diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10852 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1546,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6466701238755213,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 0.348,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.139185622660457,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 0.3305,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.881200735394447,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 0.3234,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7485825244709936,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.3036,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.870091850769131,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 0.3082,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.147534101913999,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 0.2976,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0450582796416175,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 0.3194,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.9648702987949624,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 0.3229,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.05272089122871,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 0.2646,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9192913531636706,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.2736,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.967249286697924,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.2925,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.138517274675191,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.2967,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.33936595379889,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.319,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.555705158505097,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.2793,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.181269031597184,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.2417,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.064313198312008,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.2061,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.61672727351488,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.2546,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7903212709494416,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.2521,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7762556046042675,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.1738,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6164798733780423,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.1681,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.794571211225896,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.1864,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1344038725423276,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.1685,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6772778738316547,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.1733,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5864900115307945,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.1614,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1829091283205524,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.1749,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.627918949690976,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.156,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1489057985777893,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 0.1511,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4901325902309766,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.1622,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5737914145996683,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 0.1272,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6639517888902864,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.129,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9458157448896816,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 0.1604,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0682564766093923,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.1413,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4160767147719295,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 0.1714,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9442874642653694,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.1553,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6201074618551132,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.1599,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.499964053871311,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.1375,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.392017059081343,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 0.119,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3073895290345847,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.1497,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0072381649457016,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 0.1155,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1409222987029075,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.1353,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.339746183858402,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 0.1589,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0962604851483464,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.1487,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2204015267252832,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 0.1413,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0308968159970804,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.12,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.120274606415836,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 0.1462,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.214971309457947,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.145,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.453479965060542,
+      "learning_rate": 1e-05,
+      "loss": 0.1249,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9638463123214622,
+      "learning_rate": 9.999989019140636e-06,
+      "loss": 0.0844,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0761585516779526,
+      "learning_rate": 9.999956076610778e-06,
+      "loss": 0.1446,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.061496741512052,
+      "learning_rate": 9.999901172555116e-06,
+      "loss": 0.1232,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0735089733196803,
+      "learning_rate": 9.999824307214812e-06,
+      "loss": 0.1405,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.973897420839363,
+      "learning_rate": 9.999725480927483e-06,
+      "loss": 0.1145,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.633648585469745,
+      "learning_rate": 9.99960469412721e-06,
+      "loss": 0.1133,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6460370470359849,
+      "learning_rate": 9.999461947344526e-06,
+      "loss": 0.114,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9745205343016254,
+      "learning_rate": 9.99929724120643e-06,
+      "loss": 0.1496,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0268455134518746,
+      "learning_rate": 9.99911057643636e-06,
+      "loss": 0.1067,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9753338409612728,
+      "learning_rate": 9.998901953854222e-06,
+      "loss": 0.1423,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8231456903833867,
+      "learning_rate": 9.998671374376349e-06,
+      "loss": 0.1145,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4924704043661263,
+      "learning_rate": 9.99841883901553e-06,
+      "loss": 0.1435,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9320494942976505,
+      "learning_rate": 9.998144348880984e-06,
+      "loss": 0.1188,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.599418714042963,
+      "learning_rate": 9.99784790517837e-06,
+      "loss": 0.0879,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9477442918289953,
+      "learning_rate": 9.997529509209768e-06,
+      "loss": 0.1174,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.330125781434381,
+      "learning_rate": 9.997189162373683e-06,
+      "loss": 0.164,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.166128412915446,
+      "learning_rate": 9.996826866165036e-06,
+      "loss": 0.1146,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8745142590127868,
+      "learning_rate": 9.996442622175156e-06,
+      "loss": 0.1443,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.355826899731574,
+      "learning_rate": 9.996036432091773e-06,
+      "loss": 0.0947,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2179182480709434,
+      "learning_rate": 9.995608297699018e-06,
+      "loss": 0.1541,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5146551784021653,
+      "learning_rate": 9.9951582208774e-06,
+      "loss": 0.1459,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.675961892975102,
+      "learning_rate": 9.994686203603811e-06,
+      "loss": 0.0971,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8476255656557437,
+      "learning_rate": 9.994192247951516e-06,
+      "loss": 0.1003,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.014736730942368,
+      "learning_rate": 9.993676356090139e-06,
+      "loss": 0.1097,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1036545470925554,
+      "learning_rate": 9.993138530285649e-06,
+      "loss": 0.1264,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.107461437774381,
+      "learning_rate": 9.992578772900366e-06,
+      "loss": 0.1379,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0409530279397967,
+      "learning_rate": 9.991997086392936e-06,
+      "loss": 0.1051,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7745745669427309,
+      "learning_rate": 9.991393473318326e-06,
+      "loss": 0.1139,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5490723949036367,
+      "learning_rate": 9.990767936327812e-06,
+      "loss": 0.1517,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.195736654386886,
+      "learning_rate": 9.990120478168968e-06,
+      "loss": 0.1222,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8776753484302193,
+      "learning_rate": 9.98945110168565e-06,
+      "loss": 0.1642,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3857244336207146,
+      "learning_rate": 9.988759809817995e-06,
+      "loss": 0.0998,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0118828036816483,
+      "learning_rate": 9.98804660560239e-06,
+      "loss": 0.1196,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8691147410130577,
+      "learning_rate": 9.987311492171474e-06,
+      "loss": 0.1411,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.716478047398871,
+      "learning_rate": 9.98655447275412e-06,
+      "loss": 0.1183,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.241445397670684,
+      "learning_rate": 9.985775550675415e-06,
+      "loss": 0.0781,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.637285204874327,
+      "learning_rate": 9.984974729356653e-06,
+      "loss": 0.1116,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7402339286961375,
+      "learning_rate": 9.984152012315317e-06,
+      "loss": 0.1117,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.088002718923483,
+      "learning_rate": 9.983307403165063e-06,
+      "loss": 0.1444,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.518789329085565,
+      "learning_rate": 9.982440905615705e-06,
+      "loss": 0.1158,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2952247513792705,
+      "learning_rate": 9.981552523473198e-06,
+      "loss": 0.1189,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.50747528627161,
+      "learning_rate": 9.980642260639621e-06,
+      "loss": 0.1034,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.557146005325423,
+      "learning_rate": 9.979710121113163e-06,
+      "loss": 0.1167,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.948661547281757,
+      "learning_rate": 9.978756108988098e-06,
+      "loss": 0.1696,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.230926069527611,
+      "learning_rate": 9.977780228454779e-06,
+      "loss": 0.1652,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9280522718280948,
+      "learning_rate": 9.976782483799607e-06,
+      "loss": 0.11,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.5706395089517855,
+      "learning_rate": 9.975762879405018e-06,
+      "loss": 0.1413,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9503282618719768,
+      "learning_rate": 9.974721419749466e-06,
+      "loss": 0.1221,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6897748435855646,
+      "learning_rate": 9.973658109407403e-06,
+      "loss": 0.105,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1150461944750827,
+      "learning_rate": 9.97257295304925e-06,
+      "loss": 0.1422,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.149809103068844,
+      "learning_rate": 9.971465955441386e-06,
+      "loss": 0.0991,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9774388476377238,
+      "learning_rate": 9.970337121446127e-06,
+      "loss": 0.1027,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6539049989428836,
+      "learning_rate": 9.9691864560217e-06,
+      "loss": 0.1248,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8618631406274415,
+      "learning_rate": 9.968013964222223e-06,
+      "loss": 0.0952,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6700970501837484,
+      "learning_rate": 9.96681965119768e-06,
+      "loss": 0.1595,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4203655017115824,
+      "learning_rate": 9.96560352219391e-06,
+      "loss": 0.1065,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.425921167211727,
+      "learning_rate": 9.964365582552566e-06,
+      "loss": 0.1272,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1493403207347677,
+      "learning_rate": 9.963105837711104e-06,
+      "loss": 0.0693,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9586184930220423,
+      "learning_rate": 9.961824293202758e-06,
+      "loss": 0.1232,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.016492233318853,
+      "learning_rate": 9.960520954656512e-06,
+      "loss": 0.0856,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.133900070676782,
+      "learning_rate": 9.959195827797075e-06,
+      "loss": 0.1285,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.013620495105404,
+      "learning_rate": 9.957848918444861e-06,
+      "loss": 0.1463,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.918642411411689,
+      "learning_rate": 9.956480232515959e-06,
+      "loss": 0.1226,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.743736443377982,
+      "learning_rate": 9.955089776022108e-06,
+      "loss": 0.1204,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5410707973525848,
+      "learning_rate": 9.953677555070671e-06,
+      "loss": 0.0844,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8403838218102002,
+      "learning_rate": 9.952243575864608e-06,
+      "loss": 0.1121,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8857421494208948,
+      "learning_rate": 9.950787844702447e-06,
+      "loss": 0.1447,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7670301270073683,
+      "learning_rate": 9.949310367978262e-06,
+      "loss": 0.1215,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8156401555599075,
+      "learning_rate": 9.947811152181637e-06,
+      "loss": 0.1071,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8639909815301174,
+      "learning_rate": 9.946290203897643e-06,
+      "loss": 0.1615,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.948669957321261,
+      "learning_rate": 9.944747529806811e-06,
+      "loss": 0.1239,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.093435371849378,
+      "learning_rate": 9.943183136685092e-06,
+      "loss": 0.1595,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9885745564004753,
+      "learning_rate": 9.94159703140384e-06,
+      "loss": 0.172,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0699282007190205,
+      "learning_rate": 9.939989220929772e-06,
+      "loss": 0.0966,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9316680877282852,
+      "learning_rate": 9.938359712324948e-06,
+      "loss": 0.1261,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9213513024255975,
+      "learning_rate": 9.936708512746729e-06,
+      "loss": 0.1385,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0482109640638515,
+      "learning_rate": 9.935035629447749e-06,
+      "loss": 0.1446,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7866384022513413,
+      "learning_rate": 9.93334106977589e-06,
+      "loss": 0.1218,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5840977776437648,
+      "learning_rate": 9.931624841174238e-06,
+      "loss": 0.113,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8804553233658048,
+      "learning_rate": 9.929886951181059e-06,
+      "loss": 0.124,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2666728116858486,
+      "learning_rate": 9.928127407429764e-06,
+      "loss": 0.1888,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5646682358647588,
+      "learning_rate": 9.926346217648874e-06,
+      "loss": 0.1171,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4848420454582072,
+      "learning_rate": 9.924543389661987e-06,
+      "loss": 0.088,
+      "step": 130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.020458437632709,
+      "learning_rate": 9.922718931387742e-06,
+      "loss": 0.1485,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8827184023295047,
+      "learning_rate": 9.92087285083979e-06,
+      "loss": 0.1168,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0266100825197584,
+      "learning_rate": 9.919005156126746e-06,
+      "loss": 0.146,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.235846705437004,
+      "learning_rate": 9.917115855452172e-06,
+      "loss": 0.1166,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.73904040901477,
+      "learning_rate": 9.915204957114524e-06,
+      "loss": 0.1521,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1834714487743683,
+      "learning_rate": 9.913272469507124e-06,
+      "loss": 0.1304,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8607754039829953,
+      "learning_rate": 9.911318401118124e-06,
+      "loss": 0.1205,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7738065734608983,
+      "learning_rate": 9.909342760530461e-06,
+      "loss": 0.1318,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0090370977928287,
+      "learning_rate": 9.90734555642183e-06,
+      "loss": 0.1088,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7717365204692492,
+      "learning_rate": 9.905326797564637e-06,
+      "loss": 0.1075,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9880094654910834,
+      "learning_rate": 9.903286492825965e-06,
+      "loss": 0.1027,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0750699770709584,
+      "learning_rate": 9.901224651167534e-06,
+      "loss": 0.1168,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3732135065551923,
+      "learning_rate": 9.899141281645662e-06,
+      "loss": 0.1533,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.703545439116926,
+      "learning_rate": 9.89703639341122e-06,
+      "loss": 0.1117,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0281249810083595,
+      "learning_rate": 9.894909995709607e-06,
+      "loss": 0.1188,
+      "step": 145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.883392386333987,
+      "learning_rate": 9.892762097880689e-06,
+      "loss": 0.1169,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8187730542306118,
+      "learning_rate": 9.890592709358771e-06,
+      "loss": 0.152,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2339440137327795,
+      "learning_rate": 9.888401839672554e-06,
+      "loss": 0.1311,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.349926480407424,
+      "learning_rate": 9.886189498445091e-06,
+      "loss": 0.1474,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5280358206435156,
+      "learning_rate": 9.883955695393745e-06,
+      "loss": 0.1089,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.55628950397484,
+      "learning_rate": 9.881700440330148e-06,
+      "loss": 0.1677,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.599168966333966,
+      "learning_rate": 9.879423743160154e-06,
+      "loss": 0.1218,
+      "step": 152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.065147882840751,
+      "learning_rate": 9.877125613883799e-06,
+      "loss": 0.1147,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0973331798577304,
+      "learning_rate": 9.87480606259526e-06,
+      "loss": 0.1266,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3614508711991764,
+      "learning_rate": 9.872465099482798e-06,
+      "loss": 0.1399,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2839164920052997,
+      "learning_rate": 9.870102734828733e-06,
+      "loss": 0.1515,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.191700939824119,
+      "learning_rate": 9.86771897900938e-06,
+      "loss": 0.1186,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7304607163865133,
+      "learning_rate": 9.865313842495016e-06,
+      "loss": 0.1165,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.988091314799846,
+      "learning_rate": 9.862887335849825e-06,
+      "loss": 0.1378,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8491262715856396,
+      "learning_rate": 9.86043946973186e-06,
+      "loss": 0.1293,
+      "step": 160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.529334256156627,
+      "learning_rate": 9.857970254892987e-06,
+      "loss": 0.097,
+      "step": 161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.773404734705276,
+      "learning_rate": 9.855479702178851e-06,
+      "loss": 0.1168,
+      "step": 162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4916119001719945,
+      "learning_rate": 9.852967822528814e-06,
+      "loss": 0.1687,
+      "step": 163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5393415410713551,
+      "learning_rate": 9.850434626975913e-06,
+      "loss": 0.0885,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9362575076807371,
+      "learning_rate": 9.847880126646816e-06,
+      "loss": 0.1465,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5522038167181273,
+      "learning_rate": 9.845304332761767e-06,
+      "loss": 0.119,
+      "step": 166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6333192172897075,
+      "learning_rate": 9.842707256634534e-06,
+      "loss": 0.1359,
+      "step": 167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2486156763298677,
+      "learning_rate": 9.840088909672373e-06,
+      "loss": 0.1249,
+      "step": 168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9948386846492634,
+      "learning_rate": 9.83744930337596e-06,
+      "loss": 0.1721,
+      "step": 169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5051178680333455,
+      "learning_rate": 9.834788449339359e-06,
+      "loss": 0.1419,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.089529822824879,
+      "learning_rate": 9.83210635924995e-06,
+      "loss": 0.149,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2638144924488817,
+      "learning_rate": 9.829403044888393e-06,
+      "loss": 0.1549,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4470602369375063,
+      "learning_rate": 9.82667851812858e-06,
+      "loss": 0.1178,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4740151917325277,
+      "learning_rate": 9.823932790937565e-06,
+      "loss": 0.118,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7921124188059911,
+      "learning_rate": 9.821165875375528e-06,
+      "loss": 0.1317,
+      "step": 175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7529591680637229,
+      "learning_rate": 9.818377783595712e-06,
+      "loss": 0.1209,
+      "step": 176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.680109016010429,
+      "learning_rate": 9.815568527844375e-06,
+      "loss": 0.1053,
+      "step": 177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8681169915291045,
+      "learning_rate": 9.812738120460732e-06,
+      "loss": 0.1135,
+      "step": 178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7484525794627954,
+      "learning_rate": 9.809886573876908e-06,
+      "loss": 0.1187,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9799368911132444,
+      "learning_rate": 9.807013900617874e-06,
+      "loss": 0.1148,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5209981726206028,
+      "learning_rate": 9.8041201133014e-06,
+      "loss": 0.1053,
+      "step": 181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7761665804917133,
+      "learning_rate": 9.801205224637993e-06,
+      "loss": 0.1092,
+      "step": 182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4864267110118337,
+      "learning_rate": 9.798269247430847e-06,
+      "loss": 0.1798,
+      "step": 183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6997732391026787,
+      "learning_rate": 9.79531219457578e-06,
+      "loss": 0.1074,
+      "step": 184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9235303414692586,
+      "learning_rate": 9.792334079061192e-06,
+      "loss": 0.1283,
+      "step": 185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9637623355738383,
+      "learning_rate": 9.789334913967982e-06,
+      "loss": 0.1196,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0004951230330947,
+      "learning_rate": 9.786314712469519e-06,
+      "loss": 0.1495,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.780329747363877,
+      "learning_rate": 9.783273487831564e-06,
+      "loss": 0.1545,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0685841478506255,
+      "learning_rate": 9.780211253412222e-06,
+      "loss": 0.1282,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2671215265269633,
+      "learning_rate": 9.777128022661877e-06,
+      "loss": 0.1748,
+      "step": 190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3902891616790383,
+      "learning_rate": 9.774023809123142e-06,
+      "loss": 0.1474,
+      "step": 191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.428055584599217,
+      "learning_rate": 9.770898626430786e-06,
+      "loss": 0.0995,
+      "step": 192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5921445659631617,
+      "learning_rate": 9.76775248831169e-06,
+      "loss": 0.1075,
+      "step": 193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8876348934658589,
+      "learning_rate": 9.764585408584772e-06,
+      "loss": 0.1591,
+      "step": 194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4909969670884806,
+      "learning_rate": 9.761397401160934e-06,
+      "loss": 0.0882,
+      "step": 195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6466444261010116,
+      "learning_rate": 9.758188480043e-06,
+      "loss": 0.1399,
+      "step": 196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2507322870188777,
+      "learning_rate": 9.754958659325656e-06,
+      "loss": 0.151,
+      "step": 197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9965502835375872,
+      "learning_rate": 9.751707953195386e-06,
+      "loss": 0.1243,
+      "step": 198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.967103185033387,
+      "learning_rate": 9.748436375930406e-06,
+      "loss": 0.1103,
+      "step": 199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9032906864214325,
+      "learning_rate": 9.745143941900607e-06,
+      "loss": 0.1245,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.82973826843128,
+      "learning_rate": 9.741830665567498e-06,
+      "loss": 0.1209,
+      "step": 201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1136392039623297,
+      "learning_rate": 9.738496561484118e-06,
+      "loss": 0.1063,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.656781829823401,
+      "learning_rate": 9.735141644295006e-06,
+      "loss": 0.1834,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2732162880867715,
+      "learning_rate": 9.731765928736107e-06,
+      "loss": 0.1175,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9163592371029838,
+      "learning_rate": 9.728369429634728e-06,
+      "loss": 0.1308,
+      "step": 205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9926464099145167,
+      "learning_rate": 9.724952161909456e-06,
+      "loss": 0.1055,
+      "step": 206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.212171729804821,
+      "learning_rate": 9.721514140570108e-06,
+      "loss": 0.1205,
+      "step": 207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.089447112787558,
+      "learning_rate": 9.718055380717655e-06,
+      "loss": 0.1257,
+      "step": 208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8443109675212546,
+      "learning_rate": 9.714575897544161e-06,
+      "loss": 0.1243,
+      "step": 209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8644539626190368,
+      "learning_rate": 9.71107570633271e-06,
+      "loss": 0.1027,
+      "step": 210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9358246309330898,
+      "learning_rate": 9.707554822457346e-06,
+      "loss": 0.1144,
+      "step": 211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6560610615371094,
+      "learning_rate": 9.704013261382998e-06,
+      "loss": 0.1304,
+      "step": 212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5270084771765515,
+      "learning_rate": 9.700451038665427e-06,
+      "loss": 0.0894,
+      "step": 213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5406483878286648,
+      "learning_rate": 9.696868169951133e-06,
+      "loss": 0.1333,
+      "step": 214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9375740465908717,
+      "learning_rate": 9.693264670977307e-06,
+      "loss": 0.1245,
+      "step": 215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0904332183871945,
+      "learning_rate": 9.68964055757176e-06,
+      "loss": 0.1355,
+      "step": 216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.233484010834808,
+      "learning_rate": 9.68599584565284e-06,
+      "loss": 0.1303,
+      "step": 217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7671033554601976,
+      "learning_rate": 9.682330551229378e-06,
+      "loss": 0.1539,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5736703608691354,
+      "learning_rate": 9.678644690400602e-06,
+      "loss": 0.1312,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.788108504780175,
+      "learning_rate": 9.674938279356086e-06,
+      "loss": 0.1256,
+      "step": 220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4017354768821373,
+      "learning_rate": 9.671211334375655e-06,
+      "loss": 0.1195,
+      "step": 221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.554406170607817,
+      "learning_rate": 9.667463871829336e-06,
+      "loss": 0.1152,
+      "step": 222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9776198759979988,
+      "learning_rate": 9.663695908177272e-06,
+      "loss": 0.1232,
+      "step": 223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8763841196803877,
+      "learning_rate": 9.659907459969656e-06,
+      "loss": 0.1162,
+      "step": 224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.603879538170472,
+      "learning_rate": 9.656098543846652e-06,
+      "loss": 0.0987,
+      "step": 225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5957011663071399,
+      "learning_rate": 9.652269176538332e-06,
+      "loss": 0.0967,
+      "step": 226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5595559852063516,
+      "learning_rate": 9.648419374864591e-06,
+      "loss": 0.1131,
+      "step": 227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6225818778400385,
+      "learning_rate": 9.644549155735081e-06,
+      "loss": 0.1079,
+      "step": 228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4391874862387903,
+      "learning_rate": 9.640658536149137e-06,
+      "loss": 0.111,
+      "step": 229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9059567028458015,
+      "learning_rate": 9.636747533195698e-06,
+      "loss": 0.0989,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9499035596099419,
+      "learning_rate": 9.632816164053232e-06,
+      "loss": 0.1746,
+      "step": 231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7566067913290406,
+      "learning_rate": 9.62886444598966e-06,
+      "loss": 0.1225,
+      "step": 232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4625347527200148,
+      "learning_rate": 9.624892396362293e-06,
+      "loss": 0.1066,
+      "step": 233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9096185245160564,
+      "learning_rate": 9.620900032617734e-06,
+      "loss": 0.1614,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6414618208334602,
+      "learning_rate": 9.61688737229182e-06,
+      "loss": 0.1319,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.01434795372801,
+      "learning_rate": 9.612854433009531e-06,
+      "loss": 0.0905,
+      "step": 236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8677307978193975,
+      "learning_rate": 9.608801232484923e-06,
+      "loss": 0.1145,
+      "step": 237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6100118695054884,
+      "learning_rate": 9.604727788521048e-06,
+      "loss": 0.1252,
+      "step": 238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.029485209887423,
+      "learning_rate": 9.600634119009873e-06,
+      "loss": 0.1591,
+      "step": 239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6830484776374746,
+      "learning_rate": 9.596520241932198e-06,
+      "loss": 0.0982,
+      "step": 240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8585528147339136,
+      "learning_rate": 9.592386175357589e-06,
+      "loss": 0.1297,
+      "step": 241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7993240085387165,
+      "learning_rate": 9.588231937444284e-06,
+      "loss": 0.1164,
+      "step": 242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6977256582020621,
+      "learning_rate": 9.584057546439126e-06,
+      "loss": 0.1073,
+      "step": 243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6761942593097194,
+      "learning_rate": 9.579863020677475e-06,
+      "loss": 0.1184,
+      "step": 244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8129631892657423,
+      "learning_rate": 9.575648378583129e-06,
+      "loss": 0.1019,
+      "step": 245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0313290754950706,
+      "learning_rate": 9.571413638668246e-06,
+      "loss": 0.121,
+      "step": 246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7374763184195408,
+      "learning_rate": 9.56715881953326e-06,
+      "loss": 0.1001,
+      "step": 247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6691847184224187,
+      "learning_rate": 9.562883939866797e-06,
+      "loss": 0.1373,
+      "step": 248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.755427079402434,
+      "learning_rate": 9.5585890184456e-06,
+      "loss": 0.1264,
+      "step": 249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.004942074380999,
+      "learning_rate": 9.55427407413444e-06,
+      "loss": 0.1415,
+      "step": 250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.662759577069425,
+      "learning_rate": 9.549939125886033e-06,
+      "loss": 0.1086,
+      "step": 251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.312626616306414,
+      "learning_rate": 9.545584192740965e-06,
+      "loss": 0.1526,
+      "step": 252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3728980521640644,
+      "learning_rate": 9.541209293827599e-06,
+      "loss": 0.1017,
+      "step": 253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.164712163219188,
+      "learning_rate": 9.536814448361993e-06,
+      "loss": 0.1429,
+      "step": 254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3341674550010627,
+      "learning_rate": 9.53239967564782e-06,
+      "loss": 0.0793,
+      "step": 255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9368004727513668,
+      "learning_rate": 9.527964995076279e-06,
+      "loss": 0.1179,
+      "step": 256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3974500112400439,
+      "learning_rate": 9.523510426126015e-06,
+      "loss": 0.1052,
+      "step": 257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4225137071802845,
+      "learning_rate": 9.519035988363021e-06,
+      "loss": 0.1932,
+      "step": 258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9997143412977902,
+      "learning_rate": 9.514541701440568e-06,
+      "loss": 0.1219,
+      "step": 259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.02847423848522,
+      "learning_rate": 9.510027585099107e-06,
+      "loss": 0.0878,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.89227289355829,
+      "learning_rate": 9.505493659166193e-06,
+      "loss": 0.123,
+      "step": 261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9895504729547155,
+      "learning_rate": 9.500939943556383e-06,
+      "loss": 0.1601,
+      "step": 262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.090635912690094,
+      "learning_rate": 9.496366458271165e-06,
+      "loss": 0.1193,
+      "step": 263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4116455872624623,
+      "learning_rate": 9.491773223398855e-06,
+      "loss": 0.1142,
+      "step": 264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1217316271010223,
+      "learning_rate": 9.487160259114521e-06,
+      "loss": 0.1351,
+      "step": 265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9242039435529,
+      "learning_rate": 9.482527585679886e-06,
+      "loss": 0.119,
+      "step": 266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.315115479904517,
+      "learning_rate": 9.477875223443249e-06,
+      "loss": 0.1284,
+      "step": 267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2719860315913696,
+      "learning_rate": 9.473203192839379e-06,
+      "loss": 0.1721,
+      "step": 268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5947127743255973,
+      "learning_rate": 9.468511514389442e-06,
+      "loss": 0.1036,
+      "step": 269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9584745483264658,
+      "learning_rate": 9.463800208700904e-06,
+      "loss": 0.1192,
+      "step": 270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7447628859050024,
+      "learning_rate": 9.45906929646744e-06,
+      "loss": 0.1161,
+      "step": 271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.200458714414616,
+      "learning_rate": 9.454318798468838e-06,
+      "loss": 0.0977,
+      "step": 272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.213702512044038,
+      "learning_rate": 9.449548735570922e-06,
+      "loss": 0.129,
+      "step": 273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3993828599611606,
+      "learning_rate": 9.444759128725446e-06,
+      "loss": 0.1118,
+      "step": 274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6488629327673772,
+      "learning_rate": 9.439949998970012e-06,
+      "loss": 0.1289,
+      "step": 275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8771789314272584,
+      "learning_rate": 9.435121367427969e-06,
+      "loss": 0.1407,
+      "step": 276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0141415937656255,
+      "learning_rate": 9.430273255308329e-06,
+      "loss": 0.098,
+      "step": 277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6490964523017495,
+      "learning_rate": 9.425405683905664e-06,
+      "loss": 0.1119,
+      "step": 278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.949423514590385,
+      "learning_rate": 9.420518674600019e-06,
+      "loss": 0.0949,
+      "step": 279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.301200040078636,
+      "learning_rate": 9.415612248856824e-06,
+      "loss": 0.1356,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4438448938217925,
+      "learning_rate": 9.410686428226784e-06,
+      "loss": 0.1168,
+      "step": 281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8595023110908409,
+      "learning_rate": 9.4057412343458e-06,
+      "loss": 0.1565,
+      "step": 282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.015898381538342,
+      "learning_rate": 9.40077668893486e-06,
+      "loss": 0.1257,
+      "step": 283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7641459666937271,
+      "learning_rate": 9.395792813799954e-06,
+      "loss": 0.1369,
+      "step": 284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.35003300843501,
+      "learning_rate": 9.390789630831975e-06,
+      "loss": 0.1258,
+      "step": 285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9587992320280527,
+      "learning_rate": 9.385767162006626e-06,
+      "loss": 0.1198,
+      "step": 286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3927013768818504,
+      "learning_rate": 9.380725429384311e-06,
+      "loss": 0.1038,
+      "step": 287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9346564539370021,
+      "learning_rate": 9.375664455110056e-06,
+      "loss": 0.1057,
+      "step": 288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.969752960731544,
+      "learning_rate": 9.3705842614134e-06,
+      "loss": 0.1698,
+      "step": 289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7366125913154666,
+      "learning_rate": 9.365484870608298e-06,
+      "loss": 0.1228,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0991611774972725,
+      "learning_rate": 9.360366305093029e-06,
+      "loss": 0.0868,
+      "step": 291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2938664385866487,
+      "learning_rate": 9.355228587350091e-06,
+      "loss": 0.1236,
+      "step": 292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9417985269231075,
+      "learning_rate": 9.350071739946107e-06,
+      "loss": 0.1094,
+      "step": 293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8408239731332179,
+      "learning_rate": 9.344895785531725e-06,
+      "loss": 0.1349,
+      "step": 294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6860663646872776,
+      "learning_rate": 9.339700746841514e-06,
+      "loss": 0.1159,
+      "step": 295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7887989384287633,
+      "learning_rate": 9.334486646693868e-06,
+      "loss": 0.1307,
+      "step": 296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.071047508689909,
+      "learning_rate": 9.329253507990912e-06,
+      "loss": 0.1382,
+      "step": 297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9934009243327555,
+      "learning_rate": 9.324001353718386e-06,
+      "loss": 0.1131,
+      "step": 298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.293883374927684,
+      "learning_rate": 9.318730206945558e-06,
+      "loss": 0.1875,
+      "step": 299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.186485525987884,
+      "learning_rate": 9.31344009082512e-06,
+      "loss": 0.1332,
+      "step": 300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7888241804719054,
+      "learning_rate": 9.308131028593074e-06,
+      "loss": 0.0941,
+      "step": 301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4634800844949494,
+      "learning_rate": 9.30280304356865e-06,
+      "loss": 0.1507,
+      "step": 302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0519592864111837,
+      "learning_rate": 9.29745615915419e-06,
+      "loss": 0.1696,
+      "step": 303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.16519408228547,
+      "learning_rate": 9.292090398835048e-06,
+      "loss": 0.1282,
+      "step": 304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.128764797470375,
+      "learning_rate": 9.286705786179489e-06,
+      "loss": 0.1631,
+      "step": 305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.30218954081609,
+      "learning_rate": 9.281302344838579e-06,
+      "loss": 0.1316,
+      "step": 306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9155682605173827,
+      "learning_rate": 9.275880098546092e-06,
+      "loss": 0.1542,
+      "step": 307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5614986150201413,
+      "learning_rate": 9.270439071118398e-06,
+      "loss": 0.1614,
+      "step": 308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8525768538837735,
+      "learning_rate": 9.264979286454358e-06,
+      "loss": 0.1191,
+      "step": 309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.898080870573492,
+      "learning_rate": 9.259500768535226e-06,
+      "loss": 0.1181,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.472459360187466,
+      "learning_rate": 9.254003541424534e-06,
+      "loss": 0.1516,
+      "step": 311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.305568026320411,
+      "learning_rate": 9.248487629267994e-06,
+      "loss": 0.1615,
+      "step": 312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6544217551014015,
+      "learning_rate": 9.242953056293387e-06,
+      "loss": 0.1054,
+      "step": 313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8127935644971191,
+      "learning_rate": 9.23739984681046e-06,
+      "loss": 0.1441,
+      "step": 314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9498604123511916,
+      "learning_rate": 9.231828025210821e-06,
+      "loss": 0.0988,
+      "step": 315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9854409225946612,
+      "learning_rate": 9.226237615967822e-06,
+      "loss": 0.1331,
+      "step": 316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6409608855962516,
+      "learning_rate": 9.220628643636462e-06,
+      "loss": 0.0962,
+      "step": 317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.382784821567455,
+      "learning_rate": 9.215001132853277e-06,
+      "loss": 0.1395,
+      "step": 318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.967955661382828,
+      "learning_rate": 9.20935510833623e-06,
+      "loss": 0.1418,
+      "step": 319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9420469716155404,
+      "learning_rate": 9.2036905948846e-06,
+      "loss": 0.1504,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.002899106478933,
+      "learning_rate": 9.198007617378876e-06,
+      "loss": 0.155,
+      "step": 321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.978520034374933,
+      "learning_rate": 9.192306200780652e-06,
+      "loss": 0.1147,
+      "step": 322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.861974983143129,
+      "learning_rate": 9.186586370132508e-06,
+      "loss": 0.1141,
+      "step": 323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7627717021521454,
+      "learning_rate": 9.180848150557906e-06,
+      "loss": 0.1287,
+      "step": 324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.612676555695171,
+      "learning_rate": 9.175091567261078e-06,
+      "loss": 0.1125,
+      "step": 325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.643370148198158,
+      "learning_rate": 9.169316645526919e-06,
+      "loss": 0.1505,
+      "step": 326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5430892090106256,
+      "learning_rate": 9.163523410720866e-06,
+      "loss": 0.1229,
+      "step": 327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.696558808661286,
+      "learning_rate": 9.157711888288802e-06,
+      "loss": 0.0949,
+      "step": 328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.949950682230708,
+      "learning_rate": 9.15188210375693e-06,
+      "loss": 0.1075,
+      "step": 329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.281345406537966,
+      "learning_rate": 9.146034082731668e-06,
+      "loss": 0.1504,
+      "step": 330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.786416816817829,
+      "learning_rate": 9.140167850899533e-06,
+      "loss": 0.0944,
+      "step": 331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1707974071960687,
+      "learning_rate": 9.134283434027033e-06,
+      "loss": 0.1559,
+      "step": 332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.630770248631941,
+      "learning_rate": 9.12838085796055e-06,
+      "loss": 0.1168,
+      "step": 333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0322405623333113,
+      "learning_rate": 9.122460148626227e-06,
+      "loss": 0.1527,
+      "step": 334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7322055643416592,
+      "learning_rate": 9.116521332029852e-06,
+      "loss": 0.1367,
+      "step": 335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8373532226134297,
+      "learning_rate": 9.110564434256752e-06,
+      "loss": 0.1105,
+      "step": 336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.07554187753308,
+      "learning_rate": 9.104589481471668e-06,
+      "loss": 0.1334,
+      "step": 337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2236998943249664,
+      "learning_rate": 9.098596499918648e-06,
+      "loss": 0.125,
+      "step": 338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.597186712884655,
+      "learning_rate": 9.092585515920926e-06,
+      "loss": 0.1293,
+      "step": 339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7174798604288135,
+      "learning_rate": 9.08655655588081e-06,
+      "loss": 0.1263,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2256543091224077,
+      "learning_rate": 9.080509646279564e-06,
+      "loss": 0.1771,
+      "step": 341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4758883533703013,
+      "learning_rate": 9.074444813677297e-06,
+      "loss": 0.1727,
+      "step": 342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.783793767943786,
+      "learning_rate": 9.068362084712835e-06,
+      "loss": 0.1341,
+      "step": 343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9378176129085365,
+      "learning_rate": 9.062261486103614e-06,
+      "loss": 0.1598,
+      "step": 344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7783584831433117,
+      "learning_rate": 9.056143044645564e-06,
+      "loss": 0.1352,
+      "step": 345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6546196413315133,
+      "learning_rate": 9.05000678721298e-06,
+      "loss": 0.1232,
+      "step": 346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.118380902242969,
+      "learning_rate": 9.043852740758416e-06,
+      "loss": 0.1549,
+      "step": 347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6087772904564852,
+      "learning_rate": 9.037680932312557e-06,
+      "loss": 0.0881,
+      "step": 348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7254126145695363,
+      "learning_rate": 9.031491388984108e-06,
+      "loss": 0.115,
+      "step": 349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.981321894036563,
+      "learning_rate": 9.025284137959674e-06,
+      "loss": 0.1445,
+      "step": 350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.075752065166549,
+      "learning_rate": 9.019059206503632e-06,
+      "loss": 0.1353,
+      "step": 351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7852721093628698,
+      "learning_rate": 9.012816621958018e-06,
+      "loss": 0.1359,
+      "step": 352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.400157487370791,
+      "learning_rate": 9.006556411742415e-06,
+      "loss": 0.1049,
+      "step": 353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.51114251446299,
+      "learning_rate": 9.000278603353817e-06,
+      "loss": 0.0974,
+      "step": 354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.554006360375019,
+      "learning_rate": 8.993983224366514e-06,
+      "loss": 0.1627,
+      "step": 355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.292470957192174,
+      "learning_rate": 8.987670302431977e-06,
+      "loss": 0.1205,
+      "step": 356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2727328929301023,
+      "learning_rate": 8.981339865278726e-06,
+      "loss": 0.1248,
+      "step": 357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.79429792897825,
+      "learning_rate": 8.97499194071222e-06,
+      "loss": 0.0865,
+      "step": 358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.534656293841016,
+      "learning_rate": 8.968626556614723e-06,
+      "loss": 0.1852,
+      "step": 359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.451261116593459,
+      "learning_rate": 8.962243740945194e-06,
+      "loss": 0.1735,
+      "step": 360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6099662771095622,
+      "learning_rate": 8.955843521739152e-06,
+      "loss": 0.0949,
+      "step": 361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.952880241478913,
+      "learning_rate": 8.949425927108557e-06,
+      "loss": 0.1642,
+      "step": 362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1001999836849707,
+      "learning_rate": 8.942990985241694e-06,
+      "loss": 0.1966,
+      "step": 363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7791224591481396,
+      "learning_rate": 8.936538724403036e-06,
+      "loss": 0.1094,
+      "step": 364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7248924617963337,
+      "learning_rate": 8.930069172933133e-06,
+      "loss": 0.1511,
+      "step": 365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0501675667141672,
+      "learning_rate": 8.923582359248482e-06,
+      "loss": 0.1612,
+      "step": 366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3735141875296217,
+      "learning_rate": 8.917078311841395e-06,
+      "loss": 0.1712,
+      "step": 367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0407091995824618,
+      "learning_rate": 8.91055705927988e-06,
+      "loss": 0.1648,
+      "step": 368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6480885359011928,
+      "learning_rate": 8.904018630207526e-06,
+      "loss": 0.1097,
+      "step": 369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.082826171786718,
+      "learning_rate": 8.897463053343363e-06,
+      "loss": 0.115,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0778529231041247,
+      "learning_rate": 8.89089035748173e-06,
+      "loss": 0.1412,
+      "step": 371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8870564121587825,
+      "learning_rate": 8.884300571492173e-06,
+      "loss": 0.1215,
+      "step": 372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6728688014206319,
+      "learning_rate": 8.877693724319294e-06,
+      "loss": 0.1369,
+      "step": 373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5071568023441673,
+      "learning_rate": 8.871069844982639e-06,
+      "loss": 0.1422,
+      "step": 374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8405992455653366,
+      "learning_rate": 8.864428962576558e-06,
+      "loss": 0.1413,
+      "step": 375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.04100654612223,
+      "learning_rate": 8.8577711062701e-06,
+      "loss": 0.159,
+      "step": 376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7258745698933753,
+      "learning_rate": 8.851096305306846e-06,
+      "loss": 0.0891,
+      "step": 377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4535262921515013,
+      "learning_rate": 8.844404589004825e-06,
+      "loss": 0.0743,
+      "step": 378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.983927878208087,
+      "learning_rate": 8.837695986756351e-06,
+      "loss": 0.1091,
+      "step": 379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5547966658918195,
+      "learning_rate": 8.830970528027912e-06,
+      "loss": 0.1509,
+      "step": 380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.177776323767593,
+      "learning_rate": 8.824228242360035e-06,
+      "loss": 0.0905,
+      "step": 381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8634921804601332,
+      "learning_rate": 8.817469159367159e-06,
+      "loss": 0.1439,
+      "step": 382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8253919296872063,
+      "learning_rate": 8.810693308737493e-06,
+      "loss": 0.1371,
+      "step": 383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8264289794840303,
+      "learning_rate": 8.803900720232908e-06,
+      "loss": 0.1017,
+      "step": 384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7598193012149368,
+      "learning_rate": 8.797091423688787e-06,
+      "loss": 0.104,
+      "step": 385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.6146507935852075,
+      "learning_rate": 8.790265449013899e-06,
+      "loss": 0.169,
+      "step": 386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.138258612818048,
+      "learning_rate": 8.783422826190272e-06,
+      "loss": 0.141,
+      "step": 387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.777360621027715,
+      "learning_rate": 8.776563585273057e-06,
+      "loss": 0.1147,
+      "step": 388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0918180041637924,
+      "learning_rate": 8.769687756390401e-06,
+      "loss": 0.1118,
+      "step": 389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0832589535976274,
+      "learning_rate": 8.762795369743303e-06,
+      "loss": 0.1144,
+      "step": 390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7581039464137596,
+      "learning_rate": 8.755886455605499e-06,
+      "loss": 0.1251,
+      "step": 391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4128017025241764,
+      "learning_rate": 8.748961044323312e-06,
+      "loss": 0.118,
+      "step": 392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9739614747699874,
+      "learning_rate": 8.74201916631553e-06,
+      "loss": 0.124,
+      "step": 393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.823956299069119,
+      "learning_rate": 8.735060852073267e-06,
+      "loss": 0.1307,
+      "step": 394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8653973387035925,
+      "learning_rate": 8.72808613215983e-06,
+      "loss": 0.1167,
+      "step": 395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.043078073672084,
+      "learning_rate": 8.72109503721059e-06,
+      "loss": 0.1276,
+      "step": 396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3993371483045447,
+      "learning_rate": 8.714087597932837e-06,
+      "loss": 0.0887,
+      "step": 397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.155072263044344,
+      "learning_rate": 8.70706384510565e-06,
+      "loss": 0.1027,
+      "step": 398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.863716590798003,
+      "learning_rate": 8.700023809579772e-06,
+      "loss": 0.1426,
+      "step": 399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4148553401960213,
+      "learning_rate": 8.692967522277453e-06,
+      "loss": 0.1831,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2182007745276002,
+      "learning_rate": 8.685895014192336e-06,
+      "loss": 0.1189,
+      "step": 401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7745164161829752,
+      "learning_rate": 8.678806316389308e-06,
+      "loss": 0.1264,
+      "step": 402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9954907613170605,
+      "learning_rate": 8.671701460004362e-06,
+      "loss": 0.1119,
+      "step": 403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.443494772653162,
+      "learning_rate": 8.664580476244476e-06,
+      "loss": 0.1201,
+      "step": 404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6521852062845737,
+      "learning_rate": 8.657443396387456e-06,
+      "loss": 0.1533,
+      "step": 405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7783525006636822,
+      "learning_rate": 8.650290251781806e-06,
+      "loss": 0.136,
+      "step": 406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8404598865947792,
+      "learning_rate": 8.6431210738466e-06,
+      "loss": 0.1032,
+      "step": 407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.993842867577624,
+      "learning_rate": 8.635935894071332e-06,
+      "loss": 0.1504,
+      "step": 408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.23506410525159,
+      "learning_rate": 8.628734744015781e-06,
+      "loss": 0.1362,
+      "step": 409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5101646789391827,
+      "learning_rate": 8.621517655309872e-06,
+      "loss": 0.0944,
+      "step": 410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1222112692022446,
+      "learning_rate": 8.61428465965354e-06,
+      "loss": 0.1106,
+      "step": 411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.589971462883146,
+      "learning_rate": 8.60703578881659e-06,
+      "loss": 0.1078,
+      "step": 412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8152499697587898,
+      "learning_rate": 8.599771074638552e-06,
+      "loss": 0.1465,
+      "step": 413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7515501086572345,
+      "learning_rate": 8.592490549028549e-06,
+      "loss": 0.1133,
+      "step": 414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.074545833246127,
+      "learning_rate": 8.585194243965154e-06,
+      "loss": 0.1131,
+      "step": 415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6840561699332905,
+      "learning_rate": 8.577882191496244e-06,
+      "loss": 0.1251,
+      "step": 416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8878023794880057,
+      "learning_rate": 8.570554423738865e-06,
+      "loss": 0.1415,
+      "step": 417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3218353195989114,
+      "learning_rate": 8.563210972879099e-06,
+      "loss": 0.1421,
+      "step": 418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.809764155258485,
+      "learning_rate": 8.5558518711719e-06,
+      "loss": 0.0924,
+      "step": 419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.025310033583316,
+      "learning_rate": 8.548477150940976e-06,
+      "loss": 0.1101,
+      "step": 420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.495565142634987,
+      "learning_rate": 8.541086844578632e-06,
+      "loss": 0.1063,
+      "step": 421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.00374794205199,
+      "learning_rate": 8.533680984545632e-06,
+      "loss": 0.1318,
+      "step": 422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.698220498105753,
+      "learning_rate": 8.526259603371063e-06,
+      "loss": 0.1404,
+      "step": 423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0282444847472982,
+      "learning_rate": 8.518822733652179e-06,
+      "loss": 0.1609,
+      "step": 424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4428540102326999,
+      "learning_rate": 8.511370408054269e-06,
+      "loss": 0.1111,
+      "step": 425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.823326472349433,
+      "learning_rate": 8.503902659310511e-06,
+      "loss": 0.1303,
+      "step": 426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.975642411609206,
+      "learning_rate": 8.49641952022182e-06,
+      "loss": 0.1319,
+      "step": 427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0984423008283546,
+      "learning_rate": 8.488921023656717e-06,
+      "loss": 0.1255,
+      "step": 428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6327681226804796,
+      "learning_rate": 8.481407202551179e-06,
+      "loss": 0.1183,
+      "step": 429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7233323660269233,
+      "learning_rate": 8.47387808990849e-06,
+      "loss": 0.1385,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4460372159761063,
+      "learning_rate": 8.466333718799097e-06,
+      "loss": 0.1389,
+      "step": 431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.694191543079506,
+      "learning_rate": 8.458774122360479e-06,
+      "loss": 0.1357,
+      "step": 432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.541691399674599,
+      "learning_rate": 8.451199333796974e-06,
+      "loss": 0.0873,
+      "step": 433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5747686527537694,
+      "learning_rate": 8.44360938637966e-06,
+      "loss": 0.1408,
+      "step": 434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9808406911801135,
+      "learning_rate": 8.436004313446198e-06,
+      "loss": 0.1742,
+      "step": 435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8562173966255497,
+      "learning_rate": 8.428384148400679e-06,
+      "loss": 0.1224,
+      "step": 436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8682189381449035,
+      "learning_rate": 8.420748924713489e-06,
+      "loss": 0.1185,
+      "step": 437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.076463285377218,
+      "learning_rate": 8.413098675921154e-06,
+      "loss": 0.1169,
+      "step": 438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5595921907762054,
+      "learning_rate": 8.405433435626198e-06,
+      "loss": 0.0771,
+      "step": 439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6400626849748228,
+      "learning_rate": 8.397753237496989e-06,
+      "loss": 0.1257,
+      "step": 440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9595113341341717,
+      "learning_rate": 8.390058115267599e-06,
+      "loss": 0.1308,
+      "step": 441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9753446481614831,
+      "learning_rate": 8.38234810273765e-06,
+      "loss": 0.1197,
+      "step": 442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.088782693449183,
+      "learning_rate": 8.374623233772166e-06,
+      "loss": 0.1393,
+      "step": 443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.77933848636558,
+      "learning_rate": 8.366883542301428e-06,
+      "loss": 0.1237,
+      "step": 444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3432881788234496,
+      "learning_rate": 8.35912906232082e-06,
+      "loss": 0.1008,
+      "step": 445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.542175556656122,
+      "learning_rate": 8.351359827890686e-06,
+      "loss": 0.1102,
+      "step": 446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0198766647420334,
+      "learning_rate": 8.343575873136174e-06,
+      "loss": 0.115,
+      "step": 447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9031797249550388,
+      "learning_rate": 8.335777232247086e-06,
+      "loss": 0.1523,
+      "step": 448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.179797475684397,
+      "learning_rate": 8.327963939477736e-06,
+      "loss": 0.1241,
+      "step": 449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.312455490674629,
+      "learning_rate": 8.320136029146792e-06,
+      "loss": 0.1833,
+      "step": 450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.260180433787793,
+      "learning_rate": 8.312293535637123e-06,
+      "loss": 0.1546,
+      "step": 451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6785202678147257,
+      "learning_rate": 8.304436493395663e-06,
+      "loss": 0.1455,
+      "step": 452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9858223819669532,
+      "learning_rate": 8.296564936933237e-06,
+      "loss": 0.1434,
+      "step": 453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9102543474762492,
+      "learning_rate": 8.28867890082443e-06,
+      "loss": 0.1312,
+      "step": 454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4219016246969467,
+      "learning_rate": 8.280778419707421e-06,
+      "loss": 0.1097,
+      "step": 455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4863931985205288,
+      "learning_rate": 8.27286352828384e-06,
+      "loss": 0.1032,
+      "step": 456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4262744449759,
+      "learning_rate": 8.26493426131861e-06,
+      "loss": 0.0883,
+      "step": 457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8849980135344147,
+      "learning_rate": 8.256990653639798e-06,
+      "loss": 0.1465,
+      "step": 458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.90921490609553,
+      "learning_rate": 8.24903274013846e-06,
+      "loss": 0.1027,
+      "step": 459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9345635072068623,
+      "learning_rate": 8.241060555768487e-06,
+      "loss": 0.1706,
+      "step": 460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0906273304065,
+      "learning_rate": 8.23307413554645e-06,
+      "loss": 0.1054,
+      "step": 461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7010263154623813,
+      "learning_rate": 8.225073514551458e-06,
+      "loss": 0.117,
+      "step": 462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.856999714971948,
+      "learning_rate": 8.217058727924985e-06,
+      "loss": 0.099,
+      "step": 463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.096162508384875,
+      "learning_rate": 8.20902981087073e-06,
+      "loss": 0.0992,
+      "step": 464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1897409229932543,
+      "learning_rate": 8.200986798654454e-06,
+      "loss": 0.1353,
+      "step": 465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0107497205490326,
+      "learning_rate": 8.192929726603834e-06,
+      "loss": 0.12,
+      "step": 466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6764154097639425,
+      "learning_rate": 8.184858630108301e-06,
+      "loss": 0.0961,
+      "step": 467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5132331514800432,
+      "learning_rate": 8.176773544618884e-06,
+      "loss": 0.1417,
+      "step": 468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6290482258431218,
+      "learning_rate": 8.168674505648055e-06,
+      "loss": 0.0847,
+      "step": 469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6373641890196442,
+      "learning_rate": 8.16056154876958e-06,
+      "loss": 0.0994,
+      "step": 470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.845726502290335,
+      "learning_rate": 8.152434709618355e-06,
+      "loss": 0.0841,
+      "step": 471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.155909263655658,
+      "learning_rate": 8.144294023890249e-06,
+      "loss": 0.1398,
+      "step": 472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.212072051455184,
+      "learning_rate": 8.136139527341954e-06,
+      "loss": 0.1668,
+      "step": 473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0126155869943188,
+      "learning_rate": 8.12797125579082e-06,
+      "loss": 0.1522,
+      "step": 474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4551386753433713,
+      "learning_rate": 8.119789245114704e-06,
+      "loss": 0.112,
+      "step": 475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8418803359771359,
+      "learning_rate": 8.111593531251812e-06,
+      "loss": 0.1646,
+      "step": 476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8692838890636365,
+      "learning_rate": 8.103384150200535e-06,
+      "loss": 0.1035,
+      "step": 477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9377035472958686,
+      "learning_rate": 8.095161138019294e-06,
+      "loss": 0.0809,
+      "step": 478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.587432958669322,
+      "learning_rate": 8.086924530826386e-06,
+      "loss": 0.1454,
+      "step": 479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.02896456008906,
+      "learning_rate": 8.078674364799823e-06,
+      "loss": 0.1085,
+      "step": 480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7655639368460225,
+      "learning_rate": 8.070410676177171e-06,
+      "loss": 0.1396,
+      "step": 481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1284960143825113,
+      "learning_rate": 8.062133501255388e-06,
+      "loss": 0.148,
+      "step": 482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0034337679566354,
+      "learning_rate": 8.053842876390673e-06,
+      "loss": 0.0812,
+      "step": 483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.85902216889917,
+      "learning_rate": 8.045538837998299e-06,
+      "loss": 0.1333,
+      "step": 484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7602908592648836,
+      "learning_rate": 8.03722142255246e-06,
+      "loss": 0.153,
+      "step": 485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3218830655644367,
+      "learning_rate": 8.0288906665861e-06,
+      "loss": 0.1066,
+      "step": 486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1584017599108836,
+      "learning_rate": 8.020546606690767e-06,
+      "loss": 0.1519,
+      "step": 487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7606072803807844,
+      "learning_rate": 8.012189279516435e-06,
+      "loss": 0.1137,
+      "step": 488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.888831011490011,
+      "learning_rate": 8.003818721771364e-06,
+      "loss": 0.1037,
+      "step": 489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.170836406338443,
+      "learning_rate": 7.995434970221915e-06,
+      "loss": 0.1317,
+      "step": 490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7104103028650062,
+      "learning_rate": 7.987038061692412e-06,
+      "loss": 0.1219,
+      "step": 491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.995512990844272,
+      "learning_rate": 7.97862803306496e-06,
+      "loss": 0.137,
+      "step": 492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8348429648687326,
+      "learning_rate": 7.970204921279296e-06,
+      "loss": 0.1202,
+      "step": 493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9520817239757213,
+      "learning_rate": 7.961768763332624e-06,
+      "loss": 0.1318,
+      "step": 494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8676185821725126,
+      "learning_rate": 7.953319596279447e-06,
+      "loss": 0.1976,
+      "step": 495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6485722163392649,
+      "learning_rate": 7.944857457231415e-06,
+      "loss": 0.0991,
+      "step": 496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8645910595696418,
+      "learning_rate": 7.936382383357149e-06,
+      "loss": 0.1366,
+      "step": 497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0059127758512467,
+      "learning_rate": 7.927894411882086e-06,
+      "loss": 0.1127,
+      "step": 498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.836596301752791,
+      "learning_rate": 7.919393580088317e-06,
+      "loss": 0.1194,
+      "step": 499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8460003077373996,
+      "learning_rate": 7.910879925314413e-06,
+      "loss": 0.1495,
+      "step": 500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2392969367957027,
+      "learning_rate": 7.902353484955277e-06,
+      "loss": 0.1167,
+      "step": 501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0193536716263605,
+      "learning_rate": 7.893814296461964e-06,
+      "loss": 0.1477,
+      "step": 502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0299360866994416,
+      "learning_rate": 7.885262397341524e-06,
+      "loss": 0.1795,
+      "step": 503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.148228071373966,
+      "learning_rate": 7.876697825156841e-06,
+      "loss": 0.13,
+      "step": 504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.044815106464104,
+      "learning_rate": 7.868120617526456e-06,
+      "loss": 0.1135,
+      "step": 505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8904373956408558,
+      "learning_rate": 7.859530812124416e-06,
+      "loss": 0.1556,
+      "step": 506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8046682434008035,
+      "learning_rate": 7.850928446680099e-06,
+      "loss": 0.1452,
+      "step": 507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1190190347465143,
+      "learning_rate": 7.84231355897805e-06,
+      "loss": 0.1561,
+      "step": 508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.047494027090662,
+      "learning_rate": 7.833686186857815e-06,
+      "loss": 0.1646,
+      "step": 509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8698391434382085,
+      "learning_rate": 7.825046368213782e-06,
+      "loss": 0.1391,
+      "step": 510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4357018771329595,
+      "learning_rate": 7.816394140995004e-06,
+      "loss": 0.1204,
+      "step": 511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4093119255939646,
+      "learning_rate": 7.807729543205035e-06,
+      "loss": 0.0861,
+      "step": 512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9753965478870885,
+      "learning_rate": 7.799052612901767e-06,
+      "loss": 0.1548,
+      "step": 513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.606627357271704,
+      "learning_rate": 7.790363388197263e-06,
+      "loss": 0.1268,
+      "step": 514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0943568037356117,
+      "learning_rate": 7.781661907257581e-06,
+      "loss": 0.1215,
+      "step": 515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4044794608294802,
+      "learning_rate": 7.77294820830262e-06,
+      "loss": 0.1042,
+      "step": 516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 32.00675322832298,
+      "learning_rate": 7.764222329605939e-06,
+      "loss": 0.1333,
+      "step": 517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.408529222244952,
+      "learning_rate": 7.755484309494599e-06,
+      "loss": 0.0864,
+      "step": 518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8212939593374105,
+      "learning_rate": 7.746734186348986e-06,
+      "loss": 0.1593,
+      "step": 519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.137823332099506,
+      "learning_rate": 7.737971998602648e-06,
+      "loss": 0.1586,
+      "step": 520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4537608630192507,
+      "learning_rate": 7.729197784742125e-06,
+      "loss": 0.1515,
+      "step": 521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1365319942561576,
+      "learning_rate": 7.720411583306784e-06,
+      "loss": 0.1334,
+      "step": 522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.224916240795795,
+      "learning_rate": 7.711613432888639e-06,
+      "loss": 0.117,
+      "step": 523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8956601461096838,
+      "learning_rate": 7.702803372132192e-06,
+      "loss": 0.1451,
+      "step": 524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.130808445432382,
+      "learning_rate": 7.693981439734257e-06,
+      "loss": 0.1382,
+      "step": 525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9731631849430387,
+      "learning_rate": 7.685147674443796e-06,
+      "loss": 0.1499,
+      "step": 526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4952826434696695,
+      "learning_rate": 7.676302115061742e-06,
+      "loss": 0.0755,
+      "step": 527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4755572571402387,
+      "learning_rate": 7.66744480044083e-06,
+      "loss": 0.1635,
+      "step": 528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.384504040489573,
+      "learning_rate": 7.658575769485432e-06,
+      "loss": 0.1455,
+      "step": 529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.234483049009127,
+      "learning_rate": 7.649695061151383e-06,
+      "loss": 0.1507,
+      "step": 530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3480972290543916,
+      "learning_rate": 7.640802714445803e-06,
+      "loss": 0.1164,
+      "step": 531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.449755875741938,
+      "learning_rate": 7.631898768426938e-06,
+      "loss": 0.1632,
+      "step": 532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9895634278912286,
+      "learning_rate": 7.6229832622039776e-06,
+      "loss": 0.1249,
+      "step": 533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.38742227043979,
+      "learning_rate": 7.61405623493689e-06,
+      "loss": 0.1292,
+      "step": 534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0347892998299764,
+      "learning_rate": 7.605117725836251e-06,
+      "loss": 0.119,
+      "step": 535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.390316072677058,
+      "learning_rate": 7.596167774163061e-06,
+      "loss": 0.1243,
+      "step": 536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.204263864515519,
+      "learning_rate": 7.587206419228587e-06,
+      "loss": 0.1239,
+      "step": 537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3222180687668357,
+      "learning_rate": 7.578233700394178e-06,
+      "loss": 0.0961,
+      "step": 538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8999977267746069,
+      "learning_rate": 7.569249657071102e-06,
+      "loss": 0.1176,
+      "step": 539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7389595861953973,
+      "learning_rate": 7.560254328720362e-06,
+      "loss": 0.1073,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2864263275239125,
+      "learning_rate": 7.551247754852535e-06,
+      "loss": 0.125,
+      "step": 541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.160533253201328,
+      "learning_rate": 7.542229975027591e-06,
+      "loss": 0.1067,
+      "step": 542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.328287081556125,
+      "learning_rate": 7.533201028854713e-06,
+      "loss": 0.1418,
+      "step": 543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2510136986424896,
+      "learning_rate": 7.524160955992142e-06,
+      "loss": 0.1836,
+      "step": 544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9837329996021282,
+      "learning_rate": 7.515109796146982e-06,
+      "loss": 0.1234,
+      "step": 545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7815364070607005,
+      "learning_rate": 7.506047589075041e-06,
+      "loss": 0.1016,
+      "step": 546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4046682954283596,
+      "learning_rate": 7.496974374580645e-06,
+      "loss": 0.1587,
+      "step": 547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.124077258722043,
+      "learning_rate": 7.487890192516472e-06,
+      "loss": 0.1121,
+      "step": 548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0358602644377037,
+      "learning_rate": 7.478795082783374e-06,
+      "loss": 0.0946,
+      "step": 549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.988650043863186,
+      "learning_rate": 7.469689085330196e-06,
+      "loss": 0.1421,
+      "step": 550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8578935497521591,
+      "learning_rate": 7.4605722401536106e-06,
+      "loss": 0.1546,
+      "step": 551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.735125797447667,
+      "learning_rate": 7.451444587297937e-06,
+      "loss": 0.0851,
+      "step": 552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9757497485369377,
+      "learning_rate": 7.4423061668549625e-06,
+      "loss": 0.1119,
+      "step": 553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.684954557915866,
+      "learning_rate": 7.433157018963771e-06,
+      "loss": 0.1222,
+      "step": 554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5222415804018157,
+      "learning_rate": 7.423997183810565e-06,
+      "loss": 0.0939,
+      "step": 555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9844507330670464,
+      "learning_rate": 7.41482670162849e-06,
+      "loss": 0.1415,
+      "step": 556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7027159832899936,
+      "learning_rate": 7.405645612697456e-06,
+      "loss": 0.1363,
+      "step": 557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8367490333126868,
+      "learning_rate": 7.396453957343961e-06,
+      "loss": 0.1203,
+      "step": 558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2815963303383255,
+      "learning_rate": 7.3872517759409135e-06,
+      "loss": 0.1599,
+      "step": 559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.994412011363053,
+      "learning_rate": 7.378039108907461e-06,
+      "loss": 0.1233,
+      "step": 560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9423287451389104,
+      "learning_rate": 7.368815996708801e-06,
+      "loss": 0.1244,
+      "step": 561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5723789324475477,
+      "learning_rate": 7.3595824798560135e-06,
+      "loss": 0.1762,
+      "step": 562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.512361293049007,
+      "learning_rate": 7.350338598905878e-06,
+      "loss": 0.1144,
+      "step": 563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7297775495609162,
+      "learning_rate": 7.341084394460698e-06,
+      "loss": 0.1391,
+      "step": 564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0229486702543236,
+      "learning_rate": 7.331819907168121e-06,
+      "loss": 0.1221,
+      "step": 565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.32719352344108,
+      "learning_rate": 7.3225451777209585e-06,
+      "loss": 0.1471,
+      "step": 566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.166222490887353,
+      "learning_rate": 7.313260246857013e-06,
+      "loss": 0.1238,
+      "step": 567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.114168239547472,
+      "learning_rate": 7.303965155358888e-06,
+      "loss": 0.1216,
+      "step": 568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8124445478580515,
+      "learning_rate": 7.294659944053822e-06,
+      "loss": 0.15,
+      "step": 569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2535216557090525,
+      "learning_rate": 7.285344653813505e-06,
+      "loss": 0.0811,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.047141351070446,
+      "learning_rate": 7.276019325553891e-06,
+      "loss": 0.1289,
+      "step": 571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0581370858454067,
+      "learning_rate": 7.266684000235028e-06,
+      "loss": 0.1086,
+      "step": 572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8854610758202903,
+      "learning_rate": 7.2573387188608735e-06,
+      "loss": 0.1266,
+      "step": 573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.829659439964262,
+      "learning_rate": 7.247983522479114e-06,
+      "loss": 0.1359,
+      "step": 574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.430253144187847,
+      "learning_rate": 7.238618452180991e-06,
+      "loss": 0.1242,
+      "step": 575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.3314444564097485,
+      "learning_rate": 7.22924354910111e-06,
+      "loss": 0.1509,
+      "step": 576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5681898497611435,
+      "learning_rate": 7.21985885441727e-06,
+      "loss": 0.1167,
+      "step": 577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5745929569582897,
+      "learning_rate": 7.210464409350275e-06,
+      "loss": 0.1169,
+      "step": 578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9071600113183031,
+      "learning_rate": 7.201060255163755e-06,
+      "loss": 0.1176,
+      "step": 579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5857425715812004,
+      "learning_rate": 7.191646433163992e-06,
+      "loss": 0.1221,
+      "step": 580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.882899987743563,
+      "learning_rate": 7.1822229846997246e-06,
+      "loss": 0.1043,
+      "step": 581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.738099622547025,
+      "learning_rate": 7.172789951161979e-06,
+      "loss": 0.1308,
+      "step": 582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4937477548191893,
+      "learning_rate": 7.1633473739838824e-06,
+      "loss": 0.1169,
+      "step": 583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8851301740589876,
+      "learning_rate": 7.153895294640476e-06,
+      "loss": 0.1128,
+      "step": 584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2372720900797014,
+      "learning_rate": 7.144433754648545e-06,
+      "loss": 0.1492,
+      "step": 585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.348361521037077,
+      "learning_rate": 7.134962795566425e-06,
+      "loss": 0.0779,
+      "step": 586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.048315586664474,
+      "learning_rate": 7.125482458993821e-06,
+      "loss": 0.1393,
+      "step": 587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7206749477746093,
+      "learning_rate": 7.115992786571633e-06,
+      "loss": 0.1983,
+      "step": 588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6769013220582096,
+      "learning_rate": 7.106493819981763e-06,
+      "loss": 0.1311,
+      "step": 589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8042305785903017,
+      "learning_rate": 7.096985600946937e-06,
+      "loss": 0.1074,
+      "step": 590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.252250341498427,
+      "learning_rate": 7.0874681712305236e-06,
+      "loss": 0.0803,
+      "step": 591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.934854094701551,
+      "learning_rate": 7.0779415726363446e-06,
+      "loss": 0.1035,
+      "step": 592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.888534086971409,
+      "learning_rate": 7.0684058470084946e-06,
+      "loss": 0.1445,
+      "step": 593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1608538431314646,
+      "learning_rate": 7.0588610362311595e-06,
+      "loss": 0.1585,
+      "step": 594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0745478481957553,
+      "learning_rate": 7.049307182228428e-06,
+      "loss": 0.1936,
+      "step": 595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1127380539648977,
+      "learning_rate": 7.0397443269641155e-06,
+      "loss": 0.1696,
+      "step": 596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4624486732374409,
+      "learning_rate": 7.030172512441563e-06,
+      "loss": 0.0657,
+      "step": 597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.026613587742609,
+      "learning_rate": 7.020591780703474e-06,
+      "loss": 0.1483,
+      "step": 598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9299575112371328,
+      "learning_rate": 7.011002173831714e-06,
+      "loss": 0.1451,
+      "step": 599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.977416407574812,
+      "learning_rate": 7.001403733947134e-06,
+      "loss": 0.1248,
+      "step": 600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.208409953766827,
+      "learning_rate": 6.991796503209382e-06,
+      "loss": 0.165,
+      "step": 601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.78537720018225,
+      "learning_rate": 6.982180523816715e-06,
+      "loss": 0.1323,
+      "step": 602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4678620733297283,
+      "learning_rate": 6.972555838005823e-06,
+      "loss": 0.136,
+      "step": 603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6804545502803885,
+      "learning_rate": 6.962922488051632e-06,
+      "loss": 0.2074,
+      "step": 604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.327651450853166,
+      "learning_rate": 6.953280516267129e-06,
+      "loss": 0.1505,
+      "step": 605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.751593814388165,
+      "learning_rate": 6.943629965003167e-06,
+      "loss": 0.1305,
+      "step": 606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.897509617686097,
+      "learning_rate": 6.933970876648284e-06,
+      "loss": 0.1226,
+      "step": 607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.630112664652096,
+      "learning_rate": 6.924303293628517e-06,
+      "loss": 0.1342,
+      "step": 608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7661430864522407,
+      "learning_rate": 6.914627258407214e-06,
+      "loss": 0.1188,
+      "step": 609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7454628935933312,
+      "learning_rate": 6.9049428134848475e-06,
+      "loss": 0.1326,
+      "step": 610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8984246938264262,
+      "learning_rate": 6.895250001398828e-06,
+      "loss": 0.0928,
+      "step": 611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.277357201489613,
+      "learning_rate": 6.885548864723319e-06,
+      "loss": 0.1234,
+      "step": 612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.758879562191011,
+      "learning_rate": 6.875839446069048e-06,
+      "loss": 0.1347,
+      "step": 613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8599771677057306,
+      "learning_rate": 6.866121788083118e-06,
+      "loss": 0.1581,
+      "step": 614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6561180713973451,
+      "learning_rate": 6.856395933448823e-06,
+      "loss": 0.1096,
+      "step": 615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.748117733072006,
+      "learning_rate": 6.846661924885461e-06,
+      "loss": 0.1261,
+      "step": 616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8826540597064463,
+      "learning_rate": 6.836919805148142e-06,
+      "loss": 0.1331,
+      "step": 617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.977092053423411,
+      "learning_rate": 6.827169617027607e-06,
+      "loss": 0.1258,
+      "step": 618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.168255589615196,
+      "learning_rate": 6.817411403350031e-06,
+      "loss": 0.0836,
+      "step": 619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9318484653801353,
+      "learning_rate": 6.807645206976847e-06,
+      "loss": 0.0789,
+      "step": 620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7287551190043524,
+      "learning_rate": 6.797871070804543e-06,
+      "loss": 0.1202,
+      "step": 621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4782462748125278,
+      "learning_rate": 6.788089037764487e-06,
+      "loss": 0.1157,
+      "step": 622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7883602516215786,
+      "learning_rate": 6.7782991508227295e-06,
+      "loss": 0.0899,
+      "step": 623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0292778027776923,
+      "learning_rate": 6.768501452979817e-06,
+      "loss": 0.1565,
+      "step": 624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1898092797254387,
+      "learning_rate": 6.758695987270609e-06,
+      "loss": 0.1279,
+      "step": 625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5131328560579043,
+      "learning_rate": 6.748882796764083e-06,
+      "loss": 0.1098,
+      "step": 626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.765825136606413,
+      "learning_rate": 6.739061924563141e-06,
+      "loss": 0.1024,
+      "step": 627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.013126733866634,
+      "learning_rate": 6.729233413804434e-06,
+      "loss": 0.1135,
+      "step": 628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5407548773847035,
+      "learning_rate": 6.719397307658154e-06,
+      "loss": 0.1041,
+      "step": 629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8806968993386401,
+      "learning_rate": 6.709553649327865e-06,
+      "loss": 0.1386,
+      "step": 630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6939211307688753,
+      "learning_rate": 6.699702482050298e-06,
+      "loss": 0.1329,
+      "step": 631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5237569971474656,
+      "learning_rate": 6.689843849095164e-06,
+      "loss": 0.1758,
+      "step": 632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5189326500238918,
+      "learning_rate": 6.679977793764969e-06,
+      "loss": 0.1293,
+      "step": 633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6096633555014934,
+      "learning_rate": 6.6701043593948174e-06,
+      "loss": 0.1214,
+      "step": 634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3765375159205888,
+      "learning_rate": 6.6602235893522294e-06,
+      "loss": 0.1259,
+      "step": 635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9383351279687073,
+      "learning_rate": 6.650335527036943e-06,
+      "loss": 0.1437,
+      "step": 636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1740605725514452,
+      "learning_rate": 6.640440215880726e-06,
+      "loss": 0.1197,
+      "step": 637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8553635326474456,
+      "learning_rate": 6.630537699347187e-06,
+      "loss": 0.116,
+      "step": 638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6232991847863223,
+      "learning_rate": 6.620628020931581e-06,
+      "loss": 0.1031,
+      "step": 639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5829400518191838,
+      "learning_rate": 6.6107112241606255e-06,
+      "loss": 0.0908,
+      "step": 640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0288417622221298,
+      "learning_rate": 6.600787352592297e-06,
+      "loss": 0.0982,
+      "step": 641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.984936118733807,
+      "learning_rate": 6.590856449815654e-06,
+      "loss": 0.1517,
+      "step": 642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.805427898334222,
+      "learning_rate": 6.580918559450632e-06,
+      "loss": 0.1266,
+      "step": 643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8927659434737785,
+      "learning_rate": 6.5709737251478646e-06,
+      "loss": 0.1593,
+      "step": 644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.785448329054995,
+      "learning_rate": 6.561021990588479e-06,
+      "loss": 0.1226,
+      "step": 645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.113868351962621,
+      "learning_rate": 6.551063399483919e-06,
+      "loss": 0.1294,
+      "step": 646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.950094914671727,
+      "learning_rate": 6.541097995575737e-06,
+      "loss": 0.1589,
+      "step": 647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1112170290978525,
+      "learning_rate": 6.531125822635413e-06,
+      "loss": 0.1329,
+      "step": 648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5475797916627902,
+      "learning_rate": 6.521146924464157e-06,
+      "loss": 0.1196,
+      "step": 649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9767440453636411,
+      "learning_rate": 6.511161344892721e-06,
+      "loss": 0.1109,
+      "step": 650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6265570229799957,
+      "learning_rate": 6.501169127781205e-06,
+      "loss": 0.1265,
+      "step": 651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8140713966796398,
+      "learning_rate": 6.491170317018859e-06,
+      "loss": 0.0926,
+      "step": 652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.512997483717891,
+      "learning_rate": 6.481164956523898e-06,
+      "loss": 0.1062,
+      "step": 653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5835209330529558,
+      "learning_rate": 6.4711530902433024e-06,
+      "loss": 0.1214,
+      "step": 654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1810482657565533,
+      "learning_rate": 6.461134762152634e-06,
+      "loss": 0.1884,
+      "step": 655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.751236145968338,
+      "learning_rate": 6.45111001625583e-06,
+      "loss": 0.1357,
+      "step": 656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1664736568653233,
+      "learning_rate": 6.441078896585024e-06,
+      "loss": 0.2049,
+      "step": 657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2044176541204767,
+      "learning_rate": 6.431041447200339e-06,
+      "loss": 0.1157,
+      "step": 658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8820357970464658,
+      "learning_rate": 6.4209977121897025e-06,
+      "loss": 0.1335,
+      "step": 659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5346932864729,
+      "learning_rate": 6.4109477356686545e-06,
+      "loss": 0.0982,
+      "step": 660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4430126211684466,
+      "learning_rate": 6.400891561780144e-06,
+      "loss": 0.0938,
+      "step": 661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6600799015614198,
+      "learning_rate": 6.390829234694345e-06,
+      "loss": 0.1057,
+      "step": 662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6253330956260457,
+      "learning_rate": 6.380760798608452e-06,
+      "loss": 0.0828,
+      "step": 663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9213558492054916,
+      "learning_rate": 6.370686297746504e-06,
+      "loss": 0.0921,
+      "step": 664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5362733191840696,
+      "learning_rate": 6.36060577635917e-06,
+      "loss": 0.1105,
+      "step": 665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8321777012247429,
+      "learning_rate": 6.350519278723563e-06,
+      "loss": 0.1472,
+      "step": 666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7131830343394663,
+      "learning_rate": 6.340426849143048e-06,
+      "loss": 0.1199,
+      "step": 667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5816409557797964,
+      "learning_rate": 6.330328531947045e-06,
+      "loss": 0.147,
+      "step": 668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.123874168031327,
+      "learning_rate": 6.3202243714908374e-06,
+      "loss": 0.1399,
+      "step": 669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.816343510012244,
+      "learning_rate": 6.310114412155369e-06,
+      "loss": 0.1185,
+      "step": 670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.584769637698625,
+      "learning_rate": 6.299998698347055e-06,
+      "loss": 0.0931,
+      "step": 671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9058687524566538,
+      "learning_rate": 6.28987727449759e-06,
+      "loss": 0.1248,
+      "step": 672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.548949465664935,
+      "learning_rate": 6.2797501850637465e-06,
+      "loss": 0.1294,
+      "step": 673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1149962944001177,
+      "learning_rate": 6.2696174745271795e-06,
+      "loss": 0.1329,
+      "step": 674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9168275656736968,
+      "learning_rate": 6.25947918739424e-06,
+      "loss": 0.0828,
+      "step": 675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.57954895290732,
+      "learning_rate": 6.249335368195771e-06,
+      "loss": 0.1295,
+      "step": 676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2957808654358192,
+      "learning_rate": 6.239186061486911e-06,
+      "loss": 0.1083,
+      "step": 677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1345641796436556,
+      "learning_rate": 6.229031311846902e-06,
+      "loss": 0.1097,
+      "step": 678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7952238152505315,
+      "learning_rate": 6.218871163878899e-06,
+      "loss": 0.0948,
+      "step": 679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8358217756709838,
+      "learning_rate": 6.208705662209763e-06,
+      "loss": 0.1335,
+      "step": 680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0599905715702804,
+      "learning_rate": 6.198534851489872e-06,
+      "loss": 0.1316,
+      "step": 681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7338996737841588,
+      "learning_rate": 6.188358776392921e-06,
+      "loss": 0.129,
+      "step": 682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1122584913953952,
+      "learning_rate": 6.178177481615731e-06,
+      "loss": 0.1262,
+      "step": 683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9439971478444478,
+      "learning_rate": 6.1679910118780485e-06,
+      "loss": 0.1256,
+      "step": 684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.84689942662513,
+      "learning_rate": 6.1577994119223505e-06,
+      "loss": 0.1147,
+      "step": 685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5530955100809665,
+      "learning_rate": 6.147602726513648e-06,
+      "loss": 0.1103,
+      "step": 686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2899708375245047,
+      "learning_rate": 6.137401000439286e-06,
+      "loss": 0.0754,
+      "step": 687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9394761284686224,
+      "learning_rate": 6.127194278508753e-06,
+      "loss": 0.1672,
+      "step": 688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5800890047158567,
+      "learning_rate": 6.116982605553482e-06,
+      "loss": 0.1557,
+      "step": 689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3971769786049262,
+      "learning_rate": 6.1067660264266496e-06,
+      "loss": 0.1043,
+      "step": 690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5778275295516868,
+      "learning_rate": 6.096544586002983e-06,
+      "loss": 0.0854,
+      "step": 691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8220202039345554,
+      "learning_rate": 6.0863183291785626e-06,
+      "loss": 0.1207,
+      "step": 692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3027913066854846,
+      "learning_rate": 6.076087300870622e-06,
+      "loss": 0.1567,
+      "step": 693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.696851243751135,
+      "learning_rate": 6.065851546017357e-06,
+      "loss": 0.1497,
+      "step": 694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8274017982330293,
+      "learning_rate": 6.055611109577722e-06,
+      "loss": 0.1295,
+      "step": 695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1597068887132207,
+      "learning_rate": 6.045366036531229e-06,
+      "loss": 0.1201,
+      "step": 696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.104931146165314,
+      "learning_rate": 6.035116371877766e-06,
+      "loss": 0.1434,
+      "step": 697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6266980971348304,
+      "learning_rate": 6.024862160637379e-06,
+      "loss": 0.1101,
+      "step": 698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4189796649639415,
+      "learning_rate": 6.014603447850091e-06,
+      "loss": 0.1249,
+      "step": 699
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4298746279674857,
+      "learning_rate": 6.004340278575695e-06,
+      "loss": 0.0841,
+      "step": 700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9802183561373556,
+      "learning_rate": 5.994072697893559e-06,
+      "loss": 0.1289,
+      "step": 701
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.905303404537033,
+      "learning_rate": 5.983800750902425e-06,
+      "loss": 0.1862,
+      "step": 702
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3821102495196778,
+      "learning_rate": 5.973524482720216e-06,
+      "loss": 0.0751,
+      "step": 703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.87233322122542,
+      "learning_rate": 5.963243938483834e-06,
+      "loss": 0.1401,
+      "step": 704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.338191434797121,
+      "learning_rate": 5.952959163348965e-06,
+      "loss": 0.1393,
+      "step": 705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9099485834004324,
+      "learning_rate": 5.942670202489874e-06,
+      "loss": 0.1457,
+      "step": 706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0645583006230246,
+      "learning_rate": 5.932377101099215e-06,
+      "loss": 0.1069,
+      "step": 707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5884995110913243,
+      "learning_rate": 5.922079904387826e-06,
+      "loss": 0.1397,
+      "step": 708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1415758056462293,
+      "learning_rate": 5.911778657584538e-06,
+      "loss": 0.1029,
+      "step": 709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5354706525429889,
+      "learning_rate": 5.9014734059359665e-06,
+      "loss": 0.0856,
+      "step": 710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9302822282714498,
+      "learning_rate": 5.891164194706318e-06,
+      "loss": 0.0894,
+      "step": 711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.129932164410851,
+      "learning_rate": 5.880851069177193e-06,
+      "loss": 0.1306,
+      "step": 712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.53369839366938,
+      "learning_rate": 5.870534074647382e-06,
+      "loss": 0.1148,
+      "step": 713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5416405545595093,
+      "learning_rate": 5.860213256432674e-06,
+      "loss": 0.1018,
+      "step": 714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3749980185297184,
+      "learning_rate": 5.849888659865649e-06,
+      "loss": 0.0845,
+      "step": 715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6540761654729228,
+      "learning_rate": 5.839560330295485e-06,
+      "loss": 0.0952,
+      "step": 716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.662159393449827,
+      "learning_rate": 5.829228313087756e-06,
+      "loss": 0.0994,
+      "step": 717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9503366133926616,
+      "learning_rate": 5.818892653624229e-06,
+      "loss": 0.1315,
+      "step": 718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.702587301532874,
+      "learning_rate": 5.80855339730268e-06,
+      "loss": 0.1484,
+      "step": 719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8348978334560153,
+      "learning_rate": 5.7982105895366725e-06,
+      "loss": 0.1406,
+      "step": 720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.022613796849133,
+      "learning_rate": 5.787864275755375e-06,
+      "loss": 0.1319,
+      "step": 721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5023748634115024,
+      "learning_rate": 5.777514501403353e-06,
+      "loss": 0.0929,
+      "step": 722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7013224101759017,
+      "learning_rate": 5.767161311940372e-06,
+      "loss": 0.0932,
+      "step": 723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8315351550901093,
+      "learning_rate": 5.756804752841203e-06,
+      "loss": 0.1173,
+      "step": 724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0170416119005425,
+      "learning_rate": 5.746444869595411e-06,
+      "loss": 0.1334,
+      "step": 725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8673140531034689,
+      "learning_rate": 5.736081707707163e-06,
+      "loss": 0.1484,
+      "step": 726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.210533914692524,
+      "learning_rate": 5.725715312695032e-06,
+      "loss": 0.1016,
+      "step": 727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.989411601512688,
+      "learning_rate": 5.715345730091786e-06,
+      "loss": 0.1052,
+      "step": 728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.273203843016005,
+      "learning_rate": 5.704973005444194e-06,
+      "loss": 0.1804,
+      "step": 729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.674460990852892,
+      "learning_rate": 5.6945971843128334e-06,
+      "loss": 0.09,
+      "step": 730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8538049522688833,
+      "learning_rate": 5.684218312271874e-06,
+      "loss": 0.114,
+      "step": 731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.180520132721186,
+      "learning_rate": 5.6738364349088905e-06,
+      "loss": 0.1309,
+      "step": 732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5591693338815162,
+      "learning_rate": 5.663451597824655e-06,
+      "loss": 0.1403,
+      "step": 733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7568489771713605,
+      "learning_rate": 5.653063846632945e-06,
+      "loss": 0.0832,
+      "step": 734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7500043042100533,
+      "learning_rate": 5.642673226960332e-06,
+      "loss": 0.1241,
+      "step": 735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8855691624550837,
+      "learning_rate": 5.632279784445991e-06,
+      "loss": 0.124,
+      "step": 736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.626920358067962,
+      "learning_rate": 5.621883564741494e-06,
+      "loss": 0.102,
+      "step": 737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8032329660583741,
+      "learning_rate": 5.611484613510608e-06,
+      "loss": 0.1484,
+      "step": 738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9140643898960137,
+      "learning_rate": 5.6010829764291074e-06,
+      "loss": 0.1215,
+      "step": 739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1624692381610573,
+      "learning_rate": 5.590678699184553e-06,
+      "loss": 0.1296,
+      "step": 740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.023676143260662,
+      "learning_rate": 5.580271827476111e-06,
+      "loss": 0.1355,
+      "step": 741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5620252064385884,
+      "learning_rate": 5.569862407014334e-06,
+      "loss": 0.1178,
+      "step": 742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1511386883957877,
+      "learning_rate": 5.5594504835209786e-06,
+      "loss": 0.0788,
+      "step": 743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5616956818437346,
+      "learning_rate": 5.549036102728791e-06,
+      "loss": 0.146,
+      "step": 744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8693991078131527,
+      "learning_rate": 5.538619310381313e-06,
+      "loss": 0.099,
+      "step": 745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.052175802220878,
+      "learning_rate": 5.528200152232674e-06,
+      "loss": 0.137,
+      "step": 746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7482892753111607,
+      "learning_rate": 5.5177786740474e-06,
+      "loss": 0.124,
+      "step": 747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6066071271539237,
+      "learning_rate": 5.507354921600205e-06,
+      "loss": 0.1207,
+      "step": 748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1673157599924906,
+      "learning_rate": 5.496928940675795e-06,
+      "loss": 0.0943,
+      "step": 749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.142894335501094,
+      "learning_rate": 5.486500777068659e-06,
+      "loss": 0.1771,
+      "step": 750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0460770317618135,
+      "learning_rate": 5.47607047658288e-06,
+      "loss": 0.1314,
+      "step": 751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1058657181662372,
+      "learning_rate": 5.465638085031918e-06,
+      "loss": 0.1403,
+      "step": 752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6840379754298231,
+      "learning_rate": 5.4552036482384275e-06,
+      "loss": 0.1506,
+      "step": 753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9289322283317214,
+      "learning_rate": 5.444767212034039e-06,
+      "loss": 0.1023,
+      "step": 754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.625262969875529,
+      "learning_rate": 5.434328822259167e-06,
+      "loss": 0.113,
+      "step": 755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4113122069453283,
+      "learning_rate": 5.423888524762812e-06,
+      "loss": 0.1012,
+      "step": 756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1703867341982583,
+      "learning_rate": 5.413446365402344e-06,
+      "loss": 0.1802,
+      "step": 757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4290049155356024,
+      "learning_rate": 5.403002390043319e-06,
+      "loss": 0.095,
+      "step": 758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3435851884276515,
+      "learning_rate": 5.3925566445592684e-06,
+      "loss": 0.0858,
+      "step": 759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.711060268727994,
+      "learning_rate": 5.382109174831493e-06,
+      "loss": 0.1415,
+      "step": 760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7557646547416874,
+      "learning_rate": 5.3716600267488764e-06,
+      "loss": 0.1082,
+      "step": 761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0596730304085207,
+      "learning_rate": 5.361209246207662e-06,
+      "loss": 0.1089,
+      "step": 762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.935238394520285,
+      "learning_rate": 5.350756879111278e-06,
+      "loss": 0.1137,
+      "step": 763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5074975424220682,
+      "learning_rate": 5.340302971370108e-06,
+      "loss": 0.1334,
+      "step": 764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.753317095876859,
+      "learning_rate": 5.329847568901311e-06,
+      "loss": 0.1018,
+      "step": 765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6666992189358707,
+      "learning_rate": 5.319390717628607e-06,
+      "loss": 0.1067,
+      "step": 766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1441072331379893,
+      "learning_rate": 5.308932463482082e-06,
+      "loss": 0.0976,
+      "step": 767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9819008451024653,
+      "learning_rate": 5.298472852397983e-06,
+      "loss": 0.125,
+      "step": 768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.078504932257869,
+      "learning_rate": 5.288011930318518e-06,
+      "loss": 0.1313,
+      "step": 769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7500202651291776,
+      "learning_rate": 5.277549743191653e-06,
+      "loss": 0.1256,
+      "step": 770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.957028569964894,
+      "learning_rate": 5.267086336970908e-06,
+      "loss": 0.119,
+      "step": 771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.64080169082907,
+      "learning_rate": 5.256621757615161e-06,
+      "loss": 0.1017,
+      "step": 772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.912980083382728,
+      "learning_rate": 5.2461560510884436e-06,
+      "loss": 0.132,
+      "step": 773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5189601667278445,
+      "learning_rate": 5.235689263359734e-06,
+      "loss": 0.1062,
+      "step": 774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.553830301206541,
+      "learning_rate": 5.2252214404027636e-06,
+      "loss": 0.1292,
+      "step": 775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8629058125857634,
+      "learning_rate": 5.214752628195807e-06,
+      "loss": 0.1293,
+      "step": 776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.593636492020973,
+      "learning_rate": 5.2042828727214866e-06,
+      "loss": 0.0956,
+      "step": 777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3310531956555536,
+      "learning_rate": 5.1938122199665685e-06,
+      "loss": 0.1161,
+      "step": 778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8528554576118663,
+      "learning_rate": 5.183340715921759e-06,
+      "loss": 0.1382,
+      "step": 779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.492914105028018,
+      "learning_rate": 5.172868406581502e-06,
+      "loss": 0.0887,
+      "step": 780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6396055924297854,
+      "learning_rate": 5.16239533794378e-06,
+      "loss": 0.1128,
+      "step": 781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.974365331118826,
+      "learning_rate": 5.1519215560099116e-06,
+      "loss": 0.1094,
+      "step": 782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0195657375792595,
+      "learning_rate": 5.141447106784347e-06,
+      "loss": 0.1017,
+      "step": 783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9226138716911316,
+      "learning_rate": 5.130972036274466e-06,
+      "loss": 0.1107,
+      "step": 784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8625360778194306,
+      "learning_rate": 5.120496390490382e-06,
+      "loss": 0.1265,
+      "step": 785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5463815816542692,
+      "learning_rate": 5.110020215444731e-06,
+      "loss": 0.0986,
+      "step": 786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9289825743188587,
+      "learning_rate": 5.099543557152474e-06,
+      "loss": 0.1583,
+      "step": 787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9666897967664563,
+      "learning_rate": 5.089066461630698e-06,
+      "loss": 0.1363,
+      "step": 788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9958797459896542,
+      "learning_rate": 5.0785889748984055e-06,
+      "loss": 0.1282,
+      "step": 789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9239342659219918,
+      "learning_rate": 5.06811114297632e-06,
+      "loss": 0.1207,
+      "step": 790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8061249514244604,
+      "learning_rate": 5.057633011886683e-06,
+      "loss": 0.1127,
+      "step": 791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8662644427333124,
+      "learning_rate": 5.047154627653045e-06,
+      "loss": 0.0944,
+      "step": 792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.090550754244602,
+      "learning_rate": 5.036676036300074e-06,
+      "loss": 0.1055,
+      "step": 793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7189547955473783,
+      "learning_rate": 5.026197283853345e-06,
+      "loss": 0.1271,
+      "step": 794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6763816609777438,
+      "learning_rate": 5.015718416339139e-06,
+      "loss": 0.1119,
+      "step": 795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0710333881750462,
+      "learning_rate": 5.005239479784246e-06,
+      "loss": 0.1109,
+      "step": 796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2908947072097592,
+      "learning_rate": 4.994760520215755e-06,
+      "loss": 0.0838,
+      "step": 797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.726408361017791,
+      "learning_rate": 4.984281583660864e-06,
+      "loss": 0.1009,
+      "step": 798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6042202190498385,
+      "learning_rate": 4.973802716146658e-06,
+      "loss": 0.1591,
+      "step": 799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6118800225800851,
+      "learning_rate": 4.963323963699926e-06,
+      "loss": 0.1129,
+      "step": 800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7270638214518879,
+      "learning_rate": 4.9528453723469575e-06,
+      "loss": 0.1113,
+      "step": 801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7432315697154184,
+      "learning_rate": 4.942366988113319e-06,
+      "loss": 0.0861,
+      "step": 802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0906970873027073,
+      "learning_rate": 4.931888857023682e-06,
+      "loss": 0.1382,
+      "step": 803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3501367666525494,
+      "learning_rate": 4.921411025101597e-06,
+      "loss": 0.1169,
+      "step": 804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4566720262640316,
+      "learning_rate": 4.910933538369303e-06,
+      "loss": 0.106,
+      "step": 805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5269663618460712,
+      "learning_rate": 4.900456442847528e-06,
+      "loss": 0.0916,
+      "step": 806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7453102664715303,
+      "learning_rate": 4.88997978455527e-06,
+      "loss": 0.1402,
+      "step": 807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.018716604307633,
+      "learning_rate": 4.879503609509619e-06,
+      "loss": 0.1095,
+      "step": 808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1106306694591015,
+      "learning_rate": 4.869027963725536e-06,
+      "loss": 0.1505,
+      "step": 809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1395796327180947,
+      "learning_rate": 4.858552893215655e-06,
+      "loss": 0.1141,
+      "step": 810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4202165199124486,
+      "learning_rate": 4.84807844399009e-06,
+      "loss": 0.1705,
+      "step": 811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.835373663677108,
+      "learning_rate": 4.837604662056222e-06,
+      "loss": 0.1313,
+      "step": 812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8005738743632453,
+      "learning_rate": 4.827131593418501e-06,
+      "loss": 0.1449,
+      "step": 813
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.124505896703758,
+      "learning_rate": 4.816659284078243e-06,
+      "loss": 0.1478,
+      "step": 814
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.127837289735595,
+      "learning_rate": 4.8061877800334315e-06,
+      "loss": 0.1886,
+      "step": 815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1604683050081326,
+      "learning_rate": 4.795717127278515e-06,
+      "loss": 0.0786,
+      "step": 816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9637692788927852,
+      "learning_rate": 4.785247371804195e-06,
+      "loss": 0.1063,
+      "step": 817
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7173564883561871,
+      "learning_rate": 4.77477855959724e-06,
+      "loss": 0.1149,
+      "step": 818
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5729152882720685,
+      "learning_rate": 4.764310736640267e-06,
+      "loss": 0.1293,
+      "step": 819
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6209397414677442,
+      "learning_rate": 4.753843948911556e-06,
+      "loss": 0.1285,
+      "step": 820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8314069234913113,
+      "learning_rate": 4.74337824238484e-06,
+      "loss": 0.1209,
+      "step": 821
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.570200559050954,
+      "learning_rate": 4.732913663029093e-06,
+      "loss": 0.1127,
+      "step": 822
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7368184688198605,
+      "learning_rate": 4.72245025680835e-06,
+      "loss": 0.1441,
+      "step": 823
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5220971625438804,
+      "learning_rate": 4.7119880696814835e-06,
+      "loss": 0.0893,
+      "step": 824
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8443284641960835,
+      "learning_rate": 4.701527147602018e-06,
+      "loss": 0.1247,
+      "step": 825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2951417910999217,
+      "learning_rate": 4.691067536517921e-06,
+      "loss": 0.1263,
+      "step": 826
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8763288375345808,
+      "learning_rate": 4.6806092823713944e-06,
+      "loss": 0.1207,
+      "step": 827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6884260275044232,
+      "learning_rate": 4.670152431098692e-06,
+      "loss": 0.1242,
+      "step": 828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5129931167494661,
+      "learning_rate": 4.659697028629894e-06,
+      "loss": 0.1067,
+      "step": 829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5727784925762973,
+      "learning_rate": 4.649243120888723e-06,
+      "loss": 0.0828,
+      "step": 830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1329407103752187,
+      "learning_rate": 4.6387907537923385e-06,
+      "loss": 0.0991,
+      "step": 831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7788421760508666,
+      "learning_rate": 4.628339973251126e-06,
+      "loss": 0.1151,
+      "step": 832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.413683936731647,
+      "learning_rate": 4.617890825168507e-06,
+      "loss": 0.102,
+      "step": 833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2207040797969175,
+      "learning_rate": 4.607443355440734e-06,
+      "loss": 0.1806,
+      "step": 834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3957444384519695,
+      "learning_rate": 4.596997609956682e-06,
+      "loss": 0.1311,
+      "step": 835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8035465675076767,
+      "learning_rate": 4.5865536345976576e-06,
+      "loss": 0.1434,
+      "step": 836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.819391164170477,
+      "learning_rate": 4.576111475237191e-06,
+      "loss": 0.1559,
+      "step": 837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7452355098149885,
+      "learning_rate": 4.565671177740834e-06,
+      "loss": 0.0959,
+      "step": 838
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8760725666495923,
+      "learning_rate": 4.555232787965963e-06,
+      "loss": 0.107,
+      "step": 839
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2518737780619866,
+      "learning_rate": 4.544796351761574e-06,
+      "loss": 0.0914,
+      "step": 840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2185919430350984,
+      "learning_rate": 4.534361914968083e-06,
+      "loss": 0.1812,
+      "step": 841
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.146679707787453,
+      "learning_rate": 4.523929523417122e-06,
+      "loss": 0.1183,
+      "step": 842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7114313656953086,
+      "learning_rate": 4.513499222931342e-06,
+      "loss": 0.1194,
+      "step": 843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8394742753028903,
+      "learning_rate": 4.503071059324206e-06,
+      "loss": 0.1025,
+      "step": 844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8273169411773014,
+      "learning_rate": 4.492645078399795e-06,
+      "loss": 0.0947,
+      "step": 845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7740798125372605,
+      "learning_rate": 4.482221325952602e-06,
+      "loss": 0.1431,
+      "step": 846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1192422103295256,
+      "learning_rate": 4.471799847767328e-06,
+      "loss": 0.122,
+      "step": 847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0660160064685416,
+      "learning_rate": 4.4613806896186906e-06,
+      "loss": 0.1058,
+      "step": 848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5742033350811535,
+      "learning_rate": 4.450963897271211e-06,
+      "loss": 0.0916,
+      "step": 849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8119951469136164,
+      "learning_rate": 4.440549516479022e-06,
+      "loss": 0.1163,
+      "step": 850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1159566523211066,
+      "learning_rate": 4.430137592985669e-06,
+      "loss": 0.1018,
+      "step": 851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.126873605321964,
+      "learning_rate": 4.419728172523892e-06,
+      "loss": 0.1498,
+      "step": 852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9676351837368764,
+      "learning_rate": 4.409321300815449e-06,
+      "loss": 0.1758,
+      "step": 853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.709782914468727,
+      "learning_rate": 4.398917023570894e-06,
+      "loss": 0.1008,
+      "step": 854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9285211732869538,
+      "learning_rate": 4.388515386489391e-06,
+      "loss": 0.1254,
+      "step": 855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7161777346446483,
+      "learning_rate": 4.378116435258509e-06,
+      "loss": 0.084,
+      "step": 856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.806844395754332,
+      "learning_rate": 4.36772021555401e-06,
+      "loss": 0.0986,
+      "step": 857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0837972155833735,
+      "learning_rate": 4.357326773039669e-06,
+      "loss": 0.0973,
+      "step": 858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3185486600701966,
+      "learning_rate": 4.346936153367056e-06,
+      "loss": 0.1006,
+      "step": 859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5256694435635292,
+      "learning_rate": 4.336548402175345e-06,
+      "loss": 0.0739,
+      "step": 860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.832223196645926,
+      "learning_rate": 4.326163565091112e-06,
+      "loss": 0.0987,
+      "step": 861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.002319765666875,
+      "learning_rate": 4.315781687728127e-06,
+      "loss": 0.0726,
+      "step": 862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7723246961310433,
+      "learning_rate": 4.305402815687168e-06,
+      "loss": 0.1082,
+      "step": 863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7830256916450828,
+      "learning_rate": 4.295026994555807e-06,
+      "loss": 0.0961,
+      "step": 864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8291995749358043,
+      "learning_rate": 4.284654269908216e-06,
+      "loss": 0.1442,
+      "step": 865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7253042913750083,
+      "learning_rate": 4.27428468730497e-06,
+      "loss": 0.1163,
+      "step": 866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.171182731375501,
+      "learning_rate": 4.263918292292838e-06,
+      "loss": 0.1588,
+      "step": 867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.993131212790838,
+      "learning_rate": 4.253555130404592e-06,
+      "loss": 0.1184,
+      "step": 868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9909346999767512,
+      "learning_rate": 4.243195247158798e-06,
+      "loss": 0.1361,
+      "step": 869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.985061431936418,
+      "learning_rate": 4.232838688059628e-06,
+      "loss": 0.0964,
+      "step": 870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.751213341901581,
+      "learning_rate": 4.2224854985966495e-06,
+      "loss": 0.108,
+      "step": 871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9591178958676378,
+      "learning_rate": 4.212135724244627e-06,
+      "loss": 0.1276,
+      "step": 872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7579749826323114,
+      "learning_rate": 4.20178941046333e-06,
+      "loss": 0.1062,
+      "step": 873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6097228735170386,
+      "learning_rate": 4.191446602697321e-06,
+      "loss": 0.1016,
+      "step": 874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6014944661125698,
+      "learning_rate": 4.181107346375771e-06,
+      "loss": 0.1377,
+      "step": 875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7711293693423962,
+      "learning_rate": 4.170771686912247e-06,
+      "loss": 0.1007,
+      "step": 876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7880026810514285,
+      "learning_rate": 4.160439669704516e-06,
+      "loss": 0.1143,
+      "step": 877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6987412567311313,
+      "learning_rate": 4.150111340134353e-06,
+      "loss": 0.1237,
+      "step": 878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9768502462590711,
+      "learning_rate": 4.139786743567328e-06,
+      "loss": 0.1233,
+      "step": 879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3478846945506415,
+      "learning_rate": 4.129465925352619e-06,
+      "loss": 0.1044,
+      "step": 880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9981013791009108,
+      "learning_rate": 4.11914893082281e-06,
+      "loss": 0.1339,
+      "step": 881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8517464074009944,
+      "learning_rate": 4.108835805293684e-06,
+      "loss": 0.1429,
+      "step": 882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5157512033678697,
+      "learning_rate": 4.098526594064036e-06,
+      "loss": 0.1062,
+      "step": 883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.840817457718613,
+      "learning_rate": 4.0882213424154635e-06,
+      "loss": 0.1145,
+      "step": 884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.029610584941451,
+      "learning_rate": 4.077920095612174e-06,
+      "loss": 0.1341,
+      "step": 885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6338844179144825,
+      "learning_rate": 4.067622898900788e-06,
+      "loss": 0.108,
+      "step": 886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9033531511200743,
+      "learning_rate": 4.057329797510128e-06,
+      "loss": 0.1004,
+      "step": 887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4571435474781058,
+      "learning_rate": 4.047040836651037e-06,
+      "loss": 0.1524,
+      "step": 888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5060740442891944,
+      "learning_rate": 4.036756061516166e-06,
+      "loss": 0.1106,
+      "step": 889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0788331275748475,
+      "learning_rate": 4.026475517279785e-06,
+      "loss": 0.1133,
+      "step": 890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5851053578741607,
+      "learning_rate": 4.0161992490975756e-06,
+      "loss": 0.1226,
+      "step": 891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7795177204393597,
+      "learning_rate": 4.005927302106442e-06,
+      "loss": 0.1275,
+      "step": 892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7209999961939761,
+      "learning_rate": 3.995659721424305e-06,
+      "loss": 0.1046,
+      "step": 893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5459099455260035,
+      "learning_rate": 3.98539655214991e-06,
+      "loss": 0.0984,
+      "step": 894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7475161853285592,
+      "learning_rate": 3.9751378393626224e-06,
+      "loss": 0.1262,
+      "step": 895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5464582238846574,
+      "learning_rate": 3.964883628122237e-06,
+      "loss": 0.0849,
+      "step": 896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.389196944686995,
+      "learning_rate": 3.954633963468772e-06,
+      "loss": 0.114,
+      "step": 897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7043708369854134,
+      "learning_rate": 3.944388890422281e-06,
+      "loss": 0.1255,
+      "step": 898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4523407706533134,
+      "learning_rate": 3.934148453982643e-06,
+      "loss": 0.1066,
+      "step": 899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.004405717963473,
+      "learning_rate": 3.923912699129378e-06,
+      "loss": 0.1068,
+      "step": 900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6414686497301643,
+      "learning_rate": 3.91368167082144e-06,
+      "loss": 0.1037,
+      "step": 901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.04799587934896,
+      "learning_rate": 3.903455413997018e-06,
+      "loss": 0.154,
+      "step": 902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.946993384701469,
+      "learning_rate": 3.893233973573351e-06,
+      "loss": 0.1132,
+      "step": 903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7235353612078088,
+      "learning_rate": 3.88301739444652e-06,
+      "loss": 0.1088,
+      "step": 904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6453579291204798,
+      "learning_rate": 3.872805721491247e-06,
+      "loss": 0.1041,
+      "step": 905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7804807126926472,
+      "learning_rate": 3.862598999560717e-06,
+      "loss": 0.1117,
+      "step": 906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5177028945223077,
+      "learning_rate": 3.852397273486354e-06,
+      "loss": 0.1091,
+      "step": 907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.018703368543647,
+      "learning_rate": 3.8422005880776495e-06,
+      "loss": 0.1085,
+      "step": 908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.218370435268377,
+      "learning_rate": 3.832008988121953e-06,
+      "loss": 0.1337,
+      "step": 909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0545128228188863,
+      "learning_rate": 3.82182251838427e-06,
+      "loss": 0.156,
+      "step": 910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7606510408251705,
+      "learning_rate": 3.8116412236070823e-06,
+      "loss": 0.1104,
+      "step": 911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8099072720687117,
+      "learning_rate": 3.8014651485101304e-06,
+      "loss": 0.1153,
+      "step": 912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.317637310297308,
+      "learning_rate": 3.7912943377902374e-06,
+      "loss": 0.1386,
+      "step": 913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.261715069348875,
+      "learning_rate": 3.7811288361211025e-06,
+      "loss": 0.1177,
+      "step": 914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1146627770061937,
+      "learning_rate": 3.770968688153098e-06,
+      "loss": 0.1207,
+      "step": 915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3219966429312695,
+      "learning_rate": 3.760813938513092e-06,
+      "loss": 0.1432,
+      "step": 916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5551696574986462,
+      "learning_rate": 3.7506646318042307e-06,
+      "loss": 0.1108,
+      "step": 917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6423151381409171,
+      "learning_rate": 3.740520812605759e-06,
+      "loss": 0.0809,
+      "step": 918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1270054148515642,
+      "learning_rate": 3.730382525472822e-06,
+      "loss": 0.1488,
+      "step": 919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4525654065999545,
+      "learning_rate": 3.720249814936255e-06,
+      "loss": 0.1703,
+      "step": 920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6466509751453386,
+      "learning_rate": 3.7101227255024113e-06,
+      "loss": 0.0976,
+      "step": 921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453861060575215,
+      "learning_rate": 3.700001301652947e-06,
+      "loss": 0.1087,
+      "step": 922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5946568411941369,
+      "learning_rate": 3.689885587844633e-06,
+      "loss": 0.098,
+      "step": 923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0732225644841873,
+      "learning_rate": 3.6797756285091634e-06,
+      "loss": 0.1449,
+      "step": 924
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.657934801330924,
+      "learning_rate": 3.6696714680529544e-06,
+      "loss": 0.134,
+      "step": 925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9263667651287582,
+      "learning_rate": 3.659573150856954e-06,
+      "loss": 0.1372,
+      "step": 926
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6330537523458593,
+      "learning_rate": 3.6494807212764396e-06,
+      "loss": 0.1184,
+      "step": 927
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8951926315625367,
+      "learning_rate": 3.639394223640831e-06,
+      "loss": 0.1222,
+      "step": 928
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6317636026091127,
+      "learning_rate": 3.6293137022534967e-06,
+      "loss": 0.1007,
+      "step": 929
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.108600411804529,
+      "learning_rate": 3.619239201391548e-06,
+      "loss": 0.1061,
+      "step": 930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8920524247392,
+      "learning_rate": 3.6091707653056584e-06,
+      "loss": 0.1553,
+      "step": 931
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8232758558650688,
+      "learning_rate": 3.599108438219857e-06,
+      "loss": 0.1274,
+      "step": 932
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.804319167028759,
+      "learning_rate": 3.5890522643313454e-06,
+      "loss": 0.112,
+      "step": 933
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3909948963927101,
+      "learning_rate": 3.5790022878102983e-06,
+      "loss": 0.0969,
+      "step": 934
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.44326112841619,
+      "learning_rate": 3.568958552799662e-06,
+      "loss": 0.1218,
+      "step": 935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.572623722062717,
+      "learning_rate": 3.5589211034149783e-06,
+      "loss": 0.1032,
+      "step": 936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.163157034822567,
+      "learning_rate": 3.5488899837441703e-06,
+      "loss": 0.095,
+      "step": 937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.400615824919047,
+      "learning_rate": 3.5388652378473664e-06,
+      "loss": 0.0949,
+      "step": 938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5133479004266996,
+      "learning_rate": 3.5288469097566992e-06,
+      "loss": 0.1234,
+      "step": 939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6235308351488011,
+      "learning_rate": 3.518835043476103e-06,
+      "loss": 0.0951,
+      "step": 940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9162477641836664,
+      "learning_rate": 3.508829682981143e-06,
+      "loss": 0.1454,
+      "step": 941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109715126524471,
+      "learning_rate": 3.4988308722187963e-06,
+      "loss": 0.1523,
+      "step": 942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4559210937158173,
+      "learning_rate": 3.488838655107279e-06,
+      "loss": 0.0731,
+      "step": 943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.506470505523408,
+      "learning_rate": 3.4788530755358453e-06,
+      "loss": 0.0898,
+      "step": 944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.613660972390155,
+      "learning_rate": 3.468874177364589e-06,
+      "loss": 0.0983,
+      "step": 945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5717909485298631,
+      "learning_rate": 3.458902004424265e-06,
+      "loss": 0.0973,
+      "step": 946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4205799220113384,
+      "learning_rate": 3.4489366005160823e-06,
+      "loss": 0.0959,
+      "step": 947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3451838006850796,
+      "learning_rate": 3.4389780094115206e-06,
+      "loss": 0.1076,
+      "step": 948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6388159985908506,
+      "learning_rate": 3.429026274852137e-06,
+      "loss": 0.0992,
+      "step": 949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5682048979771246,
+      "learning_rate": 3.4190814405493687e-06,
+      "loss": 0.1062,
+      "step": 950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2502678976586212,
+      "learning_rate": 3.4091435501843486e-06,
+      "loss": 0.1168,
+      "step": 951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3953877887707054,
+      "learning_rate": 3.3992126474077035e-06,
+      "loss": 0.0983,
+      "step": 952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6007212383303575,
+      "learning_rate": 3.3892887758393766e-06,
+      "loss": 0.1305,
+      "step": 953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9440119392902513,
+      "learning_rate": 3.37937197906842e-06,
+      "loss": 0.0931,
+      "step": 954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6043579363096137,
+      "learning_rate": 3.3694623006528148e-06,
+      "loss": 0.0998,
+      "step": 955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7136607779901858,
+      "learning_rate": 3.359559784119277e-06,
+      "loss": 0.1016,
+      "step": 956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5384259717082835,
+      "learning_rate": 3.349664472963059e-06,
+      "loss": 0.0989,
+      "step": 957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9497723408976853,
+      "learning_rate": 3.339776410647771e-06,
+      "loss": 0.1174,
+      "step": 958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6067390043096454,
+      "learning_rate": 3.3298956406051842e-06,
+      "loss": 0.104,
+      "step": 959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2852610652044447,
+      "learning_rate": 3.320022206235033e-06,
+      "loss": 0.1062,
+      "step": 960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5953068410830473,
+      "learning_rate": 3.3101561509048383e-06,
+      "loss": 0.0875,
+      "step": 961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6221547492274104,
+      "learning_rate": 3.3002975179497033e-06,
+      "loss": 0.1187,
+      "step": 962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6549898043169382,
+      "learning_rate": 3.290446350672134e-06,
+      "loss": 0.1013,
+      "step": 963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1405214271646877,
+      "learning_rate": 3.2806026923418473e-06,
+      "loss": 0.1245,
+      "step": 964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.812234610812394,
+      "learning_rate": 3.270766586195568e-06,
+      "loss": 0.0752,
+      "step": 965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.771591136410723,
+      "learning_rate": 3.2609380754368603e-06,
+      "loss": 0.1264,
+      "step": 966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.980251276861431,
+      "learning_rate": 3.2511172032359185e-06,
+      "loss": 0.1027,
+      "step": 967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9113584579125493,
+      "learning_rate": 3.2413040127293906e-06,
+      "loss": 0.1414,
+      "step": 968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.812466374310757,
+      "learning_rate": 3.231498547020184e-06,
+      "loss": 0.1192,
+      "step": 969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6565183224762494,
+      "learning_rate": 3.2217008491772726e-06,
+      "loss": 0.0933,
+      "step": 970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8256853684029712,
+      "learning_rate": 3.2119109622355157e-06,
+      "loss": 0.1312,
+      "step": 971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3886097887887576,
+      "learning_rate": 3.202128929195457e-06,
+      "loss": 0.1218,
+      "step": 972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8207805719268897,
+      "learning_rate": 3.1923547930231523e-06,
+      "loss": 0.0944,
+      "step": 973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.002596537052982,
+      "learning_rate": 3.1825885966499694e-06,
+      "loss": 0.1071,
+      "step": 974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.180981776214822,
+      "learning_rate": 3.172830382972394e-06,
+      "loss": 0.1026,
+      "step": 975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.132426193384201,
+      "learning_rate": 3.163080194851859e-06,
+      "loss": 0.0816,
+      "step": 976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6527864232962632,
+      "learning_rate": 3.1533380751145414e-06,
+      "loss": 0.1572,
+      "step": 977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.829324594252865,
+      "learning_rate": 3.1436040665511787e-06,
+      "loss": 0.1064,
+      "step": 978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6747234050546536,
+      "learning_rate": 3.1338782119168844e-06,
+      "loss": 0.0885,
+      "step": 979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5358562835425493,
+      "learning_rate": 3.124160553930953e-06,
+      "loss": 0.1161,
+      "step": 980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2961336183013072,
+      "learning_rate": 3.1144511352766828e-06,
+      "loss": 0.0823,
+      "step": 981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.968300020231063,
+      "learning_rate": 3.104749998601173e-06,
+      "loss": 0.1084,
+      "step": 982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2556454335910034,
+      "learning_rate": 3.0950571865151537e-06,
+      "loss": 0.0966,
+      "step": 983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7002655816514025,
+      "learning_rate": 3.0853727415927882e-06,
+      "loss": 0.1304,
+      "step": 984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.9645997532026,
+      "learning_rate": 3.075696706371484e-06,
+      "loss": 0.1041,
+      "step": 985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2613558023242604,
+      "learning_rate": 3.066029123351718e-06,
+      "loss": 0.1066,
+      "step": 986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6857551895264742,
+      "learning_rate": 3.056370034996835e-06,
+      "loss": 0.1034,
+      "step": 987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6661348639167128,
+      "learning_rate": 3.0467194837328716e-06,
+      "loss": 0.1331,
+      "step": 988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1575535840104454,
+      "learning_rate": 3.0370775119483694e-06,
+      "loss": 0.0963,
+      "step": 989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.649294756841945,
+      "learning_rate": 3.0274441619941787e-06,
+      "loss": 0.1244,
+      "step": 990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6157138126800377,
+      "learning_rate": 3.017819476183287e-06,
+      "loss": 0.1198,
+      "step": 991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6710125799816895,
+      "learning_rate": 3.0082034967906204e-06,
+      "loss": 0.0964,
+      "step": 992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5863459928289523,
+      "learning_rate": 2.998596266052866e-06,
+      "loss": 0.1032,
+      "step": 993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8074140571890533,
+      "learning_rate": 2.9889978261682873e-06,
+      "loss": 0.0888,
+      "step": 994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4876864670226553,
+      "learning_rate": 2.9794082192965272e-06,
+      "loss": 0.1023,
+      "step": 995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7127542347734488,
+      "learning_rate": 2.9698274875584393e-06,
+      "loss": 0.1246,
+      "step": 996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5660028087829367,
+      "learning_rate": 2.9602556730358865e-06,
+      "loss": 0.0858,
+      "step": 997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4787554897006587,
+      "learning_rate": 2.950692817771571e-06,
+      "loss": 0.1247,
+      "step": 998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.004753064742886,
+      "learning_rate": 2.941138963768842e-06,
+      "loss": 0.1096,
+      "step": 999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171685656988102,
+      "learning_rate": 2.931594152991506e-06,
+      "loss": 0.1273,
+      "step": 1000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0310858251944377,
+      "learning_rate": 2.9220584273636567e-06,
+      "loss": 0.1156,
+      "step": 1001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4705758430342852,
+      "learning_rate": 2.9125318287694777e-06,
+      "loss": 0.1159,
+      "step": 1002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5899197031650942,
+      "learning_rate": 2.903014399053063e-06,
+      "loss": 0.1144,
+      "step": 1003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0576756317048566,
+      "learning_rate": 2.8935061800182395e-06,
+      "loss": 0.1606,
+      "step": 1004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8486006145520921,
+      "learning_rate": 2.884007213428369e-06,
+      "loss": 0.0805,
+      "step": 1005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.902534606824819,
+      "learning_rate": 2.8745175410061822e-06,
+      "loss": 0.0827,
+      "step": 1006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6397252956915422,
+      "learning_rate": 2.8650372044335783e-06,
+      "loss": 0.09,
+      "step": 1007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.69662477582431,
+      "learning_rate": 2.8555662453514565e-06,
+      "loss": 0.105,
+      "step": 1008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5278564436727615,
+      "learning_rate": 2.846104705359527e-06,
+      "loss": 0.0943,
+      "step": 1009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6709412297375785,
+      "learning_rate": 2.836652626016121e-06,
+      "loss": 0.1357,
+      "step": 1010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3531841084884235,
+      "learning_rate": 2.8272100488380237e-06,
+      "loss": 0.0642,
+      "step": 1011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5587241550432935,
+      "learning_rate": 2.817777015300278e-06,
+      "loss": 0.0952,
+      "step": 1012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.258828525128708,
+      "learning_rate": 2.80835356683601e-06,
+      "loss": 0.1698,
+      "step": 1013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8775581455111865,
+      "learning_rate": 2.798939744836247e-06,
+      "loss": 0.1173,
+      "step": 1014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9208509554836495,
+      "learning_rate": 2.789535590649728e-06,
+      "loss": 0.1164,
+      "step": 1015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0503080097332997,
+      "learning_rate": 2.780141145582731e-06,
+      "loss": 0.1087,
+      "step": 1016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.344681587427957,
+      "learning_rate": 2.7707564508988917e-06,
+      "loss": 0.1493,
+      "step": 1017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7103934451645413,
+      "learning_rate": 2.761381547819011e-06,
+      "loss": 0.1042,
+      "step": 1018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.635862007069694,
+      "learning_rate": 2.7520164775208867e-06,
+      "loss": 0.1064,
+      "step": 1019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.136653527750661,
+      "learning_rate": 2.7426612811391295e-06,
+      "loss": 0.1268,
+      "step": 1020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.565949660217831,
+      "learning_rate": 2.733315999764974e-06,
+      "loss": 0.1289,
+      "step": 1021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.561447396904177,
+      "learning_rate": 2.7239806744461095e-06,
+      "loss": 0.1026,
+      "step": 1022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6443642185497302,
+      "learning_rate": 2.7146553461864943e-06,
+      "loss": 0.1069,
+      "step": 1023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.537615647007176,
+      "learning_rate": 2.705340055946177e-06,
+      "loss": 0.1082,
+      "step": 1024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5106720702647543,
+      "learning_rate": 2.6960348446411123e-06,
+      "loss": 0.1153,
+      "step": 1025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.635526999095943,
+      "learning_rate": 2.6867397531429895e-06,
+      "loss": 0.1184,
+      "step": 1026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9105341445360475,
+      "learning_rate": 2.6774548222790415e-06,
+      "loss": 0.1865,
+      "step": 1027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7420481882111878,
+      "learning_rate": 2.6681800928318778e-06,
+      "loss": 0.1478,
+      "step": 1028
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.512461760578711,
+      "learning_rate": 2.6589156055393023e-06,
+      "loss": 0.0799,
+      "step": 1029
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.250094678795374,
+      "learning_rate": 2.6496614010941217e-06,
+      "loss": 0.1326,
+      "step": 1030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.883689440298225,
+      "learning_rate": 2.6404175201439886e-06,
+      "loss": 0.1234,
+      "step": 1031
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9270058022648389,
+      "learning_rate": 2.6311840032912006e-06,
+      "loss": 0.1019,
+      "step": 1032
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4081318296335605,
+      "learning_rate": 2.62196089109254e-06,
+      "loss": 0.0826,
+      "step": 1033
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.6728184484530475,
+      "learning_rate": 2.6127482240590873e-06,
+      "loss": 0.1226,
+      "step": 1034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3907092679086035,
+      "learning_rate": 2.6035460426560407e-06,
+      "loss": 0.0727,
+      "step": 1035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.280540245854714,
+      "learning_rate": 2.5943543873025465e-06,
+      "loss": 0.1119,
+      "step": 1036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1251362917376775,
+      "learning_rate": 2.585173298371511e-06,
+      "loss": 0.123,
+      "step": 1037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6712555338356587,
+      "learning_rate": 2.5760028161894356e-06,
+      "loss": 0.1194,
+      "step": 1038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.492680856779564,
+      "learning_rate": 2.566842981036231e-06,
+      "loss": 0.0897,
+      "step": 1039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0679531653024457,
+      "learning_rate": 2.5576938331450384e-06,
+      "loss": 0.1405,
+      "step": 1040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3844550472485293,
+      "learning_rate": 2.548555412702065e-06,
+      "loss": 0.1385,
+      "step": 1041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8043797801884316,
+      "learning_rate": 2.5394277598463903e-06,
+      "loss": 0.1191,
+      "step": 1042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8439167974623585,
+      "learning_rate": 2.5303109146698046e-06,
+      "loss": 0.1727,
+      "step": 1043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5465929151314142,
+      "learning_rate": 2.5212049172166287e-06,
+      "loss": 0.1213,
+      "step": 1044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8983516783979946,
+      "learning_rate": 2.512109807483528e-06,
+      "loss": 0.1183,
+      "step": 1045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9065223453576003,
+      "learning_rate": 2.5030256254193576e-06,
+      "loss": 0.1327,
+      "step": 1046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0558919653887355,
+      "learning_rate": 2.493952410924961e-06,
+      "loss": 0.1207,
+      "step": 1047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4722241516028056,
+      "learning_rate": 2.4848902038530184e-06,
+      "loss": 0.0975,
+      "step": 1048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7425038137771816,
+      "learning_rate": 2.47583904400786e-06,
+      "loss": 0.1156,
+      "step": 1049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7363383932679017,
+      "learning_rate": 2.4667989711452878e-06,
+      "loss": 0.1093,
+      "step": 1050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7124149459581317,
+      "learning_rate": 2.457770024972412e-06,
+      "loss": 0.095,
+      "step": 1051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3387961439928358,
+      "learning_rate": 2.4487522451474655e-06,
+      "loss": 0.1239,
+      "step": 1052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6878226288878275,
+      "learning_rate": 2.4397456712796385e-06,
+      "loss": 0.098,
+      "step": 1053
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5795072509935049,
+      "learning_rate": 2.430750342928901e-06,
+      "loss": 0.1133,
+      "step": 1054
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8746300617437088,
+      "learning_rate": 2.4217662996058226e-06,
+      "loss": 0.1556,
+      "step": 1055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5504162897648324,
+      "learning_rate": 2.4127935807714154e-06,
+      "loss": 0.1364,
+      "step": 1056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5859801310552404,
+      "learning_rate": 2.40383222583694e-06,
+      "loss": 0.1184,
+      "step": 1057
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.635516927638879,
+      "learning_rate": 2.39488227416375e-06,
+      "loss": 0.0893,
+      "step": 1058
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1535635990621302,
+      "learning_rate": 2.3859437650631105e-06,
+      "loss": 0.102,
+      "step": 1059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.792425838316303,
+      "learning_rate": 2.377016737796024e-06,
+      "loss": 0.0752,
+      "step": 1060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2376840524407164,
+      "learning_rate": 2.368101231573066e-06,
+      "loss": 0.0874,
+      "step": 1061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8324530457880948,
+      "learning_rate": 2.3591972855541993e-06,
+      "loss": 0.1181,
+      "step": 1062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8822020144429574,
+      "learning_rate": 2.3503049388486187e-06,
+      "loss": 0.1219,
+      "step": 1063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2007033775790816,
+      "learning_rate": 2.34142423051457e-06,
+      "loss": 0.119,
+      "step": 1064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.503760887068537,
+      "learning_rate": 2.3325551995591723e-06,
+      "loss": 0.1213,
+      "step": 1065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7499648921099262,
+      "learning_rate": 2.3236978849382625e-06,
+      "loss": 0.0915,
+      "step": 1066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1132803759697496,
+      "learning_rate": 2.3148523255562065e-06,
+      "loss": 0.1555,
+      "step": 1067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3009606425237226,
+      "learning_rate": 2.3060185602657443e-06,
+      "loss": 0.0671,
+      "step": 1068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0363667534903103,
+      "learning_rate": 2.2971966278678115e-06,
+      "loss": 0.1471,
+      "step": 1069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2792526851516404,
+      "learning_rate": 2.2883865671113637e-06,
+      "loss": 0.1264,
+      "step": 1070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2635338410250414,
+      "learning_rate": 2.279588416693218e-06,
+      "loss": 0.0808,
+      "step": 1071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.827366143443778,
+      "learning_rate": 2.2708022152578775e-06,
+      "loss": 0.1169,
+      "step": 1072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7243659089253423,
+      "learning_rate": 2.262028001397355e-06,
+      "loss": 0.0842,
+      "step": 1073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.881634138113363,
+      "learning_rate": 2.2532658136510165e-06,
+      "loss": 0.1045,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.103741160610525,
+      "learning_rate": 2.244515690505403e-06,
+      "loss": 0.1071,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0469803289981146,
+      "learning_rate": 2.2357776703940613e-06,
+      "loss": 0.134,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7133679583049146,
+      "learning_rate": 2.22705179169738e-06,
+      "loss": 0.1282,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1271738970026544,
+      "learning_rate": 2.218338092742418e-06,
+      "loss": 0.1637,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6637756715555911,
+      "learning_rate": 2.2096366118027385e-06,
+      "loss": 0.0798,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6078482696889287,
+      "learning_rate": 2.200947387098232e-06,
+      "loss": 0.1207,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9300362876253738,
+      "learning_rate": 2.1922704567949643e-06,
+      "loss": 0.1139,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4856625951255353,
+      "learning_rate": 2.183605859004997e-06,
+      "loss": 0.0949,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5176974696935166,
+      "learning_rate": 2.174953631786217e-06,
+      "loss": 0.09,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.042197288042245,
+      "learning_rate": 2.166313813142185e-06,
+      "loss": 0.158,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0160094322338598,
+      "learning_rate": 2.1576864410219517e-06,
+      "loss": 0.1175,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7635899225729912,
+      "learning_rate": 2.1490715533199014e-06,
+      "loss": 0.0939,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9225389063482508,
+      "learning_rate": 2.1404691878755845e-06,
+      "loss": 0.1377,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.830475012337844,
+      "learning_rate": 2.131879382473544e-06,
+      "loss": 0.0901,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5548356207069536,
+      "learning_rate": 2.123302174843161e-06,
+      "loss": 0.107,
+      "step": 1089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7800121351929723,
+      "learning_rate": 2.114737602658476e-06,
+      "loss": 0.0943,
+      "step": 1090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8985974516513844,
+      "learning_rate": 2.1061857035380364e-06,
+      "loss": 0.1053,
+      "step": 1091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6179775706907018,
+      "learning_rate": 2.097646515044724e-06,
+      "loss": 0.0894,
+      "step": 1092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9563531354116217,
+      "learning_rate": 2.089120074685587e-06,
+      "loss": 0.1232,
+      "step": 1093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.347504434239575,
+      "learning_rate": 2.080606419911686e-06,
+      "loss": 0.1309,
+      "step": 1094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3869140849131103,
+      "learning_rate": 2.0721055881179154e-06,
+      "loss": 0.1382,
+      "step": 1095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6530719214025684,
+      "learning_rate": 2.063617616642852e-06,
+      "loss": 0.1091,
+      "step": 1096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8463484297999817,
+      "learning_rate": 2.0551425427685862e-06,
+      "loss": 0.0755,
+      "step": 1097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4804477467947055,
+      "learning_rate": 2.0466804037205527e-06,
+      "loss": 0.1261,
+      "step": 1098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9933059234486568,
+      "learning_rate": 2.0382312366673777e-06,
+      "loss": 0.1438,
+      "step": 1099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4558824986540313,
+      "learning_rate": 2.029795078720705e-06,
+      "loss": 0.0801,
+      "step": 1100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7568644774027609,
+      "learning_rate": 2.0213719669350414e-06,
+      "loss": 0.107,
+      "step": 1101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.610845221208439,
+      "learning_rate": 2.01296193830759e-06,
+      "loss": 0.1176,
+      "step": 1102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8259622221603906,
+      "learning_rate": 2.004565029778085e-06,
+      "loss": 0.1005,
+      "step": 1103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5869296681714593,
+      "learning_rate": 1.9961812782286384e-06,
+      "loss": 0.1133,
+      "step": 1104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.099922313185119,
+      "learning_rate": 1.987810720483566e-06,
+      "loss": 0.0658,
+      "step": 1105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9629552453343817,
+      "learning_rate": 1.979453393309235e-06,
+      "loss": 0.1434,
+      "step": 1106
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.050001164612842,
+      "learning_rate": 1.9711093334139013e-06,
+      "loss": 0.1201,
+      "step": 1107
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8072014706090043,
+      "learning_rate": 1.9627785774475415e-06,
+      "loss": 0.1543,
+      "step": 1108
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6558315512780304,
+      "learning_rate": 1.954461162001703e-06,
+      "loss": 0.0891,
+      "step": 1109
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.823622763564889,
+      "learning_rate": 1.946157123609329e-06,
+      "loss": 0.105,
+      "step": 1110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.601904934784887,
+      "learning_rate": 1.9378664987446135e-06,
+      "loss": 0.1156,
+      "step": 1111
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.229821956730354,
+      "learning_rate": 1.9295893238228316e-06,
+      "loss": 0.1198,
+      "step": 1112
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.893246059611001,
+      "learning_rate": 1.9213256352001777e-06,
+      "loss": 0.1559,
+      "step": 1113
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8106645848464975,
+      "learning_rate": 1.9130754691736157e-06,
+      "loss": 0.1609,
+      "step": 1114
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4879614809305632,
+      "learning_rate": 1.9048388619807085e-06,
+      "loss": 0.0842,
+      "step": 1115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7694300917166506,
+      "learning_rate": 1.896615849799467e-06,
+      "loss": 0.0707,
+      "step": 1116
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3353327364662788,
+      "learning_rate": 1.8884064687481897e-06,
+      "loss": 0.072,
+      "step": 1117
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6306354819802187,
+      "learning_rate": 1.880210754885296e-06,
+      "loss": 0.0898,
+      "step": 1118
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7707297323164894,
+      "learning_rate": 1.8720287442091823e-06,
+      "loss": 0.1469,
+      "step": 1119
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6693648130916285,
+      "learning_rate": 1.8638604726580479e-06,
+      "loss": 0.1215,
+      "step": 1120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.777331267067197,
+      "learning_rate": 1.8557059761097517e-06,
+      "loss": 0.1044,
+      "step": 1121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.2001988352604105,
+      "learning_rate": 1.8475652903816476e-06,
+      "loss": 0.1289,
+      "step": 1122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7577976787597611,
+      "learning_rate": 1.8394384512304215e-06,
+      "loss": 0.1365,
+      "step": 1123
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8901050885585722,
+      "learning_rate": 1.8313254943519482e-06,
+      "loss": 0.1306,
+      "step": 1124
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.854742043632422,
+      "learning_rate": 1.8232264553811196e-06,
+      "loss": 0.1261,
+      "step": 1125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.698941017540202,
+      "learning_rate": 1.8151413698917004e-06,
+      "loss": 0.1625,
+      "step": 1126
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7206558278103523,
+      "learning_rate": 1.8070702733961676e-06,
+      "loss": 0.0938,
+      "step": 1127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7677875559598875,
+      "learning_rate": 1.7990132013455475e-06,
+      "loss": 0.1314,
+      "step": 1128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6453012236522748,
+      "learning_rate": 1.7909701891292719e-06,
+      "loss": 0.1049,
+      "step": 1129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8889911261871926,
+      "learning_rate": 1.7829412720750172e-06,
+      "loss": 0.1181,
+      "step": 1130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8214020256704653,
+      "learning_rate": 1.774926485448543e-06,
+      "loss": 0.1221,
+      "step": 1131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.667776340671231,
+      "learning_rate": 1.7669258644535497e-06,
+      "loss": 0.1152,
+      "step": 1132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6599634490047435,
+      "learning_rate": 1.7589394442315144e-06,
+      "loss": 0.0937,
+      "step": 1133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.99996195686876,
+      "learning_rate": 1.7509672598615417e-06,
+      "loss": 0.1419,
+      "step": 1134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2118355594063708,
+      "learning_rate": 1.7430093463602021e-06,
+      "loss": 0.1449,
+      "step": 1135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8327808051263679,
+      "learning_rate": 1.7350657386813897e-06,
+      "loss": 0.1156,
+      "step": 1136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6042442726020136,
+      "learning_rate": 1.7271364717161609e-06,
+      "loss": 0.1496,
+      "step": 1137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.998649435249516,
+      "learning_rate": 1.7192215802925793e-06,
+      "loss": 0.1067,
+      "step": 1138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6141034792802256,
+      "learning_rate": 1.7113210991755713e-06,
+      "loss": 0.1178,
+      "step": 1139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9428946129686242,
+      "learning_rate": 1.7034350630667628e-06,
+      "loss": 0.1401,
+      "step": 1140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6832521348575313,
+      "learning_rate": 1.6955635066043363e-06,
+      "loss": 0.1273,
+      "step": 1141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.038854894966858,
+      "learning_rate": 1.6877064643628766e-06,
+      "loss": 0.1486,
+      "step": 1142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7326863265701336,
+      "learning_rate": 1.6798639708532088e-06,
+      "loss": 0.1034,
+      "step": 1143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.373357985004174,
+      "learning_rate": 1.6720360605222653e-06,
+      "loss": 0.0887,
+      "step": 1144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0943429087659404,
+      "learning_rate": 1.6642227677529149e-06,
+      "loss": 0.1079,
+      "step": 1145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6520799301347306,
+      "learning_rate": 1.656424126863827e-06,
+      "loss": 0.1097,
+      "step": 1146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9036469573563441,
+      "learning_rate": 1.6486401721093143e-06,
+      "loss": 0.1107,
+      "step": 1147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5580736483345345,
+      "learning_rate": 1.6408709376791798e-06,
+      "loss": 0.0925,
+      "step": 1148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.118080025975157,
+      "learning_rate": 1.6331164576985737e-06,
+      "loss": 0.0992,
+      "step": 1149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7113308881162417,
+      "learning_rate": 1.6253767662278346e-06,
+      "loss": 0.1479,
+      "step": 1150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6408114347988518,
+      "learning_rate": 1.6176518972623505e-06,
+      "loss": 0.0807,
+      "step": 1151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.632998910503062,
+      "learning_rate": 1.6099418847324022e-06,
+      "loss": 0.1413,
+      "step": 1152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.810289738754614,
+      "learning_rate": 1.6022467625030115e-06,
+      "loss": 0.1218,
+      "step": 1153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.762358379505647,
+      "learning_rate": 1.5945665643738039e-06,
+      "loss": 0.1393,
+      "step": 1154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5628602324627865,
+      "learning_rate": 1.5869013240788466e-06,
+      "loss": 0.0907,
+      "step": 1155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.95799804387182,
+      "learning_rate": 1.5792510752865114e-06,
+      "loss": 0.1213,
+      "step": 1156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6700313560006872,
+      "learning_rate": 1.5716158515993223e-06,
+      "loss": 0.134,
+      "step": 1157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.019046183613624,
+      "learning_rate": 1.563995686553803e-06,
+      "loss": 0.1817,
+      "step": 1158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7629259020298114,
+      "learning_rate": 1.556390613620341e-06,
+      "loss": 0.1636,
+      "step": 1159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9536751841576872,
+      "learning_rate": 1.548800666203028e-06,
+      "loss": 0.1116,
+      "step": 1160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0814066954978614,
+      "learning_rate": 1.5412258776395233e-06,
+      "loss": 0.143,
+      "step": 1161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.180706325778583,
+      "learning_rate": 1.5336662812009035e-06,
+      "loss": 0.0748,
+      "step": 1162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8236403642886168,
+      "learning_rate": 1.5261219100915115e-06,
+      "loss": 0.1335,
+      "step": 1163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5661440703265628,
+      "learning_rate": 1.5185927974488224e-06,
+      "loss": 0.1373,
+      "step": 1164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6203179411585926,
+      "learning_rate": 1.5110789763432832e-06,
+      "loss": 0.1257,
+      "step": 1165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.531392583201998,
+      "learning_rate": 1.5035804797781811e-06,
+      "loss": 0.1374,
+      "step": 1166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3189072111660938,
+      "learning_rate": 1.4960973406894919e-06,
+      "loss": 0.0834,
+      "step": 1167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0438901557258724,
+      "learning_rate": 1.4886295919457317e-06,
+      "loss": 0.138,
+      "step": 1168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7428867207404393,
+      "learning_rate": 1.481177266347823e-06,
+      "loss": 0.114,
+      "step": 1169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8447265154099988,
+      "learning_rate": 1.4737403966289387e-06,
+      "loss": 0.1253,
+      "step": 1170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6780724286477957,
+      "learning_rate": 1.4663190154543683e-06,
+      "loss": 0.1107,
+      "step": 1171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2162692268576656,
+      "learning_rate": 1.4589131554213704e-06,
+      "loss": 0.1187,
+      "step": 1172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1732299790030174,
+      "learning_rate": 1.451522849059025e-06,
+      "loss": 0.0794,
+      "step": 1173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5098062263871141,
+      "learning_rate": 1.444148128828101e-06,
+      "loss": 0.0916,
+      "step": 1174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.433136044460252,
+      "learning_rate": 1.4367890271209024e-06,
+      "loss": 0.0902,
+      "step": 1175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5675088273366966,
+      "learning_rate": 1.4294455762611343e-06,
+      "loss": 0.1293,
+      "step": 1176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6721476839079745,
+      "learning_rate": 1.422117808503759e-06,
+      "loss": 0.1283,
+      "step": 1177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4516772793399861,
+      "learning_rate": 1.4148057560348477e-06,
+      "loss": 0.0961,
+      "step": 1178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9838232631858521,
+      "learning_rate": 1.4075094509714526e-06,
+      "loss": 0.115,
+      "step": 1179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9665812232773578,
+      "learning_rate": 1.4002289253614492e-06,
+      "loss": 0.1168,
+      "step": 1180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6688597299685144,
+      "learning_rate": 1.3929642111834114e-06,
+      "loss": 0.1377,
+      "step": 1181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5727794307304164,
+      "learning_rate": 1.3857153403464613e-06,
+      "loss": 0.1307,
+      "step": 1182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4649402982348885,
+      "learning_rate": 1.3784823446901295e-06,
+      "loss": 0.1069,
+      "step": 1183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.72399909687252,
+      "learning_rate": 1.3712652559842205e-06,
+      "loss": 0.1293,
+      "step": 1184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.242003952713536,
+      "learning_rate": 1.364064105928668e-06,
+      "loss": 0.114,
+      "step": 1185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.705374708960644,
+      "learning_rate": 1.356878926153401e-06,
+      "loss": 0.113,
+      "step": 1186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6517476509676714,
+      "learning_rate": 1.3497097482181948e-06,
+      "loss": 0.0865,
+      "step": 1187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.651682909854867,
+      "learning_rate": 1.342556603612546e-06,
+      "loss": 0.091,
+      "step": 1188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.616992396188749,
+      "learning_rate": 1.3354195237555245e-06,
+      "loss": 0.1177,
+      "step": 1189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.973440183689998,
+      "learning_rate": 1.3282985399956372e-06,
+      "loss": 0.0745,
+      "step": 1190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2399378592144723,
+      "learning_rate": 1.3211936836106926e-06,
+      "loss": 0.0853,
+      "step": 1191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5356433036883095,
+      "learning_rate": 1.3141049858076648e-06,
+      "loss": 0.1219,
+      "step": 1192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1533362310986135,
+      "learning_rate": 1.3070324777225474e-06,
+      "loss": 0.1396,
+      "step": 1193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8636701257772759,
+      "learning_rate": 1.29997619042023e-06,
+      "loss": 0.0767,
+      "step": 1194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.219301762559893,
+      "learning_rate": 1.2929361548943503e-06,
+      "loss": 0.1646,
+      "step": 1195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4052352119926677,
+      "learning_rate": 1.2859124020671643e-06,
+      "loss": 0.1196,
+      "step": 1196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203155952648259,
+      "learning_rate": 1.2789049627894111e-06,
+      "loss": 0.1355,
+      "step": 1197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.079010510241825,
+      "learning_rate": 1.2719138678401693e-06,
+      "loss": 0.0916,
+      "step": 1198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4166843086588452,
+      "learning_rate": 1.2649391479267347e-06,
+      "loss": 0.1051,
+      "step": 1199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4052281200010173,
+      "learning_rate": 1.2579808336844711e-06,
+      "loss": 0.1011,
+      "step": 1200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1195754551679133,
+      "learning_rate": 1.2510389556766884e-06,
+      "loss": 0.1152,
+      "step": 1201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6404708356664435,
+      "learning_rate": 1.2441135443945023e-06,
+      "loss": 0.106,
+      "step": 1202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8634718218136286,
+      "learning_rate": 1.237204630256697e-06,
+      "loss": 0.084,
+      "step": 1203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8180176669915014,
+      "learning_rate": 1.2303122436096015e-06,
+      "loss": 0.1217,
+      "step": 1204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.74789232737528,
+      "learning_rate": 1.2234364147269433e-06,
+      "loss": 0.1016,
+      "step": 1205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1100851203934625,
+      "learning_rate": 1.2165771738097288e-06,
+      "loss": 0.1056,
+      "step": 1206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.137840760178262,
+      "learning_rate": 1.209734550986103e-06,
+      "loss": 0.0817,
+      "step": 1207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.016888381508961,
+      "learning_rate": 1.2029085763112142e-06,
+      "loss": 0.1414,
+      "step": 1208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.044827160753601,
+      "learning_rate": 1.1960992797670935e-06,
+      "loss": 0.0727,
+      "step": 1209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2044422944254878,
+      "learning_rate": 1.1893066912625078e-06,
+      "loss": 0.0571,
+      "step": 1210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.001239217217173,
+      "learning_rate": 1.1825308406328435e-06,
+      "loss": 0.1054,
+      "step": 1211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.898027796170711,
+      "learning_rate": 1.1757717576399658e-06,
+      "loss": 0.085,
+      "step": 1212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.866494564536938,
+      "learning_rate": 1.1690294719720891e-06,
+      "loss": 0.1069,
+      "step": 1213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.826616801921012,
+      "learning_rate": 1.1623040132436515e-06,
+      "loss": 0.1405,
+      "step": 1214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8994861648399652,
+      "learning_rate": 1.1555954109951772e-06,
+      "loss": 0.1468,
+      "step": 1215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7399303106591963,
+      "learning_rate": 1.1489036946931548e-06,
+      "loss": 0.1103,
+      "step": 1216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5618058484821695,
+      "learning_rate": 1.1422288937299037e-06,
+      "loss": 0.1147,
+      "step": 1217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2516405928444252,
+      "learning_rate": 1.1355710374234414e-06,
+      "loss": 0.0751,
+      "step": 1218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.324978806825241,
+      "learning_rate": 1.1289301550173643e-06,
+      "loss": 0.1301,
+      "step": 1219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9352463582352895,
+      "learning_rate": 1.122306275680708e-06,
+      "loss": 0.1336,
+      "step": 1220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5812187303605587,
+      "learning_rate": 1.1156994285078287e-06,
+      "loss": 0.1127,
+      "step": 1221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7723759082237054,
+      "learning_rate": 1.1091096425182719e-06,
+      "loss": 0.076,
+      "step": 1222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5841351312799044,
+      "learning_rate": 1.1025369466566394e-06,
+      "loss": 0.112,
+      "step": 1223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.529260110516267,
+      "learning_rate": 1.0959813697924743e-06,
+      "loss": 0.1328,
+      "step": 1224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.763156092795216,
+      "learning_rate": 1.0894429407201207e-06,
+      "loss": 0.115,
+      "step": 1225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5513777752156226,
+      "learning_rate": 1.082921688158608e-06,
+      "loss": 0.0888,
+      "step": 1226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8994960689856308,
+      "learning_rate": 1.0764176407515203e-06,
+      "loss": 0.1229,
+      "step": 1227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4517647924703174,
+      "learning_rate": 1.0699308270668667e-06,
+      "loss": 0.0829,
+      "step": 1228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6328827890997333,
+      "learning_rate": 1.0634612755969665e-06,
+      "loss": 0.0929,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0925694894155447,
+      "learning_rate": 1.0570090147583089e-06,
+      "loss": 0.1319,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5999398316628513,
+      "learning_rate": 1.050574072891445e-06,
+      "loss": 0.099,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0689198706320813,
+      "learning_rate": 1.044156478260851e-06,
+      "loss": 0.1131,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4542505060532096,
+      "learning_rate": 1.0377562590548067e-06,
+      "loss": 0.1119,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1296698994205494,
+      "learning_rate": 1.031373443385278e-06,
+      "loss": 0.1424,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.849387025827355,
+      "learning_rate": 1.025008059287782e-06,
+      "loss": 0.1279,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.111837408928748,
+      "learning_rate": 1.0186601347212756e-06,
+      "loss": 0.1384,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9271949909614214,
+      "learning_rate": 1.0123296975680263e-06,
+      "loss": 0.1087,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.329230313694751,
+      "learning_rate": 1.0060167756334876e-06,
+      "loss": 0.1289,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7572601092475966,
+      "learning_rate": 9.997213966461843e-07,
+      "loss": 0.1203,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1928031187601045,
+      "learning_rate": 9.934435882575849e-07,
+      "loss": 0.1436,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.89964134222432,
+      "learning_rate": 9.871833780419827e-07,
+      "loss": 0.1532,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6273782226263336,
+      "learning_rate": 9.809407934963705e-07,
+      "loss": 0.1361,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.616377832330268,
+      "learning_rate": 9.747158620403274e-07,
+      "loss": 0.0965,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7615100331467808,
+      "learning_rate": 9.685086110158926e-07,
+      "loss": 0.1186,
+      "step": 1244
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1874223303571698,
+      "learning_rate": 9.623190676874438e-07,
+      "loss": 0.0801,
+      "step": 1245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6707289299229628,
+      "learning_rate": 9.561472592415849e-07,
+      "loss": 0.1015,
+      "step": 1246
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.703153506821708,
+      "learning_rate": 9.499932127870209e-07,
+      "loss": 0.1144,
+      "step": 1247
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8021171145691988,
+      "learning_rate": 9.438569553544369e-07,
+      "loss": 0.1236,
+      "step": 1248
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5572935030368558,
+      "learning_rate": 9.377385138963868e-07,
+      "loss": 0.0957,
+      "step": 1249
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5329776444898073,
+      "learning_rate": 9.316379152871668e-07,
+      "loss": 0.118,
+      "step": 1250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8036771947979402,
+      "learning_rate": 9.255551863227041e-07,
+      "loss": 0.1209,
+      "step": 1251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3860222366876078,
+      "learning_rate": 9.194903537204363e-07,
+      "loss": 0.1009,
+      "step": 1252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.511454379513255,
+      "learning_rate": 9.134434441191908e-07,
+      "loss": 0.1137,
+      "step": 1253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4241637994362393,
+      "learning_rate": 9.074144840790755e-07,
+      "loss": 0.097,
+      "step": 1254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5641033428478222,
+      "learning_rate": 9.014035000813531e-07,
+      "loss": 0.0706,
+      "step": 1255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0497581640328453,
+      "learning_rate": 8.954105185283324e-07,
+      "loss": 0.1037,
+      "step": 1256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3381709134091255,
+      "learning_rate": 8.894355657432497e-07,
+      "loss": 0.0807,
+      "step": 1257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9154332559676057,
+      "learning_rate": 8.83478667970149e-07,
+      "loss": 0.103,
+      "step": 1258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5683413366039711,
+      "learning_rate": 8.775398513737754e-07,
+      "loss": 0.0868,
+      "step": 1259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5479216699244795,
+      "learning_rate": 8.71619142039451e-07,
+      "loss": 0.094,
+      "step": 1260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1517933989870968,
+      "learning_rate": 8.657165659729671e-07,
+      "loss": 0.1104,
+      "step": 1261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.801198378747815,
+      "learning_rate": 8.59832149100468e-07,
+      "loss": 0.1238,
+      "step": 1262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.573203527001529,
+      "learning_rate": 8.539659172683329e-07,
+      "loss": 0.0835,
+      "step": 1263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0257387197632575,
+      "learning_rate": 8.481178962430713e-07,
+      "loss": 0.1047,
+      "step": 1264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7341343767324775,
+      "learning_rate": 8.422881117111987e-07,
+      "loss": 0.0955,
+      "step": 1265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.084835594489109,
+      "learning_rate": 8.364765892791349e-07,
+      "loss": 0.1051,
+      "step": 1266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.449522355959983,
+      "learning_rate": 8.306833544730841e-07,
+      "loss": 0.0871,
+      "step": 1267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5784050635904436,
+      "learning_rate": 8.249084327389234e-07,
+      "loss": 0.0918,
+      "step": 1268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6653894058429495,
+      "learning_rate": 8.191518494420953e-07,
+      "loss": 0.1005,
+      "step": 1269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1644391236635108,
+      "learning_rate": 8.134136298674933e-07,
+      "loss": 0.1258,
+      "step": 1270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7465947018602068,
+      "learning_rate": 8.076937992193478e-07,
+      "loss": 0.0958,
+      "step": 1271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7951316250275726,
+      "learning_rate": 8.01992382621124e-07,
+      "loss": 0.1203,
+      "step": 1272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8552742016937045,
+      "learning_rate": 7.963094051154014e-07,
+      "loss": 0.0919,
+      "step": 1273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5275354336077738,
+      "learning_rate": 7.906448916637705e-07,
+      "loss": 0.0855,
+      "step": 1274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3276563751497337,
+      "learning_rate": 7.84998867146724e-07,
+      "loss": 0.0906,
+      "step": 1275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1777455379281667,
+      "learning_rate": 7.793713563635396e-07,
+      "loss": 0.1345,
+      "step": 1276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6139708215357214,
+      "learning_rate": 7.737623840321811e-07,
+      "loss": 0.1427,
+      "step": 1277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.696636849936864,
+      "learning_rate": 7.681719747891813e-07,
+      "loss": 0.1131,
+      "step": 1278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0304350301117027,
+      "learning_rate": 7.6260015318954e-07,
+      "loss": 0.0966,
+      "step": 1279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3796777684299342,
+      "learning_rate": 7.570469437066147e-07,
+      "loss": 0.1111,
+      "step": 1280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9772318815498195,
+      "learning_rate": 7.515123707320071e-07,
+      "loss": 0.0926,
+      "step": 1281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6538203453203246,
+      "learning_rate": 7.45996458575467e-07,
+      "loss": 0.1158,
+      "step": 1282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2605541600335792,
+      "learning_rate": 7.404992314647746e-07,
+      "loss": 0.0995,
+      "step": 1283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5321706969606996,
+      "learning_rate": 7.350207135456416e-07,
+      "loss": 0.1232,
+      "step": 1284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.138546185384863,
+      "learning_rate": 7.295609288816041e-07,
+      "loss": 0.0928,
+      "step": 1285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9656824274684142,
+      "learning_rate": 7.241199014539097e-07,
+      "loss": 0.1355,
+      "step": 1286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7290513121096527,
+      "learning_rate": 7.186976551614233e-07,
+      "loss": 0.1299,
+      "step": 1287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9548477721792044,
+      "learning_rate": 7.132942138205135e-07,
+      "loss": 0.1151,
+      "step": 1288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6401086744005244,
+      "learning_rate": 7.079096011649522e-07,
+      "loss": 0.1259,
+      "step": 1289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8189732666722285,
+      "learning_rate": 7.025438408458107e-07,
+      "loss": 0.0989,
+      "step": 1290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6405758936091988,
+      "learning_rate": 6.971969564313507e-07,
+      "loss": 0.1013,
+      "step": 1291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1986403624499253,
+      "learning_rate": 6.918689714069282e-07,
+      "loss": 0.119,
+      "step": 1292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6237565653249362,
+      "learning_rate": 6.865599091748826e-07,
+      "loss": 0.1073,
+      "step": 1293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7584973484983841,
+      "learning_rate": 6.81269793054442e-07,
+      "loss": 0.1034,
+      "step": 1294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6483916129764982,
+      "learning_rate": 6.759986462816142e-07,
+      "loss": 0.1214,
+      "step": 1295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6732827265357355,
+      "learning_rate": 6.707464920090895e-07,
+      "loss": 0.1001,
+      "step": 1296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9849705649558909,
+      "learning_rate": 6.65513353306132e-07,
+      "loss": 0.1374,
+      "step": 1297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7522748414091973,
+      "learning_rate": 6.602992531584873e-07,
+      "loss": 0.1043,
+      "step": 1298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.006451353252143,
+      "learning_rate": 6.551042144682763e-07,
+      "loss": 0.1329,
+      "step": 1299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8847396234732365,
+      "learning_rate": 6.49928260053893e-07,
+      "loss": 0.088,
+      "step": 1300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8113950911459717,
+      "learning_rate": 6.447714126499088e-07,
+      "loss": 0.1302,
+      "step": 1301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.076649654393772,
+      "learning_rate": 6.39633694906972e-07,
+      "loss": 0.1177,
+      "step": 1302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5934392313053296,
+      "learning_rate": 6.345151293917023e-07,
+      "loss": 0.089,
+      "step": 1303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.905616552986884,
+      "learning_rate": 6.294157385866007e-07,
+      "loss": 0.1389,
+      "step": 1304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.605376887676223,
+      "learning_rate": 6.243355448899447e-07,
+      "loss": 0.1048,
+      "step": 1305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3039710961110615,
+      "learning_rate": 6.192745706156894e-07,
+      "loss": 0.0626,
+      "step": 1306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.559131515931037,
+      "learning_rate": 6.14232837993376e-07,
+      "loss": 0.1308,
+      "step": 1307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2536324978892384,
+      "learning_rate": 6.092103691680246e-07,
+      "loss": 0.1463,
+      "step": 1308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6071289534363706,
+      "learning_rate": 6.042071862000465e-07,
+      "loss": 0.0981,
+      "step": 1309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6910196306241885,
+      "learning_rate": 5.992233110651413e-07,
+      "loss": 0.0904,
+      "step": 1310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.905303264227101,
+      "learning_rate": 5.942587656542004e-07,
+      "loss": 0.1156,
+      "step": 1311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4674031858202814,
+      "learning_rate": 5.893135717732157e-07,
+      "loss": 0.1301,
+      "step": 1312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.168788921263074,
+      "learning_rate": 5.843877511431761e-07,
+      "loss": 0.1029,
+      "step": 1313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6230153532575458,
+      "learning_rate": 5.79481325399981e-07,
+      "loss": 0.0932,
+      "step": 1314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9422581189127146,
+      "learning_rate": 5.745943160943385e-07,
+      "loss": 0.1223,
+      "step": 1315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.318704383644754,
+      "learning_rate": 5.697267446916727e-07,
+      "loss": 0.11,
+      "step": 1316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.520375714262893,
+      "learning_rate": 5.648786325720313e-07,
+      "loss": 0.0848,
+      "step": 1317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9331430191129737,
+      "learning_rate": 5.600500010299881e-07,
+      "loss": 0.0928,
+      "step": 1318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.508072126931876,
+      "learning_rate": 5.552408712745533e-07,
+      "loss": 0.0905,
+      "step": 1319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.815138714776867,
+      "learning_rate": 5.504512644290788e-07,
+      "loss": 0.1155,
+      "step": 1320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.731977717124734,
+      "learning_rate": 5.456812015311624e-07,
+      "loss": 0.1131,
+      "step": 1321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0820938949916226,
+      "learning_rate": 5.409307035325618e-07,
+      "loss": 0.0907,
+      "step": 1322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.8741872841013345,
+      "learning_rate": 5.361997912990957e-07,
+      "loss": 0.1042,
+      "step": 1323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8501234480047337,
+      "learning_rate": 5.314884856105573e-07,
+      "loss": 0.1003,
+      "step": 1324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4600368165176527,
+      "learning_rate": 5.267968071606222e-07,
+      "loss": 0.1102,
+      "step": 1325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0208377111093574,
+      "learning_rate": 5.221247765567528e-07,
+      "loss": 0.0971,
+      "step": 1326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7829392216950242,
+      "learning_rate": 5.174724143201148e-07,
+      "loss": 0.1078,
+      "step": 1327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.756667315342706,
+      "learning_rate": 5.128397408854813e-07,
+      "loss": 0.1201,
+      "step": 1328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7607102414398703,
+      "learning_rate": 5.082267766011467e-07,
+      "loss": 0.0615,
+      "step": 1329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7279397201663833,
+      "learning_rate": 5.036335417288374e-07,
+      "loss": 0.1239,
+      "step": 1330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3223287090761953,
+      "learning_rate": 4.990600564436177e-07,
+      "loss": 0.0857,
+      "step": 1331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.915742325850939,
+      "learning_rate": 4.945063408338085e-07,
+      "loss": 0.0959,
+      "step": 1332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3073053408337982,
+      "learning_rate": 4.899724149008933e-07,
+      "loss": 0.0815,
+      "step": 1333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1067359417427327,
+      "learning_rate": 4.854582985594336e-07,
+      "loss": 0.107,
+      "step": 1334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3702086647301606,
+      "learning_rate": 4.809640116369807e-07,
+      "loss": 0.0718,
+      "step": 1335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.699005590583642,
+      "learning_rate": 4.7648957387398663e-07,
+      "loss": 0.0888,
+      "step": 1336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8006153872330963,
+      "learning_rate": 4.72035004923721e-07,
+      "loss": 0.1079,
+      "step": 1337
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5752108585132742,
+      "learning_rate": 4.6760032435218094e-07,
+      "loss": 0.1028,
+      "step": 1338
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5523323756196192,
+      "learning_rate": 4.6318555163800794e-07,
+      "loss": 0.082,
+      "step": 1339
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.581535153622859,
+      "learning_rate": 4.587907061724034e-07,
+      "loss": 0.1177,
+      "step": 1340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.386615267018464,
+      "learning_rate": 4.54415807259036e-07,
+      "loss": 0.1077,
+      "step": 1341
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7675094550376917,
+      "learning_rate": 4.500608741139689e-07,
+      "loss": 0.1226,
+      "step": 1342
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.107225110659914,
+      "learning_rate": 4.4572592586556274e-07,
+      "loss": 0.1256,
+      "step": 1343
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.534264455196776,
+      "learning_rate": 4.41410981554401e-07,
+      "loss": 0.078,
+      "step": 1344
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4464350919004405,
+      "learning_rate": 4.371160601332042e-07,
+      "loss": 0.0875,
+      "step": 1345
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.171917408741793,
+      "learning_rate": 4.328411804667415e-07,
+      "loss": 0.0976,
+      "step": 1346
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.651495115764142,
+      "learning_rate": 4.2858636133175537e-07,
+      "loss": 0.0912,
+      "step": 1347
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.411017646466787,
+      "learning_rate": 4.2435162141687194e-07,
+      "loss": 0.1018,
+      "step": 1348
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6052992507261568,
+      "learning_rate": 4.2013697932252627e-07,
+      "loss": 0.1046,
+      "step": 1349
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7365312052470743,
+      "learning_rate": 4.159424535608747e-07,
+      "loss": 0.1125,
+      "step": 1350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6771179899356692,
+      "learning_rate": 4.1176806255571746e-07,
+      "loss": 0.0767,
+      "step": 1351
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9218646131425015,
+      "learning_rate": 4.0761382464241275e-07,
+      "loss": 0.1195,
+      "step": 1352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.079269658569359,
+      "learning_rate": 4.0347975806780184e-07,
+      "loss": 0.0975,
+      "step": 1353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8693170141266313,
+      "learning_rate": 3.99365880990128e-07,
+      "loss": 0.1234,
+      "step": 1354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.761622569709359,
+      "learning_rate": 3.9527221147895144e-07,
+      "loss": 0.0927,
+      "step": 1355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4891796824023398,
+      "learning_rate": 3.911987675150763e-07,
+      "loss": 0.0761,
+      "step": 1356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9547443620936422,
+      "learning_rate": 3.8714556699047054e-07,
+      "loss": 0.0985,
+      "step": 1357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.71508462863976,
+      "learning_rate": 3.831126277081809e-07,
+      "loss": 0.0972,
+      "step": 1358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1865647338107532,
+      "learning_rate": 3.7909996738226537e-07,
+      "loss": 0.0926,
+      "step": 1359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9869812144690924,
+      "learning_rate": 3.7510760363770714e-07,
+      "loss": 0.1033,
+      "step": 1360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.785541452657883,
+      "learning_rate": 3.711355540103395e-07,
+      "loss": 0.1286,
+      "step": 1361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0581948981674167,
+      "learning_rate": 3.671838359467705e-07,
+      "loss": 0.116,
+      "step": 1362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6150202050168858,
+      "learning_rate": 3.63252466804303e-07,
+      "loss": 0.0981,
+      "step": 1363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6139054545514815,
+      "learning_rate": 3.59341463850863e-07,
+      "loss": 0.0851,
+      "step": 1364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9360777170557844,
+      "learning_rate": 3.5545084426491963e-07,
+      "loss": 0.1155,
+      "step": 1365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8506126432274976,
+      "learning_rate": 3.5158062513541036e-07,
+      "loss": 0.0763,
+      "step": 1366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.968890913075809,
+      "learning_rate": 3.4773082346167e-07,
+      "loss": 0.1267,
+      "step": 1367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.673222442493309,
+      "learning_rate": 3.439014561533488e-07,
+      "loss": 0.1033,
+      "step": 1368
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0098300835751193,
+      "learning_rate": 3.400925400303451e-07,
+      "loss": 0.1317,
+      "step": 1369
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8276152882241996,
+      "learning_rate": 3.3630409182272896e-07,
+      "loss": 0.0835,
+      "step": 1370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.392229319969008,
+      "learning_rate": 3.325361281706646e-07,
+      "loss": 0.0786,
+      "step": 1371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0259190953996344,
+      "learning_rate": 3.2878866562434666e-07,
+      "loss": 0.0957,
+      "step": 1372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3994973339817043,
+      "learning_rate": 3.2506172064391605e-07,
+      "loss": 0.0706,
+      "step": 1373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9425773637759582,
+      "learning_rate": 3.2135530959939767e-07,
+      "loss": 0.1076,
+      "step": 1374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5964359732182867,
+      "learning_rate": 3.1766944877062387e-07,
+      "loss": 0.1041,
+      "step": 1375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6663126596565565,
+      "learning_rate": 3.1400415434716e-07,
+      "loss": 0.1679,
+      "step": 1376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9689921232382654,
+      "learning_rate": 3.1035944242824077e-07,
+      "loss": 0.0927,
+      "step": 1377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9496388529441202,
+      "learning_rate": 3.0673532902269323e-07,
+      "loss": 0.1552,
+      "step": 1378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.433647386785067,
+      "learning_rate": 3.0313183004886894e-07,
+      "loss": 0.0704,
+      "step": 1379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7221556497886636,
+      "learning_rate": 2.9954896133457536e-07,
+      "loss": 0.0895,
+      "step": 1380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.813340673979937,
+      "learning_rate": 2.959867386170018e-07,
+      "loss": 0.1305,
+      "step": 1381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8887367666950123,
+      "learning_rate": 2.9244517754265624e-07,
+      "loss": 0.1116,
+      "step": 1382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.724113628132934,
+      "learning_rate": 2.889242936672915e-07,
+      "loss": 0.1058,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4971417466355235,
+      "learning_rate": 2.8542410245583965e-07,
+      "loss": 0.1513,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.944450344851451,
+      "learning_rate": 2.819446192823455e-07,
+      "loss": 0.1277,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.107987311880698,
+      "learning_rate": 2.784858594298928e-07,
+      "loss": 0.1376,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9120369570817841,
+      "learning_rate": 2.7504783809054524e-07,
+      "loss": 0.103,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9866422051792438,
+      "learning_rate": 2.7163057036527386e-07,
+      "loss": 0.0783,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7235673574630885,
+      "learning_rate": 2.682340712638931e-07,
+      "loss": 0.1052,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5155437133847502,
+      "learning_rate": 2.6485835570499496e-07,
+      "loss": 0.1111,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4465984926363034,
+      "learning_rate": 2.615034385158821e-07,
+      "loss": 0.1096,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9410060852090447,
+      "learning_rate": 2.581693344325048e-07,
+      "loss": 0.0983,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.021116437716017,
+      "learning_rate": 2.5485605809939253e-07,
+      "loss": 0.1355,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.18020026790015,
+      "learning_rate": 2.5156362406959564e-07,
+      "loss": 0.078,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0017179172037247,
+      "learning_rate": 2.4829204680461626e-07,
+      "loss": 0.0933,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1818146072860545,
+      "learning_rate": 2.450413406743446e-07,
+      "loss": 0.1163,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5819291262050605,
+      "learning_rate": 2.4181151995700146e-07,
+      "loss": 0.1078,
+      "step": 1397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.736708968598104,
+      "learning_rate": 2.386025988390678e-07,
+      "loss": 0.1283,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9280851996846244,
+      "learning_rate": 2.3541459141522894e-07,
+      "loss": 0.1502,
+      "step": 1399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.035864827975166,
+      "learning_rate": 2.322475116883105e-07,
+      "loss": 0.1535,
+      "step": 1400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.091516257516916,
+      "learning_rate": 2.2910137356921325e-07,
+      "loss": 0.1316,
+      "step": 1401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7329390530290005,
+      "learning_rate": 2.2597619087685874e-07,
+      "loss": 0.1017,
+      "step": 1402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.648710415442816,
+      "learning_rate": 2.2287197733812383e-07,
+      "loss": 0.1234,
+      "step": 1403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8801767205230242,
+      "learning_rate": 2.197887465877796e-07,
+      "loss": 0.106,
+      "step": 1404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4917355080321695,
+      "learning_rate": 2.1672651216843698e-07,
+      "loss": 0.1018,
+      "step": 1405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5157369357883248,
+      "learning_rate": 2.1368528753048234e-07,
+      "loss": 0.0852,
+      "step": 1406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5686151676879427,
+      "learning_rate": 2.1066508603201862e-07,
+      "loss": 0.08,
+      "step": 1407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5800995132819375,
+      "learning_rate": 2.0766592093880934e-07,
+      "loss": 0.108,
+      "step": 1408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8731292418109589,
+      "learning_rate": 2.046878054242196e-07,
+      "loss": 0.1165,
+      "step": 1409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.625115047806522,
+      "learning_rate": 2.017307525691542e-07,
+      "loss": 0.122,
+      "step": 1410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8551925821603361,
+      "learning_rate": 1.9879477536200786e-07,
+      "loss": 0.0919,
+      "step": 1411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5373585296146006,
+      "learning_rate": 1.9587988669860113e-07,
+      "loss": 0.1327,
+      "step": 1412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6368123757454525,
+      "learning_rate": 1.9298609938212641e-07,
+      "loss": 0.1024,
+      "step": 1413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5102918071695266,
+      "learning_rate": 1.9011342612309248e-07,
+      "loss": 0.1014,
+      "step": 1414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.928399186896058,
+      "learning_rate": 1.8726187953926833e-07,
+      "loss": 0.1018,
+      "step": 1415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6015556491974345,
+      "learning_rate": 1.8443147215562617e-07,
+      "loss": 0.0999,
+      "step": 1416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0479629218603264,
+      "learning_rate": 1.8162221640428912e-07,
+      "loss": 0.107,
+      "step": 1417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.336348686630352,
+      "learning_rate": 1.7883412462447292e-07,
+      "loss": 0.1599,
+      "step": 1418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6875419627180652,
+      "learning_rate": 1.760672090624349e-07,
+      "loss": 0.1338,
+      "step": 1419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3436701858234987,
+      "learning_rate": 1.733214818714213e-07,
+      "loss": 0.1008,
+      "step": 1420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0254065183126375,
+      "learning_rate": 1.7059695511160712e-07,
+      "loss": 0.1186,
+      "step": 1421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7256603704001379,
+      "learning_rate": 1.678936407500531e-07,
+      "loss": 0.1033,
+      "step": 1422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.847803376270568,
+      "learning_rate": 1.6521155066064276e-07,
+      "loss": 0.0877,
+      "step": 1423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5500932148117843,
+      "learning_rate": 1.6255069662403865e-07,
+      "loss": 0.0858,
+      "step": 1424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.18652420380819,
+      "learning_rate": 1.599110903276274e-07,
+      "loss": 0.1018,
+      "step": 1425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6021945415397538,
+      "learning_rate": 1.5729274336546584e-07,
+      "loss": 0.1168,
+      "step": 1426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3055029012521335,
+      "learning_rate": 1.5469566723823493e-07,
+      "loss": 0.1058,
+      "step": 1427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.68487165673662,
+      "learning_rate": 1.5211987335318424e-07,
+      "loss": 0.1075,
+      "step": 1428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9419405460987873,
+      "learning_rate": 1.4956537302408703e-07,
+      "loss": 0.0961,
+      "step": 1429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6832498033387395,
+      "learning_rate": 1.4703217747118748e-07,
+      "loss": 0.1008,
+      "step": 1430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0905006447579084,
+      "learning_rate": 1.4452029782114962e-07,
+      "loss": 0.0761,
+      "step": 1431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6638727887584686,
+      "learning_rate": 1.4202974510701352e-07,
+      "loss": 0.1254,
+      "step": 1432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8394491514748785,
+      "learning_rate": 1.3956053026814242e-07,
+      "loss": 0.0913,
+      "step": 1433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.695184906190245,
+      "learning_rate": 1.371126641501763e-07,
+      "loss": 0.1401,
+      "step": 1434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.090555292874061,
+      "learning_rate": 1.346861575049857e-07,
+      "loss": 0.0738,
+      "step": 1435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4368996437936123,
+      "learning_rate": 1.3228102099062055e-07,
+      "loss": 0.098,
+      "step": 1436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.10070004440555,
+      "learning_rate": 1.298972651712682e-07,
+      "loss": 0.111,
+      "step": 1437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7671896818019424,
+      "learning_rate": 1.2753490051720273e-07,
+      "loss": 0.1162,
+      "step": 1438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8504396251495214,
+      "learning_rate": 1.251939374047423e-07,
+      "loss": 0.0936,
+      "step": 1439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9012722618600095,
+      "learning_rate": 1.2287438611620185e-07,
+      "loss": 0.1283,
+      "step": 1440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.843998247236543,
+      "learning_rate": 1.2057625683984776e-07,
+      "loss": 0.1705,
+      "step": 1441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7252084874321323,
+      "learning_rate": 1.1829955966985385e-07,
+      "loss": 0.1007,
+      "step": 1442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5529314859786916,
+      "learning_rate": 1.1604430460625593e-07,
+      "loss": 0.0603,
+      "step": 1443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.963948826434045,
+      "learning_rate": 1.1381050155491013e-07,
+      "loss": 0.1292,
+      "step": 1444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8390891040629755,
+      "learning_rate": 1.1159816032744742e-07,
+      "loss": 0.1179,
+      "step": 1445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.485864391698009,
+      "learning_rate": 1.094072906412308e-07,
+      "loss": 0.1418,
+      "step": 1446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9376692960274544,
+      "learning_rate": 1.0723790211931318e-07,
+      "loss": 0.0898,
+      "step": 1447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.41456782126224,
+      "learning_rate": 1.0509000429039462e-07,
+      "loss": 0.1479,
+      "step": 1448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.689576167778868,
+      "learning_rate": 1.0296360658878013e-07,
+      "loss": 0.1314,
+      "step": 1449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7379030872650063,
+      "learning_rate": 1.0085871835434025e-07,
+      "loss": 0.1169,
+      "step": 1450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9275633567974124,
+      "learning_rate": 9.87753488324672e-08,
+      "loss": 0.09,
+      "step": 1451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3179852772495493,
+      "learning_rate": 9.671350717403605e-08,
+      "loss": 0.1221,
+      "step": 1452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4098404189265352,
+      "learning_rate": 9.46732024353636e-08,
+      "loss": 0.087,
+      "step": 1453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2314538937236406,
+      "learning_rate": 9.265444357817066e-08,
+      "loss": 0.117,
+      "step": 1454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0505181997424025,
+      "learning_rate": 9.065723946953986e-08,
+      "loss": 0.1495,
+      "step": 1455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7905067116279876,
+      "learning_rate": 8.868159888187733e-08,
+      "loss": 0.1193,
+      "step": 1456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.81799273893074,
+      "learning_rate": 8.672753049287664e-08,
+      "loss": 0.1203,
+      "step": 1457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.890868232522128,
+      "learning_rate": 8.479504288547769e-08,
+      "loss": 0.1222,
+      "step": 1458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5163383931955312,
+      "learning_rate": 8.288414454782901e-08,
+      "loss": 0.0799,
+      "step": 1459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3767272905844417,
+      "learning_rate": 8.099484387325496e-08,
+      "loss": 0.1081,
+      "step": 1460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7287779946935522,
+      "learning_rate": 7.912714916021247e-08,
+      "loss": 0.1233,
+      "step": 1461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6267375792799417,
+      "learning_rate": 7.728106861225881e-08,
+      "loss": 0.15,
+      "step": 1462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9028146859955775,
+      "learning_rate": 7.545661033801388e-08,
+      "loss": 0.1214,
+      "step": 1463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4298206498200186,
+      "learning_rate": 7.365378235112686e-08,
+      "loss": 0.0675,
+      "step": 1464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8210087293673047,
+      "learning_rate": 7.187259257023682e-08,
+      "loss": 0.1221,
+      "step": 1465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9812606141566764,
+      "learning_rate": 7.011304881894166e-08,
+      "loss": 0.1239,
+      "step": 1466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.785272114417367,
+      "learning_rate": 6.837515882576307e-08,
+      "loss": 0.1048,
+      "step": 1467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.215464996164515,
+      "learning_rate": 6.665893022411107e-08,
+      "loss": 0.1775,
+      "step": 1468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.203008183595577,
+      "learning_rate": 6.496437055225069e-08,
+      "loss": 0.1211,
+      "step": 1469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8000601715434712,
+      "learning_rate": 6.329148725327194e-08,
+      "loss": 0.1299,
+      "step": 1470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5637951655305367,
+      "learning_rate": 6.164028767505214e-08,
+      "loss": 0.0834,
+      "step": 1471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2653968543895533,
+      "learning_rate": 6.001077907022812e-08,
+      "loss": 0.085,
+      "step": 1472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5580858576013026,
+      "learning_rate": 5.840296859616179e-08,
+      "loss": 0.1167,
+      "step": 1473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.185376329825769,
+      "learning_rate": 5.68168633149091e-08,
+      "loss": 0.1184,
+      "step": 1474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2645285125903305,
+      "learning_rate": 5.525247019319002e-08,
+      "loss": 0.129,
+      "step": 1475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7106023406707447,
+      "learning_rate": 5.3709796102356385e-08,
+      "loss": 0.0788,
+      "step": 1476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3380603961586757,
+      "learning_rate": 5.218884781836297e-08,
+      "loss": 0.1284,
+      "step": 1477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.512720892675652,
+      "learning_rate": 5.068963202173871e-08,
+      "loss": 0.1356,
+      "step": 1478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4871931979530277,
+      "learning_rate": 4.92121552975533e-08,
+      "loss": 0.1214,
+      "step": 1479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7640991609880372,
+      "learning_rate": 4.775642413539339e-08,
+      "loss": 0.1143,
+      "step": 1480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7031338085396708,
+      "learning_rate": 4.63224449293298e-08,
+      "loss": 0.1065,
+      "step": 1481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3246872570618125,
+      "learning_rate": 4.4910223977892574e-08,
+      "loss": 0.0979,
+      "step": 1482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6097338260347043,
+      "learning_rate": 4.351976748404208e-08,
+      "loss": 0.1167,
+      "step": 1483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0636912682383777,
+      "learning_rate": 4.215108155514014e-08,
+      "loss": 0.1263,
+      "step": 1484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.961857851727143,
+      "learning_rate": 4.080417220292676e-08,
+      "loss": 0.1154,
+      "step": 1485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3164583272195585,
+      "learning_rate": 3.9479045343489565e-08,
+      "loss": 0.0697,
+      "step": 1486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.567057715269293,
+      "learning_rate": 3.817570679724269e-08,
+      "loss": 0.0815,
+      "step": 1487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8112893939559034,
+      "learning_rate": 3.6894162288896864e-08,
+      "loss": 0.1435,
+      "step": 1488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8971429929466814,
+      "learning_rate": 3.56344174474349e-08,
+      "loss": 0.1706,
+      "step": 1489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.924060975672454,
+      "learning_rate": 3.439647780609068e-08,
+      "loss": 0.1051,
+      "step": 1490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6477043584675648,
+      "learning_rate": 3.318034880231968e-08,
+      "loss": 0.1199,
+      "step": 1491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.518455162450937,
+      "learning_rate": 3.198603577777848e-08,
+      "loss": 0.095,
+      "step": 1492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.778985124002772,
+      "learning_rate": 3.081354397830083e-08,
+      "loss": 0.091,
+      "step": 1493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5009023960275205,
+      "learning_rate": 2.9662878553873286e-08,
+      "loss": 0.0883,
+      "step": 1494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5496874067735975,
+      "learning_rate": 2.853404455861519e-08,
+      "loss": 0.094,
+      "step": 1495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7150317244996973,
+      "learning_rate": 2.742704695075149e-08,
+      "loss": 0.1335,
+      "step": 1496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5255335098132747,
+      "learning_rate": 2.634189059259773e-08,
+      "loss": 0.0768,
+      "step": 1497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6059908970378205,
+      "learning_rate": 2.5278580250533426e-08,
+      "loss": 0.0926,
+      "step": 1498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1013819398925992,
+      "learning_rate": 2.4237120594982623e-08,
+      "loss": 0.0849,
+      "step": 1499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.064549400690215,
+      "learning_rate": 2.3217516200394474e-08,
+      "loss": 0.1496,
+      "step": 1500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.503585036914316,
+      "learning_rate": 2.221977154522159e-08,
+      "loss": 0.1173,
+      "step": 1501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6924026786355353,
+      "learning_rate": 2.124389101190172e-08,
+      "loss": 0.126,
+      "step": 1502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3013906000929985,
+      "learning_rate": 2.028987888683831e-08,
+      "loss": 0.0807,
+      "step": 1503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.561526266894261,
+      "learning_rate": 1.935773936037888e-08,
+      "loss": 0.1269,
+      "step": 1504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8204583155645853,
+      "learning_rate": 1.8447476526802787e-08,
+      "loss": 0.095,
+      "step": 1505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6129658981176747,
+      "learning_rate": 1.755909438429515e-08,
+      "loss": 0.1419,
+      "step": 1506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9525112868271206,
+      "learning_rate": 1.6692596834937402e-08,
+      "loss": 0.1011,
+      "step": 1507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5166569607658893,
+      "learning_rate": 1.584798768468343e-08,
+      "loss": 0.1127,
+      "step": 1508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9131305609595923,
+      "learning_rate": 1.502527064334791e-08,
+      "loss": 0.1484,
+      "step": 1509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9633908750208668,
+      "learning_rate": 1.422444932458633e-08,
+      "loss": 0.1188,
+      "step": 1510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7438826669306713,
+      "learning_rate": 1.3445527245881108e-08,
+      "loss": 0.1087,
+      "step": 1511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.407030042161492,
+      "learning_rate": 1.2688507828526042e-08,
+      "loss": 0.1052,
+      "step": 1512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.677863053038089,
+      "learning_rate": 1.1953394397610784e-08,
+      "loss": 0.1456,
+      "step": 1513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8106192549070852,
+      "learning_rate": 1.1240190182005838e-08,
+      "loss": 0.0885,
+      "step": 1514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9341430041133112,
+      "learning_rate": 1.0548898314349798e-08,
+      "loss": 0.138,
+      "step": 1515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.710589815924763,
+      "learning_rate": 9.879521831033801e-09,
+      "loss": 0.1021,
+      "step": 1516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3493401007508228,
+      "learning_rate": 9.232063672188762e-09,
+      "loss": 0.0788,
+      "step": 1517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.562870179187617,
+      "learning_rate": 8.606526681674831e-09,
+      "loss": 0.0565,
+      "step": 1518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5218941371791326,
+      "learning_rate": 8.002913607064732e-09,
+      "loss": 0.0959,
+      "step": 1519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2566841880869064,
+      "learning_rate": 7.421227099634887e-09,
+      "loss": 0.067,
+      "step": 1520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.673469964690565,
+      "learning_rate": 6.861469714352087e-09,
+      "loss": 0.086,
+      "step": 1521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4357324182592992,
+      "learning_rate": 6.323643909863508e-09,
+      "loss": 0.0783,
+      "step": 1522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7508482265946752,
+      "learning_rate": 5.8077520484844944e-09,
+      "loss": 0.1297,
+      "step": 1523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5920805207057862,
+      "learning_rate": 5.3137963961896745e-09,
+      "loss": 0.1059,
+      "step": 1524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.701684710269744,
+      "learning_rate": 4.841779122601864e-09,
+      "loss": 0.0931,
+      "step": 1525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8670628864440295,
+      "learning_rate": 4.391702300983735e-09,
+      "loss": 0.1095,
+      "step": 1526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.908885162509405,
+      "learning_rate": 3.963567908227273e-09,
+      "loss": 0.1092,
+      "step": 1527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5062237753340408,
+      "learning_rate": 3.5573778248454426e-09,
+      "loss": 0.1203,
+      "step": 1528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7524116046366505,
+      "learning_rate": 3.173133834964981e-09,
+      "loss": 0.1093,
+      "step": 1529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9006153233468437,
+      "learning_rate": 2.8108376263175087e-09,
+      "loss": 0.1049,
+      "step": 1530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1102330560966145,
+      "learning_rate": 2.470490790232871e-09,
+      "loss": 0.1352,
+      "step": 1531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8097652022371047,
+      "learning_rate": 2.1520948216308124e-09,
+      "loss": 0.1154,
+      "step": 1532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6090194528912374,
+      "learning_rate": 1.8556511190165326e-09,
+      "loss": 0.111,
+      "step": 1533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6317356695738183,
+      "learning_rate": 1.5811609844718078e-09,
+      "loss": 0.1025,
+      "step": 1534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.896746595936969,
+      "learning_rate": 1.3286256236522133e-09,
+      "loss": 0.1372,
+      "step": 1535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2497007471454076,
+      "learning_rate": 1.0980461457799075e-09,
+      "loss": 0.1063,
+      "step": 1536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1785740989098141,
+      "learning_rate": 8.894235636386362e-10,
+      "loss": 0.0622,
+      "step": 1537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.056741697011482,
+      "learning_rate": 7.027587935715119e-10,
+      "loss": 0.1564,
+      "step": 1538
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.0200483834115825,
+      "learning_rate": 5.380526554737974e-10,
+      "loss": 0.0995,
+      "step": 1539
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7186699717455354,
+      "learning_rate": 3.9530587279124065e-10,
+      "loss": 0.0835,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8843448685818118,
+      "learning_rate": 2.7451907251729857e-10,
+      "loss": 0.1151,
+      "step": 1541
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.363354272019609,
+      "learning_rate": 1.756927851881418e-10,
+      "loss": 0.1365,
+      "step": 1542
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.555450224617764,
+      "learning_rate": 9.882744488431961e-11,
+      "loss": 0.1124,
+      "step": 1543
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.729701488199868,
+      "learning_rate": 4.392338922354356e-11,
+      "loss": 0.1046,
+      "step": 1544
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2949144400623056,
+      "learning_rate": 1.098085936401816e-11,
+      "loss": 0.1159,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.637260805128306,
+      "learning_rate": 0.0,
+      "loss": 0.1204,
+      "step": 1546
+    },
+    {
+      "epoch": 1.0,
+      "step": 1546,
+      "total_flos": 71888985692160.0,
+      "train_loss": 0.12158192890572131,
+      "train_runtime": 29405.3322,
+      "train_samples_per_second": 4.207,
+      "train_steps_per_second": 0.053
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1546,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 71888985692160.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}