 2.3016,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5199449793672627,
+      "grad_norm": 0.4939787358250652,
+      "learning_rate": 0.0004922020365182968,
+      "loss": 2.2407,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5204034846400734,
+      "grad_norm": 0.46131045651254576,
+      "learning_rate": 0.0004914594424244897,
+      "loss": 2.2691,
+      "step": 1135
+    },
+    {
+      "epoch": 0.520861989912884,
+      "grad_norm": 0.47742176760667787,
+      "learning_rate": 0.0004907168671743376,
+      "loss": 2.259,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5213204951856947,
+      "grad_norm": 0.49123180685873113,
+      "learning_rate": 0.0004899743124062387,
+      "loss": 2.1935,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5217790004585052,
+      "grad_norm": 0.4882970435449315,
+      "learning_rate": 0.0004892317797585456,
+      "loss": 2.2259,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5222375057313159,
+      "grad_norm": 0.47905423082773696,
+      "learning_rate": 0.0004884892708695623,
+      "loss": 2.1873,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5226960110041265,
+      "grad_norm": 0.49563670459275067,
+      "learning_rate": 0.0004877467873775402,
+      "loss": 2.2882,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5231545162769372,
+      "grad_norm": 0.5417077398498199,
+      "learning_rate": 0.00048700433092067473,
+      "loss": 2.328,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5236130215497479,
+      "grad_norm": 0.45395176593339054,
+      "learning_rate": 0.0004862619031371019,
+      "loss": 2.2782,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5240715268225584,
+      "grad_norm": 0.5004205048422604,
+      "learning_rate": 0.0004855195056648942,
+      "loss": 2.2551,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5245300320953691,
+      "grad_norm": 0.4826036261214723,
+      "learning_rate": 0.00048477714014205734,
+      "loss": 2.2741,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5249885373681797,
+      "grad_norm": 0.4679995878102019,
+      "learning_rate": 0.00048403480820652644,
+      "loss": 2.3661,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5254470426409904,
+      "grad_norm": 0.4468036520096377,
+      "learning_rate": 0.0004832925114961629,
+      "loss": 2.2967,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5259055479138011,
+      "grad_norm": 0.46128321014340196,
+      "learning_rate": 0.0004825502516487497,
+      "loss": 2.3257,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5263640531866116,
+      "grad_norm": 0.4609425582850656,
+      "learning_rate": 0.00048180803030198896,
+      "loss": 2.3869,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5268225584594223,
+      "grad_norm": 0.4697287092536518,
+      "learning_rate": 0.0004810658490934979,
+      "loss": 2.3475,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5272810637322329,
+      "grad_norm": 0.44860469621960164,
+      "learning_rate": 0.000480323709660805,
+      "loss": 2.359,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5277395690050436,
+      "grad_norm": 0.4963930597056019,
+      "learning_rate": 0.0004795816136413467,
+      "loss": 2.2343,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5281980742778541,
+      "grad_norm": 0.4456143578287099,
+      "learning_rate": 0.00047883956267246353,
+      "loss": 2.3323,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5286565795506648,
+      "grad_norm": 0.4472882157611263,
+      "learning_rate": 0.00047809755839139657,
+      "loss": 2.3438,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5291150848234755,
+      "grad_norm": 0.4596158400216166,
+      "learning_rate": 0.0004773556024352841,
+      "loss": 2.3358,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5295735900962861,
+      "grad_norm": 0.8317860899197641,
+      "learning_rate": 0.00047661369644115754,
+      "loss": 2.1786,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5300320953690968,
+      "grad_norm": 0.48786710604375705,
+      "learning_rate": 0.0004758718420459383,
+      "loss": 2.224,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5304906006419073,
+      "grad_norm": 0.45204436817222715,
+      "learning_rate": 0.0004751300408864339,
+      "loss": 2.2152,
+      "step": 1157
+    },
+    {
+      "epoch": 0.530949105914718,
+      "grad_norm": 0.46899604052308247,
+      "learning_rate": 0.00047438829459933414,
+      "loss": 2.2999,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5314076111875287,
+      "grad_norm": 0.4880321859289823,
+      "learning_rate": 0.0004736466048212082,
+      "loss": 2.3392,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5318661164603393,
+      "grad_norm": 0.4842066541985399,
+      "learning_rate": 0.0004729049731885002,
+      "loss": 2.2793,
+      "step": 1160
+    },
+    {
+      "epoch": 0.53232462173315,
+      "grad_norm": 0.4532097313847725,
+      "learning_rate": 0.000472163401337526,
+      "loss": 2.2546,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5327831270059605,
+      "grad_norm": 0.4455969636777601,
+      "learning_rate": 0.00047142189090446985,
+      "loss": 2.2304,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5332416322787712,
+      "grad_norm": 0.43860106702426876,
+      "learning_rate": 0.0004706804435253802,
+      "loss": 2.3806,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5337001375515819,
+      "grad_norm": 0.452343364721177,
+      "learning_rate": 0.0004699390608361665,
+      "loss": 2.3987,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5341586428243925,
+      "grad_norm": 0.4814803882534715,
+      "learning_rate": 0.0004691977444725955,
+      "loss": 2.2813,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5346171480972031,
+      "grad_norm": 0.5210560538012846,
+      "learning_rate": 0.0004684564960702877,
+      "loss": 2.2958,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5350756533700137,
+      "grad_norm": 0.4746722343599108,
+      "learning_rate": 0.0004677153172647131,
+      "loss": 2.2962,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5355341586428244,
+      "grad_norm": 0.4721345288053883,
+      "learning_rate": 0.00046697420969118894,
+      "loss": 2.3531,
+      "step": 1168
+    },
+    {
+      "epoch": 0.535992663915635,
+      "grad_norm": 0.4816470758326327,
+      "learning_rate": 0.00046623317498487466,
+      "loss": 2.2548,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5364511691884457,
+      "grad_norm": 0.4813007484441071,
+      "learning_rate": 0.0004654922147807694,
+      "loss": 2.381,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5369096744612563,
+      "grad_norm": 0.457281421187799,
+      "learning_rate": 0.00046475133071370757,
+      "loss": 2.2383,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5373681797340669,
+      "grad_norm": 0.4961059530796062,
+      "learning_rate": 0.00046401052441835574,
+      "loss": 2.2988,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5378266850068776,
+      "grad_norm": 0.4642022390990009,
+      "learning_rate": 0.000463269797529209,
+      "loss": 2.1822,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5382851902796882,
+      "grad_norm": 0.5194100952588628,
+      "learning_rate": 0.00046252915168058697,
+      "loss": 2.3257,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5387436955524989,
+      "grad_norm": 0.47779969697141556,
+      "learning_rate": 0.0004617885885066305,
+      "loss": 2.2725,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5392022008253095,
+      "grad_norm": 0.47822718709684875,
+      "learning_rate": 0.0004610481096412984,
+      "loss": 2.296,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5396607060981201,
+      "grad_norm": 0.5036046214447458,
+      "learning_rate": 0.000460307716718363,
+      "loss": 2.2865,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5401192113709308,
+      "grad_norm": 0.4748132526813028,
+      "learning_rate": 0.0004595674113714074,
+      "loss": 2.2725,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5405777166437414,
+      "grad_norm": 0.4724072373555433,
+      "learning_rate": 0.0004588271952338212,
+      "loss": 2.3067,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5410362219165521,
+      "grad_norm": 0.5994840156575252,
+      "learning_rate": 0.00045808706993879714,
+      "loss": 2.3705,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5414947271893626,
+      "grad_norm": 0.44910640814888547,
+      "learning_rate": 0.00045734703711932767,
+      "loss": 2.2902,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5419532324621733,
+      "grad_norm": 0.4770164232122006,
+      "learning_rate": 0.0004566070984082013,
+      "loss": 2.1914,
+      "step": 1182
+    },
+    {
+      "epoch": 0.542411737734984,
+      "grad_norm": 0.47950346102154734,
+      "learning_rate": 0.00045586725543799865,
+      "loss": 2.213,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5428702430077946,
+      "grad_norm": 0.49877130237806677,
+      "learning_rate": 0.00045512750984108937,
+      "loss": 2.3503,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5433287482806052,
+      "grad_norm": 0.48961801889028833,
+      "learning_rate": 0.000454387863249628,
+      "loss": 2.363,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5437872535534158,
+      "grad_norm": 0.5037655580839661,
+      "learning_rate": 0.00045364831729555096,
+      "loss": 2.3015,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5442457588262265,
+      "grad_norm": 0.5658926604324885,
+      "learning_rate": 0.0004529088736105721,
+      "loss": 2.3231,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5447042640990372,
+      "grad_norm": 0.4662761866707688,
+      "learning_rate": 0.0004521695338261802,
+      "loss": 2.3177,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5451627693718478,
+      "grad_norm": 0.5581753823401705,
+      "learning_rate": 0.0004514302995736344,
+      "loss": 2.323,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5456212746446584,
+      "grad_norm": 0.44701933313369846,
+      "learning_rate": 0.0004506911724839613,
+      "loss": 2.2934,
+      "step": 1190
+    },
+    {
+      "epoch": 0.546079779917469,
+      "grad_norm": 0.47190119742862174,
+      "learning_rate": 0.0004499521541879508,
+      "loss": 2.2163,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5465382851902797,
+      "grad_norm": 0.4478460594576136,
+      "learning_rate": 0.00044921324631615303,
+      "loss": 2.285,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5469967904630904,
+      "grad_norm": 0.4702100834006392,
+      "learning_rate": 0.0004484744504988742,
+      "loss": 2.2494,
+      "step": 1193
+    },
+    {
+      "epoch": 0.547455295735901,
+      "grad_norm": 0.49812120252312553,
+      "learning_rate": 0.00044773576836617336,
+      "loss": 2.2414,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5479138010087116,
+      "grad_norm": 0.4749471943442546,
+      "learning_rate": 0.0004469972015478588,
+      "loss": 2.286,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5483723062815222,
+      "grad_norm": 0.4412235549447814,
+      "learning_rate": 0.0004462587516734844,
+      "loss": 2.2453,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5488308115543329,
+      "grad_norm": 0.46587209156346415,
+      "learning_rate": 0.00044552042037234596,
+      "loss": 2.3289,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5492893168271435,
+      "grad_norm": 0.44574348451273405,
+      "learning_rate": 0.00044478220927347774,
+      "loss": 2.2776,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5497478220999541,
+      "grad_norm": 0.44363136091265204,
+      "learning_rate": 0.00044404412000564875,
+      "loss": 2.2303,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5502063273727648,
+      "grad_norm": 0.4465900373763079,
+      "learning_rate": 0.000443306154197359,
+      "loss": 2.2636,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5506648326455754,
+      "grad_norm": 0.4565065073559317,
+      "learning_rate": 0.00044256831347683646,
+      "loss": 2.2505,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5511233379183861,
+      "grad_norm": 0.5184610510372265,
+      "learning_rate": 0.0004418305994720328,
+      "loss": 2.2523,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5515818431911967,
+      "grad_norm": 0.44737133826295306,
+      "learning_rate": 0.0004410930138106203,
+      "loss": 2.2561,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5520403484640073,
+      "grad_norm": 0.4526634340879452,
+      "learning_rate": 0.000440355558119988,
+      "loss": 2.3563,
+      "step": 1204
+    },
+    {
+      "epoch": 0.552498853736818,
+      "grad_norm": 0.46304802404036344,
+      "learning_rate": 0.00043961823402723814,
+      "loss": 2.2647,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5529573590096286,
+      "grad_norm": 0.47022581706607264,
+      "learning_rate": 0.0004388810431591829,
+      "loss": 2.3424,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5534158642824393,
+      "grad_norm": 0.46255795253273835,
+      "learning_rate": 0.0004381439871423398,
+      "loss": 2.2743,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5538743695552499,
+      "grad_norm": 0.44246854027775573,
+      "learning_rate": 0.00043740706760292966,
+      "loss": 2.28,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5543328748280605,
+      "grad_norm": 0.4617871871111375,
+      "learning_rate": 0.0004366702861668716,
+      "loss": 2.3006,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5547913801008711,
+      "grad_norm": 0.44491936859931175,
+      "learning_rate": 0.00043593364445978036,
+      "loss": 2.3148,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5552498853736818,
+      "grad_norm": 0.4772659463697155,
+      "learning_rate": 0.0004351971441069622,
+      "loss": 2.2327,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5557083906464925,
+      "grad_norm": 0.4856912110000148,
+      "learning_rate": 0.0004344607867334116,
+      "loss": 2.2424,
+      "step": 1212
+    },
+    {
+      "epoch": 0.556166895919303,
+      "grad_norm": 0.45168632599046266,
+      "learning_rate": 0.00043372457396380766,
+      "loss": 2.1999,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5566254011921137,
+      "grad_norm": 0.4995996536721345,
+      "learning_rate": 0.00043298850742251013,
+      "loss": 2.4345,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5570839064649243,
+      "grad_norm": 0.44654106552435224,
+      "learning_rate": 0.0004322525887335563,
+      "loss": 2.2291,
+      "step": 1215
+    },
+    {
+      "epoch": 0.557542411737735,
+      "grad_norm": 0.5012341477550766,
+      "learning_rate": 0.00043151681952065734,
+      "loss": 2.353,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5580009170105457,
+      "grad_norm": 0.49441976164355383,
+      "learning_rate": 0.00043078120140719456,
+      "loss": 2.3831,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5584594222833562,
+      "grad_norm": 0.4705858805516597,
+      "learning_rate": 0.0004300457360162158,
+      "loss": 2.2716,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5589179275561669,
+      "grad_norm": 0.46699531354500057,
+      "learning_rate": 0.0004293104249704319,
+      "loss": 2.2316,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5593764328289775,
+      "grad_norm": 0.4503275411777988,
+      "learning_rate": 0.00042857526989221355,
+      "loss": 2.3544,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5598349381017882,
+      "grad_norm": 0.45116269914577006,
+      "learning_rate": 0.00042784027240358674,
+      "loss": 2.3172,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5602934433745989,
+      "grad_norm": 0.46289921025269964,
+      "learning_rate": 0.0004271054341262301,
+      "loss": 2.2379,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5607519486474094,
+      "grad_norm": 0.4181893200642839,
+      "learning_rate": 0.000426370756681471,
+      "loss": 2.2471,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5612104539202201,
+      "grad_norm": 0.45353625165006106,
+      "learning_rate": 0.0004256362416902817,
+      "loss": 2.29,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5616689591930307,
+      "grad_norm": 0.44793514010138036,
+      "learning_rate": 0.00042490189077327637,
+      "loss": 2.2358,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5621274644658414,
+      "grad_norm": 0.45174662419597356,
+      "learning_rate": 0.00042416770555070703,
+      "loss": 2.2889,
+      "step": 1226
+    },
+    {
+      "epoch": 0.562585969738652,
+      "grad_norm": 0.45722402323832895,
+      "learning_rate": 0.00042343368764246,
+      "loss": 2.1806,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5630444750114626,
+      "grad_norm": 0.46879350663979025,
+      "learning_rate": 0.0004226998386680524,
+      "loss": 2.3638,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5635029802842733,
+      "grad_norm": 0.49934612906042825,
+      "learning_rate": 0.000421966160246629,
+      "loss": 2.3903,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5639614855570839,
+      "grad_norm": 0.4869974348820216,
+      "learning_rate": 0.00042123265399695783,
+      "loss": 2.2997,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5644199908298946,
+      "grad_norm": 0.4722111905561648,
+      "learning_rate": 0.0004204993215374273,
+      "loss": 2.4123,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5648784961027051,
+      "grad_norm": 0.4528186147385469,
+      "learning_rate": 0.00041976616448604226,
+      "loss": 2.3239,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5653370013755158,
+      "grad_norm": 0.463710771088582,
+      "learning_rate": 0.00041903318446042076,
+      "loss": 2.2783,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5657955066483265,
+      "grad_norm": 0.4694074724898153,
+      "learning_rate": 0.00041830038307778984,
+      "loss": 2.3641,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5662540119211371,
+      "grad_norm": 0.4930448749247346,
+      "learning_rate": 0.0004175677619549828,
+      "loss": 2.4141,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5667125171939478,
+      "grad_norm": 0.45543786414667337,
+      "learning_rate": 0.000416835322708435,
+      "loss": 2.1975,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5671710224667583,
+      "grad_norm": 0.46336031211692497,
+      "learning_rate": 0.00041610306695418056,
+      "loss": 2.2491,
+      "step": 1237
+    },
+    {
+      "epoch": 0.567629527739569,
+      "grad_norm": 0.5055259580183835,
+      "learning_rate": 0.0004153709963078488,
+      "loss": 2.331,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5680880330123796,
+      "grad_norm": 0.4432605643014693,
+      "learning_rate": 0.0004146391123846606,
+      "loss": 2.2909,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5685465382851903,
+      "grad_norm": 0.46303808578642497,
+      "learning_rate": 0.0004139074167994249,
+      "loss": 2.2833,
+      "step": 1240
+    },
+    {
+      "epoch": 0.569005043558001,
+      "grad_norm": 0.44440703609196647,
+      "learning_rate": 0.00041317591116653486,
+      "loss": 2.3033,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5694635488308115,
+      "grad_norm": 0.43864767223520157,
+      "learning_rate": 0.0004124445970999648,
+      "loss": 2.2964,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5699220541036222,
+      "grad_norm": 0.5033447397826637,
+      "learning_rate": 0.00041171347621326627,
+      "loss": 2.2908,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5703805593764328,
+      "grad_norm": 0.48327333801545824,
+      "learning_rate": 0.00041098255011956465,
+      "loss": 2.3362,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5708390646492435,
+      "grad_norm": 0.4684568611114492,
+      "learning_rate": 0.00041025182043155547,
+      "loss": 2.3514,
+      "step": 1245
+    },
+    {
+      "epoch": 0.571297569922054,
+      "grad_norm": 0.4863889452234541,
+      "learning_rate": 0.000409521288761501,
+      "loss": 2.3526,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5717560751948647,
+      "grad_norm": 0.46195897134562397,
+      "learning_rate": 0.00040879095672122646,
+      "loss": 2.2702,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5722145804676754,
+      "grad_norm": 0.6476131101437701,
+      "learning_rate": 0.0004080608259221167,
+      "loss": 2.2801,
+      "step": 1248
+    },
+    {
+      "epoch": 0.572673085740486,
+      "grad_norm": 0.4304262483730132,
+      "learning_rate": 0.0004073308979751126,
+      "loss": 2.2218,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5731315910132967,
+      "grad_norm": 0.4408040464541368,
+      "learning_rate": 0.0004066011744907074,
+      "loss": 2.1732,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5735900962861072,
+      "grad_norm": 0.46860639662282794,
+      "learning_rate": 0.00040587165707894326,
+      "loss": 2.2947,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5740486015589179,
+      "grad_norm": 0.43739398801189217,
+      "learning_rate": 0.0004051423473494076,
+      "loss": 2.2563,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5745071068317286,
+      "grad_norm": 0.45509333019783876,
+      "learning_rate": 0.0004044132469112299,
+      "loss": 2.2853,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5749656121045392,
+      "grad_norm": 0.5101285166654109,
+      "learning_rate": 0.00040368435737307733,
+      "loss": 2.3896,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5754241173773499,
+      "grad_norm": 0.4343777066320545,
+      "learning_rate": 0.00040295568034315224,
+      "loss": 2.2266,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5758826226501604,
+      "grad_norm": 0.44372952864835413,
+      "learning_rate": 0.0004022272174291878,
+      "loss": 2.3538,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5763411279229711,
+      "grad_norm": 0.497151118588466,
+      "learning_rate": 0.0004014989702384449,
+      "loss": 2.2439,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5767996331957818,
+      "grad_norm": 0.46309248782061896,
+      "learning_rate": 0.00040077094037770843,
+      "loss": 2.2671,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5772581384685924,
+      "grad_norm": 0.48187122888363476,
+      "learning_rate": 0.0004000431294532838,
+      "loss": 2.2591,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5777166437414031,
+      "grad_norm": 0.46769542630926314,
+      "learning_rate": 0.0003993155390709935,
+      "loss": 2.2727,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5781751490142136,
+      "grad_norm": 0.43850100516582596,
+      "learning_rate": 0.0003985881708361729,
+      "loss": 2.3242,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5786336542870243,
+      "grad_norm": 0.46935942070599584,
+      "learning_rate": 0.00039786102635366784,
+      "loss": 2.2424,
+      "step": 1262
+    },
+    {
+      "epoch": 0.579092159559835,
+      "grad_norm": 0.5011987321333168,
+      "learning_rate": 0.0003971341072278302,
+      "loss": 2.3351,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5795506648326456,
+      "grad_norm": 0.45733164109339225,
+      "learning_rate": 0.00039640741506251457,
+      "loss": 2.3272,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5800091701054562,
+      "grad_norm": 0.4771924413727776,
+      "learning_rate": 0.00039568095146107495,
+      "loss": 2.2507,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5804676753782668,
+      "grad_norm": 0.4980245746500559,
+      "learning_rate": 0.00039495471802636096,
+      "loss": 2.4184,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5809261806510775,
+      "grad_norm": 0.49617989727960987,
+      "learning_rate": 0.0003942287163607145,
+      "loss": 2.2634,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5813846859238881,
+      "grad_norm": 0.4842339498263633,
+      "learning_rate": 0.0003935029480659658,
+      "loss": 2.316,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5818431911966988,
+      "grad_norm": 0.44850706175575833,
+      "learning_rate": 0.00039277741474343054,
+      "loss": 2.3619,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5823016964695094,
+      "grad_norm": 0.4469526452478849,
+      "learning_rate": 0.0003920521179939057,
+      "loss": 2.289,
+      "step": 1270
+    },
+    {
+      "epoch": 0.58276020174232,
+      "grad_norm": 0.4588154112745075,
+      "learning_rate": 0.00039132705941766644,
+      "loss": 2.3699,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5832187070151307,
+      "grad_norm": 0.514637028682608,
+      "learning_rate": 0.0003906022406144624,
+      "loss": 2.2422,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5836772122879413,
+      "grad_norm": 0.4506057981308666,
+      "learning_rate": 0.0003898776631835143,
+      "loss": 2.3586,
+      "step": 1273
+    },
+    {
+      "epoch": 0.584135717560752,
+      "grad_norm": 0.4521917150951977,
+      "learning_rate": 0.00038915332872350994,
+      "loss": 2.177,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5845942228335625,
+      "grad_norm": 0.4470240849659399,
+      "learning_rate": 0.00038842923883260135,
+      "loss": 2.2219,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5850527281063732,
+      "grad_norm": 0.4652523649804786,
+      "learning_rate": 0.00038770539510840093,
+      "loss": 2.3244,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5855112333791839,
+      "grad_norm": 0.4684528842195352,
+      "learning_rate": 0.00038698179914797783,
+      "loss": 2.338,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5859697386519945,
+      "grad_norm": 0.4733803834676486,
+      "learning_rate": 0.0003862584525478545,
+      "loss": 2.3488,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5864282439248051,
+      "grad_norm": 0.4874715409507715,
+      "learning_rate": 0.00038553535690400353,
+      "loss": 2.3594,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5868867491976157,
+      "grad_norm": 0.5005287544819179,
+      "learning_rate": 0.00038481251381184355,
+      "loss": 2.254,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5873452544704264,
+      "grad_norm": 0.47220677205993106,
+      "learning_rate": 0.00038408992486623584,
+      "loss": 2.2743,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5878037597432371,
+      "grad_norm": 0.4467714848681405,
+      "learning_rate": 0.00038336759166148117,
+      "loss": 2.3537,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5882622650160477,
+      "grad_norm": 0.4499946980424808,
+      "learning_rate": 0.0003826455157913159,
+      "loss": 2.1494,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5887207702888583,
+      "grad_norm": 0.47449522598206345,
+      "learning_rate": 0.00038192369884890886,
+      "loss": 2.3316,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5891792755616689,
+      "grad_norm": 0.47639719803341457,
+      "learning_rate": 0.00038120214242685723,
+      "loss": 2.2588,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5896377808344796,
+      "grad_norm": 0.4615036121707951,
+      "learning_rate": 0.00038048084811718373,
+      "loss": 2.2462,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5900962861072903,
+      "grad_norm": 0.47942259906542645,
+      "learning_rate": 0.0003797598175113327,
+      "loss": 2.2884,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5905547913801009,
+      "grad_norm": 0.43151612746488166,
+      "learning_rate": 0.0003790390522001662,
+      "loss": 2.2234,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5910132966529115,
+      "grad_norm": 0.4590465798988853,
+      "learning_rate": 0.0003783185537739615,
+      "loss": 2.3135,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5914718019257221,
+      "grad_norm": 0.47834812025452045,
+      "learning_rate": 0.00037759832382240697,
+      "loss": 2.3475,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5919303071985328,
+      "grad_norm": 0.46237791615206203,
+      "learning_rate": 0.00037687836393459826,
+      "loss": 2.2626,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5923888124713435,
+      "grad_norm": 0.4812321306988998,
+      "learning_rate": 0.0003761586756990354,
+      "loss": 2.2898,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5928473177441541,
+      "grad_norm": 0.44227940003841143,
+      "learning_rate": 0.0003754392607036191,
+      "loss": 2.324,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5933058230169647,
+      "grad_norm": 0.46649907662920376,
+      "learning_rate": 0.0003747201205356472,
+      "loss": 2.3317,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5937643282897753,
+      "grad_norm": 0.4887046937686202,
+      "learning_rate": 0.0003740012567818111,
+      "loss": 2.4105,
+      "step": 1295
+    },
+    {
+      "epoch": 0.594222833562586,
+      "grad_norm": 0.4456482339801651,
+      "learning_rate": 0.0003732826710281922,
+      "loss": 2.2274,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5946813388353966,
+      "grad_norm": 0.4481615949049684,
+      "learning_rate": 0.0003725643648602588,
+      "loss": 2.2273,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5951398441082072,
+      "grad_norm": 0.4491411195154847,
+      "learning_rate": 0.0003718463398628621,
+      "loss": 2.297,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5955983493810179,
+      "grad_norm": 0.43583926059368494,
+      "learning_rate": 0.0003711285976202331,
+      "loss": 2.3297,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5960568546538285,
+      "grad_norm": 0.4557574224085634,
+      "learning_rate": 0.0003704111397159787,
+      "loss": 2.2698,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5965153599266392,
+      "grad_norm": 0.46132129730050275,
+      "learning_rate": 0.0003696939677330788,
+      "loss": 2.3288,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5969738651994498,
+      "grad_norm": 0.46041253470671156,
+      "learning_rate": 0.00036897708325388213,
+      "loss": 2.2336,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5974323704722604,
+      "grad_norm": 0.45419628055317673,
+      "learning_rate": 0.0003682604878601034,
+      "loss": 2.2543,
+      "step": 1303
+    },
+    {
+      "epoch": 0.597890875745071,
+      "grad_norm": 0.6071161410309122,
+      "learning_rate": 0.000367544183132819,
+      "loss": 2.2089,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5983493810178817,
+      "grad_norm": 0.45661098361178243,
+      "learning_rate": 0.0003668281706524645,
+      "loss": 2.2187,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5988078862906924,
+      "grad_norm": 0.4463414671658297,
+      "learning_rate": 0.0003661124519988304,
+      "loss": 2.2923,
+      "step": 1306
+    },
+    {
+      "epoch": 0.599266391563503,
+      "grad_norm": 0.482657441162913,
+      "learning_rate": 0.00036539702875105893,
+      "loss": 2.2304,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5997248968363136,
+      "grad_norm": 0.46827982503553583,
+      "learning_rate": 0.0003646819024876406,
+      "loss": 2.2873,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6001834021091242,
+      "grad_norm": 0.5053398919724338,
+      "learning_rate": 0.0003639670747864107,
+      "loss": 2.3967,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6006419073819349,
+      "grad_norm": 0.4702315916760791,
+      "learning_rate": 0.00036325254722454584,
+      "loss": 2.3284,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6011004126547456,
+      "grad_norm": 0.4779687760111233,
+      "learning_rate": 0.00036253832137855997,
+      "loss": 2.309,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6015589179275561,
+      "grad_norm": 0.4644976808288744,
+      "learning_rate": 0.00036182439882430183,
+      "loss": 2.2945,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6020174232003668,
+      "grad_norm": 0.48567778362082203,
+      "learning_rate": 0.00036111078113695096,
+      "loss": 2.1981,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6024759284731774,
+      "grad_norm": 0.4537053608880063,
+      "learning_rate": 0.0003603974698910139,
+      "loss": 2.203,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6029344337459881,
+      "grad_norm": 0.4743544973188022,
+      "learning_rate": 0.0003596844666603214,
+      "loss": 2.2521,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6033929390187988,
+      "grad_norm": 0.456742513239187,
+      "learning_rate": 0.0003589717730180245,
+      "loss": 2.3099,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6038514442916093,
+      "grad_norm": 0.42848658787701027,
+      "learning_rate": 0.00035825939053659117,
+      "loss": 2.3107,
+      "step": 1317
+    },
+    {
+      "epoch": 0.60430994956442,
+      "grad_norm": 0.42900777231493803,
+      "learning_rate": 0.00035754732078780273,
+      "loss": 2.2146,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6047684548372306,
+      "grad_norm": 0.4527210871451394,
+      "learning_rate": 0.00035683556534275076,
+      "loss": 2.3612,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6052269601100413,
+      "grad_norm": 0.4975094644626279,
+      "learning_rate": 0.00035612412577183303,
+      "loss": 2.2911,
+      "step": 1320
+    },
+    {
+      "epoch": 0.605685465382852,
+      "grad_norm": 0.46348615494067763,
+      "learning_rate": 0.00035541300364475063,
+      "loss": 2.2388,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6061439706556625,
+      "grad_norm": 0.4715993272996897,
+      "learning_rate": 0.0003547022005305043,
+      "loss": 2.336,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6066024759284732,
+      "grad_norm": 0.46636924998249646,
+      "learning_rate": 0.0003539917179973907,
+      "loss": 2.3418,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6070609812012838,
+      "grad_norm": 0.4470752781877185,
+      "learning_rate": 0.00035328155761299917,
+      "loss": 2.2705,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6075194864740945,
+      "grad_norm": 0.502046980785031,
+      "learning_rate": 0.0003525717209442085,
+      "loss": 2.2048,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6079779917469051,
+      "grad_norm": 0.43527786886684017,
+      "learning_rate": 0.00035186220955718306,
+      "loss": 2.0925,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6084364970197157,
+      "grad_norm": 0.4733164444712461,
+      "learning_rate": 0.0003511530250173696,
+      "loss": 2.2867,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6088950022925264,
+      "grad_norm": 0.4621992711361381,
+      "learning_rate": 0.00035044416888949364,
+      "loss": 2.2456,
+      "step": 1328
+    },
+    {
+      "epoch": 0.609353507565337,
+      "grad_norm": 0.4503725157212825,
+      "learning_rate": 0.0003497356427375562,
+      "loss": 2.2848,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6098120128381477,
+      "grad_norm": 0.46170707371090736,
+      "learning_rate": 0.00034902744812483034,
+      "loss": 2.3082,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6102705181109582,
+      "grad_norm": 0.44296787809461907,
+      "learning_rate": 0.00034831958661385714,
+      "loss": 2.2938,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6107290233837689,
+      "grad_norm": 0.44364300278119734,
+      "learning_rate": 0.0003476120597664434,
+      "loss": 2.2443,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6111875286565795,
+      "grad_norm": 0.44202571707794264,
+      "learning_rate": 0.00034690486914365704,
+      "loss": 2.3347,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6116460339293902,
+      "grad_norm": 0.46436364209816927,
+      "learning_rate": 0.00034619801630582435,
+      "loss": 2.3076,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6121045392022009,
+      "grad_norm": 0.4735431285137109,
+      "learning_rate": 0.00034549150281252633,
+      "loss": 2.2691,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6125630444750114,
+      "grad_norm": 0.45392072701922054,
+      "learning_rate": 0.0003447853302225952,
+      "loss": 2.2888,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6130215497478221,
+      "grad_norm": 0.47620631757207926,
+      "learning_rate": 0.00034407950009411126,
+      "loss": 2.1849,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6134800550206327,
+      "grad_norm": 0.4700078887595244,
+      "learning_rate": 0.00034337401398439873,
+      "loss": 2.2404,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6139385602934434,
+      "grad_norm": 0.47136081197514834,
+      "learning_rate": 0.00034266887345002305,
+      "loss": 2.2058,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6143970655662541,
+      "grad_norm": 0.4596426083976514,
+      "learning_rate": 0.0003419640800467874,
+      "loss": 2.3175,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6148555708390646,
+      "grad_norm": 0.4661709613554954,
+      "learning_rate": 0.0003412596353297288,
+      "loss": 2.3053,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6153140761118753,
+      "grad_norm": 0.47341537695363745,
+      "learning_rate": 0.00034055554085311493,
+      "loss": 2.2808,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6157725813846859,
+      "grad_norm": 0.47775513688868526,
+      "learning_rate": 0.00033985179817044105,
+      "loss": 2.3656,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6162310866574966,
+      "grad_norm": 0.4722617752933491,
+      "learning_rate": 0.0003391484088344257,
+      "loss": 2.2843,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6166895919303071,
+      "grad_norm": 0.45587928567297487,
+      "learning_rate": 0.00033844537439700807,
+      "loss": 2.4054,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6171480972031178,
+      "grad_norm": 0.49587837337249807,
+      "learning_rate": 0.00033774269640934445,
+      "loss": 2.3148,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6176066024759285,
+      "grad_norm": 0.45402891336759116,
+      "learning_rate": 0.0003370403764218045,
+      "loss": 2.2329,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6180651077487391,
+      "grad_norm": 0.46257062317709835,
+      "learning_rate": 0.000336338415983968,
+      "loss": 2.2183,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6185236130215498,
+      "grad_norm": 0.4360147872718794,
+      "learning_rate": 0.00033563681664462155,
+      "loss": 2.2144,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6189821182943603,
+      "grad_norm": 0.48479409434901827,
+      "learning_rate": 0.000334935579951755,
+      "loss": 2.4215,
+      "step": 1350
+    },
+    {
+      "epoch": 0.619440623567171,
+      "grad_norm": 0.4550298521425519,
+      "learning_rate": 0.0003342347074525578,
+      "loss": 2.2515,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6198991288399817,
+      "grad_norm": 0.4481838312735244,
+      "learning_rate": 0.0003335342006934161,
+      "loss": 2.3162,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6203576341127923,
+      "grad_norm": 0.44182275049469893,
+      "learning_rate": 0.00033283406121990914,
+      "loss": 2.2849,
+      "step": 1353
+    },
+    {
+      "epoch": 0.620816139385603,
+      "grad_norm": 0.4830996777415071,
+      "learning_rate": 0.0003321342905768057,
+      "loss": 2.2224,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6212746446584135,
+      "grad_norm": 0.46846751304130063,
+      "learning_rate": 0.00033143489030806086,
+      "loss": 2.2604,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6217331499312242,
+      "grad_norm": 0.4901038234447172,
+      "learning_rate": 0.00033073586195681227,
+      "loss": 2.2279,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6221916552040349,
+      "grad_norm": 0.48095207090772885,
+      "learning_rate": 0.00033003720706537736,
+      "loss": 2.1278,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6226501604768455,
+      "grad_norm": 0.5211650220133498,
+      "learning_rate": 0.0003293389271752492,
+      "loss": 2.2534,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6231086657496562,
+      "grad_norm": 0.5031250317366958,
+      "learning_rate": 0.00032864102382709374,
+      "loss": 2.3927,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6235671710224667,
+      "grad_norm": 0.4821713229851666,
+      "learning_rate": 0.000327943498560746,
+      "loss": 2.2879,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6240256762952774,
+      "grad_norm": 0.4727067699949811,
+      "learning_rate": 0.00032724635291520694,
+      "loss": 2.3222,
+      "step": 1361
+    },
+    {
+      "epoch": 0.624484181568088,
+      "grad_norm": 0.5030452204378784,
+      "learning_rate": 0.00032654958842863967,
+      "loss": 2.2767,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6249426868408987,
+      "grad_norm": 0.46969221632926755,
+      "learning_rate": 0.0003258532066383667,
+      "loss": 2.1645,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6254011921137093,
+      "grad_norm": 0.4895053019135574,
+      "learning_rate": 0.000325157209080866,
+      "loss": 2.2059,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6258596973865199,
+      "grad_norm": 0.4613771647621272,
+      "learning_rate": 0.00032446159729176743,
+      "loss": 2.3132,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6263182026593306,
+      "grad_norm": 0.5016644672728751,
+      "learning_rate": 0.0003237663728058502,
+      "loss": 2.2077,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6267767079321412,
+      "grad_norm": 0.4656001591416693,
+      "learning_rate": 0.0003230715371570389,
+      "loss": 2.3084,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6272352132049519,
+      "grad_norm": 0.4658359290300327,
+      "learning_rate": 0.00032237709187839996,
+      "loss": 2.2048,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6276937184777625,
+      "grad_norm": 0.4695140733230978,
+      "learning_rate": 0.0003216830385021388,
+      "loss": 2.1885,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6281522237505731,
+      "grad_norm": 0.45996663091055623,
+      "learning_rate": 0.0003209893785595959,
+      "loss": 2.3646,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6286107290233838,
+      "grad_norm": 0.4587460079002363,
+      "learning_rate": 0.00032029611358124366,
+      "loss": 2.1996,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6290692342961944,
+      "grad_norm": 0.4913303631065698,
+      "learning_rate": 0.00031960324509668336,
+      "loss": 2.2494,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6295277395690051,
+      "grad_norm": 0.5048556345868549,
+      "learning_rate": 0.0003189107746346412,
+      "loss": 2.3132,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6299862448418156,
+      "grad_norm": 0.4551788446805768,
+      "learning_rate": 0.0003182187037229653,
+      "loss": 2.2499,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6304447501146263,
+      "grad_norm": 0.45483187865928265,
+      "learning_rate": 0.0003175270338886221,
+      "loss": 2.3398,
+      "step": 1375
+    },
+    {
+      "epoch": 0.630903255387437,
+      "grad_norm": 0.48353998274516286,
+      "learning_rate": 0.00031683576665769345,
+      "loss": 2.312,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6313617606602476,
+      "grad_norm": 0.4802040113619127,
+      "learning_rate": 0.0003161449035553724,
+      "loss": 2.1867,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6318202659330582,
+      "grad_norm": 0.49824523516892616,
+      "learning_rate": 0.00031545444610596077,
+      "loss": 2.3004,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6322787712058688,
+      "grad_norm": 0.5143268807123631,
+      "learning_rate": 0.000314764395832865,
+      "loss": 2.2758,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6327372764786795,
+      "grad_norm": 0.4468577741184305,
+      "learning_rate": 0.0003140747542585934,
+      "loss": 2.1959,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6331957817514902,
+      "grad_norm": 0.45610559008908313,
+      "learning_rate": 0.00031338552290475266,
+      "loss": 2.3092,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6336542870243008,
+      "grad_norm": 0.45896558913379737,
+      "learning_rate": 0.00031269670329204396,
+      "loss": 2.2273,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6341127922971114,
+      "grad_norm": 0.4844156421734518,
+      "learning_rate": 0.0003120082969402604,
+      "loss": 2.3486,
+      "step": 1383
+    },
+    {
+      "epoch": 0.634571297569922,
+      "grad_norm": 0.615898484415851,
+      "learning_rate": 0.00031132030536828314,
+      "loss": 2.3202,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6350298028427327,
+      "grad_norm": 0.48209293054313346,
+      "learning_rate": 0.00031063273009407805,
+      "loss": 2.2714,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6354883081155434,
+      "grad_norm": 0.4602166943386828,
+      "learning_rate": 0.00030994557263469265,
+      "loss": 2.2779,
+      "step": 1386
+    },
+    {
+      "epoch": 0.635946813388354,
+      "grad_norm": 0.480059985552595,
+      "learning_rate": 0.0003092588345062526,
+      "loss": 2.2602,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6364053186611646,
+      "grad_norm": 0.45795005411782935,
+      "learning_rate": 0.0003085725172239582,
+      "loss": 2.3007,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6368638239339752,
+      "grad_norm": 0.4676085808263615,
+      "learning_rate": 0.0003078866223020815,
+      "loss": 2.2739,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6373223292067859,
+      "grad_norm": 0.4541679092873809,
+      "learning_rate": 0.0003072011512539624,
+      "loss": 2.2595,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6377808344795965,
+      "grad_norm": 0.4791342134435867,
+      "learning_rate": 0.00030651610559200574,
+      "loss": 2.3504,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6382393397524071,
+      "grad_norm": 0.4998018642396482,
+      "learning_rate": 0.00030583148682767757,
+      "loss": 2.2635,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6386978450252178,
+      "grad_norm": 0.48897152377202197,
+      "learning_rate": 0.00030514729647150243,
+      "loss": 2.2793,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6391563502980284,
+      "grad_norm": 0.46993931046208787,
+      "learning_rate": 0.0003044635360330592,
+      "loss": 2.2223,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6396148555708391,
+      "grad_norm": 0.5107645057037883,
+      "learning_rate": 0.00030378020702097845,
+      "loss": 2.2911,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6400733608436497,
+      "grad_norm": 0.4720939550607026,
+      "learning_rate": 0.000303097310942939,
+      "loss": 2.3416,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6405318661164603,
+      "grad_norm": 0.4658269000787192,
+      "learning_rate": 0.0003024148493056641,
+      "loss": 2.2473,
+      "step": 1397
+    },
+    {
+      "epoch": 0.640990371389271,
+      "grad_norm": 0.45658411067567123,
+      "learning_rate": 0.00030173282361491865,
+      "loss": 2.1991,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6414488766620816,
+      "grad_norm": 0.4647644767013233,
+      "learning_rate": 0.0003010512353755057,
+      "loss": 2.2141,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6419073819348923,
+      "grad_norm": 0.46643518806585726,
+      "learning_rate": 0.00030037008609126313,
+      "loss": 2.2568,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6423658872077029,
+      "grad_norm": 0.4757563839513519,
+      "learning_rate": 0.0002996893772650602,
+      "loss": 2.2225,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6428243924805135,
+      "grad_norm": 0.4445514998397275,
+      "learning_rate": 0.0002990091103987945,
+      "loss": 2.3519,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6432828977533241,
+      "grad_norm": 0.5091552370830954,
+      "learning_rate": 0.0002983292869933886,
+      "loss": 2.3622,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6437414030261348,
+      "grad_norm": 0.46295405483443763,
+      "learning_rate": 0.0002976499085487862,
+      "loss": 2.2864,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6441999082989455,
+      "grad_norm": 0.4448900552229659,
+      "learning_rate": 0.00029697097656394963,
+      "loss": 2.2335,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6446584135717561,
+      "grad_norm": 0.4815629847704046,
+      "learning_rate": 0.00029629249253685595,
+      "loss": 2.1832,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6451169188445667,
+      "grad_norm": 0.44215168293977547,
+      "learning_rate": 0.00029561445796449416,
+      "loss": 2.3575,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6455754241173773,
+      "grad_norm": 0.45626022551640283,
+      "learning_rate": 0.0002949368743428612,
+      "loss": 2.2309,
+      "step": 1408
+    },
+    {
+      "epoch": 0.646033929390188,
+      "grad_norm": 0.45245123108324575,
+      "learning_rate": 0.0002942597431669593,
+      "loss": 2.2643,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6464924346629987,
+      "grad_norm": 0.4495870082382914,
+      "learning_rate": 0.0002935830659307924,
+      "loss": 2.3034,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6469509399358092,
+      "grad_norm": 0.44685919160470483,
+      "learning_rate": 0.0002929068441273629,
+      "loss": 2.2485,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6474094452086199,
+      "grad_norm": 0.5128009227561882,
+      "learning_rate": 0.0002922310792486681,
+      "loss": 2.2995,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6478679504814305,
+      "grad_norm": 0.43394144155525544,
+      "learning_rate": 0.00029155577278569745,
+      "loss": 2.2427,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6483264557542412,
+      "grad_norm": 0.4890396653559723,
+      "learning_rate": 0.00029088092622842895,
+      "loss": 2.3293,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6487849610270519,
+      "grad_norm": 0.4576756926868447,
+      "learning_rate": 0.00029020654106582544,
+      "loss": 2.342,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6492434662998624,
+      "grad_norm": 0.4555474247458495,
+      "learning_rate": 0.0002895326187858326,
+      "loss": 2.2477,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6497019715726731,
+      "grad_norm": 0.46591127305265856,
+      "learning_rate": 0.00028885916087537377,
+      "loss": 2.3133,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6501604768454837,
+      "grad_norm": 0.5604113683334142,
+      "learning_rate": 0.00028818616882034877,
+      "loss": 2.3386,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6506189821182944,
+      "grad_norm": 0.43639769545924506,
+      "learning_rate": 0.0002875136441056286,
+      "loss": 2.2082,
+      "step": 1419
+    },
+    {
+      "epoch": 0.651077487391105,
+      "grad_norm": 0.4899283594808847,
+      "learning_rate": 0.000286841588215054,
+      "loss": 2.3176,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6515359926639156,
+      "grad_norm": 0.4605292534368965,
+      "learning_rate": 0.0002861700026314308,
+      "loss": 2.2601,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6519944979367263,
+      "grad_norm": 0.47362059142381296,
+      "learning_rate": 0.00028549888883652686,
+      "loss": 2.2363,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6524530032095369,
+      "grad_norm": 0.4865403035658126,
+      "learning_rate": 0.00028482824831107,
+      "loss": 2.3473,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6529115084823476,
+      "grad_norm": 0.5169827259788238,
+      "learning_rate": 0.000284158082534743,
+      "loss": 2.2251,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6533700137551581,
+      "grad_norm": 0.46392379666287253,
+      "learning_rate": 0.00028348839298618177,
+      "loss": 2.2567,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6538285190279688,
+      "grad_norm": 0.462758954576732,
+      "learning_rate": 0.0002828191811429709,
+      "loss": 2.2593,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6542870243007795,
+      "grad_norm": 0.457991246938549,
+      "learning_rate": 0.00028215044848164164,
+      "loss": 2.3267,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6547455295735901,
+      "grad_norm": 0.4830570867685424,
+      "learning_rate": 0.00028148219647766747,
+      "loss": 2.4058,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6552040348464008,
+      "grad_norm": 0.46039931223038366,
+      "learning_rate": 0.00028081442660546124,
+      "loss": 2.2162,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6556625401192113,
+      "grad_norm": 0.47741242569700226,
+      "learning_rate": 0.0002801471403383728,
+      "loss": 2.1807,
+      "step": 1430
+    },
+    {
+      "epoch": 0.656121045392022,
+      "grad_norm": 0.4809834924428443,
+      "learning_rate": 0.00027948033914868415,
+      "loss": 2.2718,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6565795506648326,
+      "grad_norm": 0.4754804130877837,
+      "learning_rate": 0.00027881402450760775,
+      "loss": 2.3237,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6570380559376433,
+      "grad_norm": 0.4693252809565253,
+      "learning_rate": 0.00027814819788528165,
+      "loss": 2.323,
+      "step": 1433
+    },
+    {
+      "epoch": 0.657496561210454,
+      "grad_norm": 0.49464205557769975,
+      "learning_rate": 0.00027748286075076836,
+      "loss": 2.2417,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6579550664832645,
+      "grad_norm": 0.4851718320116103,
+      "learning_rate": 0.00027681801457204937,
+      "loss": 2.2104,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6584135717560752,
+      "grad_norm": 0.4886313080053568,
+      "learning_rate": 0.00027615366081602306,
+      "loss": 2.3394,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6588720770288858,
+      "grad_norm": 0.5197112945179782,
+      "learning_rate": 0.0002754898009485021,
+      "loss": 2.2248,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6593305823016965,
+      "grad_norm": 0.4781264810172998,
+      "learning_rate": 0.0002748264364342085,
+      "loss": 2.2305,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6597890875745072,
+      "grad_norm": 0.49095662313701494,
+      "learning_rate": 0.00027416356873677204,
+      "loss": 2.2354,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6602475928473177,
+      "grad_norm": 0.5491087448940488,
+      "learning_rate": 0.0002735011993187258,
+      "loss": 2.2279,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6607060981201284,
+      "grad_norm": 0.45182079925107377,
+      "learning_rate": 0.0002728393296415042,
+      "loss": 2.4031,
+      "step": 1441
+    },
+    {
+      "epoch": 0.661164603392939,
+      "grad_norm": 0.4600299517169964,
+      "learning_rate": 0.00027217796116543817,
+      "loss": 2.277,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6616231086657497,
+      "grad_norm": 0.5003073303569041,
+      "learning_rate": 0.0002715170953497532,
+      "loss": 2.1929,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6620816139385602,
+      "grad_norm": 0.49377052720547626,
+      "learning_rate": 0.00027085673365256614,
+      "loss": 2.2472,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6625401192113709,
+      "grad_norm": 0.46811277991308226,
+      "learning_rate": 0.00027019687753088075,
+      "loss": 2.3244,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6629986244841816,
+      "grad_norm": 0.515062310056156,
+      "learning_rate": 0.00026953752844058597,
+      "loss": 2.2615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6634571297569922,
+      "grad_norm": 0.5061323752964065,
+      "learning_rate": 0.0002688786878364516,
+      "loss": 2.3411,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6639156350298029,
+      "grad_norm": 0.4745904470817687,
+      "learning_rate": 0.00026822035717212597,
+      "loss": 2.3179,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6643741403026134,
+      "grad_norm": 0.47890362555070987,
+      "learning_rate": 0.00026756253790013193,
+      "loss": 2.2028,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6648326455754241,
+      "grad_norm": 0.48074699743894056,
+      "learning_rate": 0.0002669052314718641,
+      "loss": 2.2316,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6652911508482348,
+      "grad_norm": 0.4648195030496216,
+      "learning_rate": 0.0002662484393375855,
+      "loss": 2.2171,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6657496561210454,
+      "grad_norm": 0.5090358288100331,
+      "learning_rate": 0.00026559216294642446,
+      "loss": 2.2783,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6662081613938561,
+      "grad_norm": 0.459204784441275,
+      "learning_rate": 0.0002649364037463718,
+      "loss": 2.3247,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.45980414296671246,
+      "learning_rate": 0.0002642811631842764,
+      "loss": 2.2584,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6671251719394773,
+      "grad_norm": 0.6331134006433501,
+      "learning_rate": 0.0002636264427058439,
+      "loss": 2.2449,
+      "step": 1455
+    },
+    {
+      "epoch": 0.667583677212288,
+      "grad_norm": 0.4711181004232556,
+      "learning_rate": 0.00026297224375563123,
+      "loss": 2.2627,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6680421824850986,
+      "grad_norm": 0.4614702149094049,
+      "learning_rate": 0.00026231856777704575,
+      "loss": 2.2938,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6685006877579092,
+      "grad_norm": 0.48159759146313824,
+      "learning_rate": 0.00026166541621234026,
+      "loss": 2.2391,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6689591930307198,
+      "grad_norm": 0.47908111629525457,
+      "learning_rate": 0.00026101279050261045,
+      "loss": 2.1421,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6694176983035305,
+      "grad_norm": 0.515784838495538,
+      "learning_rate": 0.00026036069208779247,
+      "loss": 2.3312,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6698762035763411,
+      "grad_norm": 0.4624733637446395,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 2.2308,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6703347088491518,
+      "grad_norm": 0.4700029115663184,
+      "learning_rate": 0.00025905808289681365,
+      "loss": 2.3046,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6707932141219624,
+      "grad_norm": 0.4722933634771388,
+      "learning_rate": 0.0002584075749946946,
+      "loss": 2.2924,
+      "step": 1463
+    },
+    {
+      "epoch": 0.671251719394773,
+      "grad_norm": 0.46105409328064784,
+      "learning_rate": 0.00025775760013556424,
+      "loss": 2.2348,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6717102246675837,
+      "grad_norm": 0.43683485549276646,
+      "learning_rate": 0.0002571081597535095,
+      "loss": 2.3184,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6721687299403943,
+      "grad_norm": 0.5095238642141188,
+      "learning_rate": 0.00025645925528143776,
+      "loss": 2.3366,
+      "step": 1466
+    },
+    {
+      "epoch": 0.672627235213205,
+      "grad_norm": 0.49421165355383917,
+      "learning_rate": 0.0002558108881510747,
+      "loss": 2.3217,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6730857404860155,
+      "grad_norm": 0.4809158330545061,
+      "learning_rate": 0.00025516305979295963,
+      "loss": 2.1852,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6735442457588262,
+      "grad_norm": 0.4685012696395692,
+      "learning_rate": 0.0002545157716364439,
+      "loss": 2.3226,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6740027510316369,
+      "grad_norm": 0.4727200874358837,
+      "learning_rate": 0.00025386902510968624,
+      "loss": 2.2182,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6744612563044475,
+      "grad_norm": 0.45260983825028855,
+      "learning_rate": 0.00025322282163965095,
+      "loss": 2.3588,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6749197615772582,
+      "grad_norm": 0.4492940785129958,
+      "learning_rate": 0.00025257716265210384,
+      "loss": 2.2415,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6753782668500687,
+      "grad_norm": 0.5015150751234094,
+      "learning_rate": 0.0002519320495716091,
+      "loss": 2.1973,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6758367721228794,
+      "grad_norm": 0.4772522799963434,
+      "learning_rate": 0.00025128748382152716,
+      "loss": 2.3736,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6762952773956901,
+      "grad_norm": 0.4464257902618613,
+      "learning_rate": 0.00025064346682401016,
+      "loss": 2.2181,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6767537826685007,
+      "grad_norm": 0.4933997230895835,
+      "learning_rate": 0.0002500000000000001,
+      "loss": 2.2933,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6772122879413113,
+      "grad_norm": 0.47554957886796406,
+      "learning_rate": 0.0002493570847692246,
+      "loss": 2.3102,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6776707932141219,
+      "grad_norm": 0.4593998352663516,
+      "learning_rate": 0.00024871472255019424,
+      "loss": 2.2102,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6781292984869326,
+      "grad_norm": 0.45454070125761153,
+      "learning_rate": 0.00024807291476019994,
+      "loss": 2.2497,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6785878037597433,
+      "grad_norm": 0.49539825612962857,
+      "learning_rate": 0.00024743166281530877,
+      "loss": 2.2399,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6790463090325539,
+      "grad_norm": 0.44702040999266235,
+      "learning_rate": 0.000246790968130362,
+      "loss": 2.2636,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6795048143053645,
+      "grad_norm": 0.482363011942832,
+      "learning_rate": 0.0002461508321189706,
+      "loss": 2.206,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6799633195781751,
+      "grad_norm": 0.44106687574461356,
+      "learning_rate": 0.00024551125619351385,
+      "loss": 2.2538,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6804218248509858,
+      "grad_norm": 0.47528315379783986,
+      "learning_rate": 0.00024487224176513453,
+      "loss": 2.2416,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6808803301237965,
+      "grad_norm": 0.46512431254912123,
+      "learning_rate": 0.0002442337902437365,
+      "loss": 2.3167,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6813388353966071,
+      "grad_norm": 0.4954225257712541,
+      "learning_rate": 0.0002435959030379824,
+      "loss": 2.2408,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6817973406694177,
+      "grad_norm": 0.47032237148004513,
+      "learning_rate": 0.00024295858155528888,
+      "loss": 2.263,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6822558459422283,
+      "grad_norm": 0.8224543085422215,
+      "learning_rate": 0.00024232182720182523,
+      "loss": 2.298,
+      "step": 1488
+    },
+    {
+      "epoch": 0.682714351215039,
+      "grad_norm": 0.4698561175724695,
+      "learning_rate": 0.00024168564138250855,
+      "loss": 2.2962,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6831728564878496,
+      "grad_norm": 0.44897361019390497,
+      "learning_rate": 0.00024105002550100246,
+      "loss": 2.2686,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6836313617606602,
+      "grad_norm": 0.44800626541935423,
+      "learning_rate": 0.00024041498095971254,
+      "loss": 2.2825,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6840898670334709,
+      "grad_norm": 0.479512038834823,
+      "learning_rate": 0.0002397805091597835,
+      "loss": 2.3159,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6845483723062815,
+      "grad_norm": 0.4623959706037454,
+      "learning_rate": 0.0002391466115010973,
+      "loss": 2.2925,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6850068775790922,
+      "grad_norm": 0.47624248184274387,
+      "learning_rate": 0.00023851328938226808,
+      "loss": 2.3029,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6854653828519028,
+      "grad_norm": 0.4732529663077768,
+      "learning_rate": 0.00023788054420064109,
+      "loss": 2.222,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6859238881247134,
+      "grad_norm": 0.4675456631836174,
+      "learning_rate": 0.00023724837735228773,
+      "loss": 2.2076,
+      "step": 1496
+    },
+    {
+      "epoch": 0.686382393397524,
+      "grad_norm": 0.4704260296159902,
+      "learning_rate": 0.00023661679023200422,
+      "loss": 2.3138,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6868408986703347,
+      "grad_norm": 0.45489779417378445,
+      "learning_rate": 0.00023598578423330714,
+      "loss": 2.2334,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6872994039431454,
+      "grad_norm": 0.4568592997726249,
+      "learning_rate": 0.00023535536074843083,
+      "loss": 2.2386,
+      "step": 1499
+    },
+    {
+      "epoch": 0.687757909215956,
+      "grad_norm": 0.4888912110440222,
+      "learning_rate": 0.00023472552116832502,
+      "loss": 2.3266,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6882164144887666,
+      "grad_norm": 0.450873118755475,
+      "learning_rate": 0.0002340962668826503,
+      "loss": 2.2735,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6886749197615772,
+      "grad_norm": 0.4889712731906452,
+      "learning_rate": 0.00023346759927977663,
+      "loss": 2.3356,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6891334250343879,
+      "grad_norm": 0.44299004667440856,
+      "learning_rate": 0.0002328395197467789,
+      "loss": 2.3034,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6895919303071986,
+      "grad_norm": 0.4644325491461807,
+      "learning_rate": 0.00023221202966943515,
+      "loss": 2.2978,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6900504355800092,
+      "grad_norm": 0.4722568854534182,
+      "learning_rate": 0.0002315851304322223,
+      "loss": 2.2742,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6905089408528198,
+      "grad_norm": 0.48446724454575046,
+      "learning_rate": 0.0002309588234183137,
+      "loss": 2.3354,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6909674461256304,
+      "grad_norm": 0.4747363639806262,
+      "learning_rate": 0.00023033311000957653,
+      "loss": 2.2633,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6914259513984411,
+      "grad_norm": 0.4840801892284972,
+      "learning_rate": 0.00022970799158656758,
+      "loss": 2.2827,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6918844566712518,
+      "grad_norm": 0.4613632876346254,
+      "learning_rate": 0.0002290834695285316,
+      "loss": 2.2838,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6923429619440623,
+      "grad_norm": 0.43426547020304906,
+      "learning_rate": 0.00022845954521339678,
+      "loss": 2.3399,
+      "step": 1510
+    },
+    {
+      "epoch": 0.692801467216873,
+      "grad_norm": 0.4584181930655538,
+      "learning_rate": 0.0002278362200177732,
+      "loss": 2.3013,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6932599724896836,
+      "grad_norm": 0.4533238622007884,
+      "learning_rate": 0.00022721349531694852,
+      "loss": 2.2163,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6937184777624943,
+      "grad_norm": 0.46841434317463215,
+      "learning_rate": 0.0002265913724848855,
+      "loss": 2.2388,
+      "step": 1513
+    },
+    {
+      "epoch": 0.694176983035305,
+      "grad_norm": 0.4807477209977391,
+      "learning_rate": 0.00022596985289421946,
+      "loss": 2.2965,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6946354883081155,
+      "grad_norm": 0.5030444162486121,
+      "learning_rate": 0.00022534893791625405,
+      "loss": 2.2536,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6950939935809262,
+      "grad_norm": 0.4897936999182432,
+      "learning_rate": 0.00022472862892095968,
+      "loss": 2.249,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6955524988537368,
+      "grad_norm": 0.48395000656417686,
+      "learning_rate": 0.00022410892727696896,
+      "loss": 2.3496,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6960110041265475,
+      "grad_norm": 0.4453661265730055,
+      "learning_rate": 0.0002234898343515751,
+      "loss": 2.2852,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6964695093993581,
+      "grad_norm": 0.4675382508697846,
+      "learning_rate": 0.00022287135151072792,
+      "loss": 2.2186,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6969280146721687,
+      "grad_norm": 0.4457250560286611,
+      "learning_rate": 0.00022225348011903096,
+      "loss": 2.2426,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6973865199449794,
+      "grad_norm": 0.47858012942134426,
+      "learning_rate": 0.0002216362215397393,
+      "loss": 2.2045,
+      "step": 1521
+    },
+    {
+      "epoch": 0.69784502521779,
+      "grad_norm": 0.45826579097428055,
+      "learning_rate": 0.00022101957713475522,
+      "loss": 2.2144,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6983035304906007,
+      "grad_norm": 0.456439243865162,
+      "learning_rate": 0.00022040354826462666,
+      "loss": 2.2362,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6987620357634112,
+      "grad_norm": 0.4607690009670779,
+      "learning_rate": 0.0002197881362885426,
+      "loss": 2.2558,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6992205410362219,
+      "grad_norm": 0.5006569199109505,
+      "learning_rate": 0.0002191733425643318,
+      "loss": 2.175,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6996790463090325,
+      "grad_norm": 0.46934490054410716,
+      "learning_rate": 0.00021855916844845826,
+      "loss": 2.2292,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7001375515818432,
+      "grad_norm": 0.45007615040625065,
+      "learning_rate": 0.00021794561529601898,
+      "loss": 2.1864,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7005960568546539,
+      "grad_norm": 0.45445910447321475,
+      "learning_rate": 0.00021733268446074138,
+      "loss": 2.1921,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7010545621274644,
+      "grad_norm": 0.4965225449475197,
+      "learning_rate": 0.00021672037729497917,
+      "loss": 2.2142,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7015130674002751,
+      "grad_norm": 0.5021719658440579,
+      "learning_rate": 0.0002161086951497106,
+      "loss": 2.3184,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7019715726730857,
+      "grad_norm": 0.5015651999206677,
+      "learning_rate": 0.00021549763937453442,
+      "loss": 2.333,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7024300779458964,
+      "grad_norm": 0.4830235629188072,
+      "learning_rate": 0.00021488721131766736,
+      "loss": 2.2904,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7028885832187071,
+      "grad_norm": 0.4518716469051841,
+      "learning_rate": 0.00021427741232594183,
+      "loss": 2.2007,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7033470884915176,
+      "grad_norm": 0.48698126785867085,
+      "learning_rate": 0.0002136682437448013,
+      "loss": 2.3073,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7038055937643283,
+      "grad_norm": 0.5001095651977056,
+      "learning_rate": 0.0002130597069182994,
+      "loss": 2.1586,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7042640990371389,
+      "grad_norm": 0.49241806724576465,
+      "learning_rate": 0.0002124518031890948,
+      "loss": 2.2478,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7047226043099496,
+      "grad_norm": 0.4724162510633656,
+      "learning_rate": 0.0002118445338984502,
+      "loss": 2.2815,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7051811095827603,
+      "grad_norm": 0.5102971485285926,
+      "learning_rate": 0.00021123790038622808,
+      "loss": 2.3095,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7056396148555708,
+      "grad_norm": 0.48482209263097076,
+      "learning_rate": 0.0002106319039908879,
+      "loss": 2.3838,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7060981201283815,
+      "grad_norm": 0.45059663929774535,
+      "learning_rate": 0.00021002654604948412,
+      "loss": 2.3001,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7065566254011921,
+      "grad_norm": 0.4703763306192407,
+      "learning_rate": 0.00020942182789766172,
+      "loss": 2.2561,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7070151306740028,
+      "grad_norm": 0.47797026368922035,
+      "learning_rate": 0.00020881775086965492,
+      "loss": 2.2578,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7074736359468133,
+      "grad_norm": 0.5003162967776387,
+      "learning_rate": 0.00020821431629828246,
+      "loss": 2.232,
+      "step": 1543
+    },
+    {
+      "epoch": 0.707932141219624,
+      "grad_norm": 0.4741314286285168,
+      "learning_rate": 0.00020761152551494643,
+      "loss": 2.2716,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7083906464924347,
+      "grad_norm": 0.45565337320004934,
+      "learning_rate": 0.00020700937984962798,
+      "loss": 2.2232,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7088491517652453,
+      "grad_norm": 0.5444457416291061,
+      "learning_rate": 0.0002064078806308848,
+      "loss": 2.3752,
+      "step": 1546
+    },
+    {
+      "epoch": 0.709307657038056,
+      "grad_norm": 0.49589690579677,
+      "learning_rate": 0.00020580702918584882,
+      "loss": 2.3843,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7097661623108665,
+      "grad_norm": 1.3688608688448363,
+      "learning_rate": 0.000205206826840222,
+      "loss": 2.1817,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7102246675836772,
+      "grad_norm": 0.4926004477429723,
+      "learning_rate": 0.0002046072749182751,
+      "loss": 2.339,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7106831728564879,
+      "grad_norm": 0.4719852631016075,
+      "learning_rate": 0.00020400837474284273,
+      "loss": 2.284,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7111416781292985,
+      "grad_norm": 0.5073446112748211,
+      "learning_rate": 0.0002034101276353224,
+      "loss": 2.2873,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7116001834021092,
+      "grad_norm": 0.4706779325834857,
+      "learning_rate": 0.00020281253491567027,
+      "loss": 2.3124,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7120586886749197,
+      "grad_norm": 0.4561624536218782,
+      "learning_rate": 0.0002022155979023984,
+      "loss": 2.2241,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7125171939477304,
+      "grad_norm": 0.47581009030136695,
+      "learning_rate": 0.000201619317912573,
+      "loss": 2.2605,
+      "step": 1554
+    },
+    {
+      "epoch": 0.712975699220541,
+      "grad_norm": 0.4901829410924124,
+      "learning_rate": 0.00020102369626180962,
+      "loss": 2.221,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7134342044933517,
+      "grad_norm": 0.4578390206948753,
+      "learning_rate": 0.0002004287342642721,
+      "loss": 2.2273,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7138927097661623,
+      "grad_norm": 0.5181621689740282,
+      "learning_rate": 0.00019983443323266824,
+      "loss": 2.3989,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7143512150389729,
+      "grad_norm": 0.4639378480046995,
+      "learning_rate": 0.00019924079447824805,
+      "loss": 2.2605,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7148097203117836,
+      "grad_norm": 0.47268287553157107,
+      "learning_rate": 0.00019864781931079977,
+      "loss": 2.2064,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7152682255845942,
+      "grad_norm": 0.440667424987599,
+      "learning_rate": 0.00019805550903864773,
+      "loss": 2.2512,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7157267308574049,
+      "grad_norm": 0.4661432091537637,
+      "learning_rate": 0.00019746386496864948,
+      "loss": 2.2208,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7161852361302155,
+      "grad_norm": 0.4788140509806164,
+      "learning_rate": 0.00019687288840619226,
+      "loss": 2.2566,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7166437414030261,
+      "grad_norm": 0.47640800206895806,
+      "learning_rate": 0.0001962825806551911,
+      "loss": 2.2211,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7171022466758368,
+      "grad_norm": 0.4765529069755778,
+      "learning_rate": 0.0001956929430180846,
+      "loss": 2.2015,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7175607519486474,
+      "grad_norm": 0.45468938541287723,
+      "learning_rate": 0.00019510397679583374,
+      "loss": 2.2602,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7180192572214581,
+      "grad_norm": 0.5230770996554427,
+      "learning_rate": 0.0001945156832879174,
+      "loss": 2.2049,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7184777624942686,
+      "grad_norm": 0.5085599128374265,
+      "learning_rate": 0.00019392806379233036,
+      "loss": 2.2729,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7189362677670793,
+      "grad_norm": 0.5208540683085926,
+      "learning_rate": 0.00019334111960558065,
+      "loss": 2.1522,
+      "step": 1568
+    },
+    {
+      "epoch": 0.71939477303989,
+      "grad_norm": 0.6230600108657323,
+      "learning_rate": 0.00019275485202268573,
+      "loss": 2.328,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7198532783127006,
+      "grad_norm": 0.5092402712832784,
+      "learning_rate": 0.00019216926233717085,
+      "loss": 2.1661,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7203117835855112,
+      "grad_norm": 0.47053136662191014,
+      "learning_rate": 0.00019158435184106498,
+      "loss": 2.3019,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7207702888583218,
+      "grad_norm": 0.487844814756633,
+      "learning_rate": 0.00019100012182489905,
+      "loss": 2.2573,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7212287941311325,
+      "grad_norm": 0.45515189665215355,
+      "learning_rate": 0.00019041657357770226,
+      "loss": 2.2698,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7216872994039432,
+      "grad_norm": 0.5111944225523117,
+      "learning_rate": 0.00018983370838699943,
+      "loss": 2.2584,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7221458046767538,
+      "grad_norm": 0.48091156134632107,
+      "learning_rate": 0.00018925152753880892,
+      "loss": 2.1729,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7226043099495644,
+      "grad_norm": 0.45104897735168814,
+      "learning_rate": 0.00018867003231763847,
+      "loss": 2.284,
+      "step": 1576
+    },
+    {
+      "epoch": 0.723062815222375,
+      "grad_norm": 0.45377097517231324,
+      "learning_rate": 0.00018808922400648375,
+      "loss": 2.2561,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7235213204951857,
+      "grad_norm": 0.45474467959538617,
+      "learning_rate": 0.00018750910388682428,
+      "loss": 2.2565,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7239798257679964,
+      "grad_norm": 0.46031815052975,
+      "learning_rate": 0.00018692967323862125,
+      "loss": 2.2354,
+      "step": 1579
+    },
+    {
+      "epoch": 0.724438331040807,
+      "grad_norm": 0.4801028307659971,
+      "learning_rate": 0.00018635093334031517,
+      "loss": 2.2349,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7248968363136176,
+      "grad_norm": 0.447114686817742,
+      "learning_rate": 0.00018577288546882165,
+      "loss": 2.1689,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7253553415864282,
+      "grad_norm": 0.4578142120435913,
+      "learning_rate": 0.00018519553089953023,
+      "loss": 2.2598,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7258138468592389,
+      "grad_norm": 0.5228792824092746,
+      "learning_rate": 0.0001846188709063001,
+      "loss": 2.3782,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7262723521320495,
+      "grad_norm": 0.4568165285921803,
+      "learning_rate": 0.00018404290676145857,
+      "loss": 2.2716,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7267308574048602,
+      "grad_norm": 0.48670737266063185,
+      "learning_rate": 0.00018346763973579722,
+      "loss": 2.2231,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7271893626776708,
+      "grad_norm": 0.5017825339579983,
+      "learning_rate": 0.00018289307109856939,
+      "loss": 2.2611,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7276478679504814,
+      "grad_norm": 0.5209162701446077,
+      "learning_rate": 0.0001823192021174882,
+      "loss": 2.2445,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7281063732232921,
+      "grad_norm": 0.5179296191828859,
+      "learning_rate": 0.0001817460340587223,
+      "loss": 2.2951,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7285648784961027,
+      "grad_norm": 0.509907855182608,
+      "learning_rate": 0.00018117356818689445,
+      "loss": 2.219,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7290233837689133,
+      "grad_norm": 0.4942985239277642,
+      "learning_rate": 0.00018060180576507756,
+      "loss": 2.289,
+      "step": 1590
+    },
+    {
+      "epoch": 0.729481889041724,
+      "grad_norm": 0.47675780943796703,
+      "learning_rate": 0.00018003074805479313,
+      "loss": 2.3591,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7299403943145346,
+      "grad_norm": 0.5084555133435279,
+      "learning_rate": 0.00017946039631600724,
+      "loss": 2.2227,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7303988995873453,
+      "grad_norm": 0.465379919272908,
+      "learning_rate": 0.00017889075180712837,
+      "loss": 2.3211,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7308574048601559,
+      "grad_norm": 0.49678035332668175,
+      "learning_rate": 0.00017832181578500512,
+      "loss": 2.2787,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7313159101329665,
+      "grad_norm": 0.4802013919106934,
+      "learning_rate": 0.0001777535895049221,
+      "loss": 2.1894,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7317744154057771,
+      "grad_norm": 0.49024778562417837,
+      "learning_rate": 0.0001771860742205988,
+      "loss": 2.2869,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7322329206785878,
+      "grad_norm": 0.47872038384564247,
+      "learning_rate": 0.00017661927118418525,
+      "loss": 2.2129,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7326914259513985,
+      "grad_norm": 0.5320465441832274,
+      "learning_rate": 0.00017605318164626066,
+      "loss": 2.235,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7331499312242091,
+      "grad_norm": 0.4761048002691075,
+      "learning_rate": 0.00017548780685582949,
+      "loss": 2.3496,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7336084364970197,
+      "grad_norm": 0.490891284579756,
+      "learning_rate": 0.00017492314806031922,
+      "loss": 2.3452,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7340669417698303,
+      "grad_norm": 0.47233192468124263,
+      "learning_rate": 0.00017435920650557806,
+      "loss": 2.1259,
+      "step": 1601
+    },
+    {
+      "epoch": 0.734525447042641,
+      "grad_norm": 0.4599755848234981,
+      "learning_rate": 0.00017379598343587112,
+      "loss": 2.257,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7349839523154517,
+      "grad_norm": 0.499183005232666,
+      "learning_rate": 0.00017323348009387878,
+      "loss": 2.3295,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7354424575882622,
+      "grad_norm": 0.47468048475948316,
+      "learning_rate": 0.0001726716977206929,
+      "loss": 2.2284,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7359009628610729,
+      "grad_norm": 0.5144032372860707,
+      "learning_rate": 0.00017211063755581525,
+      "loss": 2.2502,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7363594681338835,
+      "grad_norm": 0.4683886731071831,
+      "learning_rate": 0.0001715503008371536,
+      "loss": 2.2655,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7368179734066942,
+      "grad_norm": 0.4630587498594915,
+      "learning_rate": 0.0001709906888010196,
+      "loss": 2.2848,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7372764786795049,
+      "grad_norm": 0.4944431301092377,
+      "learning_rate": 0.00017043180268212638,
+      "loss": 2.2481,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7377349839523154,
+      "grad_norm": 0.6639990008371849,
+      "learning_rate": 0.00016987364371358481,
+      "loss": 2.2173,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7381934892251261,
+      "grad_norm": 0.5016467919053781,
+      "learning_rate": 0.00016931621312690214,
+      "loss": 2.2837,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7386519944979367,
+      "grad_norm": 0.4794904333478769,
+      "learning_rate": 0.00016875951215197777,
+      "loss": 2.2801,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7391104997707474,
+      "grad_norm": 0.49938258826039994,
+      "learning_rate": 0.00016820354201710214,
+      "loss": 2.2336,
+      "step": 1612
+    },
+    {
+      "epoch": 0.739569005043558,
+      "grad_norm": 0.47980256933695686,
+      "learning_rate": 0.00016764830394895203,
+      "loss": 2.3011,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7400275103163686,
+      "grad_norm": 0.5720612385539089,
+      "learning_rate": 0.00016709379917259027,
+      "loss": 2.2578,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7404860155891793,
+      "grad_norm": 0.4694239180222222,
+      "learning_rate": 0.00016654002891146091,
+      "loss": 2.2371,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7409445208619899,
+      "grad_norm": 0.4651275918190467,
+      "learning_rate": 0.00016598699438738764,
+      "loss": 2.2247,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7414030261348006,
+      "grad_norm": 0.5093439322432295,
+      "learning_rate": 0.00016543469682057105,
+      "loss": 2.3098,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7418615314076112,
+      "grad_norm": 0.4943179007630498,
+      "learning_rate": 0.00016488313742958526,
+      "loss": 2.2856,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7423200366804218,
+      "grad_norm": 0.49272874856026244,
+      "learning_rate": 0.00016433231743137646,
+      "loss": 2.3249,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7427785419532325,
+      "grad_norm": 0.44983315860673406,
+      "learning_rate": 0.0001637822380412584,
+      "loss": 2.2051,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7432370472260431,
+      "grad_norm": 0.4688816000067551,
+      "learning_rate": 0.00016323290047291195,
+      "loss": 2.2137,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7436955524988538,
+      "grad_norm": 0.4744162717340211,
+      "learning_rate": 0.0001626843059383803,
+      "loss": 2.3348,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7441540577716643,
+      "grad_norm": 0.5171875573176961,
+      "learning_rate": 0.00016213645564806752,
+      "loss": 2.3119,
+      "step": 1623
+    },
+    {
+      "epoch": 0.744612563044475,
+      "grad_norm": 0.4978511025829592,
+      "learning_rate": 0.0001615893508107359,
+      "loss": 2.1522,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7450710683172856,
+      "grad_norm": 0.44461358809251283,
+      "learning_rate": 0.00016104299263350252,
+      "loss": 2.2684,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7455295735900963,
+      "grad_norm": 0.49664605059974776,
+      "learning_rate": 0.00016049738232183758,
+      "loss": 2.2311,
+      "step": 1626
+    },
+    {
+      "epoch": 0.745988078862907,
+      "grad_norm": 0.4649755705224005,
+      "learning_rate": 0.0001599525210795606,
+      "loss": 2.1808,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7464465841357175,
+      "grad_norm": 0.4680125404637998,
+      "learning_rate": 0.00015940841010883889,
+      "loss": 2.2728,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7469050894085282,
+      "grad_norm": 0.4891115811670911,
+      "learning_rate": 0.00015886505061018413,
+      "loss": 2.2597,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7473635946813388,
+      "grad_norm": 0.47702680786692697,
+      "learning_rate": 0.0001583224437824498,
+      "loss": 2.214,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7478220999541495,
+      "grad_norm": 0.4573830120470832,
+      "learning_rate": 0.0001577805908228293,
+      "loss": 2.3158,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7482806052269602,
+      "grad_norm": 0.4939726000347658,
+      "learning_rate": 0.00015723949292685191,
+      "loss": 2.2389,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7487391104997707,
+      "grad_norm": 0.5150610226342482,
+      "learning_rate": 0.0001566991512883818,
+      "loss": 2.2104,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7491976157725814,
+      "grad_norm": 0.4740801091507419,
+      "learning_rate": 0.00015615956709961378,
+      "loss": 2.3683,
+      "step": 1634
+    },
+    {
+      "epoch": 0.749656121045392,
+      "grad_norm": 0.4821542299553622,
+      "learning_rate": 0.00015562074155107215,
+      "loss": 2.2151,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7501146263182027,
+      "grad_norm": 0.47787220098712785,
+      "learning_rate": 0.0001550826758316068,
+      "loss": 2.3172,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7505731315910132,
+      "grad_norm": 0.4977239672263947,
+      "learning_rate": 0.00015454537112839122,
+      "loss": 2.3591,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7510316368638239,
+      "grad_norm": 0.5047338197163657,
+      "learning_rate": 0.00015400882862692033,
+      "loss": 2.2898,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7514901421366346,
+      "grad_norm": 0.5132959984085916,
+      "learning_rate": 0.00015347304951100665,
+      "loss": 2.2374,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7519486474094452,
+      "grad_norm": 0.4797746982711949,
+      "learning_rate": 0.00015293803496277907,
+      "loss": 2.316,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7524071526822559,
+      "grad_norm": 0.47462839610440993,
+      "learning_rate": 0.00015240378616267886,
+      "loss": 2.2906,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7528656579550664,
+      "grad_norm": 0.48757488752242883,
+      "learning_rate": 0.00015187030428945843,
+      "loss": 2.1454,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7533241632278771,
+      "grad_norm": 0.44678428089509487,
+      "learning_rate": 0.0001513375905201776,
+      "loss": 2.248,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7537826685006878,
+      "grad_norm": 0.46976803992519206,
+      "learning_rate": 0.00015080564603020142,
+      "loss": 2.2472,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7542411737734984,
+      "grad_norm": 0.5344427254857236,
+      "learning_rate": 0.0001502744719931982,
+      "loss": 2.1831,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7546996790463091,
+      "grad_norm": 0.45587617928263924,
+      "learning_rate": 0.00014974406958113558,
+      "loss": 2.2804,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7551581843191196,
+      "grad_norm": 0.49489922447204643,
+      "learning_rate": 0.00014921443996427947,
+      "loss": 2.218,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7556166895919303,
+      "grad_norm": 0.48269447499386464,
+      "learning_rate": 0.0001486855843111901,
+      "loss": 2.349,
+      "step": 1648
+    },
+    {
+      "epoch": 0.756075194864741,
+      "grad_norm": 0.4693615960500264,
+      "learning_rate": 0.0001481575037887201,
+      "loss": 2.2949,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7565337001375516,
+      "grad_norm": 0.45422670757521155,
+      "learning_rate": 0.00014763019956201253,
+      "loss": 2.2867,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7569922054103623,
+      "grad_norm": 0.4635855187237622,
+      "learning_rate": 0.0001471036727944966,
+      "loss": 2.2086,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7574507106831728,
+      "grad_norm": 0.43905558035529346,
+      "learning_rate": 0.0001465779246478872,
+      "loss": 2.1713,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7579092159559835,
+      "grad_norm": 0.4773865837365777,
+      "learning_rate": 0.00014605295628218045,
+      "loss": 2.17,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7583677212287941,
+      "grad_norm": 0.4766854917490475,
+      "learning_rate": 0.0001455287688556527,
+      "loss": 2.2331,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7588262265016048,
+      "grad_norm": 0.4977873998135065,
+      "learning_rate": 0.00014500536352485673,
+      "loss": 2.2162,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7592847317744154,
+      "grad_norm": 0.47387186523783836,
+      "learning_rate": 0.00014448274144461965,
+      "loss": 2.2327,
+      "step": 1656
+    },
+    {
+      "epoch": 0.759743237047226,
+      "grad_norm": 0.4808693816829195,
+      "learning_rate": 0.00014396090376804112,
+      "loss": 2.3143,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7602017423200367,
+      "grad_norm": 0.4764966071511441,
+      "learning_rate": 0.00014343985164648926,
+      "loss": 2.163,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7606602475928473,
+      "grad_norm": 0.4834701324769076,
+      "learning_rate": 0.00014291958622959973,
+      "loss": 2.4016,
+      "step": 1659
+    },
+    {
+      "epoch": 0.761118752865658,
+      "grad_norm": 0.5094652252229327,
+      "learning_rate": 0.00014240010866527176,
+      "loss": 2.3212,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7615772581384685,
+      "grad_norm": 0.4502760037515098,
+      "learning_rate": 0.00014188142009966686,
+      "loss": 2.259,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7620357634112792,
+      "grad_norm": 0.47905565948573114,
+      "learning_rate": 0.0001413635216772053,
+      "loss": 2.2606,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7624942686840899,
+      "grad_norm": 0.4769407348445002,
+      "learning_rate": 0.000140846414540564,
+      "loss": 2.3101,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7629527739569005,
+      "grad_norm": 0.4671079364088291,
+      "learning_rate": 0.00014033009983067452,
+      "loss": 2.2231,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7634112792297112,
+      "grad_norm": 0.5197865465686196,
+      "learning_rate": 0.00013981457868671927,
+      "loss": 2.2084,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7638697845025217,
+      "grad_norm": 0.4697065687435049,
+      "learning_rate": 0.0001392998522461305,
+      "loss": 2.2205,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7643282897753324,
+      "grad_norm": 0.4771111133194987,
+      "learning_rate": 0.00013878592164458635,
+      "loss": 2.2941,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7647867950481431,
+      "grad_norm": 0.4632757248376126,
+      "learning_rate": 0.00013827278801600978,
+      "loss": 2.2436,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7652453003209537,
+      "grad_norm": 0.4806646828529233,
+      "learning_rate": 0.0001377604524925647,
+      "loss": 2.2564,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7657038055937643,
+      "grad_norm": 0.46220275401701993,
+      "learning_rate": 0.00013724891620465424,
+      "loss": 2.219,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7661623108665749,
+      "grad_norm": 0.4688068647215289,
+      "learning_rate": 0.0001367381802809185,
+      "loss": 2.2541,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7666208161393856,
+      "grad_norm": 0.48643843877123843,
+      "learning_rate": 0.00013622824584823113,
+      "loss": 2.2382,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7670793214121963,
+      "grad_norm": 0.4819742711465602,
+      "learning_rate": 0.00013571911403169795,
+      "loss": 2.2756,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7675378266850069,
+      "grad_norm": 0.5227982430457984,
+      "learning_rate": 0.0001352107859546533,
+      "loss": 2.3185,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7679963319578175,
+      "grad_norm": 0.48167581944957794,
+      "learning_rate": 0.00013470326273865886,
+      "loss": 2.2786,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7684548372306281,
+      "grad_norm": 0.4806608369987207,
+      "learning_rate": 0.00013419654550349985,
+      "loss": 2.1906,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7689133425034388,
+      "grad_norm": 0.5113680494100289,
+      "learning_rate": 0.00013369063536718346,
+      "loss": 2.1993,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7693718477762495,
+      "grad_norm": 0.45260666055059917,
+      "learning_rate": 0.00013318553344593632,
+      "loss": 2.1865,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7698303530490601,
+      "grad_norm": 0.4994839255486534,
+      "learning_rate": 0.00013268124085420136,
+      "loss": 2.2226,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7702888583218707,
+      "grad_norm": 0.49612041391407213,
+      "learning_rate": 0.0001321777587046364,
+      "loss": 2.2539,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7707473635946813,
+      "grad_norm": 0.49125804495123854,
+      "learning_rate": 0.00013167508810811059,
+      "loss": 2.2614,
+      "step": 1681
+    },
+    {
+      "epoch": 0.771205868867492,
+      "grad_norm": 0.48874571167699954,
+      "learning_rate": 0.0001311732301737029,
+      "loss": 2.2676,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7716643741403026,
+      "grad_norm": 0.4781487283270805,
+      "learning_rate": 0.0001306721860086991,
+      "loss": 2.2644,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7721228794131133,
+      "grad_norm": 0.47827598862720117,
+      "learning_rate": 0.00013017195671858928,
+      "loss": 2.1943,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7725813846859239,
+      "grad_norm": 0.48736807023010303,
+      "learning_rate": 0.0001296725434070661,
+      "loss": 2.2072,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7730398899587345,
+      "grad_norm": 0.4431423840993422,
+      "learning_rate": 0.00012917394717602121,
+      "loss": 2.2227,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7734983952315452,
+      "grad_norm": 0.45976707854077625,
+      "learning_rate": 0.00012867616912554426,
+      "loss": 2.3485,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7739569005043558,
+      "grad_norm": 0.46957013374116885,
+      "learning_rate": 0.00012817921035391882,
+      "loss": 2.221,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7744154057771664,
+      "grad_norm": 0.4939062317448751,
+      "learning_rate": 0.00012768307195762168,
+      "loss": 2.2852,
+      "step": 1689
+    },
+    {
+      "epoch": 0.774873911049977,
+      "grad_norm": 0.49445198237924043,
+      "learning_rate": 0.00012718775503131908,
+      "loss": 2.2303,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7753324163227877,
+      "grad_norm": 0.46303626728876335,
+      "learning_rate": 0.0001266932606678646,
+      "loss": 2.1902,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7757909215955984,
+      "grad_norm": 0.5036198609402923,
+      "learning_rate": 0.00012619958995829756,
+      "loss": 2.2302,
+      "step": 1692
+    },
+    {
+      "epoch": 0.776249426868409,
+      "grad_norm": 0.4685706733122915,
+      "learning_rate": 0.0001257067439918394,
+      "loss": 2.2098,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7767079321412196,
+      "grad_norm": 0.5067225605745731,
+      "learning_rate": 0.00012521472385589234,
+      "loss": 2.2802,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7771664374140302,
+      "grad_norm": 0.4664855602508808,
+      "learning_rate": 0.00012472353063603626,
+      "loss": 2.2409,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7776249426868409,
+      "grad_norm": 0.553374339745492,
+      "learning_rate": 0.0001242331654160263,
+      "loss": 2.3034,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7780834479596516,
+      "grad_norm": 0.5033390513333899,
+      "learning_rate": 0.0001237436292777914,
+      "loss": 2.294,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7785419532324622,
+      "grad_norm": 0.49756695681385865,
+      "learning_rate": 0.00012325492330143061,
+      "loss": 2.269,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7790004585052728,
+      "grad_norm": 0.47461423324425717,
+      "learning_rate": 0.00012276704856521175,
+      "loss": 2.2986,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7794589637780834,
+      "grad_norm": 0.49166941410373144,
+      "learning_rate": 0.00012228000614556816,
+      "loss": 2.1644,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7799174690508941,
+      "grad_norm": 0.461120162903064,
+      "learning_rate": 0.00012179379711709738,
+      "loss": 2.1295,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7803759743237048,
+      "grad_norm": 0.4511990385475929,
+      "learning_rate": 0.0001213084225525577,
+      "loss": 2.3841,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7808344795965153,
+      "grad_norm": 0.47198057524211773,
+      "learning_rate": 0.00012082388352286627,
+      "loss": 2.2378,
+      "step": 1703
+    },
+    {
+      "epoch": 0.781292984869326,
+      "grad_norm": 0.4500513147745038,
+      "learning_rate": 0.00012034018109709716,
+      "loss": 2.2656,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7817514901421366,
+      "grad_norm": 0.4574819845082376,
+      "learning_rate": 0.00011985731634247809,
+      "loss": 2.3103,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7822099954149473,
+      "grad_norm": 0.5048553982307074,
+      "learning_rate": 0.00011937529032438904,
+      "loss": 2.258,
+      "step": 1706
+    },
+    {
+      "epoch": 0.782668500687758,
+      "grad_norm": 0.47830316439630766,
+      "learning_rate": 0.00011889410410635887,
+      "loss": 2.2449,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7831270059605685,
+      "grad_norm": 0.46186161310546925,
+      "learning_rate": 0.0001184137587500641,
+      "loss": 2.2418,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7835855112333792,
+      "grad_norm": 0.5077258480858357,
+      "learning_rate": 0.00011793425531532564,
+      "loss": 2.2067,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7840440165061898,
+      "grad_norm": 0.5249597316243132,
+      "learning_rate": 0.00011745559486010671,
+      "loss": 2.3484,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7845025217790005,
+      "grad_norm": 0.45602325894276846,
+      "learning_rate": 0.00011697777844051105,
+      "loss": 2.2166,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7849610270518111,
+      "grad_norm": 0.458726989454394,
+      "learning_rate": 0.00011650080711077964,
+      "loss": 2.2358,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7854195323246217,
+      "grad_norm": 0.480482107433286,
+      "learning_rate": 0.00011602468192328936,
+      "loss": 2.287,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7858780375974324,
+      "grad_norm": 0.5256685163290534,
+      "learning_rate": 0.00011554940392854973,
+      "loss": 2.2445,
+      "step": 1714
+    },
+    {
+      "epoch": 0.786336542870243,
+      "grad_norm": 0.5362224693887908,
+      "learning_rate": 0.00011507497417520146,
+      "loss": 2.3039,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7867950481430537,
+      "grad_norm": 0.5124119111569784,
+      "learning_rate": 0.00011460139371001339,
+      "loss": 2.2763,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7872535534158643,
+      "grad_norm": 0.5562316488000781,
+      "learning_rate": 0.00011412866357788049,
+      "loss": 2.258,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7877120586886749,
+      "grad_norm": 0.47979704385302063,
+      "learning_rate": 0.00011365678482182207,
+      "loss": 2.2513,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7881705639614855,
+      "grad_norm": 0.4960090233275527,
+      "learning_rate": 0.0001131857584829783,
+      "loss": 2.2605,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7886290692342962,
+      "grad_norm": 0.5191313422986287,
+      "learning_rate": 0.0001127155856006093,
+      "loss": 2.2073,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7890875745071069,
+      "grad_norm": 0.4591399951148828,
+      "learning_rate": 0.00011224626721209141,
+      "loss": 2.2631,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7895460797799174,
+      "grad_norm": 0.49187576761975005,
+      "learning_rate": 0.0001117778043529164,
+      "loss": 2.278,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7900045850527281,
+      "grad_norm": 0.5343912179163182,
+      "learning_rate": 0.0001113101980566879,
+      "loss": 2.2639,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7904630903255387,
+      "grad_norm": 0.46232445059783345,
+      "learning_rate": 0.00011084344935511958,
+      "loss": 2.2179,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7909215955983494,
+      "grad_norm": 0.47283903072449923,
+      "learning_rate": 0.00011037755927803345,
+      "loss": 2.2606,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7913801008711601,
+      "grad_norm": 0.4777694154610554,
+      "learning_rate": 0.00010991252885335651,
+      "loss": 2.2466,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7918386061439706,
+      "grad_norm": 0.5983240514062375,
+      "learning_rate": 0.00010944835910711958,
+      "loss": 2.1915,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7922971114167813,
+      "grad_norm": 0.5301434393124634,
+      "learning_rate": 0.00010898505106345396,
+      "loss": 2.296,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7927556166895919,
+      "grad_norm": 0.5081890104241117,
+      "learning_rate": 0.00010852260574459022,
+      "loss": 2.2186,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7932141219624026,
+      "grad_norm": 0.5103682091100672,
+      "learning_rate": 0.00010806102417085512,
+      "loss": 2.2906,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7936726272352133,
+      "grad_norm": 0.4781885268658402,
+      "learning_rate": 0.00010760030736066951,
+      "loss": 2.2837,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7941311325080238,
+      "grad_norm": 0.49266605892172516,
+      "learning_rate": 0.00010714045633054687,
+      "loss": 2.2842,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7945896377808345,
+      "grad_norm": 0.48325515029865623,
+      "learning_rate": 0.00010668147209508971,
+      "loss": 2.1881,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7950481430536451,
+      "grad_norm": 0.5004911790159121,
+      "learning_rate": 0.00010622335566698877,
+      "loss": 2.3588,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7955066483264558,
+      "grad_norm": 0.47723679530068125,
+      "learning_rate": 0.00010576610805701942,
+      "loss": 2.2838,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7959651535992663,
+      "grad_norm": 0.4953203499259206,
+      "learning_rate": 0.00010530973027404073,
+      "loss": 2.2388,
+      "step": 1736
+    },
+    {
+      "epoch": 0.796423658872077,
+      "grad_norm": 0.5421264477849382,
+      "learning_rate": 0.00010485422332499212,
+      "loss": 2.2695,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7968821641448877,
+      "grad_norm": 0.47581955554105876,
+      "learning_rate": 0.00010439958821489165,
+      "loss": 2.2706,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7973406694176983,
+      "grad_norm": 0.4884791018657579,
+      "learning_rate": 0.00010394582594683428,
+      "loss": 2.3249,
+      "step": 1739
+    },
+    {
+      "epoch": 0.797799174690509,
+      "grad_norm": 0.4757614763990529,
+      "learning_rate": 0.0001034929375219884,
+      "loss": 2.1998,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7982576799633195,
+      "grad_norm": 0.5095031633999362,
+      "learning_rate": 0.00010304092393959514,
+      "loss": 2.1933,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7987161852361302,
+      "grad_norm": 0.4682900955434661,
+      "learning_rate": 0.00010258978619696468,
+      "loss": 2.2495,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7991746905089409,
+      "grad_norm": 0.505583683624658,
+      "learning_rate": 0.00010213952528947551,
+      "loss": 2.2984,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7996331957817515,
+      "grad_norm": 0.515973394157164,
+      "learning_rate": 0.00010169014221057089,
+      "loss": 2.2349,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8000917010545622,
+      "grad_norm": 0.48498925217798494,
+      "learning_rate": 0.00010124163795175734,
+      "loss": 2.2352,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8005502063273727,
+      "grad_norm": 0.4692787389650759,
+      "learning_rate": 0.00010079401350260287,
+      "loss": 2.2767,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8010087116001834,
+      "grad_norm": 0.46323486434338446,
+      "learning_rate": 0.00010034726985073362,
+      "loss": 2.2926,
+      "step": 1747
+    },
+    {
+      "epoch": 0.801467216872994,
+      "grad_norm": 0.581598361097335,
+      "learning_rate": 9.9901407981833e-05,
+      "loss": 2.2778,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8019257221458047,
+      "grad_norm": 0.45735458506992915,
+      "learning_rate": 9.94564288796384e-05,
+      "loss": 2.1924,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8023842274186153,
+      "grad_norm": 0.4739910949556759,
+      "learning_rate": 9.901233352593953e-05,
+      "loss": 2.2783,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8028427326914259,
+      "grad_norm": 0.47641382734451604,
+      "learning_rate": 9.856912290057668e-05,
+      "loss": 2.1936,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8033012379642366,
+      "grad_norm": 0.4810933239491401,
+      "learning_rate": 9.812679798143748e-05,
+      "loss": 2.251,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8037597432370472,
+      "grad_norm": 0.4933820568444858,
+      "learning_rate": 9.768535974445586e-05,
+      "loss": 2.2852,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8042182485098579,
+      "grad_norm": 0.5112446668688165,
+      "learning_rate": 9.724480916360906e-05,
+      "loss": 2.2615,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8046767537826685,
+      "grad_norm": 0.5455792008425354,
+      "learning_rate": 9.68051472109162e-05,
+      "loss": 2.3911,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8051352590554791,
+      "grad_norm": 0.5173536416640167,
+      "learning_rate": 9.636637485643529e-05,
+      "loss": 2.2298,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8055937643282898,
+      "grad_norm": 0.4922923766807779,
+      "learning_rate": 9.592849306826174e-05,
+      "loss": 2.2153,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8060522696011004,
+      "grad_norm": 0.49585359098450205,
+      "learning_rate": 9.549150281252633e-05,
+      "loss": 2.3699,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8065107748739111,
+      "grad_norm": 0.4855256314159026,
+      "learning_rate": 9.505540505339223e-05,
+      "loss": 2.18,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8069692801467216,
+      "grad_norm": 0.4982535758431046,
+      "learning_rate": 9.4620200753054e-05,
+      "loss": 2.2389,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8074277854195323,
+      "grad_norm": 0.49760774696263865,
+      "learning_rate": 9.418589087173441e-05,
+      "loss": 2.2347,
+      "step": 1761
+    },
+    {
+      "epoch": 0.807886290692343,
+      "grad_norm": 0.4709347898275405,
+      "learning_rate": 9.375247636768325e-05,
+      "loss": 2.1712,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8083447959651536,
+      "grad_norm": 0.5087801616525308,
+      "learning_rate": 9.331995819717443e-05,
+      "loss": 2.296,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8088033012379643,
+      "grad_norm": 0.5203519354265439,
+      "learning_rate": 9.288833731450419e-05,
+      "loss": 2.244,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8092618065107748,
+      "grad_norm": 0.47916496664901914,
+      "learning_rate": 9.245761467198948e-05,
+      "loss": 2.2343,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8097203117835855,
+      "grad_norm": 0.5049754397666188,
+      "learning_rate": 9.20277912199648e-05,
+      "loss": 2.1416,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8101788170563962,
+      "grad_norm": 0.47724385159045607,
+      "learning_rate": 9.159886790678123e-05,
+      "loss": 2.248,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8106373223292068,
+      "grad_norm": 0.462437288123017,
+      "learning_rate": 9.11708456788033e-05,
+      "loss": 2.1825,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8110958276020174,
+      "grad_norm": 0.4600461073095763,
+      "learning_rate": 9.074372548040793e-05,
+      "loss": 2.1583,
+      "step": 1769
+    },
+    {
+      "epoch": 0.811554332874828,
+      "grad_norm": 0.4753292390176522,
+      "learning_rate": 9.031750825398145e-05,
+      "loss": 2.2602,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8120128381476387,
+      "grad_norm": 0.47288552215570895,
+      "learning_rate": 8.98921949399179e-05,
+      "loss": 2.3017,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8124713434204494,
+      "grad_norm": 0.5137775180154532,
+      "learning_rate": 8.94677864766173e-05,
+      "loss": 2.2344,
+      "step": 1772
+    },
+    {
+      "epoch": 0.81292984869326,
+      "grad_norm": 0.4907603470006631,
+      "learning_rate": 8.904428380048269e-05,
+      "loss": 2.2735,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8133883539660706,
+      "grad_norm": 0.5126444491632961,
+      "learning_rate": 8.862168784591929e-05,
+      "loss": 2.2564,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8138468592388812,
+      "grad_norm": 0.48954172924088435,
+      "learning_rate": 8.819999954533115e-05,
+      "loss": 2.2352,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8143053645116919,
+      "grad_norm": 0.4690432887659008,
+      "learning_rate": 8.777921982911996e-05,
+      "loss": 2.1584,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8147638697845025,
+      "grad_norm": 0.5107039512276659,
+      "learning_rate": 8.735934962568253e-05,
+      "loss": 2.2434,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8152223750573132,
+      "grad_norm": 0.47822459966176245,
+      "learning_rate": 8.694038986140945e-05,
+      "loss": 2.2716,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8156808803301238,
+      "grad_norm": 0.49826931672555286,
+      "learning_rate": 8.652234146068206e-05,
+      "loss": 2.2787,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8161393856029344,
+      "grad_norm": 0.47026342948560357,
+      "learning_rate": 8.610520534587086e-05,
+      "loss": 2.1821,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8165978908757451,
+      "grad_norm": 0.48420887163203735,
+      "learning_rate": 8.568898243733397e-05,
+      "loss": 2.2316,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8170563961485557,
+      "grad_norm": 0.510220714794586,
+      "learning_rate": 8.527367365341409e-05,
+      "loss": 2.2797,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8175149014213663,
+      "grad_norm": 0.5363533646853594,
+      "learning_rate": 8.485927991043757e-05,
+      "loss": 2.3293,
+      "step": 1783
+    },
+    {
+      "epoch": 0.817973406694177,
+      "grad_norm": 0.5100471734863302,
+      "learning_rate": 8.444580212271125e-05,
+      "loss": 2.3055,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8184319119669876,
+      "grad_norm": 0.46589698163056376,
+      "learning_rate": 8.403324120252159e-05,
+      "loss": 2.2926,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8188904172397983,
+      "grad_norm": 0.5426322337801304,
+      "learning_rate": 8.362159806013175e-05,
+      "loss": 2.2351,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8193489225126089,
+      "grad_norm": 0.4772221646379862,
+      "learning_rate": 8.321087360377988e-05,
+      "loss": 2.2115,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8198074277854195,
+      "grad_norm": 0.5125833059759716,
+      "learning_rate": 8.280106873967752e-05,
+      "loss": 2.3452,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8202659330582301,
+      "grad_norm": 0.4711789579180109,
+      "learning_rate": 8.239218437200679e-05,
+      "loss": 2.273,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8207244383310408,
+      "grad_norm": 0.4869707186044968,
+      "learning_rate": 8.198422140291939e-05,
+      "loss": 2.2293,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8211829436038515,
+      "grad_norm": 0.5096421437044084,
+      "learning_rate": 8.157718073253351e-05,
+      "loss": 2.3058,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8216414488766621,
+      "grad_norm": 0.4858157235776389,
+      "learning_rate": 8.117106325893287e-05,
+      "loss": 2.2243,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8220999541494727,
+      "grad_norm": 0.5287324543697557,
+      "learning_rate": 8.076586987816404e-05,
+      "loss": 2.2217,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8225584594222833,
+      "grad_norm": 0.4912277634963918,
+      "learning_rate": 8.036160148423449e-05,
+      "loss": 2.239,
+      "step": 1794
+    },
+    {
+      "epoch": 0.823016964695094,
+      "grad_norm": 0.49005630881117496,
+      "learning_rate": 7.995825896911141e-05,
+      "loss": 2.2808,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8234754699679047,
+      "grad_norm": 0.46298457564923,
+      "learning_rate": 7.955584322271853e-05,
+      "loss": 2.2358,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8239339752407153,
+      "grad_norm": 0.5091162131341147,
+      "learning_rate": 7.915435513293523e-05,
+      "loss": 2.2704,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8243924805135259,
+      "grad_norm": 0.5138198125956981,
+      "learning_rate": 7.875379558559387e-05,
+      "loss": 2.1027,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8248509857863365,
+      "grad_norm": 0.46558985379983775,
+      "learning_rate": 7.835416546447838e-05,
+      "loss": 2.2293,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8253094910591472,
+      "grad_norm": 0.480567719134681,
+      "learning_rate": 7.795546565132167e-05,
+      "loss": 2.2522,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8257679963319579,
+      "grad_norm": 0.5831062001846804,
+      "learning_rate": 7.755769702580412e-05,
+      "loss": 2.2654,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8262265016047684,
+      "grad_norm": 0.49740202029338987,
+      "learning_rate": 7.716086046555193e-05,
+      "loss": 2.2279,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8266850068775791,
+      "grad_norm": 0.486761403898445,
+      "learning_rate": 7.676495684613432e-05,
+      "loss": 2.2866,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8271435121503897,
+      "grad_norm": 0.45633144285064153,
+      "learning_rate": 7.636998704106252e-05,
+      "loss": 2.1892,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8276020174232004,
+      "grad_norm": 0.47620978222612903,
+      "learning_rate": 7.597595192178702e-05,
+      "loss": 2.2177,
+      "step": 1805
+    },
+    {
+      "epoch": 0.828060522696011,
+      "grad_norm": 0.4923931838732233,
+      "learning_rate": 7.558285235769646e-05,
+      "loss": 2.2246,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8285190279688216,
+      "grad_norm": 0.4881270668263674,
+      "learning_rate": 7.519068921611494e-05,
+      "loss": 2.278,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8289775332416323,
+      "grad_norm": 0.49581051059081943,
+      "learning_rate": 7.479946336230047e-05,
+      "loss": 2.242,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8294360385144429,
+      "grad_norm": 0.5057066661375778,
+      "learning_rate": 7.440917565944349e-05,
+      "loss": 2.281,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8298945437872536,
+      "grad_norm": 0.4872343073687879,
+      "learning_rate": 7.4019826968664e-05,
+      "loss": 2.2077,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8303530490600642,
+      "grad_norm": 0.4741174520211262,
+      "learning_rate": 7.363141814901053e-05,
+      "loss": 2.1969,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8308115543328748,
+      "grad_norm": 0.4768688799262972,
+      "learning_rate": 7.32439500574577e-05,
+      "loss": 2.2504,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8312700596056855,
+      "grad_norm": 0.4845846142228576,
+      "learning_rate": 7.285742354890473e-05,
+      "loss": 2.1885,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8317285648784961,
+      "grad_norm": 0.49091759956388475,
+      "learning_rate": 7.247183947617325e-05,
+      "loss": 2.2266,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8321870701513068,
+      "grad_norm": 0.4799323330692874,
+      "learning_rate": 7.20871986900053e-05,
+      "loss": 2.2345,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8326455754241173,
+      "grad_norm": 0.49228018903089765,
+      "learning_rate": 7.170350203906218e-05,
+      "loss": 2.2092,
+      "step": 1816
+    },
+    {
+      "epoch": 0.833104080696928,
+      "grad_norm": 0.5080521923403886,
+      "learning_rate": 7.132075036992158e-05,
+      "loss": 2.2175,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8335625859697386,
+      "grad_norm": 0.6096880873410149,
+      "learning_rate": 7.093894452707666e-05,
+      "loss": 2.1855,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8340210912425493,
+      "grad_norm": 0.4884336064584524,
+      "learning_rate": 7.055808535293334e-05,
+      "loss": 2.1542,
+      "step": 1819
+    },
+    {
+      "epoch": 0.83447959651536,
+      "grad_norm": 0.6030120073522733,
+      "learning_rate": 7.017817368780888e-05,
+      "loss": 2.2767,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8349381017881705,
+      "grad_norm": 0.49439122784476597,
+      "learning_rate": 6.979921036993042e-05,
+      "loss": 2.2243,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8353966070609812,
+      "grad_norm": 0.5191507673869431,
+      "learning_rate": 6.942119623543202e-05,
+      "loss": 2.3236,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8358551123337918,
+      "grad_norm": 0.48500207706053816,
+      "learning_rate": 6.904413211835414e-05,
+      "loss": 2.2441,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8363136176066025,
+      "grad_norm": 0.503645337972468,
+      "learning_rate": 6.866801885064056e-05,
+      "loss": 2.2668,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8367721228794132,
+      "grad_norm": 0.5110016634608799,
+      "learning_rate": 6.829285726213769e-05,
+      "loss": 2.2248,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8372306281522237,
+      "grad_norm": 0.529141207225735,
+      "learning_rate": 6.79186481805918e-05,
+      "loss": 2.3263,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8376891334250344,
+      "grad_norm": 0.5170618129351389,
+      "learning_rate": 6.754539243164754e-05,
+      "loss": 2.2312,
+      "step": 1827
+    },
+    {
+      "epoch": 0.838147638697845,
+      "grad_norm": 0.4741888066662591,
+      "learning_rate": 6.717309083884654e-05,
+      "loss": 2.2711,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8386061439706557,
+      "grad_norm": 0.4802200842490468,
+      "learning_rate": 6.680174422362468e-05,
+      "loss": 2.1745,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8390646492434664,
+      "grad_norm": 0.500268118164079,
+      "learning_rate": 6.643135340531136e-05,
+      "loss": 2.2615,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8395231545162769,
+      "grad_norm": 0.500559695196355,
+      "learning_rate": 6.606191920112664e-05,
+      "loss": 2.3467,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8399816597890876,
+      "grad_norm": 0.5962538803603356,
+      "learning_rate": 6.569344242618036e-05,
+      "loss": 2.2061,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8404401650618982,
+      "grad_norm": 0.5062443550591345,
+      "learning_rate": 6.532592389346958e-05,
+      "loss": 2.3479,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8408986703347089,
+      "grad_norm": 0.4636079118818432,
+      "learning_rate": 6.495936441387713e-05,
+      "loss": 2.1465,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8413571756075194,
+      "grad_norm": 0.5328406306483462,
+      "learning_rate": 6.459376479617013e-05,
+      "loss": 2.219,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8418156808803301,
+      "grad_norm": 0.46572827683534923,
+      "learning_rate": 6.422912584699752e-05,
+      "loss": 2.2701,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8422741861531408,
+      "grad_norm": 0.5003381666824435,
+      "learning_rate": 6.386544837088904e-05,
+      "loss": 2.1906,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8427326914259514,
+      "grad_norm": 0.4942057942588186,
+      "learning_rate": 6.350273317025251e-05,
+      "loss": 2.1341,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8431911966987621,
+      "grad_norm": 0.5132440049177521,
+      "learning_rate": 6.314098104537324e-05,
+      "loss": 2.2452,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8436497019715726,
+      "grad_norm": 0.47396857856237573,
+      "learning_rate": 6.278019279441122e-05,
+      "loss": 2.237,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8441082072443833,
+      "grad_norm": 0.4722601715231321,
+      "learning_rate": 6.242036921339972e-05,
+      "loss": 2.2325,
+      "step": 1841
+    },
+    {
+      "epoch": 0.844566712517194,
+      "grad_norm": 0.4898049522660517,
+      "learning_rate": 6.206151109624402e-05,
+      "loss": 2.2279,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8450252177900046,
+      "grad_norm": 0.4974287397559667,
+      "learning_rate": 6.170361923471868e-05,
+      "loss": 2.1892,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8454837230628153,
+      "grad_norm": 0.47872582874109265,
+      "learning_rate": 6.134669441846691e-05,
+      "loss": 2.2317,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8459422283356258,
+      "grad_norm": 0.48848927203253917,
+      "learning_rate": 6.099073743499772e-05,
+      "loss": 2.258,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8464007336084365,
+      "grad_norm": 0.4759199200520246,
+      "learning_rate": 6.063574906968511e-05,
+      "loss": 2.2102,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8468592388812471,
+      "grad_norm": 0.664723385622916,
+      "learning_rate": 6.028173010576582e-05,
+      "loss": 2.1719,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8473177441540578,
+      "grad_norm": 0.49405786043560773,
+      "learning_rate": 5.9928681324337544e-05,
+      "loss": 2.2585,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8477762494268684,
+      "grad_norm": 0.4741050472451257,
+      "learning_rate": 5.957660350435773e-05,
+      "loss": 2.19,
+      "step": 1849
+    },
+    {
+      "epoch": 0.848234754699679,
+      "grad_norm": 0.49567035428822714,
+      "learning_rate": 5.922549742264122e-05,
+      "loss": 2.286,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8486932599724897,
+      "grad_norm": 0.47999800832412504,
+      "learning_rate": 5.8875363853859166e-05,
+      "loss": 2.1974,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8491517652453003,
+      "grad_norm": 0.47844592190100393,
+      "learning_rate": 5.852620357053651e-05,
+      "loss": 2.2584,
+      "step": 1852
+    },
+    {
+      "epoch": 0.849610270518111,
+      "grad_norm": 0.5217513252131173,
+      "learning_rate": 5.8178017343051336e-05,
+      "loss": 2.1571,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8500687757909215,
+      "grad_norm": 0.46451760506482204,
+      "learning_rate": 5.783080593963219e-05,
+      "loss": 2.2294,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8505272810637322,
+      "grad_norm": 0.5064522167166243,
+      "learning_rate": 5.748457012635683e-05,
+      "loss": 2.2304,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8509857863365429,
+      "grad_norm": 0.4538097414259151,
+      "learning_rate": 5.713931066715078e-05,
+      "loss": 2.1187,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8514442916093535,
+      "grad_norm": 0.5023490264294365,
+      "learning_rate": 5.679502832378497e-05,
+      "loss": 2.322,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8519027968821642,
+      "grad_norm": 0.4808820157682433,
+      "learning_rate": 5.645172385587482e-05,
+      "loss": 2.187,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8523613021549747,
+      "grad_norm": 0.4488504893194128,
+      "learning_rate": 5.6109398020877834e-05,
+      "loss": 2.1611,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8528198074277854,
+      "grad_norm": 0.5455640340560308,
+      "learning_rate": 5.576805157409265e-05,
+      "loss": 2.1997,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8532783127005961,
+      "grad_norm": 0.494086725797245,
+      "learning_rate": 5.542768526865677e-05,
+      "loss": 2.1796,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8537368179734067,
+      "grad_norm": 0.5298454694212393,
+      "learning_rate": 5.508829985554509e-05,
+      "loss": 2.3289,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8541953232462174,
+      "grad_norm": 0.49601700543038646,
+      "learning_rate": 5.474989608356856e-05,
+      "loss": 2.3063,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8546538285190279,
+      "grad_norm": 0.4677355221830177,
+      "learning_rate": 5.441247469937194e-05,
+      "loss": 2.1962,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8551123337918386,
+      "grad_norm": 0.45684253966945626,
+      "learning_rate": 5.407603644743286e-05,
+      "loss": 2.2125,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8555708390646493,
+      "grad_norm": 0.4659721872423064,
+      "learning_rate": 5.374058207005944e-05,
+      "loss": 2.1169,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8560293443374599,
+      "grad_norm": 0.48956347076578666,
+      "learning_rate": 5.3406112307389066e-05,
+      "loss": 2.2057,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8564878496102705,
+      "grad_norm": 0.4791507341585217,
+      "learning_rate": 5.3072627897386926e-05,
+      "loss": 2.2686,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8569463548830811,
+      "grad_norm": 0.474574973901403,
+      "learning_rate": 5.27401295758439e-05,
+      "loss": 2.2369,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8574048601558918,
+      "grad_norm": 0.5139634857996955,
+      "learning_rate": 5.2408618076375315e-05,
+      "loss": 2.3536,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8578633654287025,
+      "grad_norm": 0.4530570795275264,
+      "learning_rate": 5.207809413041914e-05,
+      "loss": 2.3028,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8583218707015131,
+      "grad_norm": 0.4876753895186355,
+      "learning_rate": 5.174855846723459e-05,
+      "loss": 2.2355,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8587803759743237,
+      "grad_norm": 0.5287823375548467,
+      "learning_rate": 5.1420011813900104e-05,
+      "loss": 2.2436,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8592388812471343,
+      "grad_norm": 0.4792614444029872,
+      "learning_rate": 5.109245489531211e-05,
+      "loss": 2.2253,
+      "step": 1874
+    },
+    {
+      "epoch": 0.859697386519945,
+      "grad_norm": 0.47198936870185176,
+      "learning_rate": 5.0765888434183446e-05,
+      "loss": 2.1351,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8601558917927556,
+      "grad_norm": 0.5017809123666681,
+      "learning_rate": 5.0440313151041364e-05,
+      "loss": 2.289,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8606143970655663,
+      "grad_norm": 0.46578401748042375,
+      "learning_rate": 5.011572976422657e-05,
+      "loss": 2.1541,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8610729023383769,
+      "grad_norm": 0.46085824464376274,
+      "learning_rate": 4.9792138989890825e-05,
+      "loss": 2.2448,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8615314076111875,
+      "grad_norm": 0.634019517916116,
+      "learning_rate": 4.9469541541996234e-05,
+      "loss": 2.2217,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8619899128839982,
+      "grad_norm": 0.4822772129872547,
+      "learning_rate": 4.914793813231305e-05,
+      "loss": 2.2641,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8624484181568088,
+      "grad_norm": 0.489403685554443,
+      "learning_rate": 4.882732947041818e-05,
+      "loss": 2.2752,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8629069234296194,
+      "grad_norm": 0.45959756819921815,
+      "learning_rate": 4.850771626369416e-05,
+      "loss": 2.2705,
+      "step": 1882
+    },
+    {
+      "epoch": 0.86336542870243,
+      "grad_norm": 0.5015020210447435,
+      "learning_rate": 4.818909921732662e-05,
+      "loss": 2.2704,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8638239339752407,
+      "grad_norm": 0.4904793258202689,
+      "learning_rate": 4.787147903430383e-05,
+      "loss": 2.2244,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8642824392480514,
+      "grad_norm": 0.47235929602991766,
+      "learning_rate": 4.755485641541424e-05,
+      "loss": 2.2234,
+      "step": 1885
+    },
+    {
+      "epoch": 0.864740944520862,
+      "grad_norm": 0.47555935528858884,
+      "learning_rate": 4.723923205924557e-05,
+      "loss": 2.2023,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8651994497936726,
+      "grad_norm": 0.4433283373549399,
+      "learning_rate": 4.6924606662182736e-05,
+      "loss": 2.2856,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8656579550664832,
+      "grad_norm": 0.47789684120134784,
+      "learning_rate": 4.6610980918406596e-05,
+      "loss": 2.2024,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8661164603392939,
+      "grad_norm": 0.4836629760230387,
+      "learning_rate": 4.629835551989276e-05,
+      "loss": 2.2812,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8665749656121046,
+      "grad_norm": 0.4748349510967027,
+      "learning_rate": 4.5986731156409224e-05,
+      "loss": 2.2229,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8670334708849152,
+      "grad_norm": 0.4744694806778088,
+      "learning_rate": 4.567610851551568e-05,
+      "loss": 2.342,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8674919761577258,
+      "grad_norm": 0.5231028220154214,
+      "learning_rate": 4.536648828256146e-05,
+      "loss": 2.2541,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8679504814305364,
+      "grad_norm": 0.5896781636067812,
+      "learning_rate": 4.505787114068433e-05,
+      "loss": 2.2303,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8684089867033471,
+      "grad_norm": 0.47516580501614214,
+      "learning_rate": 4.4750257770808764e-05,
+      "loss": 2.3237,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8688674919761578,
+      "grad_norm": 0.46341529769616446,
+      "learning_rate": 4.444364885164448e-05,
+      "loss": 2.2136,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8693259972489684,
+      "grad_norm": 0.4843345253131042,
+      "learning_rate": 4.413804505968533e-05,
+      "loss": 2.2459,
+      "step": 1896
+    },
+    {
+      "epoch": 0.869784502521779,
+      "grad_norm": 0.5652856392110602,
+      "learning_rate": 4.3833447069206944e-05,
+      "loss": 2.3274,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8702430077945896,
+      "grad_norm": 0.4575113907225181,
+      "learning_rate": 4.352985555226635e-05,
+      "loss": 2.3086,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8707015130674003,
+      "grad_norm": 0.47134611290442835,
+      "learning_rate": 4.322727117869951e-05,
+      "loss": 2.2126,
+      "step": 1899
+    },
+    {
+      "epoch": 0.871160018340211,
+      "grad_norm": 0.4692229640328368,
+      "learning_rate": 4.29256946161205e-05,
+      "loss": 2.1866,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8716185236130215,
+      "grad_norm": 0.46135568023204165,
+      "learning_rate": 4.262512652991968e-05,
+      "loss": 2.1823,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8720770288858322,
+      "grad_norm": 0.46033843825017423,
+      "learning_rate": 4.2325567583262113e-05,
+      "loss": 2.1876,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8725355341586428,
+      "grad_norm": 0.5450658548187013,
+      "learning_rate": 4.2027018437086895e-05,
+      "loss": 2.1879,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8729940394314535,
+      "grad_norm": 0.5066801319319982,
+      "learning_rate": 4.172947975010449e-05,
+      "loss": 2.1453,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8734525447042641,
+      "grad_norm": 0.5264043750613489,
+      "learning_rate": 4.143295217879645e-05,
+      "loss": 2.3426,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8739110499770747,
+      "grad_norm": 0.5011234186712479,
+      "learning_rate": 4.113743637741296e-05,
+      "loss": 2.2602,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8743695552498854,
+      "grad_norm": 0.5162565946774744,
+      "learning_rate": 4.084293299797226e-05,
+      "loss": 2.2826,
+      "step": 1907
+    },
+    {
+      "epoch": 0.874828060522696,
+      "grad_norm": 0.4916370735179604,
+      "learning_rate": 4.054944269025862e-05,
+      "loss": 2.1767,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8752865657955067,
+      "grad_norm": 0.5202655302164014,
+      "learning_rate": 4.025696610182095e-05,
+      "loss": 2.1591,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8757450710683173,
+      "grad_norm": 0.5111877545162299,
+      "learning_rate": 3.996550387797187e-05,
+      "loss": 2.1952,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8762035763411279,
+      "grad_norm": 0.48774871718527696,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 2.1456,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8766620816139385,
+      "grad_norm": 0.47251574404699914,
+      "learning_rate": 3.9385625094097154e-05,
+      "loss": 2.2654,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8771205868867492,
+      "grad_norm": 0.4835520213758781,
+      "learning_rate": 3.909720981350034e-05,
+      "loss": 2.1493,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8775790921595599,
+      "grad_norm": 0.4913503184141984,
+      "learning_rate": 3.880981145634704e-05,
+      "loss": 2.3353,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8780375974323704,
+      "grad_norm": 0.48104113325154507,
+      "learning_rate": 3.852343065674507e-05,
+      "loss": 2.2824,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8784961027051811,
+      "grad_norm": 0.47087376010029214,
+      "learning_rate": 3.8238068046557276e-05,
+      "loss": 2.1877,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8789546079779917,
+      "grad_norm": 0.4943960843144546,
+      "learning_rate": 3.795372425540006e-05,
+      "loss": 2.1839,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8794131132508024,
+      "grad_norm": 0.5018999518254544,
+      "learning_rate": 3.76703999106418e-05,
+      "loss": 2.1684,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8798716185236131,
+      "grad_norm": 0.45881820483540325,
+      "learning_rate": 3.7388095637401754e-05,
+      "loss": 2.1592,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8803301237964236,
+      "grad_norm": 0.44872774158599626,
+      "learning_rate": 3.7106812058548376e-05,
+      "loss": 2.1737,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8807886290692343,
+      "grad_norm": 0.48348785606509914,
+      "learning_rate": 3.682654979469807e-05,
+      "loss": 2.2175,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8812471343420449,
+      "grad_norm": 0.4726471046625043,
+      "learning_rate": 3.654730946421403e-05,
+      "loss": 2.2612,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8817056396148556,
+      "grad_norm": 0.48249031630640105,
+      "learning_rate": 3.6269091683204466e-05,
+      "loss": 2.2045,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8821641448876663,
+      "grad_norm": 0.51276260320396,
+      "learning_rate": 3.5991897065521693e-05,
+      "loss": 2.3204,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8826226501604768,
+      "grad_norm": 0.4915488244179048,
+      "learning_rate": 3.571572622276026e-05,
+      "loss": 2.2983,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8830811554332875,
+      "grad_norm": 0.49820624903311367,
+      "learning_rate": 3.544057976425619e-05,
+      "loss": 2.3476,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8835396607060981,
+      "grad_norm": 0.47263268981490736,
+      "learning_rate": 3.5166458297085146e-05,
+      "loss": 2.3298,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8839981659789088,
+      "grad_norm": 0.4859511712733128,
+      "learning_rate": 3.489336242606111e-05,
+      "loss": 2.2579,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8844566712517193,
+      "grad_norm": 0.46860683186987234,
+      "learning_rate": 3.462129275373577e-05,
+      "loss": 2.1881,
+      "step": 1929
+    },
+    {
+      "epoch": 0.88491517652453,
+      "grad_norm": 0.46389145561783324,
+      "learning_rate": 3.4350249880395924e-05,
+      "loss": 2.2464,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8853736817973407,
+      "grad_norm": 0.47165148193622036,
+      "learning_rate": 3.408023440406355e-05,
+      "loss": 2.1898,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8858321870701513,
+      "grad_norm": 0.5707964048552328,
+      "learning_rate": 3.381124692049331e-05,
+      "loss": 2.2761,
+      "step": 1932
+    },
+    {
+      "epoch": 0.886290692342962,
+      "grad_norm": 0.4888024413140837,
+      "learning_rate": 3.354328802317197e-05,
+      "loss": 2.2062,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8867491976157725,
+      "grad_norm": 0.47773427364285126,
+      "learning_rate": 3.327635830331677e-05,
+      "loss": 2.199,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8872077028885832,
+      "grad_norm": 0.49438608169380494,
+      "learning_rate": 3.3010458349874206e-05,
+      "loss": 2.2488,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8876662081613939,
+      "grad_norm": 0.5051920082842116,
+      "learning_rate": 3.2745588749518775e-05,
+      "loss": 2.305,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8881247134342045,
+      "grad_norm": 0.4754989369916751,
+      "learning_rate": 3.248175008665161e-05,
+      "loss": 2.1691,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8885832187070152,
+      "grad_norm": 0.48565673912022456,
+      "learning_rate": 3.221894294339911e-05,
+      "loss": 2.2365,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8890417239798257,
+      "grad_norm": 0.4806038775037684,
+      "learning_rate": 3.1957167899611836e-05,
+      "loss": 2.1984,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8895002292526364,
+      "grad_norm": 0.4624645539744969,
+      "learning_rate": 3.169642553286334e-05,
+      "loss": 2.2616,
+      "step": 1940
+    },
+    {
+      "epoch": 0.889958734525447,
+      "grad_norm": 0.4799777023103763,
+      "learning_rate": 3.143671641844831e-05,
+      "loss": 2.2764,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8904172397982577,
+      "grad_norm": 0.5081135712420585,
+      "learning_rate": 3.117804112938205e-05,
+      "loss": 2.2003,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8908757450710684,
+      "grad_norm": 0.48029880309322126,
+      "learning_rate": 3.092040023639869e-05,
+      "loss": 2.2637,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8913342503438789,
+      "grad_norm": 0.4752967543242551,
+      "learning_rate": 3.066379430795002e-05,
+      "loss": 2.2138,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8917927556166896,
+      "grad_norm": 0.47041641959850794,
+      "learning_rate": 3.040822391020459e-05,
+      "loss": 2.2605,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8922512608895002,
+      "grad_norm": 1.1321737502821747,
+      "learning_rate": 3.0153689607045842e-05,
+      "loss": 2.1699,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8927097661623109,
+      "grad_norm": 0.4603648784807076,
+      "learning_rate": 2.9900191960071545e-05,
+      "loss": 2.19,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8931682714351215,
+      "grad_norm": 0.5041696844958067,
+      "learning_rate": 2.9647731528591848e-05,
+      "loss": 2.2374,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8936267767079321,
+      "grad_norm": 0.4899765391923433,
+      "learning_rate": 2.9396308869628795e-05,
+      "loss": 2.246,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8940852819807428,
+      "grad_norm": 0.4699701228695544,
+      "learning_rate": 2.914592453791448e-05,
+      "loss": 2.2432,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8945437872535534,
+      "grad_norm": 0.48167854810069055,
+      "learning_rate": 2.8896579085889994e-05,
+      "loss": 2.2324,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8950022925263641,
+      "grad_norm": 0.4696020647367482,
+      "learning_rate": 2.86482730637046e-05,
+      "loss": 2.2034,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8954607977991746,
+      "grad_norm": 0.4822151713988078,
+      "learning_rate": 2.840100701921383e-05,
+      "loss": 2.2446,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8959193030719853,
+      "grad_norm": 0.468337017539627,
+      "learning_rate": 2.8154781497978898e-05,
+      "loss": 2.2732,
+      "step": 1954
+    },
+    {
+      "epoch": 0.896377808344796,
+      "grad_norm": 0.504925063853735,
+      "learning_rate": 2.7909597043265013e-05,
+      "loss": 2.2233,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8968363136176066,
+      "grad_norm": 0.4792012696118311,
+      "learning_rate": 2.7665454196040662e-05,
+      "loss": 2.2209,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8972948188904173,
+      "grad_norm": 0.4697386684067184,
+      "learning_rate": 2.7422353494975905e-05,
+      "loss": 2.2155,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8977533241632278,
+      "grad_norm": 0.46534564955498786,
+      "learning_rate": 2.7180295476441573e-05,
+      "loss": 2.2187,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8982118294360385,
+      "grad_norm": 0.49489210340979684,
+      "learning_rate": 2.6939280674508016e-05,
+      "loss": 2.22,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8986703347088492,
+      "grad_norm": 0.47837261524984703,
+      "learning_rate": 2.669930962094358e-05,
+      "loss": 2.2233,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8991288399816598,
+      "grad_norm": 0.4775197250940464,
+      "learning_rate": 2.6460382845214126e-05,
+      "loss": 2.315,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8995873452544704,
+      "grad_norm": 0.47059777609856135,
+      "learning_rate": 2.6222500874481025e-05,
+      "loss": 2.3114,
+      "step": 1962
+    },
+    {
+      "epoch": 0.900045850527281,
+      "grad_norm": 0.5061237281253961,
+      "learning_rate": 2.5985664233600827e-05,
+      "loss": 2.2367,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9005043558000917,
+      "grad_norm": 0.5034583255472078,
+      "learning_rate": 2.574987344512336e-05,
+      "loss": 2.2101,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9009628610729024,
+      "grad_norm": 0.4825266791076565,
+      "learning_rate": 2.5515129029290984e-05,
+      "loss": 2.2637,
+      "step": 1965
+    },
+    {
+      "epoch": 0.901421366345713,
+      "grad_norm": 0.4803960092606593,
+      "learning_rate": 2.5281431504037556e-05,
+      "loss": 2.2227,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9018798716185236,
+      "grad_norm": 0.511910223446175,
+      "learning_rate": 2.504878138498684e-05,
+      "loss": 2.235,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9023383768913342,
+      "grad_norm": 0.6039671983103355,
+      "learning_rate": 2.48171791854519e-05,
+      "loss": 2.2455,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9027968821641449,
+      "grad_norm": 0.471494911832845,
+      "learning_rate": 2.4586625416433473e-05,
+      "loss": 2.2702,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9032553874369555,
+      "grad_norm": 0.47226623592009837,
+      "learning_rate": 2.435712058661921e-05,
+      "loss": 2.2597,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9037138927097662,
+      "grad_norm": 0.4926724476318695,
+      "learning_rate": 2.4128665202382327e-05,
+      "loss": 2.1889,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9041723979825768,
+      "grad_norm": 0.4707731624194087,
+      "learning_rate": 2.3901259767780515e-05,
+      "loss": 2.2048,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9046309032553874,
+      "grad_norm": 0.4842931670859483,
+      "learning_rate": 2.367490478455514e-05,
+      "loss": 2.3174,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9050894085281981,
+      "grad_norm": 0.51806510382227,
+      "learning_rate": 2.3449600752129597e-05,
+      "loss": 2.2343,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9055479138010087,
+      "grad_norm": 0.49458020621053383,
+      "learning_rate": 2.3225348167608685e-05,
+      "loss": 2.2162,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9060064190738194,
+      "grad_norm": 0.5005469646010208,
+      "learning_rate": 2.3002147525777118e-05,
+      "loss": 2.3174,
+      "step": 1976
+    },
+    {
+      "epoch": 0.90646492434663,
+      "grad_norm": 0.4924288464978994,
+      "learning_rate": 2.2779999319098856e-05,
+      "loss": 2.1908,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9069234296194406,
+      "grad_norm": 0.5152706985570376,
+      "learning_rate": 2.255890403771571e-05,
+      "loss": 2.2703,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9073819348922513,
+      "grad_norm": 0.4599526370670357,
+      "learning_rate": 2.233886216944614e-05,
+      "loss": 2.2071,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9078404401650619,
+      "grad_norm": 0.47818504416946117,
+      "learning_rate": 2.211987419978484e-05,
+      "loss": 2.3093,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9082989454378725,
+      "grad_norm": 0.4807359771987218,
+      "learning_rate": 2.1901940611900705e-05,
+      "loss": 2.3156,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9087574507106831,
+      "grad_norm": 0.4822390983260801,
+      "learning_rate": 2.168506188663666e-05,
+      "loss": 2.2853,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9092159559834938,
+      "grad_norm": 0.5051829264194683,
+      "learning_rate": 2.1469238502507925e-05,
+      "loss": 2.2577,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9096744612563045,
+      "grad_norm": 0.4504983902187097,
+      "learning_rate": 2.125447093570154e-05,
+      "loss": 2.1911,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9101329665291151,
+      "grad_norm": 0.4766405131632672,
+      "learning_rate": 2.1040759660074793e-05,
+      "loss": 2.2795,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9105914718019257,
+      "grad_norm": 0.5259163557822212,
+      "learning_rate": 2.0828105147154273e-05,
+      "loss": 2.2373,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9110499770747363,
+      "grad_norm": 0.7139793393024482,
+      "learning_rate": 2.061650786613545e-05,
+      "loss": 2.244,
+      "step": 1987
+    },
+    {
+      "epoch": 0.911508482347547,
+      "grad_norm": 0.6171644993461826,
+      "learning_rate": 2.040596828388058e-05,
+      "loss": 2.2617,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9119669876203577,
+      "grad_norm": 0.5238325477348673,
+      "learning_rate": 2.019648686491865e-05,
+      "loss": 2.2009,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9124254928931683,
+      "grad_norm": 0.4944402402200901,
+      "learning_rate": 1.9988064071443767e-05,
+      "loss": 2.1821,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9128839981659789,
+      "grad_norm": 0.5110438394692683,
+      "learning_rate": 1.9780700363314253e-05,
+      "loss": 2.3047,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9133425034387895,
+      "grad_norm": 0.46537610732900225,
+      "learning_rate": 1.957439619805196e-05,
+      "loss": 2.2249,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9138010087116002,
+      "grad_norm": 0.5117797100906453,
+      "learning_rate": 1.9369152030840554e-05,
+      "loss": 2.1537,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9142595139844109,
+      "grad_norm": 0.47085589337727163,
+      "learning_rate": 1.916496831452552e-05,
+      "loss": 2.2582,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9147180192572214,
+      "grad_norm": 0.48692188559818506,
+      "learning_rate": 1.8961845499611998e-05,
+      "loss": 2.1468,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9151765245300321,
+      "grad_norm": 0.4955959779554118,
+      "learning_rate": 1.8759784034264925e-05,
+      "loss": 2.2779,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9156350298028427,
+      "grad_norm": 0.4559796402261251,
+      "learning_rate": 1.855878436430708e-05,
+      "loss": 2.302,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9160935350756534,
+      "grad_norm": 0.7508132134360899,
+      "learning_rate": 1.835884693321871e-05,
+      "loss": 2.2011,
+      "step": 1998
+    },
+    {
+      "epoch": 0.916552040348464,
+      "grad_norm": 0.4809998939749503,
+      "learning_rate": 1.8159972182136386e-05,
+      "loss": 2.3174,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9170105456212746,
+      "grad_norm": 0.49339328693303086,
+      "learning_rate": 1.7962160549851945e-05,
+      "loss": 2.2501,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9174690508940853,
+      "grad_norm": 0.4825901756751267,
+      "learning_rate": 1.776541247281177e-05,
+      "loss": 2.2675,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9179275561668959,
+      "grad_norm": 0.46623026991439587,
+      "learning_rate": 1.7569728385115224e-05,
+      "loss": 2.2038,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9183860614397066,
+      "grad_norm": 0.48092860728472075,
+      "learning_rate": 1.7375108718514665e-05,
+      "loss": 2.2057,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9188445667125172,
+      "grad_norm": 0.4934424476216192,
+      "learning_rate": 1.7181553902413438e-05,
+      "loss": 2.2536,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9193030719853278,
+      "grad_norm": 0.4889234581282474,
+      "learning_rate": 1.698906436386577e-05,
+      "loss": 2.2499,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9197615772581385,
+      "grad_norm": 0.613105162090966,
+      "learning_rate": 1.679764052757532e-05,
+      "loss": 2.3265,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9202200825309491,
+      "grad_norm": 0.47380577006757213,
+      "learning_rate": 1.6607282815894464e-05,
+      "loss": 2.1969,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9206785878037598,
+      "grad_norm": 0.4679321338919368,
+      "learning_rate": 1.6417991648823405e-05,
+      "loss": 2.1868,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9211370930765704,
+      "grad_norm": 0.4827791639967281,
+      "learning_rate": 1.6229767444008835e-05,
+      "loss": 2.2141,
+      "step": 2009
+    },
+    {
+      "epoch": 0.921595598349381,
+      "grad_norm": 0.47425119621515166,
+      "learning_rate": 1.604261061674378e-05,
+      "loss": 2.3071,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9220541036221916,
+      "grad_norm": 0.6218905278943773,
+      "learning_rate": 1.5856521579965865e-05,
+      "loss": 2.1858,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9225126088950023,
+      "grad_norm": 0.49187572643807737,
+      "learning_rate": 1.5671500744256938e-05,
+      "loss": 2.3168,
+      "step": 2012
+    },
+    {
+      "epoch": 0.922971114167813,
+      "grad_norm": 0.47089387784868136,
+      "learning_rate": 1.5487548517841953e-05,
+      "loss": 2.1897,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9234296194406235,
+      "grad_norm": 0.4846790022115859,
+      "learning_rate": 1.530466530658814e-05,
+      "loss": 2.3195,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9238881247134342,
+      "grad_norm": 0.49189612242282466,
+      "learning_rate": 1.5122851514004054e-05,
+      "loss": 2.2877,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9243466299862448,
+      "grad_norm": 0.5242862556774294,
+      "learning_rate": 1.4942107541238703e-05,
+      "loss": 2.2958,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9248051352590555,
+      "grad_norm": 1.2048308415937161,
+      "learning_rate": 1.4762433787080809e-05,
+      "loss": 2.1689,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9252636405318662,
+      "grad_norm": 0.49486155183725317,
+      "learning_rate": 1.4583830647957541e-05,
+      "loss": 2.2763,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9257221458046767,
+      "grad_norm": 0.7833217806843479,
+      "learning_rate": 1.4406298517934068e-05,
+      "loss": 2.1084,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9261806510774874,
+      "grad_norm": 0.46110094307484145,
+      "learning_rate": 1.4229837788712562e-05,
+      "loss": 2.1692,
+      "step": 2020
+    },
+    {
+      "epoch": 0.926639156350298,
+      "grad_norm": 0.46812121194596334,
+      "learning_rate": 1.4054448849631085e-05,
+      "loss": 2.2827,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9270976616231087,
+      "grad_norm": 0.46474950775724927,
+      "learning_rate": 1.3880132087663145e-05,
+      "loss": 2.1917,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9275561668959194,
+      "grad_norm": 0.47966159451999807,
+      "learning_rate": 1.3706887887416419e-05,
+      "loss": 2.225,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9280146721687299,
+      "grad_norm": 0.46939677901962534,
+      "learning_rate": 1.3534716631132316e-05,
+      "loss": 2.1881,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9284731774415406,
+      "grad_norm": 0.47657729796686354,
+      "learning_rate": 1.3363618698684853e-05,
+      "loss": 2.2119,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9289316827143512,
+      "grad_norm": 0.4701187580537652,
+      "learning_rate": 1.3193594467579728e-05,
+      "loss": 2.187,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9293901879871619,
+      "grad_norm": 0.47361120993222494,
+      "learning_rate": 1.3024644312954026e-05,
+      "loss": 2.245,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9298486932599724,
+      "grad_norm": 0.4572531640760547,
+      "learning_rate": 1.2856768607574564e-05,
+      "loss": 2.175,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9303071985327831,
+      "grad_norm": 0.5350564531654919,
+      "learning_rate": 1.2689967721837947e-05,
+      "loss": 2.2596,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9307657038055938,
+      "grad_norm": 0.4856229804889852,
+      "learning_rate": 1.2524242023769006e-05,
+      "loss": 2.2764,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9312242090784044,
+      "grad_norm": 0.4766058397665891,
+      "learning_rate": 1.2359591879020526e-05,
+      "loss": 2.2941,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9316827143512151,
+      "grad_norm": 0.5065042158047997,
+      "learning_rate": 1.2196017650872081e-05,
+      "loss": 2.2456,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9321412196240256,
+      "grad_norm": 0.5127042251902137,
+      "learning_rate": 1.2033519700229367e-05,
+      "loss": 2.1689,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9325997248968363,
+      "grad_norm": 0.4972462178920916,
+      "learning_rate": 1.1872098385623586e-05,
+      "loss": 2.3028,
+      "step": 2034
+    },
+    {
+      "epoch": 0.933058230169647,
+      "grad_norm": 0.5781041301443213,
+      "learning_rate": 1.1711754063210289e-05,
+      "loss": 2.2461,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9335167354424576,
+      "grad_norm": 0.48475675766893794,
+      "learning_rate": 1.155248708676887e-05,
+      "loss": 2.1493,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9339752407152683,
+      "grad_norm": 0.5059312291644701,
+      "learning_rate": 1.1394297807701737e-05,
+      "loss": 2.226,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9344337459880788,
+      "grad_norm": 0.4951932284877858,
+      "learning_rate": 1.1237186575033254e-05,
+      "loss": 2.2947,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9348922512608895,
+      "grad_norm": 0.5498280219585822,
+      "learning_rate": 1.1081153735409522e-05,
+      "loss": 2.2531,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9353507565337001,
+      "grad_norm": 0.49072988798971356,
+      "learning_rate": 1.0926199633097156e-05,
+      "loss": 2.2849,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9358092618065108,
+      "grad_norm": 0.48951758492926656,
+      "learning_rate": 1.0772324609982787e-05,
+      "loss": 2.2484,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9362677670793215,
+      "grad_norm": 0.46872170618715575,
+      "learning_rate": 1.0619529005571893e-05,
+      "loss": 2.254,
+      "step": 2042
+    },
+    {
+      "epoch": 0.936726272352132,
+      "grad_norm": 0.5513939662136935,
+      "learning_rate": 1.0467813156988748e-05,
+      "loss": 2.2968,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9371847776249427,
+      "grad_norm": 0.4651249365182297,
+      "learning_rate": 1.0317177398975031e-05,
+      "loss": 2.2296,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9376432828977533,
+      "grad_norm": 0.5389786483156581,
+      "learning_rate": 1.0167622063889326e-05,
+      "loss": 2.1767,
+      "step": 2045
+    },
+    {
+      "epoch": 0.938101788170564,
+      "grad_norm": 0.4611133943469783,
+      "learning_rate": 1.0019147481706625e-05,
+      "loss": 2.2202,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9385602934433745,
+      "grad_norm": 0.5704041788636699,
+      "learning_rate": 9.871753980017051e-06,
+      "loss": 2.2102,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9390187987161852,
+      "grad_norm": 0.5831982281285109,
+      "learning_rate": 9.725441884025855e-06,
+      "loss": 2.2528,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9394773039889959,
+      "grad_norm": 0.5009197171253204,
+      "learning_rate": 9.580211516551862e-06,
+      "loss": 2.2074,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9399358092618065,
+      "grad_norm": 0.4539422414584854,
+      "learning_rate": 9.436063198027589e-06,
+      "loss": 2.2298,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9403943145346172,
+      "grad_norm": 0.4969716435532869,
+      "learning_rate": 9.292997246497959e-06,
+      "loss": 2.2738,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9408528198074277,
+      "grad_norm": 0.4760045652297781,
+      "learning_rate": 9.151013977619693e-06,
+      "loss": 2.1768,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9413113250802384,
+      "grad_norm": 0.4685373929291693,
+      "learning_rate": 9.010113704661038e-06,
+      "loss": 2.1916,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9417698303530491,
+      "grad_norm": 0.5056370400525466,
+      "learning_rate": 8.870296738500316e-06,
+      "loss": 2.221,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9422283356258597,
+      "grad_norm": 0.45200032234618864,
+      "learning_rate": 8.731563387626095e-06,
+      "loss": 2.29,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9426868408986704,
+      "grad_norm": 0.4523759220997389,
+      "learning_rate": 8.59391395813569e-06,
+      "loss": 2.199,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9431453461714809,
+      "grad_norm": 0.451802877867888,
+      "learning_rate": 8.457348753735328e-06,
+      "loss": 2.1533,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9436038514442916,
+      "grad_norm": 0.4657603756113442,
+      "learning_rate": 8.321868075738593e-06,
+      "loss": 2.3033,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9440623567171023,
+      "grad_norm": 0.4818573884708117,
+      "learning_rate": 8.187472223066371e-06,
+      "loss": 2.2433,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9445208619899129,
+      "grad_norm": 0.47056687564986355,
+      "learning_rate": 8.054161492246136e-06,
+      "loss": 2.219,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9449793672627235,
+      "grad_norm": 0.4849733762369186,
+      "learning_rate": 7.921936177411049e-06,
+      "loss": 2.3026,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9454378725355341,
+      "grad_norm": 0.496718964151624,
+      "learning_rate": 7.790796570299463e-06,
+      "loss": 2.23,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9458963778083448,
+      "grad_norm": 0.5107373825224678,
+      "learning_rate": 7.660742960254207e-06,
+      "loss": 2.2389,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9463548830811555,
+      "grad_norm": 0.4764869199584925,
+      "learning_rate": 7.531775634222138e-06,
+      "loss": 2.1937,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9468133883539661,
+      "grad_norm": 0.5001976235675933,
+      "learning_rate": 7.403894876753192e-06,
+      "loss": 2.187,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9472718936267767,
+      "grad_norm": 0.5110251520848073,
+      "learning_rate": 7.277100970000061e-06,
+      "loss": 2.2268,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9477303988995873,
+      "grad_norm": 0.48122504555884416,
+      "learning_rate": 7.151394193717408e-06,
+      "loss": 2.2231,
+      "step": 2067
+    },
+    {
+      "epoch": 0.948188904172398,
+      "grad_norm": 0.4892674960769469,
+      "learning_rate": 7.026774825261151e-06,
+      "loss": 2.306,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9486474094452086,
+      "grad_norm": 0.8509776456759013,
+      "learning_rate": 6.903243139588233e-06,
+      "loss": 2.1752,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9491059147180193,
+      "grad_norm": 0.479527904769022,
+      "learning_rate": 6.780799409255522e-06,
+      "loss": 2.2861,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9495644199908299,
+      "grad_norm": 0.5024995811860097,
+      "learning_rate": 6.659443904419637e-06,
+      "loss": 2.322,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9500229252636405,
+      "grad_norm": 0.480083438402163,
+      "learning_rate": 6.539176892836008e-06,
+      "loss": 2.2316,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9504814305364512,
+      "grad_norm": 0.5087182910767835,
+      "learning_rate": 6.4199986398585375e-06,
+      "loss": 2.278,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9509399358092618,
+      "grad_norm": 0.4947159270251409,
+      "learning_rate": 6.3019094084388884e-06,
+      "loss": 2.3201,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9513984410820725,
+      "grad_norm": 0.5195434141348381,
+      "learning_rate": 6.18490945912592e-06,
+      "loss": 2.1681,
+      "step": 2075
+    },
+    {
+      "epoch": 0.951856946354883,
+      "grad_norm": 0.5510060809169997,
+      "learning_rate": 6.068999050065249e-06,
+      "loss": 2.2209,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9523154516276937,
+      "grad_norm": 0.4628043627538752,
+      "learning_rate": 5.9541784369983586e-06,
+      "loss": 2.1537,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9527739569005044,
+      "grad_norm": 0.5222559611475667,
+      "learning_rate": 5.840447873262433e-06,
+      "loss": 2.2091,
+      "step": 2078
+    },
+    {
+      "epoch": 0.953232462173315,
+      "grad_norm": 0.4789008973757523,
+      "learning_rate": 5.727807609789471e-06,
+      "loss": 2.2557,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9536909674461256,
+      "grad_norm": 0.4830033256545644,
+      "learning_rate": 5.616257895105892e-06,
+      "loss": 2.2428,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9541494727189362,
+      "grad_norm": 0.5437902877276044,
+      "learning_rate": 5.505798975331933e-06,
+      "loss": 2.3223,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9546079779917469,
+      "grad_norm": 0.5922296532242108,
+      "learning_rate": 5.396431094181198e-06,
+      "loss": 2.268,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9550664832645576,
+      "grad_norm": 0.47241725105622123,
+      "learning_rate": 5.288154492960107e-06,
+      "loss": 2.2443,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9555249885373682,
+      "grad_norm": 0.4946992530864159,
+      "learning_rate": 5.1809694105671155e-06,
+      "loss": 2.162,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9559834938101788,
+      "grad_norm": 0.47507856057184256,
+      "learning_rate": 5.074876083492441e-06,
+      "loss": 2.1766,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9564419990829894,
+      "grad_norm": 0.473273952939663,
+      "learning_rate": 4.96987474581767e-06,
+      "loss": 2.299,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9569005043558001,
+      "grad_norm": 0.4692147059973741,
+      "learning_rate": 4.865965629214819e-06,
+      "loss": 2.1671,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9573590096286108,
+      "grad_norm": 0.5045730968532742,
+      "learning_rate": 4.763148962946218e-06,
+      "loss": 2.2324,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9578175149014214,
+      "grad_norm": 0.5352608955183099,
+      "learning_rate": 4.661424973863681e-06,
+      "loss": 2.2531,
+      "step": 2089
+    },
+    {
+      "epoch": 0.958276020174232,
+      "grad_norm": 0.5147137858290269,
+      "learning_rate": 4.560793886408398e-06,
+      "loss": 2.2014,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9587345254470426,
+      "grad_norm": 0.539184953768462,
+      "learning_rate": 4.461255922609986e-06,
+      "loss": 2.2937,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9591930307198533,
+      "grad_norm": 0.5093696241811828,
+      "learning_rate": 4.362811302086267e-06,
+      "loss": 2.1839,
+      "step": 2092
+    },
+    {
+      "epoch": 0.959651535992664,
+      "grad_norm": 0.46679849342540475,
+      "learning_rate": 4.265460242042885e-06,
+      "loss": 2.2467,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9601100412654745,
+      "grad_norm": 0.47338636588587013,
+      "learning_rate": 4.169202957272522e-06,
+      "loss": 2.1776,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9605685465382852,
+      "grad_norm": 0.46961555636495617,
+      "learning_rate": 4.074039660154738e-06,
+      "loss": 2.1743,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9610270518110958,
+      "grad_norm": 0.47539003506547844,
+      "learning_rate": 3.9799705606551325e-06,
+      "loss": 2.2721,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9614855570839065,
+      "grad_norm": 0.4999803660540667,
+      "learning_rate": 3.886995866325294e-06,
+      "loss": 2.2783,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9619440623567171,
+      "grad_norm": 0.47994818886921925,
+      "learning_rate": 3.795115782302072e-06,
+      "loss": 2.2486,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9624025676295277,
+      "grad_norm": 0.47776981007513014,
+      "learning_rate": 3.704330511307197e-06,
+      "loss": 2.2233,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9628610729023384,
+      "grad_norm": 0.8753224251397906,
+      "learning_rate": 3.614640253646828e-06,
+      "loss": 2.2519,
+      "step": 2100
+    },
+    {
+      "epoch": 0.963319578175149,
+      "grad_norm": 0.48587553705478076,
+      "learning_rate": 3.5260452072110594e-06,
+      "loss": 2.2893,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9637780834479597,
+      "grad_norm": 0.49906475697179775,
+      "learning_rate": 3.4385455674737498e-06,
+      "loss": 2.2319,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9642365887207703,
+      "grad_norm": 0.47409626540204175,
+      "learning_rate": 3.3521415274915256e-06,
+      "loss": 2.2029,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9646950939935809,
+      "grad_norm": 0.507410994544119,
+      "learning_rate": 3.2668332779041133e-06,
+      "loss": 2.2587,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9651535992663915,
+      "grad_norm": 0.49370280195647526,
+      "learning_rate": 3.1826210069332838e-06,
+      "loss": 2.2373,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9656121045392022,
+      "grad_norm": 0.4683371669311861,
+      "learning_rate": 3.0995049003826324e-06,
+      "loss": 2.3235,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9660706098120129,
+      "grad_norm": 0.49005037809617036,
+      "learning_rate": 3.017485141637355e-06,
+      "loss": 2.1704,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9665291150848234,
+      "grad_norm": 0.4926618242318962,
+      "learning_rate": 2.9365619116636376e-06,
+      "loss": 2.1736,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9669876203576341,
+      "grad_norm": 0.5099892147620418,
+      "learning_rate": 2.856735389008269e-06,
+      "loss": 2.2411,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9674461256304447,
+      "grad_norm": 0.4735437701528052,
+      "learning_rate": 2.778005749798307e-06,
+      "loss": 2.201,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9679046309032554,
+      "grad_norm": 0.46602086682181454,
+      "learning_rate": 2.700373167740744e-06,
+      "loss": 2.2719,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9683631361760661,
+      "grad_norm": 0.4682260886092651,
+      "learning_rate": 2.62383781412201e-06,
+      "loss": 2.2231,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9688216414488766,
+      "grad_norm": 0.4699825660645056,
+      "learning_rate": 2.5483998578076373e-06,
+      "loss": 2.2434,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9692801467216873,
+      "grad_norm": 0.47715113661904784,
+      "learning_rate": 2.4740594652418736e-06,
+      "loss": 2.25,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9697386519944979,
+      "grad_norm": 0.46620575147752025,
+      "learning_rate": 2.4008168004472917e-06,
+      "loss": 2.2225,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9701971572673086,
+      "grad_norm": 0.6165980911549875,
+      "learning_rate": 2.3286720250246253e-06,
+      "loss": 2.1792,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9706556625401193,
+      "grad_norm": 0.48615404181720057,
+      "learning_rate": 2.2576252981520994e-06,
+      "loss": 2.2249,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9711141678129298,
+      "grad_norm": 0.47746694981811577,
+      "learning_rate": 2.1876767765853233e-06,
+      "loss": 2.1984,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9715726730857405,
+      "grad_norm": 0.4805411723020939,
+      "learning_rate": 2.118826614656788e-06,
+      "loss": 2.2283,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9720311783585511,
+      "grad_norm": 0.47606164744480656,
+      "learning_rate": 2.051074964275701e-06,
+      "loss": 2.2279,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9724896836313618,
+      "grad_norm": 0.49199679938718227,
+      "learning_rate": 1.984421974927375e-06,
+      "loss": 2.2518,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9729481889041725,
+      "grad_norm": 0.5109894358726804,
+      "learning_rate": 1.9188677936731734e-06,
+      "loss": 2.2928,
+      "step": 2122
+    },
+    {
+      "epoch": 0.973406694176983,
+      "grad_norm": 0.49700707556100876,
+      "learning_rate": 1.8544125651501208e-06,
+      "loss": 2.293,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9738651994497937,
+      "grad_norm": 0.5039313111662157,
+      "learning_rate": 1.7910564315704035e-06,
+      "loss": 2.2508,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9743237047226043,
+      "grad_norm": 0.49018026322576974,
+      "learning_rate": 1.7287995327214257e-06,
+      "loss": 2.2841,
+      "step": 2125
+    },
+    {
+      "epoch": 0.974782209995415,
+      "grad_norm": 0.4616998379615357,
+      "learning_rate": 1.6676420059649754e-06,
+      "loss": 2.226,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9752407152682255,
+      "grad_norm": 0.49223802913609876,
+      "learning_rate": 1.6075839862374485e-06,
+      "loss": 2.1571,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9756992205410362,
+      "grad_norm": 0.5571972406389564,
+      "learning_rate": 1.5486256060492366e-06,
+      "loss": 2.1761,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9761577258138469,
+      "grad_norm": 0.47137569766677706,
+      "learning_rate": 1.4907669954844495e-06,
+      "loss": 2.1535,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9766162310866575,
+      "grad_norm": 0.4524143013573851,
+      "learning_rate": 1.434008282200805e-06,
+      "loss": 2.1885,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9770747363594682,
+      "grad_norm": 0.5111550325101623,
+      "learning_rate": 1.3783495914291844e-06,
+      "loss": 2.3057,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9775332416322787,
+      "grad_norm": 0.48679968808564633,
+      "learning_rate": 1.3237910459734104e-06,
+      "loss": 2.3131,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9779917469050894,
+      "grad_norm": 1.005634850840659,
+      "learning_rate": 1.270332766210025e-06,
+      "loss": 2.3066,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9784502521779,
+      "grad_norm": 0.48318757999833156,
+      "learning_rate": 1.2179748700879012e-06,
+      "loss": 2.2842,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9789087574507107,
+      "grad_norm": 0.45648176769850846,
+      "learning_rate": 1.1667174731280205e-06,
+      "loss": 2.2393,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9793672627235214,
+      "grad_norm": 0.46946780558431905,
+      "learning_rate": 1.1165606884234182e-06,
+      "loss": 2.2179,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9798257679963319,
+      "grad_norm": 0.5148279453371226,
+      "learning_rate": 1.0675046266386268e-06,
+      "loss": 2.2555,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9802842732691426,
+      "grad_norm": 0.5949436178839725,
+      "learning_rate": 1.019549396009567e-06,
+      "loss": 2.315,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9807427785419532,
+      "grad_norm": 1.0515277773918728,
+      "learning_rate": 9.726951023434348e-07,
+      "loss": 2.2568,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9812012838147639,
+      "grad_norm": 0.49121232061020753,
+      "learning_rate": 9.269418490182591e-07,
+      "loss": 2.3125,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9816597890875745,
+      "grad_norm": 0.46921824001618945,
+      "learning_rate": 8.822897369827332e-07,
+      "loss": 2.2565,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9821182943603851,
+      "grad_norm": 0.5140963462257429,
+      "learning_rate": 8.387388647561611e-07,
+      "loss": 2.1836,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9825767996331958,
+      "grad_norm": 0.4801328629112835,
+      "learning_rate": 7.962893284279016e-07,
+      "loss": 2.1803,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9830353049060064,
+      "grad_norm": 0.5156845693694972,
+      "learning_rate": 7.549412216574791e-07,
+      "loss": 2.3647,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9834938101788171,
+      "grad_norm": 0.5245483100512751,
+      "learning_rate": 7.146946356743067e-07,
+      "loss": 2.1774,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9839523154516276,
+      "grad_norm": 0.5058380324048,
+      "learning_rate": 6.755496592773524e-07,
+      "loss": 2.1874,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9844108207244383,
+      "grad_norm": 0.5169847976511661,
+      "learning_rate": 6.375063788349733e-07,
+      "loss": 2.2827,
+      "step": 2147
+    },
+    {
+      "epoch": 0.984869325997249,
+      "grad_norm": 0.4990946745462514,
+      "learning_rate": 6.005648782848594e-07,
+      "loss": 2.2068,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9853278312700596,
+      "grad_norm": 0.48507249175564,
+      "learning_rate": 5.647252391337565e-07,
+      "loss": 2.3341,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9857863365428703,
+      "grad_norm": 0.4726584501944319,
+      "learning_rate": 5.299875404572441e-07,
+      "loss": 2.2529,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9862448418156808,
+      "grad_norm": 0.4773214693808581,
+      "learning_rate": 4.963518588996796e-07,
+      "loss": 2.1597,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9867033470884915,
+      "grad_norm": 0.5040051226729652,
+      "learning_rate": 4.638182686738657e-07,
+      "loss": 2.2082,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9871618523613022,
+      "grad_norm": 0.5277211290448849,
+      "learning_rate": 4.3238684156110543e-07,
+      "loss": 2.1694,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9876203576341128,
+      "grad_norm": 0.46833877468957535,
+      "learning_rate": 4.020576469108139e-07,
+      "loss": 2.2057,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9880788629069235,
+      "grad_norm": 0.4679988660588894,
+      "learning_rate": 3.7283075164046274e-07,
+      "loss": 2.2738,
+      "step": 2155
+    },
+    {
+      "epoch": 0.988537368179734,
+      "grad_norm": 0.4704504935285761,
+      "learning_rate": 3.4470622023557995e-07,
+      "loss": 2.2224,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9889958734525447,
+      "grad_norm": 0.49313504256422325,
+      "learning_rate": 3.176841147492504e-07,
+      "loss": 2.2146,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9894543787253554,
+      "grad_norm": 0.4744582903386652,
+      "learning_rate": 2.9176449480244895e-07,
+      "loss": 2.1457,
+      "step": 2158
+    },
+    {
+      "epoch": 0.989912883998166,
+      "grad_norm": 0.47358293885220726,
+      "learning_rate": 2.6694741758342967e-07,
+      "loss": 2.2867,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9903713892709766,
+      "grad_norm": 0.4678229713145958,
+      "learning_rate": 2.432329378478926e-07,
+      "loss": 2.227,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9908298945437872,
+      "grad_norm": 0.4811567517539918,
+      "learning_rate": 2.2062110791892798e-07,
+      "loss": 2.2409,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9912883998165979,
+      "grad_norm": 0.5214847951344429,
+      "learning_rate": 1.9911197768662792e-07,
+      "loss": 2.244,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9917469050894085,
+      "grad_norm": 0.4705826074052618,
+      "learning_rate": 1.7870559460814173e-07,
+      "loss": 2.282,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9922054103622192,
+      "grad_norm": 0.4737222195083807,
+      "learning_rate": 1.5940200370750947e-07,
+      "loss": 2.2634,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9926639156350298,
+      "grad_norm": 0.5074703921094397,
+      "learning_rate": 1.4120124757577291e-07,
+      "loss": 2.3258,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9931224209078404,
+      "grad_norm": 0.4824749689576653,
+      "learning_rate": 1.2410336637047603e-07,
+      "loss": 2.144,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9935809261806511,
+      "grad_norm": 0.50741618055411,
+      "learning_rate": 1.081083978159425e-07,
+      "loss": 2.3164,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9940394314534617,
+      "grad_norm": 0.48763871999861574,
+      "learning_rate": 9.321637720310915e-08,
+      "loss": 2.1825,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9944979367262724,
+      "grad_norm": 0.4860433447023315,
+      "learning_rate": 7.942733738924845e-08,
+      "loss": 2.2673,
+      "step": 2169
+    },
+    {
+      "epoch": 0.994956441999083,
+      "grad_norm": 0.44671639042288186,
+      "learning_rate": 6.6741308798135e-08,
+      "loss": 2.2052,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9954149472718936,
+      "grad_norm": 0.4736440075947356,
+      "learning_rate": 5.5158319419934546e-08,
+      "loss": 2.348,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9958734525447043,
+      "grad_norm": 0.46632326924363793,
+      "learning_rate": 4.4678394810981906e-08,
+      "loss": 2.2195,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9963319578175149,
+      "grad_norm": 0.6288829778428912,
+      "learning_rate": 3.5301558093947527e-08,
+      "loss": 2.3073,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9967904630903255,
+      "grad_norm": 0.4653217054594382,
+      "learning_rate": 2.7027829957559925e-08,
+      "loss": 2.2729,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9972489683631361,
+      "grad_norm": 0.8257914023848946,
+      "learning_rate": 1.985722865682771e-08,
+      "loss": 2.2326,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9977074736359468,
+      "grad_norm": 0.46759213381371456,
+      "learning_rate": 1.3789770012762048e-08,
+      "loss": 2.1731,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9981659789087575,
+      "grad_norm": 0.5044342121882323,
+      "learning_rate": 8.825467412376665e-09,
+      "loss": 2.2304,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9986244841815681,
+      "grad_norm": 0.49705088042046464,
+      "learning_rate": 4.96433180879885e-09,
+      "loss": 2.2552,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9990829894543787,
+      "grad_norm": 0.4883270419821403,
+      "learning_rate": 2.206371721158451e-09,
+      "loss": 2.2514,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9995414947271893,
+      "grad_norm": 0.4695567477879236,
+      "learning_rate": 5.515932345323549e-10,
+      "loss": 2.1771,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46548952159977236,
+      "learning_rate": 0.0,
+      "loss": 2.2621,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0,
+      "step": 2181,
+      "total_flos": 4.316830046879744e+17,
+      "train_loss": 2.397268701725627,
+      "train_runtime": 7112.1207,
+      "train_samples_per_second": 78.476,
+      "train_steps_per_second": 0.307
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2181,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 24000,
+  "total_flos": 4.316830046879744e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}