diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13105 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.995983935742972,
+  "eval_steps": 500,
+  "global_step": 1866,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001606425702811245,
+      "grad_norm": 55.425456981828255,
+      "learning_rate": 0.0,
+      "loss": 11.0914,
+      "step": 1
+    },
+    {
+      "epoch": 0.00321285140562249,
+      "grad_norm": 55.94565165274405,
+      "learning_rate": 2.6737967914438503e-07,
+      "loss": 11.072,
+      "step": 2
+    },
+    {
+      "epoch": 0.004819277108433735,
+      "grad_norm": 55.85336231763491,
+      "learning_rate": 5.347593582887701e-07,
+      "loss": 11.1735,
+      "step": 3
+    },
+    {
+      "epoch": 0.00642570281124498,
+      "grad_norm": 57.38149685878335,
+      "learning_rate": 8.021390374331552e-07,
+      "loss": 11.1121,
+      "step": 4
+    },
+    {
+      "epoch": 0.008032128514056224,
+      "grad_norm": 53.734533962442185,
+      "learning_rate": 1.0695187165775401e-06,
+      "loss": 11.1462,
+      "step": 5
+    },
+    {
+      "epoch": 0.00963855421686747,
+      "grad_norm": 56.856357722018124,
+      "learning_rate": 1.3368983957219252e-06,
+      "loss": 11.0398,
+      "step": 6
+    },
+    {
+      "epoch": 0.011244979919678716,
+      "grad_norm": 63.94394813269617,
+      "learning_rate": 1.6042780748663105e-06,
+      "loss": 10.6269,
+      "step": 7
+    },
+    {
+      "epoch": 0.01285140562248996,
+      "grad_norm": 59.69388329116394,
+      "learning_rate": 1.8716577540106951e-06,
+      "loss": 10.7207,
+      "step": 8
+    },
+    {
+      "epoch": 0.014457831325301205,
+      "grad_norm": 65.08501918743596,
+      "learning_rate": 2.1390374331550802e-06,
+      "loss": 10.6603,
+      "step": 9
+    },
+    {
+      "epoch": 0.01606425702811245,
+      "grad_norm": 90.6773969037033,
+      "learning_rate": 2.4064171122994653e-06,
+      "loss": 9.3899,
+      "step": 10
+    },
+    {
+      "epoch": 0.017670682730923693,
+      "grad_norm": 100.40100812807319,
+      "learning_rate": 2.6737967914438504e-06,
+      "loss": 9.067,
+      "step": 11
+    },
+    {
+      "epoch": 0.01927710843373494,
+      "grad_norm": 102.64034759457971,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 8.8161,
+      "step": 12
+    },
+    {
+      "epoch": 0.020883534136546186,
+      "grad_norm": 63.9375793920087,
+      "learning_rate": 3.208556149732621e-06,
+      "loss": 3.6047,
+      "step": 13
+    },
+    {
+      "epoch": 0.02248995983935743,
+      "grad_norm": 55.14440358777841,
+      "learning_rate": 3.4759358288770056e-06,
+      "loss": 3.2931,
+      "step": 14
+    },
+    {
+      "epoch": 0.024096385542168676,
+      "grad_norm": 38.756404211435886,
+      "learning_rate": 3.7433155080213903e-06,
+      "loss": 2.633,
+      "step": 15
+    },
+    {
+      "epoch": 0.02570281124497992,
+      "grad_norm": 32.624445944408805,
+      "learning_rate": 4.010695187165775e-06,
+      "loss": 2.347,
+      "step": 16
+    },
+    {
+      "epoch": 0.027309236947791166,
+      "grad_norm": 6.743482970548288,
+      "learning_rate": 4.2780748663101604e-06,
+      "loss": 1.455,
+      "step": 17
+    },
+    {
+      "epoch": 0.02891566265060241,
+      "grad_norm": 4.939939222809209,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.2889,
+      "step": 18
+    },
+    {
+      "epoch": 0.030522088353413655,
+      "grad_norm": 4.037313717639722,
+      "learning_rate": 4.812834224598931e-06,
+      "loss": 1.2466,
+      "step": 19
+    },
+    {
+      "epoch": 0.0321285140562249,
+      "grad_norm": 2.8978325234814135,
+      "learning_rate": 5.080213903743316e-06,
+      "loss": 1.1727,
+      "step": 20
+    },
+    {
+      "epoch": 0.033734939759036145,
+      "grad_norm": 2.4292778275916858,
+      "learning_rate": 5.347593582887701e-06,
+      "loss": 1.1498,
+      "step": 21
+    },
+    {
+      "epoch": 0.035341365461847386,
+      "grad_norm": 2.0750809208626078,
+      "learning_rate": 5.614973262032086e-06,
+      "loss": 1.1131,
+      "step": 22
+    },
+    {
+      "epoch": 0.036947791164658635,
+      "grad_norm": 1.38050141256256,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.969,
+      "step": 23
+    },
+    {
+      "epoch": 0.03855421686746988,
+      "grad_norm": 76.44314104688057,
+      "learning_rate": 6.149732620320856e-06,
+      "loss": 1.0004,
+      "step": 24
+    },
+    {
+      "epoch": 0.040160642570281124,
+      "grad_norm": 20.7727103372233,
+      "learning_rate": 6.417112299465242e-06,
+      "loss": 0.9133,
+      "step": 25
+    },
+    {
+      "epoch": 0.04176706827309237,
+      "grad_norm": 2.4294991120332226,
+      "learning_rate": 6.684491978609626e-06,
+      "loss": 0.937,
+      "step": 26
+    },
+    {
+      "epoch": 0.043373493975903614,
+      "grad_norm": 1.2141284131568895,
+      "learning_rate": 6.951871657754011e-06,
+      "loss": 0.868,
+      "step": 27
+    },
+    {
+      "epoch": 0.04497991967871486,
+      "grad_norm": 1.066098750215553,
+      "learning_rate": 7.2192513368983955e-06,
+      "loss": 0.8941,
+      "step": 28
+    },
+    {
+      "epoch": 0.046586345381526104,
+      "grad_norm": 0.8592318109384306,
+      "learning_rate": 7.4866310160427806e-06,
+      "loss": 0.8108,
+      "step": 29
+    },
+    {
+      "epoch": 0.04819277108433735,
+      "grad_norm": 0.8600957750100187,
+      "learning_rate": 7.754010695187166e-06,
+      "loss": 0.8223,
+      "step": 30
+    },
+    {
+      "epoch": 0.04979919678714859,
+      "grad_norm": 0.7679060410994321,
+      "learning_rate": 8.02139037433155e-06,
+      "loss": 0.7769,
+      "step": 31
+    },
+    {
+      "epoch": 0.05140562248995984,
+      "grad_norm": 0.8453908066226332,
+      "learning_rate": 8.288770053475937e-06,
+      "loss": 0.7396,
+      "step": 32
+    },
+    {
+      "epoch": 0.05301204819277108,
+      "grad_norm": 0.8000437895485839,
+      "learning_rate": 8.556149732620321e-06,
+      "loss": 0.7268,
+      "step": 33
+    },
+    {
+      "epoch": 0.05461847389558233,
+      "grad_norm": 0.5961380612903852,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.7434,
+      "step": 34
+    },
+    {
+      "epoch": 0.05622489959839357,
+      "grad_norm": 0.7023665082451603,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.7352,
+      "step": 35
+    },
+    {
+      "epoch": 0.05783132530120482,
+      "grad_norm": 0.6795471325789469,
+      "learning_rate": 9.358288770053477e-06,
+      "loss": 0.692,
+      "step": 36
+    },
+    {
+      "epoch": 0.05943775100401606,
+      "grad_norm": 0.6267269031521756,
+      "learning_rate": 9.625668449197861e-06,
+      "loss": 0.7032,
+      "step": 37
+    },
+    {
+      "epoch": 0.06104417670682731,
+      "grad_norm": 0.5038575166572804,
+      "learning_rate": 9.893048128342247e-06,
+      "loss": 0.6996,
+      "step": 38
+    },
+    {
+      "epoch": 0.06265060240963856,
+      "grad_norm": 0.5695324600161978,
+      "learning_rate": 1.0160427807486631e-05,
+      "loss": 0.7056,
+      "step": 39
+    },
+    {
+      "epoch": 0.0642570281124498,
+      "grad_norm": 0.5077700319850176,
+      "learning_rate": 1.0427807486631017e-05,
+      "loss": 0.6679,
+      "step": 40
+    },
+    {
+      "epoch": 0.06586345381526104,
+      "grad_norm": 0.4743877655225355,
+      "learning_rate": 1.0695187165775402e-05,
+      "loss": 0.6673,
+      "step": 41
+    },
+    {
+      "epoch": 0.06746987951807229,
+      "grad_norm": 0.5244542863377575,
+      "learning_rate": 1.0962566844919786e-05,
+      "loss": 0.6671,
+      "step": 42
+    },
+    {
+      "epoch": 0.06907630522088354,
+      "grad_norm": 0.419545810897641,
+      "learning_rate": 1.1229946524064172e-05,
+      "loss": 0.6545,
+      "step": 43
+    },
+    {
+      "epoch": 0.07068273092369477,
+      "grad_norm": 0.37783127229950486,
+      "learning_rate": 1.1497326203208558e-05,
+      "loss": 0.6576,
+      "step": 44
+    },
+    {
+      "epoch": 0.07228915662650602,
+      "grad_norm": 0.4178580350161024,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.643,
+      "step": 45
+    },
+    {
+      "epoch": 0.07389558232931727,
+      "grad_norm": 0.43845663225475845,
+      "learning_rate": 1.2032085561497326e-05,
+      "loss": 0.6346,
+      "step": 46
+    },
+    {
+      "epoch": 0.07550200803212852,
+      "grad_norm": 0.44070415442299,
+      "learning_rate": 1.2299465240641712e-05,
+      "loss": 0.6688,
+      "step": 47
+    },
+    {
+      "epoch": 0.07710843373493977,
+      "grad_norm": 0.3669910492352237,
+      "learning_rate": 1.2566844919786098e-05,
+      "loss": 0.6454,
+      "step": 48
+    },
+    {
+      "epoch": 0.078714859437751,
+      "grad_norm": 0.362801619015921,
+      "learning_rate": 1.2834224598930484e-05,
+      "loss": 0.6185,
+      "step": 49
+    },
+    {
+      "epoch": 0.08032128514056225,
+      "grad_norm": 0.36003109712794723,
+      "learning_rate": 1.3101604278074866e-05,
+      "loss": 0.6082,
+      "step": 50
+    },
+    {
+      "epoch": 0.0819277108433735,
+      "grad_norm": 0.3790196621978306,
+      "learning_rate": 1.3368983957219252e-05,
+      "loss": 0.6196,
+      "step": 51
+    },
+    {
+      "epoch": 0.08353413654618475,
+      "grad_norm": 0.32881533358374376,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.6379,
+      "step": 52
+    },
+    {
+      "epoch": 0.08514056224899598,
+      "grad_norm": 0.2897536087349657,
+      "learning_rate": 1.3903743315508022e-05,
+      "loss": 0.6166,
+      "step": 53
+    },
+    {
+      "epoch": 0.08674698795180723,
+      "grad_norm": 0.3347710929583827,
+      "learning_rate": 1.4171122994652408e-05,
+      "loss": 0.5807,
+      "step": 54
+    },
+    {
+      "epoch": 0.08835341365461848,
+      "grad_norm": 0.3393972794432162,
+      "learning_rate": 1.4438502673796791e-05,
+      "loss": 0.5832,
+      "step": 55
+    },
+    {
+      "epoch": 0.08995983935742972,
+      "grad_norm": 0.29350372494642607,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.5927,
+      "step": 56
+    },
+    {
+      "epoch": 0.09156626506024096,
+      "grad_norm": 0.3252911723032582,
+      "learning_rate": 1.4973262032085561e-05,
+      "loss": 0.619,
+      "step": 57
+    },
+    {
+      "epoch": 0.09317269076305221,
+      "grad_norm": 0.3149875163427349,
+      "learning_rate": 1.5240641711229947e-05,
+      "loss": 0.6133,
+      "step": 58
+    },
+    {
+      "epoch": 0.09477911646586346,
+      "grad_norm": 0.3108517007068939,
+      "learning_rate": 1.5508021390374333e-05,
+      "loss": 0.582,
+      "step": 59
+    },
+    {
+      "epoch": 0.0963855421686747,
+      "grad_norm": 0.28641822750047513,
+      "learning_rate": 1.5775401069518716e-05,
+      "loss": 0.6001,
+      "step": 60
+    },
+    {
+      "epoch": 0.09799196787148594,
+      "grad_norm": 0.26760169263753647,
+      "learning_rate": 1.60427807486631e-05,
+      "loss": 0.5834,
+      "step": 61
+    },
+    {
+      "epoch": 0.09959839357429719,
+      "grad_norm": 0.2982920467349389,
+      "learning_rate": 1.6310160427807487e-05,
+      "loss": 0.5869,
+      "step": 62
+    },
+    {
+      "epoch": 0.10120481927710843,
+      "grad_norm": 0.2671097597027451,
+      "learning_rate": 1.6577540106951873e-05,
+      "loss": 0.5573,
+      "step": 63
+    },
+    {
+      "epoch": 0.10281124497991968,
+      "grad_norm": 0.2676703133591729,
+      "learning_rate": 1.684491978609626e-05,
+      "loss": 0.5798,
+      "step": 64
+    },
+    {
+      "epoch": 0.10441767068273092,
+      "grad_norm": 0.26495729621046843,
+      "learning_rate": 1.7112299465240642e-05,
+      "loss": 0.5617,
+      "step": 65
+    },
+    {
+      "epoch": 0.10602409638554217,
+      "grad_norm": 0.26227499331646464,
+      "learning_rate": 1.7379679144385028e-05,
+      "loss": 0.5616,
+      "step": 66
+    },
+    {
+      "epoch": 0.10763052208835341,
+      "grad_norm": 0.2648120721870536,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.5634,
+      "step": 67
+    },
+    {
+      "epoch": 0.10923694779116466,
+      "grad_norm": 0.26215156916312604,
+      "learning_rate": 1.7914438502673796e-05,
+      "loss": 0.5684,
+      "step": 68
+    },
+    {
+      "epoch": 0.1108433734939759,
+      "grad_norm": 0.27157578337184746,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.5913,
+      "step": 69
+    },
+    {
+      "epoch": 0.11244979919678715,
+      "grad_norm": 0.27163024369026484,
+      "learning_rate": 1.8449197860962568e-05,
+      "loss": 0.566,
+      "step": 70
+    },
+    {
+      "epoch": 0.1140562248995984,
+      "grad_norm": 0.26979630288410955,
+      "learning_rate": 1.8716577540106954e-05,
+      "loss": 0.5656,
+      "step": 71
+    },
+    {
+      "epoch": 0.11566265060240964,
+      "grad_norm": 0.2633626931673852,
+      "learning_rate": 1.898395721925134e-05,
+      "loss": 0.5393,
+      "step": 72
+    },
+    {
+      "epoch": 0.11726907630522089,
+      "grad_norm": 0.24409011774218942,
+      "learning_rate": 1.9251336898395722e-05,
+      "loss": 0.5563,
+      "step": 73
+    },
+    {
+      "epoch": 0.11887550200803212,
+      "grad_norm": 0.25395135149818904,
+      "learning_rate": 1.951871657754011e-05,
+      "loss": 0.5489,
+      "step": 74
+    },
+    {
+      "epoch": 0.12048192771084337,
+      "grad_norm": 0.2550278467892386,
+      "learning_rate": 1.9786096256684494e-05,
+      "loss": 0.5614,
+      "step": 75
+    },
+    {
+      "epoch": 0.12208835341365462,
+      "grad_norm": 0.38496356216000965,
+      "learning_rate": 2.0053475935828877e-05,
+      "loss": 0.5704,
+      "step": 76
+    },
+    {
+      "epoch": 0.12369477911646587,
+      "grad_norm": 0.2567139059670428,
+      "learning_rate": 2.0320855614973263e-05,
+      "loss": 0.5459,
+      "step": 77
+    },
+    {
+      "epoch": 0.12530120481927712,
+      "grad_norm": 0.25038957425452213,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.5454,
+      "step": 78
+    },
+    {
+      "epoch": 0.12690763052208837,
+      "grad_norm": 0.23452705979435273,
+      "learning_rate": 2.0855614973262035e-05,
+      "loss": 0.5492,
+      "step": 79
+    },
+    {
+      "epoch": 0.1285140562248996,
+      "grad_norm": 0.24323287410165828,
+      "learning_rate": 2.112299465240642e-05,
+      "loss": 0.5301,
+      "step": 80
+    },
+    {
+      "epoch": 0.13012048192771083,
+      "grad_norm": 0.26334529962321623,
+      "learning_rate": 2.1390374331550803e-05,
+      "loss": 0.523,
+      "step": 81
+    },
+    {
+      "epoch": 0.13172690763052208,
+      "grad_norm": 0.22329323094624642,
+      "learning_rate": 2.165775401069519e-05,
+      "loss": 0.5592,
+      "step": 82
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.2474076406178327,
+      "learning_rate": 2.192513368983957e-05,
+      "loss": 0.5437,
+      "step": 83
+    },
+    {
+      "epoch": 0.13493975903614458,
+      "grad_norm": 0.24355733442576052,
+      "learning_rate": 2.2192513368983957e-05,
+      "loss": 0.542,
+      "step": 84
+    },
+    {
+      "epoch": 0.13654618473895583,
+      "grad_norm": 0.2577406108651445,
+      "learning_rate": 2.2459893048128343e-05,
+      "loss": 0.5287,
+      "step": 85
+    },
+    {
+      "epoch": 0.13815261044176708,
+      "grad_norm": 0.2512287959204346,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.5527,
+      "step": 86
+    },
+    {
+      "epoch": 0.13975903614457832,
+      "grad_norm": 0.24412043857047824,
+      "learning_rate": 2.2994652406417115e-05,
+      "loss": 0.5288,
+      "step": 87
+    },
+    {
+      "epoch": 0.14136546184738955,
+      "grad_norm": 0.24004859620662336,
+      "learning_rate": 2.32620320855615e-05,
+      "loss": 0.521,
+      "step": 88
+    },
+    {
+      "epoch": 0.1429718875502008,
+      "grad_norm": 0.2685831450802228,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.553,
+      "step": 89
+    },
+    {
+      "epoch": 0.14457831325301204,
+      "grad_norm": 0.22208558557827937,
+      "learning_rate": 2.379679144385027e-05,
+      "loss": 0.5273,
+      "step": 90
+    },
+    {
+      "epoch": 0.1461847389558233,
+      "grad_norm": 0.25800455920009974,
+      "learning_rate": 2.4064171122994652e-05,
+      "loss": 0.5292,
+      "step": 91
+    },
+    {
+      "epoch": 0.14779116465863454,
+      "grad_norm": 0.26150864060963513,
+      "learning_rate": 2.4331550802139038e-05,
+      "loss": 0.5344,
+      "step": 92
+    },
+    {
+      "epoch": 0.1493975903614458,
+      "grad_norm": 0.2503576394708978,
+      "learning_rate": 2.4598930481283424e-05,
+      "loss": 0.558,
+      "step": 93
+    },
+    {
+      "epoch": 0.15100401606425704,
+      "grad_norm": 0.2723029245292207,
+      "learning_rate": 2.4866310160427807e-05,
+      "loss": 0.5225,
+      "step": 94
+    },
+    {
+      "epoch": 0.15261044176706828,
+      "grad_norm": 0.2677703323128444,
+      "learning_rate": 2.5133689839572196e-05,
+      "loss": 0.5051,
+      "step": 95
+    },
+    {
+      "epoch": 0.15421686746987953,
+      "grad_norm": 0.2750945129448115,
+      "learning_rate": 2.5401069518716582e-05,
+      "loss": 0.538,
+      "step": 96
+    },
+    {
+      "epoch": 0.15582329317269075,
+      "grad_norm": 0.26505576602859154,
+      "learning_rate": 2.5668449197860968e-05,
+      "loss": 0.5183,
+      "step": 97
+    },
+    {
+      "epoch": 0.157429718875502,
+      "grad_norm": 0.28571711863479626,
+      "learning_rate": 2.5935828877005347e-05,
+      "loss": 0.4988,
+      "step": 98
+    },
+    {
+      "epoch": 0.15903614457831325,
+      "grad_norm": 0.2806910619639985,
+      "learning_rate": 2.6203208556149733e-05,
+      "loss": 0.5251,
+      "step": 99
+    },
+    {
+      "epoch": 0.1606425702811245,
+      "grad_norm": 0.2738624463968397,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.5186,
+      "step": 100
+    },
+    {
+      "epoch": 0.16224899598393575,
+      "grad_norm": 0.3061159832393734,
+      "learning_rate": 2.6737967914438505e-05,
+      "loss": 0.5365,
+      "step": 101
+    },
+    {
+      "epoch": 0.163855421686747,
+      "grad_norm": 0.24613646526070415,
+      "learning_rate": 2.700534759358289e-05,
+      "loss": 0.516,
+      "step": 102
+    },
+    {
+      "epoch": 0.16546184738955824,
+      "grad_norm": 0.29918830859130285,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.5269,
+      "step": 103
+    },
+    {
+      "epoch": 0.1670682730923695,
+      "grad_norm": 0.2954328993387257,
+      "learning_rate": 2.754010695187166e-05,
+      "loss": 0.5518,
+      "step": 104
+    },
+    {
+      "epoch": 0.1686746987951807,
+      "grad_norm": 0.2839510653050071,
+      "learning_rate": 2.7807486631016045e-05,
+      "loss": 0.5218,
+      "step": 105
+    },
+    {
+      "epoch": 0.17028112449799196,
+      "grad_norm": 0.2880614346246641,
+      "learning_rate": 2.807486631016043e-05,
+      "loss": 0.5333,
+      "step": 106
+    },
+    {
+      "epoch": 0.1718875502008032,
+      "grad_norm": 0.2724366108411872,
+      "learning_rate": 2.8342245989304817e-05,
+      "loss": 0.538,
+      "step": 107
+    },
+    {
+      "epoch": 0.17349397590361446,
+      "grad_norm": 0.2589102891755998,
+      "learning_rate": 2.8609625668449196e-05,
+      "loss": 0.5103,
+      "step": 108
+    },
+    {
+      "epoch": 0.1751004016064257,
+      "grad_norm": 0.2785106034490977,
+      "learning_rate": 2.8877005347593582e-05,
+      "loss": 0.5262,
+      "step": 109
+    },
+    {
+      "epoch": 0.17670682730923695,
+      "grad_norm": 0.282537837948337,
+      "learning_rate": 2.9144385026737968e-05,
+      "loss": 0.5457,
+      "step": 110
+    },
+    {
+      "epoch": 0.1783132530120482,
+      "grad_norm": 0.29229050692264746,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.5176,
+      "step": 111
+    },
+    {
+      "epoch": 0.17991967871485945,
+      "grad_norm": 0.28743219881720417,
+      "learning_rate": 2.9679144385026743e-05,
+      "loss": 0.5373,
+      "step": 112
+    },
+    {
+      "epoch": 0.18152610441767067,
+      "grad_norm": 0.29397840241802925,
+      "learning_rate": 2.9946524064171122e-05,
+      "loss": 0.5105,
+      "step": 113
+    },
+    {
+      "epoch": 0.18313253012048192,
+      "grad_norm": 0.2556368263796127,
+      "learning_rate": 3.0213903743315508e-05,
+      "loss": 0.5078,
+      "step": 114
+    },
+    {
+      "epoch": 0.18473895582329317,
+      "grad_norm": 0.2585019833123394,
+      "learning_rate": 3.0481283422459894e-05,
+      "loss": 0.5084,
+      "step": 115
+    },
+    {
+      "epoch": 0.18634538152610441,
+      "grad_norm": 0.3141816091328503,
+      "learning_rate": 3.0748663101604283e-05,
+      "loss": 0.4973,
+      "step": 116
+    },
+    {
+      "epoch": 0.18795180722891566,
+      "grad_norm": 0.25636118440595357,
+      "learning_rate": 3.1016042780748666e-05,
+      "loss": 0.514,
+      "step": 117
+    },
+    {
+      "epoch": 0.1895582329317269,
+      "grad_norm": 0.2876061400917617,
+      "learning_rate": 3.128342245989305e-05,
+      "loss": 0.5213,
+      "step": 118
+    },
+    {
+      "epoch": 0.19116465863453816,
+      "grad_norm": 0.34868750999339,
+      "learning_rate": 3.155080213903743e-05,
+      "loss": 0.5071,
+      "step": 119
+    },
+    {
+      "epoch": 0.1927710843373494,
+      "grad_norm": 0.3015203596477017,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.5435,
+      "step": 120
+    },
+    {
+      "epoch": 0.19437751004016066,
+      "grad_norm": 0.35119720538788096,
+      "learning_rate": 3.20855614973262e-05,
+      "loss": 0.522,
+      "step": 121
+    },
+    {
+      "epoch": 0.19598393574297188,
+      "grad_norm": 0.3477752312107028,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.5123,
+      "step": 122
+    },
+    {
+      "epoch": 0.19759036144578312,
+      "grad_norm": 0.29202999960570547,
+      "learning_rate": 3.2620320855614975e-05,
+      "loss": 0.5003,
+      "step": 123
+    },
+    {
+      "epoch": 0.19919678714859437,
+      "grad_norm": 0.35278041495114454,
+      "learning_rate": 3.288770053475936e-05,
+      "loss": 0.53,
+      "step": 124
+    },
+    {
+      "epoch": 0.20080321285140562,
+      "grad_norm": 0.3142151416610959,
+      "learning_rate": 3.3155080213903747e-05,
+      "loss": 0.4988,
+      "step": 125
+    },
+    {
+      "epoch": 0.20240963855421687,
+      "grad_norm": 0.2996067451911012,
+      "learning_rate": 3.342245989304813e-05,
+      "loss": 0.5052,
+      "step": 126
+    },
+    {
+      "epoch": 0.20401606425702812,
+      "grad_norm": 0.29176643730247437,
+      "learning_rate": 3.368983957219252e-05,
+      "loss": 0.5114,
+      "step": 127
+    },
+    {
+      "epoch": 0.20562248995983937,
+      "grad_norm": 0.30502573267866906,
+      "learning_rate": 3.39572192513369e-05,
+      "loss": 0.519,
+      "step": 128
+    },
+    {
+      "epoch": 0.20722891566265061,
+      "grad_norm": 0.3602982254540244,
+      "learning_rate": 3.4224598930481284e-05,
+      "loss": 0.4895,
+      "step": 129
+    },
+    {
+      "epoch": 0.20883534136546184,
+      "grad_norm": 0.3631477481903701,
+      "learning_rate": 3.4491978609625666e-05,
+      "loss": 0.5127,
+      "step": 130
+    },
+    {
+      "epoch": 0.21044176706827308,
+      "grad_norm": 0.30786368287839666,
+      "learning_rate": 3.4759358288770055e-05,
+      "loss": 0.4864,
+      "step": 131
+    },
+    {
+      "epoch": 0.21204819277108433,
+      "grad_norm": 0.43529055398718036,
+      "learning_rate": 3.5026737967914445e-05,
+      "loss": 0.5196,
+      "step": 132
+    },
+    {
+      "epoch": 0.21365461847389558,
+      "grad_norm": 0.3322195243932612,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.5045,
+      "step": 133
+    },
+    {
+      "epoch": 0.21526104417670683,
+      "grad_norm": 0.3158713023648389,
+      "learning_rate": 3.556149732620321e-05,
+      "loss": 0.5044,
+      "step": 134
+    },
+    {
+      "epoch": 0.21686746987951808,
+      "grad_norm": 0.5110582437494932,
+      "learning_rate": 3.582887700534759e-05,
+      "loss": 0.5133,
+      "step": 135
+    },
+    {
+      "epoch": 0.21847389558232932,
+      "grad_norm": 0.40979685812262173,
+      "learning_rate": 3.609625668449198e-05,
+      "loss": 0.522,
+      "step": 136
+    },
+    {
+      "epoch": 0.22008032128514057,
+      "grad_norm": 0.297856947051714,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.5046,
+      "step": 137
+    },
+    {
+      "epoch": 0.2216867469879518,
+      "grad_norm": 0.40746351543085246,
+      "learning_rate": 3.6631016042780753e-05,
+      "loss": 0.4896,
+      "step": 138
+    },
+    {
+      "epoch": 0.22329317269076304,
+      "grad_norm": 0.34013676583539143,
+      "learning_rate": 3.6898395721925136e-05,
+      "loss": 0.4802,
+      "step": 139
+    },
+    {
+      "epoch": 0.2248995983935743,
+      "grad_norm": 0.32168863523210933,
+      "learning_rate": 3.716577540106952e-05,
+      "loss": 0.521,
+      "step": 140
+    },
+    {
+      "epoch": 0.22650602409638554,
+      "grad_norm": 0.4805805507577635,
+      "learning_rate": 3.743315508021391e-05,
+      "loss": 0.5087,
+      "step": 141
+    },
+    {
+      "epoch": 0.2281124497991968,
+      "grad_norm": 0.4250197536795807,
+      "learning_rate": 3.770053475935829e-05,
+      "loss": 0.4834,
+      "step": 142
+    },
+    {
+      "epoch": 0.22971887550200804,
+      "grad_norm": 0.32342622533227194,
+      "learning_rate": 3.796791443850268e-05,
+      "loss": 0.4782,
+      "step": 143
+    },
+    {
+      "epoch": 0.23132530120481928,
+      "grad_norm": 0.507215711582196,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.5074,
+      "step": 144
+    },
+    {
+      "epoch": 0.23293172690763053,
+      "grad_norm": 0.3529654058092393,
+      "learning_rate": 3.8502673796791445e-05,
+      "loss": 0.4936,
+      "step": 145
+    },
+    {
+      "epoch": 0.23453815261044178,
+      "grad_norm": 0.3652515364897037,
+      "learning_rate": 3.877005347593583e-05,
+      "loss": 0.4939,
+      "step": 146
+    },
+    {
+      "epoch": 0.236144578313253,
+      "grad_norm": 0.4676716273810472,
+      "learning_rate": 3.903743315508022e-05,
+      "loss": 0.5188,
+      "step": 147
+    },
+    {
+      "epoch": 0.23775100401606425,
+      "grad_norm": 0.2695003287936738,
+      "learning_rate": 3.93048128342246e-05,
+      "loss": 0.5001,
+      "step": 148
+    },
+    {
+      "epoch": 0.2393574297188755,
+      "grad_norm": 0.4286990558745755,
+      "learning_rate": 3.957219251336899e-05,
+      "loss": 0.4756,
+      "step": 149
+    },
+    {
+      "epoch": 0.24096385542168675,
+      "grad_norm": 0.323374315472829,
+      "learning_rate": 3.983957219251337e-05,
+      "loss": 0.4936,
+      "step": 150
+    },
+    {
+      "epoch": 0.242570281124498,
+      "grad_norm": 0.3297390702151962,
+      "learning_rate": 4.0106951871657754e-05,
+      "loss": 0.48,
+      "step": 151
+    },
+    {
+      "epoch": 0.24417670682730924,
+      "grad_norm": 0.400561308543879,
+      "learning_rate": 4.037433155080214e-05,
+      "loss": 0.5215,
+      "step": 152
+    },
+    {
+      "epoch": 0.2457831325301205,
+      "grad_norm": 0.32182070624523856,
+      "learning_rate": 4.0641711229946525e-05,
+      "loss": 0.4903,
+      "step": 153
+    },
+    {
+      "epoch": 0.24738955823293174,
+      "grad_norm": 0.3388648123975653,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.507,
+      "step": 154
+    },
+    {
+      "epoch": 0.24899598393574296,
+      "grad_norm": 0.3793550770714379,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.4987,
+      "step": 155
+    },
+    {
+      "epoch": 0.25060240963855424,
+      "grad_norm": 0.2589296639211757,
+      "learning_rate": 4.144385026737968e-05,
+      "loss": 0.4902,
+      "step": 156
+    },
+    {
+      "epoch": 0.25220883534136546,
+      "grad_norm": 0.35110307031028004,
+      "learning_rate": 4.171122994652407e-05,
+      "loss": 0.4842,
+      "step": 157
+    },
+    {
+      "epoch": 0.25381526104417673,
+      "grad_norm": 0.2981715234028958,
+      "learning_rate": 4.197860962566845e-05,
+      "loss": 0.4898,
+      "step": 158
+    },
+    {
+      "epoch": 0.25542168674698795,
+      "grad_norm": 0.2950222833943252,
+      "learning_rate": 4.224598930481284e-05,
+      "loss": 0.494,
+      "step": 159
+    },
+    {
+      "epoch": 0.2570281124497992,
+      "grad_norm": 0.3257791358359003,
+      "learning_rate": 4.251336898395722e-05,
+      "loss": 0.4968,
+      "step": 160
+    },
+    {
+      "epoch": 0.25863453815261045,
+      "grad_norm": 0.30157818953474835,
+      "learning_rate": 4.2780748663101606e-05,
+      "loss": 0.4917,
+      "step": 161
+    },
+    {
+      "epoch": 0.26024096385542167,
+      "grad_norm": 0.3575465752435161,
+      "learning_rate": 4.304812834224599e-05,
+      "loss": 0.478,
+      "step": 162
+    },
+    {
+      "epoch": 0.26184738955823295,
+      "grad_norm": 0.31606947548536135,
+      "learning_rate": 4.331550802139038e-05,
+      "loss": 0.5003,
+      "step": 163
+    },
+    {
+      "epoch": 0.26345381526104417,
+      "grad_norm": 0.305902408372958,
+      "learning_rate": 4.358288770053476e-05,
+      "loss": 0.496,
+      "step": 164
+    },
+    {
+      "epoch": 0.26506024096385544,
+      "grad_norm": 0.32628577417972293,
+      "learning_rate": 4.385026737967914e-05,
+      "loss": 0.4685,
+      "step": 165
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.30821982802871856,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.4836,
+      "step": 166
+    },
+    {
+      "epoch": 0.26827309236947794,
+      "grad_norm": 0.4087606869865735,
+      "learning_rate": 4.4385026737967915e-05,
+      "loss": 0.4841,
+      "step": 167
+    },
+    {
+      "epoch": 0.26987951807228916,
+      "grad_norm": 0.3551707415888165,
+      "learning_rate": 4.4652406417112304e-05,
+      "loss": 0.4831,
+      "step": 168
+    },
+    {
+      "epoch": 0.2714859437751004,
+      "grad_norm": 0.41179453792210313,
+      "learning_rate": 4.491978609625669e-05,
+      "loss": 0.4961,
+      "step": 169
+    },
+    {
+      "epoch": 0.27309236947791166,
+      "grad_norm": 0.4108455418244695,
+      "learning_rate": 4.518716577540107e-05,
+      "loss": 0.5044,
+      "step": 170
+    },
+    {
+      "epoch": 0.2746987951807229,
+      "grad_norm": 0.2896753775037168,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.5103,
+      "step": 171
+    },
+    {
+      "epoch": 0.27630522088353415,
+      "grad_norm": 0.3758505698298356,
+      "learning_rate": 4.572192513368984e-05,
+      "loss": 0.5014,
+      "step": 172
+    },
+    {
+      "epoch": 0.2779116465863454,
+      "grad_norm": 0.29588516880323346,
+      "learning_rate": 4.598930481283423e-05,
+      "loss": 0.4885,
+      "step": 173
+    },
+    {
+      "epoch": 0.27951807228915665,
+      "grad_norm": 0.3147164922030141,
+      "learning_rate": 4.625668449197861e-05,
+      "loss": 0.4745,
+      "step": 174
+    },
+    {
+      "epoch": 0.28112449799196787,
+      "grad_norm": 0.3573489846573228,
+      "learning_rate": 4.6524064171123e-05,
+      "loss": 0.4842,
+      "step": 175
+    },
+    {
+      "epoch": 0.2827309236947791,
+      "grad_norm": 0.28544250545044564,
+      "learning_rate": 4.679144385026738e-05,
+      "loss": 0.4914,
+      "step": 176
+    },
+    {
+      "epoch": 0.28433734939759037,
+      "grad_norm": 0.37401209553179027,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.5144,
+      "step": 177
+    },
+    {
+      "epoch": 0.2859437751004016,
+      "grad_norm": 0.301831763757364,
+      "learning_rate": 4.732620320855615e-05,
+      "loss": 0.5014,
+      "step": 178
+    },
+    {
+      "epoch": 0.28755020080321286,
+      "grad_norm": 0.3211966241616229,
+      "learning_rate": 4.759358288770054e-05,
+      "loss": 0.4773,
+      "step": 179
+    },
+    {
+      "epoch": 0.2891566265060241,
+      "grad_norm": 0.32307860861614235,
+      "learning_rate": 4.786096256684492e-05,
+      "loss": 0.4859,
+      "step": 180
+    },
+    {
+      "epoch": 0.29076305220883536,
+      "grad_norm": 0.2828104733952161,
+      "learning_rate": 4.8128342245989304e-05,
+      "loss": 0.4655,
+      "step": 181
+    },
+    {
+      "epoch": 0.2923694779116466,
+      "grad_norm": 0.33271877867676497,
+      "learning_rate": 4.8395721925133694e-05,
+      "loss": 0.4765,
+      "step": 182
+    },
+    {
+      "epoch": 0.29397590361445786,
+      "grad_norm": 0.3260829000408802,
+      "learning_rate": 4.8663101604278076e-05,
+      "loss": 0.514,
+      "step": 183
+    },
+    {
+      "epoch": 0.2955823293172691,
+      "grad_norm": 0.3493692790433782,
+      "learning_rate": 4.8930481283422465e-05,
+      "loss": 0.4988,
+      "step": 184
+    },
+    {
+      "epoch": 0.2971887550200803,
+      "grad_norm": 0.32066390126257877,
+      "learning_rate": 4.919786096256685e-05,
+      "loss": 0.4717,
+      "step": 185
+    },
+    {
+      "epoch": 0.2987951807228916,
+      "grad_norm": 0.30294593832444744,
+      "learning_rate": 4.946524064171123e-05,
+      "loss": 0.4987,
+      "step": 186
+    },
+    {
+      "epoch": 0.3004016064257028,
+      "grad_norm": 0.28922416659203676,
+      "learning_rate": 4.973262032085561e-05,
+      "loss": 0.5041,
+      "step": 187
+    },
+    {
+      "epoch": 0.30200803212851407,
+      "grad_norm": 0.30600524056610523,
+      "learning_rate": 5e-05,
+      "loss": 0.4886,
+      "step": 188
+    },
+    {
+      "epoch": 0.3036144578313253,
+      "grad_norm": 0.31028179281939255,
+      "learning_rate": 4.997022036926742e-05,
+      "loss": 0.4908,
+      "step": 189
+    },
+    {
+      "epoch": 0.30522088353413657,
+      "grad_norm": 0.3598166672734698,
+      "learning_rate": 4.994044073853485e-05,
+      "loss": 0.4882,
+      "step": 190
+    },
+    {
+      "epoch": 0.3068273092369478,
+      "grad_norm": 0.3327095437744154,
+      "learning_rate": 4.9910661107802266e-05,
+      "loss": 0.4721,
+      "step": 191
+    },
+    {
+      "epoch": 0.30843373493975906,
+      "grad_norm": 0.33567656707078364,
+      "learning_rate": 4.9880881477069685e-05,
+      "loss": 0.5082,
+      "step": 192
+    },
+    {
+      "epoch": 0.3100401606425703,
+      "grad_norm": 0.30629964580097546,
+      "learning_rate": 4.9851101846337103e-05,
+      "loss": 0.4787,
+      "step": 193
+    },
+    {
+      "epoch": 0.3116465863453815,
+      "grad_norm": 0.4036803078419087,
+      "learning_rate": 4.982132221560453e-05,
+      "loss": 0.4864,
+      "step": 194
+    },
+    {
+      "epoch": 0.3132530120481928,
+      "grad_norm": 0.28293884051805157,
+      "learning_rate": 4.979154258487195e-05,
+      "loss": 0.4511,
+      "step": 195
+    },
+    {
+      "epoch": 0.314859437751004,
+      "grad_norm": 0.37218469846708774,
+      "learning_rate": 4.9761762954139374e-05,
+      "loss": 0.4775,
+      "step": 196
+    },
+    {
+      "epoch": 0.3164658634538153,
+      "grad_norm": 0.33300859531440663,
+      "learning_rate": 4.973198332340679e-05,
+      "loss": 0.4686,
+      "step": 197
+    },
+    {
+      "epoch": 0.3180722891566265,
+      "grad_norm": 0.3032456489439544,
+      "learning_rate": 4.970220369267422e-05,
+      "loss": 0.5017,
+      "step": 198
+    },
+    {
+      "epoch": 0.3196787148594378,
+      "grad_norm": 0.3382732850661887,
+      "learning_rate": 4.967242406194164e-05,
+      "loss": 0.4827,
+      "step": 199
+    },
+    {
+      "epoch": 0.321285140562249,
+      "grad_norm": 0.27738294797571994,
+      "learning_rate": 4.9642644431209056e-05,
+      "loss": 0.4825,
+      "step": 200
+    },
+    {
+      "epoch": 0.3228915662650602,
+      "grad_norm": 0.34032502157492933,
+      "learning_rate": 4.9612864800476475e-05,
+      "loss": 0.4684,
+      "step": 201
+    },
+    {
+      "epoch": 0.3244979919678715,
+      "grad_norm": 0.30558882899916945,
+      "learning_rate": 4.95830851697439e-05,
+      "loss": 0.4708,
+      "step": 202
+    },
+    {
+      "epoch": 0.3261044176706827,
+      "grad_norm": 0.33223396277492057,
+      "learning_rate": 4.955330553901132e-05,
+      "loss": 0.4721,
+      "step": 203
+    },
+    {
+      "epoch": 0.327710843373494,
+      "grad_norm": 0.35686857251795423,
+      "learning_rate": 4.952352590827874e-05,
+      "loss": 0.5025,
+      "step": 204
+    },
+    {
+      "epoch": 0.3293172690763052,
+      "grad_norm": 0.3144168287969755,
+      "learning_rate": 4.949374627754616e-05,
+      "loss": 0.4891,
+      "step": 205
+    },
+    {
+      "epoch": 0.3309236947791165,
+      "grad_norm": 0.36830817356433687,
+      "learning_rate": 4.946396664681358e-05,
+      "loss": 0.4821,
+      "step": 206
+    },
+    {
+      "epoch": 0.3325301204819277,
+      "grad_norm": 0.33504115804761303,
+      "learning_rate": 4.9434187016081e-05,
+      "loss": 0.4582,
+      "step": 207
+    },
+    {
+      "epoch": 0.334136546184739,
+      "grad_norm": 0.3374864014399646,
+      "learning_rate": 4.940440738534842e-05,
+      "loss": 0.4583,
+      "step": 208
+    },
+    {
+      "epoch": 0.3357429718875502,
+      "grad_norm": 0.32010693614220614,
+      "learning_rate": 4.9374627754615846e-05,
+      "loss": 0.476,
+      "step": 209
+    },
+    {
+      "epoch": 0.3373493975903614,
+      "grad_norm": 0.3921796646269442,
+      "learning_rate": 4.9344848123883265e-05,
+      "loss": 0.4856,
+      "step": 210
+    },
+    {
+      "epoch": 0.3389558232931727,
+      "grad_norm": 0.3583231598969687,
+      "learning_rate": 4.9315068493150684e-05,
+      "loss": 0.4861,
+      "step": 211
+    },
+    {
+      "epoch": 0.3405622489959839,
+      "grad_norm": 0.3197993185697981,
+      "learning_rate": 4.928528886241811e-05,
+      "loss": 0.4685,
+      "step": 212
+    },
+    {
+      "epoch": 0.3421686746987952,
+      "grad_norm": 0.3447752045592113,
+      "learning_rate": 4.9255509231685535e-05,
+      "loss": 0.478,
+      "step": 213
+    },
+    {
+      "epoch": 0.3437751004016064,
+      "grad_norm": 0.3070127440753511,
+      "learning_rate": 4.9225729600952954e-05,
+      "loss": 0.4906,
+      "step": 214
+    },
+    {
+      "epoch": 0.3453815261044177,
+      "grad_norm": 0.32721230883556446,
+      "learning_rate": 4.919594997022037e-05,
+      "loss": 0.4623,
+      "step": 215
+    },
+    {
+      "epoch": 0.3469879518072289,
+      "grad_norm": 0.3032610735626838,
+      "learning_rate": 4.916617033948779e-05,
+      "loss": 0.4727,
+      "step": 216
+    },
+    {
+      "epoch": 0.3485943775100402,
+      "grad_norm": 0.36030842044565586,
+      "learning_rate": 4.913639070875522e-05,
+      "loss": 0.4948,
+      "step": 217
+    },
+    {
+      "epoch": 0.3502008032128514,
+      "grad_norm": 0.3578112671065048,
+      "learning_rate": 4.9106611078022636e-05,
+      "loss": 0.5095,
+      "step": 218
+    },
+    {
+      "epoch": 0.35180722891566263,
+      "grad_norm": 0.29519131738013554,
+      "learning_rate": 4.9076831447290055e-05,
+      "loss": 0.4938,
+      "step": 219
+    },
+    {
+      "epoch": 0.3534136546184739,
+      "grad_norm": 0.3647464193576054,
+      "learning_rate": 4.9047051816557474e-05,
+      "loss": 0.4679,
+      "step": 220
+    },
+    {
+      "epoch": 0.3550200803212851,
+      "grad_norm": 0.28402742866330677,
+      "learning_rate": 4.90172721858249e-05,
+      "loss": 0.4889,
+      "step": 221
+    },
+    {
+      "epoch": 0.3566265060240964,
+      "grad_norm": 0.35954979625647304,
+      "learning_rate": 4.898749255509232e-05,
+      "loss": 0.4645,
+      "step": 222
+    },
+    {
+      "epoch": 0.3582329317269076,
+      "grad_norm": 0.32086331610979696,
+      "learning_rate": 4.895771292435974e-05,
+      "loss": 0.468,
+      "step": 223
+    },
+    {
+      "epoch": 0.3598393574297189,
+      "grad_norm": 0.3436497487869299,
+      "learning_rate": 4.892793329362716e-05,
+      "loss": 0.4791,
+      "step": 224
+    },
+    {
+      "epoch": 0.3614457831325301,
+      "grad_norm": 0.3487700049222101,
+      "learning_rate": 4.889815366289458e-05,
+      "loss": 0.4736,
+      "step": 225
+    },
+    {
+      "epoch": 0.36305220883534134,
+      "grad_norm": 0.31931610499502094,
+      "learning_rate": 4.8868374032162e-05,
+      "loss": 0.4513,
+      "step": 226
+    },
+    {
+      "epoch": 0.3646586345381526,
+      "grad_norm": 0.3402642317758613,
+      "learning_rate": 4.8838594401429426e-05,
+      "loss": 0.4464,
+      "step": 227
+    },
+    {
+      "epoch": 0.36626506024096384,
+      "grad_norm": 0.3420992789401369,
+      "learning_rate": 4.8808814770696845e-05,
+      "loss": 0.4391,
+      "step": 228
+    },
+    {
+      "epoch": 0.3678714859437751,
+      "grad_norm": 0.31932568158310554,
+      "learning_rate": 4.877903513996427e-05,
+      "loss": 0.4725,
+      "step": 229
+    },
+    {
+      "epoch": 0.36947791164658633,
+      "grad_norm": 0.3208368137704536,
+      "learning_rate": 4.874925550923169e-05,
+      "loss": 0.478,
+      "step": 230
+    },
+    {
+      "epoch": 0.3710843373493976,
+      "grad_norm": 0.3461690668941253,
+      "learning_rate": 4.871947587849911e-05,
+      "loss": 0.4954,
+      "step": 231
+    },
+    {
+      "epoch": 0.37269076305220883,
+      "grad_norm": 0.30920224902950433,
+      "learning_rate": 4.8689696247766534e-05,
+      "loss": 0.4585,
+      "step": 232
+    },
+    {
+      "epoch": 0.3742971887550201,
+      "grad_norm": 0.2968003609562749,
+      "learning_rate": 4.865991661703395e-05,
+      "loss": 0.46,
+      "step": 233
+    },
+    {
+      "epoch": 0.3759036144578313,
+      "grad_norm": 0.3822740069427355,
+      "learning_rate": 4.863013698630137e-05,
+      "loss": 0.4684,
+      "step": 234
+    },
+    {
+      "epoch": 0.37751004016064255,
+      "grad_norm": 0.28524554577386263,
+      "learning_rate": 4.860035735556879e-05,
+      "loss": 0.4593,
+      "step": 235
+    },
+    {
+      "epoch": 0.3791164658634538,
+      "grad_norm": 0.3972725136832213,
+      "learning_rate": 4.8570577724836216e-05,
+      "loss": 0.4696,
+      "step": 236
+    },
+    {
+      "epoch": 0.38072289156626504,
+      "grad_norm": 0.3054040969686469,
+      "learning_rate": 4.8540798094103635e-05,
+      "loss": 0.4606,
+      "step": 237
+    },
+    {
+      "epoch": 0.3823293172690763,
+      "grad_norm": 0.4114177747549895,
+      "learning_rate": 4.8511018463371054e-05,
+      "loss": 0.4704,
+      "step": 238
+    },
+    {
+      "epoch": 0.38393574297188754,
+      "grad_norm": 0.35567661029453534,
+      "learning_rate": 4.848123883263847e-05,
+      "loss": 0.4867,
+      "step": 239
+    },
+    {
+      "epoch": 0.3855421686746988,
+      "grad_norm": 0.4375083408829605,
+      "learning_rate": 4.84514592019059e-05,
+      "loss": 0.4777,
+      "step": 240
+    },
+    {
+      "epoch": 0.38714859437751004,
+      "grad_norm": 0.355462219685631,
+      "learning_rate": 4.842167957117332e-05,
+      "loss": 0.4755,
+      "step": 241
+    },
+    {
+      "epoch": 0.3887550200803213,
+      "grad_norm": 0.3868787344983709,
+      "learning_rate": 4.839189994044074e-05,
+      "loss": 0.4789,
+      "step": 242
+    },
+    {
+      "epoch": 0.39036144578313253,
+      "grad_norm": 0.36707388510401423,
+      "learning_rate": 4.836212030970816e-05,
+      "loss": 0.4716,
+      "step": 243
+    },
+    {
+      "epoch": 0.39196787148594375,
+      "grad_norm": 0.40507444450377067,
+      "learning_rate": 4.833234067897559e-05,
+      "loss": 0.4979,
+      "step": 244
+    },
+    {
+      "epoch": 0.39357429718875503,
+      "grad_norm": 0.41739786566121945,
+      "learning_rate": 4.8302561048243006e-05,
+      "loss": 0.4548,
+      "step": 245
+    },
+    {
+      "epoch": 0.39518072289156625,
+      "grad_norm": 0.3215725343767495,
+      "learning_rate": 4.8272781417510425e-05,
+      "loss": 0.4658,
+      "step": 246
+    },
+    {
+      "epoch": 0.3967871485943775,
+      "grad_norm": 0.39159227056027734,
+      "learning_rate": 4.824300178677785e-05,
+      "loss": 0.4591,
+      "step": 247
+    },
+    {
+      "epoch": 0.39839357429718875,
+      "grad_norm": 0.3084653643465216,
+      "learning_rate": 4.821322215604527e-05,
+      "loss": 0.4773,
+      "step": 248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38462131026029667,
+      "learning_rate": 4.818344252531269e-05,
+      "loss": 0.4822,
+      "step": 249
+    },
+    {
+      "epoch": 0.40160642570281124,
+      "grad_norm": 0.277347167260431,
+      "learning_rate": 4.815366289458011e-05,
+      "loss": 0.4794,
+      "step": 250
+    },
+    {
+      "epoch": 0.40321285140562246,
+      "grad_norm": 0.3530776967566136,
+      "learning_rate": 4.812388326384753e-05,
+      "loss": 0.4797,
+      "step": 251
+    },
+    {
+      "epoch": 0.40481927710843374,
+      "grad_norm": 0.32131209625672547,
+      "learning_rate": 4.809410363311495e-05,
+      "loss": 0.4656,
+      "step": 252
+    },
+    {
+      "epoch": 0.40642570281124496,
+      "grad_norm": 0.28272814685756176,
+      "learning_rate": 4.806432400238237e-05,
+      "loss": 0.4661,
+      "step": 253
+    },
+    {
+      "epoch": 0.40803212851405624,
+      "grad_norm": 0.3060278443090253,
+      "learning_rate": 4.803454437164979e-05,
+      "loss": 0.4656,
+      "step": 254
+    },
+    {
+      "epoch": 0.40963855421686746,
+      "grad_norm": 0.297979403125008,
+      "learning_rate": 4.8004764740917215e-05,
+      "loss": 0.4731,
+      "step": 255
+    },
+    {
+      "epoch": 0.41124497991967873,
+      "grad_norm": 0.2858924170890451,
+      "learning_rate": 4.7974985110184634e-05,
+      "loss": 0.4703,
+      "step": 256
+    },
+    {
+      "epoch": 0.41285140562248995,
+      "grad_norm": 0.2504985080692676,
+      "learning_rate": 4.794520547945205e-05,
+      "loss": 0.4528,
+      "step": 257
+    },
+    {
+      "epoch": 0.41445783132530123,
+      "grad_norm": 0.2624859429254907,
+      "learning_rate": 4.791542584871948e-05,
+      "loss": 0.4698,
+      "step": 258
+    },
+    {
+      "epoch": 0.41606425702811245,
+      "grad_norm": 0.2642208647062898,
+      "learning_rate": 4.7885646217986904e-05,
+      "loss": 0.4663,
+      "step": 259
+    },
+    {
+      "epoch": 0.41767068273092367,
+      "grad_norm": 0.27787708065649286,
+      "learning_rate": 4.785586658725432e-05,
+      "loss": 0.4824,
+      "step": 260
+    },
+    {
+      "epoch": 0.41927710843373495,
+      "grad_norm": 0.26698431800536987,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.4648,
+      "step": 261
+    },
+    {
+      "epoch": 0.42088353413654617,
+      "grad_norm": 0.28516211486766074,
+      "learning_rate": 4.779630732578916e-05,
+      "loss": 0.4507,
+      "step": 262
+    },
+    {
+      "epoch": 0.42248995983935744,
+      "grad_norm": 0.31276639788022564,
+      "learning_rate": 4.7766527695056586e-05,
+      "loss": 0.4667,
+      "step": 263
+    },
+    {
+      "epoch": 0.42409638554216866,
+      "grad_norm": 0.32259830537808853,
+      "learning_rate": 4.7736748064324005e-05,
+      "loss": 0.47,
+      "step": 264
+    },
+    {
+      "epoch": 0.42570281124497994,
+      "grad_norm": 0.3118440200761508,
+      "learning_rate": 4.7706968433591424e-05,
+      "loss": 0.4617,
+      "step": 265
+    },
+    {
+      "epoch": 0.42730923694779116,
+      "grad_norm": 0.2813632911379893,
+      "learning_rate": 4.767718880285885e-05,
+      "loss": 0.4573,
+      "step": 266
+    },
+    {
+      "epoch": 0.42891566265060244,
+      "grad_norm": 0.29321679658663186,
+      "learning_rate": 4.764740917212627e-05,
+      "loss": 0.4715,
+      "step": 267
+    },
+    {
+      "epoch": 0.43052208835341366,
+      "grad_norm": 0.22982181765765133,
+      "learning_rate": 4.761762954139369e-05,
+      "loss": 0.4409,
+      "step": 268
+    },
+    {
+      "epoch": 0.4321285140562249,
+      "grad_norm": 0.3006542887568241,
+      "learning_rate": 4.7587849910661106e-05,
+      "loss": 0.452,
+      "step": 269
+    },
+    {
+      "epoch": 0.43373493975903615,
+      "grad_norm": 0.29535878661480397,
+      "learning_rate": 4.755807027992853e-05,
+      "loss": 0.4778,
+      "step": 270
+    },
+    {
+      "epoch": 0.4353413654618474,
+      "grad_norm": 0.28772801070152304,
+      "learning_rate": 4.752829064919595e-05,
+      "loss": 0.4649,
+      "step": 271
+    },
+    {
+      "epoch": 0.43694779116465865,
+      "grad_norm": 0.2898158678126117,
+      "learning_rate": 4.749851101846337e-05,
+      "loss": 0.4686,
+      "step": 272
+    },
+    {
+      "epoch": 0.43855421686746987,
+      "grad_norm": 0.3234748814438192,
+      "learning_rate": 4.7468731387730795e-05,
+      "loss": 0.4675,
+      "step": 273
+    },
+    {
+      "epoch": 0.44016064257028115,
+      "grad_norm": 0.3492017937374251,
+      "learning_rate": 4.743895175699822e-05,
+      "loss": 0.4619,
+      "step": 274
+    },
+    {
+      "epoch": 0.44176706827309237,
+      "grad_norm": 0.3814195544285374,
+      "learning_rate": 4.740917212626564e-05,
+      "loss": 0.4438,
+      "step": 275
+    },
+    {
+      "epoch": 0.4433734939759036,
+      "grad_norm": 0.3585304162384897,
+      "learning_rate": 4.737939249553306e-05,
+      "loss": 0.4683,
+      "step": 276
+    },
+    {
+      "epoch": 0.44497991967871486,
+      "grad_norm": 0.3956072839161844,
+      "learning_rate": 4.734961286480048e-05,
+      "loss": 0.4601,
+      "step": 277
+    },
+    {
+      "epoch": 0.4465863453815261,
+      "grad_norm": 0.3911749897730973,
+      "learning_rate": 4.73198332340679e-05,
+      "loss": 0.4771,
+      "step": 278
+    },
+    {
+      "epoch": 0.44819277108433736,
+      "grad_norm": 0.3089732227604252,
+      "learning_rate": 4.729005360333532e-05,
+      "loss": 0.454,
+      "step": 279
+    },
+    {
+      "epoch": 0.4497991967871486,
+      "grad_norm": 0.39150802953689284,
+      "learning_rate": 4.726027397260274e-05,
+      "loss": 0.4565,
+      "step": 280
+    },
+    {
+      "epoch": 0.45140562248995986,
+      "grad_norm": 0.29879509492587897,
+      "learning_rate": 4.723049434187016e-05,
+      "loss": 0.4594,
+      "step": 281
+    },
+    {
+      "epoch": 0.4530120481927711,
+      "grad_norm": 0.36082290350674623,
+      "learning_rate": 4.7200714711137585e-05,
+      "loss": 0.4713,
+      "step": 282
+    },
+    {
+      "epoch": 0.45461847389558235,
+      "grad_norm": 0.35711098350802967,
+      "learning_rate": 4.7170935080405004e-05,
+      "loss": 0.4793,
+      "step": 283
+    },
+    {
+      "epoch": 0.4562248995983936,
+      "grad_norm": 0.2941919814125454,
+      "learning_rate": 4.714115544967242e-05,
+      "loss": 0.463,
+      "step": 284
+    },
+    {
+      "epoch": 0.4578313253012048,
+      "grad_norm": 0.33048721274799503,
+      "learning_rate": 4.711137581893985e-05,
+      "loss": 0.4801,
+      "step": 285
+    },
+    {
+      "epoch": 0.45943775100401607,
+      "grad_norm": 0.30614689618327984,
+      "learning_rate": 4.708159618820727e-05,
+      "loss": 0.473,
+      "step": 286
+    },
+    {
+      "epoch": 0.4610441767068273,
+      "grad_norm": 0.31807094823468063,
+      "learning_rate": 4.7051816557474686e-05,
+      "loss": 0.4736,
+      "step": 287
+    },
+    {
+      "epoch": 0.46265060240963857,
+      "grad_norm": 0.3790486412902381,
+      "learning_rate": 4.7022036926742105e-05,
+      "loss": 0.4693,
+      "step": 288
+    },
+    {
+      "epoch": 0.4642570281124498,
+      "grad_norm": 0.2811084565359757,
+      "learning_rate": 4.699225729600954e-05,
+      "loss": 0.4818,
+      "step": 289
+    },
+    {
+      "epoch": 0.46586345381526106,
+      "grad_norm": 0.3257486496872889,
+      "learning_rate": 4.6962477665276956e-05,
+      "loss": 0.4733,
+      "step": 290
+    },
+    {
+      "epoch": 0.4674698795180723,
+      "grad_norm": 0.26923234391566453,
+      "learning_rate": 4.6932698034544375e-05,
+      "loss": 0.4695,
+      "step": 291
+    },
+    {
+      "epoch": 0.46907630522088356,
+      "grad_norm": 0.37080256454192484,
+      "learning_rate": 4.6902918403811794e-05,
+      "loss": 0.4569,
+      "step": 292
+    },
+    {
+      "epoch": 0.4706827309236948,
+      "grad_norm": 0.30586107586982886,
+      "learning_rate": 4.687313877307922e-05,
+      "loss": 0.4654,
+      "step": 293
+    },
+    {
+      "epoch": 0.472289156626506,
+      "grad_norm": 0.36870299166682075,
+      "learning_rate": 4.684335914234664e-05,
+      "loss": 0.4407,
+      "step": 294
+    },
+    {
+      "epoch": 0.4738955823293173,
+      "grad_norm": 0.28377325722930746,
+      "learning_rate": 4.681357951161406e-05,
+      "loss": 0.4674,
+      "step": 295
+    },
+    {
+      "epoch": 0.4755020080321285,
+      "grad_norm": 0.2953311214919539,
+      "learning_rate": 4.6783799880881476e-05,
+      "loss": 0.4641,
+      "step": 296
+    },
+    {
+      "epoch": 0.4771084337349398,
+      "grad_norm": 0.3117438841349859,
+      "learning_rate": 4.67540202501489e-05,
+      "loss": 0.4601,
+      "step": 297
+    },
+    {
+      "epoch": 0.478714859437751,
+      "grad_norm": 0.256254866185249,
+      "learning_rate": 4.672424061941632e-05,
+      "loss": 0.4488,
+      "step": 298
+    },
+    {
+      "epoch": 0.48032128514056227,
+      "grad_norm": 0.3212937484737991,
+      "learning_rate": 4.669446098868374e-05,
+      "loss": 0.463,
+      "step": 299
+    },
+    {
+      "epoch": 0.4819277108433735,
+      "grad_norm": 0.2605989012252139,
+      "learning_rate": 4.6664681357951165e-05,
+      "loss": 0.4488,
+      "step": 300
+    },
+    {
+      "epoch": 0.4835341365461847,
+      "grad_norm": 0.36926616999987727,
+      "learning_rate": 4.6634901727218584e-05,
+      "loss": 0.4678,
+      "step": 301
+    },
+    {
+      "epoch": 0.485140562248996,
+      "grad_norm": 0.27259057184527724,
+      "learning_rate": 4.6605122096486e-05,
+      "loss": 0.4675,
+      "step": 302
+    },
+    {
+      "epoch": 0.4867469879518072,
+      "grad_norm": 0.332589334481269,
+      "learning_rate": 4.657534246575342e-05,
+      "loss": 0.4431,
+      "step": 303
+    },
+    {
+      "epoch": 0.4883534136546185,
+      "grad_norm": 0.28195976257236244,
+      "learning_rate": 4.654556283502085e-05,
+      "loss": 0.45,
+      "step": 304
+    },
+    {
+      "epoch": 0.4899598393574297,
+      "grad_norm": 0.3311412168331424,
+      "learning_rate": 4.651578320428827e-05,
+      "loss": 0.4865,
+      "step": 305
+    },
+    {
+      "epoch": 0.491566265060241,
+      "grad_norm": 0.2547234216247052,
+      "learning_rate": 4.648600357355569e-05,
+      "loss": 0.4507,
+      "step": 306
+    },
+    {
+      "epoch": 0.4931726907630522,
+      "grad_norm": 0.2993718374438959,
+      "learning_rate": 4.645622394282311e-05,
+      "loss": 0.4731,
+      "step": 307
+    },
+    {
+      "epoch": 0.4947791164658635,
+      "grad_norm": 0.25532758970228964,
+      "learning_rate": 4.6426444312090536e-05,
+      "loss": 0.4286,
+      "step": 308
+    },
+    {
+      "epoch": 0.4963855421686747,
+      "grad_norm": 0.3178442593811434,
+      "learning_rate": 4.6396664681357955e-05,
+      "loss": 0.4434,
+      "step": 309
+    },
+    {
+      "epoch": 0.4979919678714859,
+      "grad_norm": 0.3057512261403026,
+      "learning_rate": 4.6366885050625374e-05,
+      "loss": 0.4562,
+      "step": 310
+    },
+    {
+      "epoch": 0.4995983935742972,
+      "grad_norm": 0.3030624790962943,
+      "learning_rate": 4.633710541989279e-05,
+      "loss": 0.4863,
+      "step": 311
+    },
+    {
+      "epoch": 0.5012048192771085,
+      "grad_norm": 0.2979287221067525,
+      "learning_rate": 4.630732578916022e-05,
+      "loss": 0.4421,
+      "step": 312
+    },
+    {
+      "epoch": 0.5028112449799197,
+      "grad_norm": 0.31752192755216657,
+      "learning_rate": 4.627754615842764e-05,
+      "loss": 0.471,
+      "step": 313
+    },
+    {
+      "epoch": 0.5044176706827309,
+      "grad_norm": 0.2659253763839392,
+      "learning_rate": 4.6247766527695056e-05,
+      "loss": 0.4595,
+      "step": 314
+    },
+    {
+      "epoch": 0.5060240963855421,
+      "grad_norm": 0.3255884377064916,
+      "learning_rate": 4.6217986896962475e-05,
+      "loss": 0.4571,
+      "step": 315
+    },
+    {
+      "epoch": 0.5076305220883535,
+      "grad_norm": 0.2882963346908339,
+      "learning_rate": 4.61882072662299e-05,
+      "loss": 0.4629,
+      "step": 316
+    },
+    {
+      "epoch": 0.5092369477911647,
+      "grad_norm": 0.30168124050054673,
+      "learning_rate": 4.615842763549732e-05,
+      "loss": 0.4647,
+      "step": 317
+    },
+    {
+      "epoch": 0.5108433734939759,
+      "grad_norm": 0.3068238071113629,
+      "learning_rate": 4.612864800476474e-05,
+      "loss": 0.4602,
+      "step": 318
+    },
+    {
+      "epoch": 0.5124497991967871,
+      "grad_norm": 0.26522047223534434,
+      "learning_rate": 4.6098868374032164e-05,
+      "loss": 0.4504,
+      "step": 319
+    },
+    {
+      "epoch": 0.5140562248995983,
+      "grad_norm": 0.332945130105212,
+      "learning_rate": 4.606908874329959e-05,
+      "loss": 0.4427,
+      "step": 320
+    },
+    {
+      "epoch": 0.5156626506024097,
+      "grad_norm": 0.34353665143726153,
+      "learning_rate": 4.603930911256701e-05,
+      "loss": 0.4529,
+      "step": 321
+    },
+    {
+      "epoch": 0.5172690763052209,
+      "grad_norm": 0.35658461465036295,
+      "learning_rate": 4.600952948183443e-05,
+      "loss": 0.4475,
+      "step": 322
+    },
+    {
+      "epoch": 0.5188755020080321,
+      "grad_norm": 0.31267607053571417,
+      "learning_rate": 4.5979749851101846e-05,
+      "loss": 0.4453,
+      "step": 323
+    },
+    {
+      "epoch": 0.5204819277108433,
+      "grad_norm": 0.27347014445709883,
+      "learning_rate": 4.594997022036927e-05,
+      "loss": 0.4601,
+      "step": 324
+    },
+    {
+      "epoch": 0.5220883534136547,
+      "grad_norm": 0.3477235744237479,
+      "learning_rate": 4.592019058963669e-05,
+      "loss": 0.4543,
+      "step": 325
+    },
+    {
+      "epoch": 0.5236947791164659,
+      "grad_norm": 0.332561502304534,
+      "learning_rate": 4.589041095890411e-05,
+      "loss": 0.4442,
+      "step": 326
+    },
+    {
+      "epoch": 0.5253012048192771,
+      "grad_norm": 0.2809644940810856,
+      "learning_rate": 4.5860631328171535e-05,
+      "loss": 0.4535,
+      "step": 327
+    },
+    {
+      "epoch": 0.5269076305220883,
+      "grad_norm": 0.3576928200471685,
+      "learning_rate": 4.5830851697438954e-05,
+      "loss": 0.4749,
+      "step": 328
+    },
+    {
+      "epoch": 0.5285140562248996,
+      "grad_norm": 0.2703626548975125,
+      "learning_rate": 4.580107206670637e-05,
+      "loss": 0.4638,
+      "step": 329
+    },
+    {
+      "epoch": 0.5301204819277109,
+      "grad_norm": 0.3636772731014924,
+      "learning_rate": 4.577129243597379e-05,
+      "loss": 0.4788,
+      "step": 330
+    },
+    {
+      "epoch": 0.5317269076305221,
+      "grad_norm": 0.26520757679211987,
+      "learning_rate": 4.574151280524122e-05,
+      "loss": 0.4634,
+      "step": 331
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.3402903862345153,
+      "learning_rate": 4.5711733174508637e-05,
+      "loss": 0.4703,
+      "step": 332
+    },
+    {
+      "epoch": 0.5349397590361445,
+      "grad_norm": 0.27480422946520894,
+      "learning_rate": 4.5681953543776055e-05,
+      "loss": 0.4555,
+      "step": 333
+    },
+    {
+      "epoch": 0.5365461847389559,
+      "grad_norm": 0.2797169554300838,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 0.4534,
+      "step": 334
+    },
+    {
+      "epoch": 0.5381526104417671,
+      "grad_norm": 0.269091260800782,
+      "learning_rate": 4.562239428231091e-05,
+      "loss": 0.4494,
+      "step": 335
+    },
+    {
+      "epoch": 0.5397590361445783,
+      "grad_norm": 0.27190306586871754,
+      "learning_rate": 4.5592614651578326e-05,
+      "loss": 0.4522,
+      "step": 336
+    },
+    {
+      "epoch": 0.5413654618473895,
+      "grad_norm": 0.31334311480552396,
+      "learning_rate": 4.5562835020845744e-05,
+      "loss": 0.4475,
+      "step": 337
+    },
+    {
+      "epoch": 0.5429718875502008,
+      "grad_norm": 0.2845123465716489,
+      "learning_rate": 4.553305539011316e-05,
+      "loss": 0.4533,
+      "step": 338
+    },
+    {
+      "epoch": 0.5445783132530121,
+      "grad_norm": 0.3345213805522309,
+      "learning_rate": 4.550327575938059e-05,
+      "loss": 0.4546,
+      "step": 339
+    },
+    {
+      "epoch": 0.5461847389558233,
+      "grad_norm": 0.3299456597920992,
+      "learning_rate": 4.547349612864801e-05,
+      "loss": 0.4549,
+      "step": 340
+    },
+    {
+      "epoch": 0.5477911646586345,
+      "grad_norm": 0.31320544671321066,
+      "learning_rate": 4.5443716497915427e-05,
+      "loss": 0.448,
+      "step": 341
+    },
+    {
+      "epoch": 0.5493975903614458,
+      "grad_norm": 0.31255596857244,
+      "learning_rate": 4.541393686718285e-05,
+      "loss": 0.4694,
+      "step": 342
+    },
+    {
+      "epoch": 0.551004016064257,
+      "grad_norm": 0.25871783218229927,
+      "learning_rate": 4.538415723645027e-05,
+      "loss": 0.4543,
+      "step": 343
+    },
+    {
+      "epoch": 0.5526104417670683,
+      "grad_norm": 0.304269822025114,
+      "learning_rate": 4.535437760571769e-05,
+      "loss": 0.4507,
+      "step": 344
+    },
+    {
+      "epoch": 0.5542168674698795,
+      "grad_norm": 0.2656090175745257,
+      "learning_rate": 4.532459797498511e-05,
+      "loss": 0.4502,
+      "step": 345
+    },
+    {
+      "epoch": 0.5558232931726907,
+      "grad_norm": 0.2907242019133886,
+      "learning_rate": 4.5294818344252534e-05,
+      "loss": 0.4587,
+      "step": 346
+    },
+    {
+      "epoch": 0.557429718875502,
+      "grad_norm": 0.28206137237730466,
+      "learning_rate": 4.526503871351995e-05,
+      "loss": 0.4351,
+      "step": 347
+    },
+    {
+      "epoch": 0.5590361445783133,
+      "grad_norm": 0.28586719298856994,
+      "learning_rate": 4.523525908278737e-05,
+      "loss": 0.4681,
+      "step": 348
+    },
+    {
+      "epoch": 0.5606425702811245,
+      "grad_norm": 0.28463207414012487,
+      "learning_rate": 4.520547945205479e-05,
+      "loss": 0.4487,
+      "step": 349
+    },
+    {
+      "epoch": 0.5622489959839357,
+      "grad_norm": 0.28954112182874076,
+      "learning_rate": 4.5175699821322223e-05,
+      "loss": 0.4486,
+      "step": 350
+    },
+    {
+      "epoch": 0.563855421686747,
+      "grad_norm": 0.26885455709526246,
+      "learning_rate": 4.514592019058964e-05,
+      "loss": 0.4299,
+      "step": 351
+    },
+    {
+      "epoch": 0.5654618473895582,
+      "grad_norm": 0.2396609192525509,
+      "learning_rate": 4.511614055985706e-05,
+      "loss": 0.4486,
+      "step": 352
+    },
+    {
+      "epoch": 0.5670682730923695,
+      "grad_norm": 0.26759074416116824,
+      "learning_rate": 4.508636092912448e-05,
+      "loss": 0.4616,
+      "step": 353
+    },
+    {
+      "epoch": 0.5686746987951807,
+      "grad_norm": 0.24351279249189578,
+      "learning_rate": 4.5056581298391906e-05,
+      "loss": 0.4469,
+      "step": 354
+    },
+    {
+      "epoch": 0.570281124497992,
+      "grad_norm": 0.299896902235091,
+      "learning_rate": 4.5026801667659325e-05,
+      "loss": 0.4651,
+      "step": 355
+    },
+    {
+      "epoch": 0.5718875502008032,
+      "grad_norm": 0.25008325396684505,
+      "learning_rate": 4.499702203692674e-05,
+      "loss": 0.4741,
+      "step": 356
+    },
+    {
+      "epoch": 0.5734939759036145,
+      "grad_norm": 0.27752152084274156,
+      "learning_rate": 4.496724240619416e-05,
+      "loss": 0.4428,
+      "step": 357
+    },
+    {
+      "epoch": 0.5751004016064257,
+      "grad_norm": 0.28830683956457576,
+      "learning_rate": 4.493746277546159e-05,
+      "loss": 0.4503,
+      "step": 358
+    },
+    {
+      "epoch": 0.576706827309237,
+      "grad_norm": 0.2803688735487852,
+      "learning_rate": 4.490768314472901e-05,
+      "loss": 0.4577,
+      "step": 359
+    },
+    {
+      "epoch": 0.5783132530120482,
+      "grad_norm": 0.2782744621460348,
+      "learning_rate": 4.4877903513996426e-05,
+      "loss": 0.4612,
+      "step": 360
+    },
+    {
+      "epoch": 0.5799196787148594,
+      "grad_norm": 0.2587035764907811,
+      "learning_rate": 4.484812388326385e-05,
+      "loss": 0.422,
+      "step": 361
+    },
+    {
+      "epoch": 0.5815261044176707,
+      "grad_norm": 0.3402519963361588,
+      "learning_rate": 4.481834425253127e-05,
+      "loss": 0.4531,
+      "step": 362
+    },
+    {
+      "epoch": 0.5831325301204819,
+      "grad_norm": 0.4367512894100571,
+      "learning_rate": 4.478856462179869e-05,
+      "loss": 0.4525,
+      "step": 363
+    },
+    {
+      "epoch": 0.5847389558232932,
+      "grad_norm": 0.3058357640519284,
+      "learning_rate": 4.475878499106611e-05,
+      "loss": 0.4635,
+      "step": 364
+    },
+    {
+      "epoch": 0.5863453815261044,
+      "grad_norm": 0.28012373506975213,
+      "learning_rate": 4.4729005360333533e-05,
+      "loss": 0.4578,
+      "step": 365
+    },
+    {
+      "epoch": 0.5879518072289157,
+      "grad_norm": 0.2659693307554147,
+      "learning_rate": 4.469922572960096e-05,
+      "loss": 0.443,
+      "step": 366
+    },
+    {
+      "epoch": 0.5895582329317269,
+      "grad_norm": 0.2631611308004764,
+      "learning_rate": 4.466944609886838e-05,
+      "loss": 0.4509,
+      "step": 367
+    },
+    {
+      "epoch": 0.5911646586345382,
+      "grad_norm": 0.26249207008050757,
+      "learning_rate": 4.46396664681358e-05,
+      "loss": 0.4595,
+      "step": 368
+    },
+    {
+      "epoch": 0.5927710843373494,
+      "grad_norm": 0.2765644480893888,
+      "learning_rate": 4.460988683740322e-05,
+      "loss": 0.4515,
+      "step": 369
+    },
+    {
+      "epoch": 0.5943775100401606,
+      "grad_norm": 0.24009658771752188,
+      "learning_rate": 4.458010720667064e-05,
+      "loss": 0.446,
+      "step": 370
+    },
+    {
+      "epoch": 0.5959839357429719,
+      "grad_norm": 0.27183227934106,
+      "learning_rate": 4.455032757593806e-05,
+      "loss": 0.4305,
+      "step": 371
+    },
+    {
+      "epoch": 0.5975903614457831,
+      "grad_norm": 0.2517239356470154,
+      "learning_rate": 4.452054794520548e-05,
+      "loss": 0.4537,
+      "step": 372
+    },
+    {
+      "epoch": 0.5991967871485944,
+      "grad_norm": 0.30072025208230435,
+      "learning_rate": 4.4490768314472905e-05,
+      "loss": 0.468,
+      "step": 373
+    },
+    {
+      "epoch": 0.6008032128514056,
+      "grad_norm": 0.25818282402507275,
+      "learning_rate": 4.4460988683740323e-05,
+      "loss": 0.4406,
+      "step": 374
+    },
+    {
+      "epoch": 0.6024096385542169,
+      "grad_norm": 0.2851206120266935,
+      "learning_rate": 4.443120905300774e-05,
+      "loss": 0.4749,
+      "step": 375
+    },
+    {
+      "epoch": 0.6040160642570281,
+      "grad_norm": 0.3297131575662275,
+      "learning_rate": 4.440142942227517e-05,
+      "loss": 0.4398,
+      "step": 376
+    },
+    {
+      "epoch": 0.6056224899598394,
+      "grad_norm": 0.2824428795313683,
+      "learning_rate": 4.437164979154259e-05,
+      "loss": 0.4612,
+      "step": 377
+    },
+    {
+      "epoch": 0.6072289156626506,
+      "grad_norm": 0.33993463600355445,
+      "learning_rate": 4.4341870160810006e-05,
+      "loss": 0.437,
+      "step": 378
+    },
+    {
+      "epoch": 0.6088353413654618,
+      "grad_norm": 0.307216587987155,
+      "learning_rate": 4.4312090530077425e-05,
+      "loss": 0.4501,
+      "step": 379
+    },
+    {
+      "epoch": 0.6104417670682731,
+      "grad_norm": 0.34827036194744154,
+      "learning_rate": 4.428231089934485e-05,
+      "loss": 0.4679,
+      "step": 380
+    },
+    {
+      "epoch": 0.6120481927710844,
+      "grad_norm": 0.3504294248069219,
+      "learning_rate": 4.4252531268612276e-05,
+      "loss": 0.443,
+      "step": 381
+    },
+    {
+      "epoch": 0.6136546184738956,
+      "grad_norm": 0.33883468625080354,
+      "learning_rate": 4.4222751637879695e-05,
+      "loss": 0.4497,
+      "step": 382
+    },
+    {
+      "epoch": 0.6152610441767068,
+      "grad_norm": 0.30376665359188365,
+      "learning_rate": 4.4192972007147114e-05,
+      "loss": 0.4533,
+      "step": 383
+    },
+    {
+      "epoch": 0.6168674698795181,
+      "grad_norm": 0.3434907934945615,
+      "learning_rate": 4.416319237641454e-05,
+      "loss": 0.4524,
+      "step": 384
+    },
+    {
+      "epoch": 0.6184738955823293,
+      "grad_norm": 0.2816956482632934,
+      "learning_rate": 4.413341274568196e-05,
+      "loss": 0.433,
+      "step": 385
+    },
+    {
+      "epoch": 0.6200803212851406,
+      "grad_norm": 0.32605794938456106,
+      "learning_rate": 4.410363311494938e-05,
+      "loss": 0.4472,
+      "step": 386
+    },
+    {
+      "epoch": 0.6216867469879518,
+      "grad_norm": 0.3125757813939004,
+      "learning_rate": 4.4073853484216796e-05,
+      "loss": 0.4484,
+      "step": 387
+    },
+    {
+      "epoch": 0.623293172690763,
+      "grad_norm": 0.28211080101504216,
+      "learning_rate": 4.404407385348422e-05,
+      "loss": 0.4485,
+      "step": 388
+    },
+    {
+      "epoch": 0.6248995983935743,
+      "grad_norm": 0.3567349794579873,
+      "learning_rate": 4.401429422275164e-05,
+      "loss": 0.4761,
+      "step": 389
+    },
+    {
+      "epoch": 0.6265060240963856,
+      "grad_norm": 0.29512054372110924,
+      "learning_rate": 4.398451459201906e-05,
+      "loss": 0.4696,
+      "step": 390
+    },
+    {
+      "epoch": 0.6281124497991968,
+      "grad_norm": 0.2783488849608133,
+      "learning_rate": 4.395473496128648e-05,
+      "loss": 0.4618,
+      "step": 391
+    },
+    {
+      "epoch": 0.629718875502008,
+      "grad_norm": 0.31557267040914205,
+      "learning_rate": 4.3924955330553904e-05,
+      "loss": 0.4745,
+      "step": 392
+    },
+    {
+      "epoch": 0.6313253012048192,
+      "grad_norm": 0.2770239046196387,
+      "learning_rate": 4.389517569982132e-05,
+      "loss": 0.4454,
+      "step": 393
+    },
+    {
+      "epoch": 0.6329317269076306,
+      "grad_norm": 0.24316677828872768,
+      "learning_rate": 4.386539606908874e-05,
+      "loss": 0.4385,
+      "step": 394
+    },
+    {
+      "epoch": 0.6345381526104418,
+      "grad_norm": 0.35557179231752045,
+      "learning_rate": 4.383561643835617e-05,
+      "loss": 0.4468,
+      "step": 395
+    },
+    {
+      "epoch": 0.636144578313253,
+      "grad_norm": 0.24743033676911177,
+      "learning_rate": 4.380583680762359e-05,
+      "loss": 0.4559,
+      "step": 396
+    },
+    {
+      "epoch": 0.6377510040160642,
+      "grad_norm": 0.32902571391845864,
+      "learning_rate": 4.377605717689101e-05,
+      "loss": 0.444,
+      "step": 397
+    },
+    {
+      "epoch": 0.6393574297188755,
+      "grad_norm": 0.2922383486142178,
+      "learning_rate": 4.374627754615843e-05,
+      "loss": 0.4467,
+      "step": 398
+    },
+    {
+      "epoch": 0.6409638554216868,
+      "grad_norm": 0.31920179525012243,
+      "learning_rate": 4.371649791542585e-05,
+      "loss": 0.4639,
+      "step": 399
+    },
+    {
+      "epoch": 0.642570281124498,
+      "grad_norm": 0.2631355475269553,
+      "learning_rate": 4.3686718284693275e-05,
+      "loss": 0.4428,
+      "step": 400
+    },
+    {
+      "epoch": 0.6441767068273092,
+      "grad_norm": 0.28345993830303334,
+      "learning_rate": 4.3656938653960694e-05,
+      "loss": 0.4369,
+      "step": 401
+    },
+    {
+      "epoch": 0.6457831325301204,
+      "grad_norm": 0.26637720130752807,
+      "learning_rate": 4.362715902322811e-05,
+      "loss": 0.4623,
+      "step": 402
+    },
+    {
+      "epoch": 0.6473895582329318,
+      "grad_norm": 0.3203791905662697,
+      "learning_rate": 4.359737939249554e-05,
+      "loss": 0.4289,
+      "step": 403
+    },
+    {
+      "epoch": 0.648995983935743,
+      "grad_norm": 0.22842574861228213,
+      "learning_rate": 4.356759976176296e-05,
+      "loss": 0.4277,
+      "step": 404
+    },
+    {
+      "epoch": 0.6506024096385542,
+      "grad_norm": 0.3067343141644802,
+      "learning_rate": 4.3537820131030376e-05,
+      "loss": 0.4433,
+      "step": 405
+    },
+    {
+      "epoch": 0.6522088353413654,
+      "grad_norm": 0.2812065783428807,
+      "learning_rate": 4.3508040500297795e-05,
+      "loss": 0.4322,
+      "step": 406
+    },
+    {
+      "epoch": 0.6538152610441768,
+      "grad_norm": 0.3040061933492447,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.437,
+      "step": 407
+    },
+    {
+      "epoch": 0.655421686746988,
+      "grad_norm": 0.2987941211140326,
+      "learning_rate": 4.344848123883264e-05,
+      "loss": 0.4555,
+      "step": 408
+    },
+    {
+      "epoch": 0.6570281124497992,
+      "grad_norm": 0.2659280792695334,
+      "learning_rate": 4.341870160810006e-05,
+      "loss": 0.431,
+      "step": 409
+    },
+    {
+      "epoch": 0.6586345381526104,
+      "grad_norm": 0.28023761684807197,
+      "learning_rate": 4.3388921977367484e-05,
+      "loss": 0.4706,
+      "step": 410
+    },
+    {
+      "epoch": 0.6602409638554216,
+      "grad_norm": 0.2963346211131788,
+      "learning_rate": 4.33591423466349e-05,
+      "loss": 0.4563,
+      "step": 411
+    },
+    {
+      "epoch": 0.661847389558233,
+      "grad_norm": 0.23247955721717886,
+      "learning_rate": 4.332936271590233e-05,
+      "loss": 0.4297,
+      "step": 412
+    },
+    {
+      "epoch": 0.6634538152610442,
+      "grad_norm": 0.29491691232014927,
+      "learning_rate": 4.329958308516975e-05,
+      "loss": 0.4274,
+      "step": 413
+    },
+    {
+      "epoch": 0.6650602409638554,
+      "grad_norm": 0.25617212995433053,
+      "learning_rate": 4.3269803454437166e-05,
+      "loss": 0.4564,
+      "step": 414
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.23564356472993905,
+      "learning_rate": 4.324002382370459e-05,
+      "loss": 0.4276,
+      "step": 415
+    },
+    {
+      "epoch": 0.668273092369478,
+      "grad_norm": 0.29822994568037164,
+      "learning_rate": 4.321024419297201e-05,
+      "loss": 0.4456,
+      "step": 416
+    },
+    {
+      "epoch": 0.6698795180722892,
+      "grad_norm": 0.27049585970987766,
+      "learning_rate": 4.318046456223943e-05,
+      "loss": 0.4378,
+      "step": 417
+    },
+    {
+      "epoch": 0.6714859437751004,
+      "grad_norm": 0.24936723553806212,
+      "learning_rate": 4.3150684931506855e-05,
+      "loss": 0.4616,
+      "step": 418
+    },
+    {
+      "epoch": 0.6730923694779116,
+      "grad_norm": 0.2711234211903887,
+      "learning_rate": 4.3120905300774274e-05,
+      "loss": 0.4564,
+      "step": 419
+    },
+    {
+      "epoch": 0.6746987951807228,
+      "grad_norm": 0.2545386509478158,
+      "learning_rate": 4.309112567004169e-05,
+      "loss": 0.4395,
+      "step": 420
+    },
+    {
+      "epoch": 0.6763052208835342,
+      "grad_norm": 0.2584633324804562,
+      "learning_rate": 4.306134603930911e-05,
+      "loss": 0.4498,
+      "step": 421
+    },
+    {
+      "epoch": 0.6779116465863454,
+      "grad_norm": 0.2719252003406635,
+      "learning_rate": 4.303156640857654e-05,
+      "loss": 0.4382,
+      "step": 422
+    },
+    {
+      "epoch": 0.6795180722891566,
+      "grad_norm": 0.24135698151855509,
+      "learning_rate": 4.3001786777843956e-05,
+      "loss": 0.4347,
+      "step": 423
+    },
+    {
+      "epoch": 0.6811244979919678,
+      "grad_norm": 0.268141549049486,
+      "learning_rate": 4.2972007147111375e-05,
+      "loss": 0.4357,
+      "step": 424
+    },
+    {
+      "epoch": 0.6827309236947792,
+      "grad_norm": 0.2693180226887475,
+      "learning_rate": 4.2942227516378794e-05,
+      "loss": 0.4683,
+      "step": 425
+    },
+    {
+      "epoch": 0.6843373493975904,
+      "grad_norm": 0.27574435135497044,
+      "learning_rate": 4.291244788564622e-05,
+      "loss": 0.4281,
+      "step": 426
+    },
+    {
+      "epoch": 0.6859437751004016,
+      "grad_norm": 0.2655247693683116,
+      "learning_rate": 4.2882668254913645e-05,
+      "loss": 0.4408,
+      "step": 427
+    },
+    {
+      "epoch": 0.6875502008032128,
+      "grad_norm": 0.23605713593282013,
+      "learning_rate": 4.2852888624181064e-05,
+      "loss": 0.4264,
+      "step": 428
+    },
+    {
+      "epoch": 0.689156626506024,
+      "grad_norm": 0.2668756850543248,
+      "learning_rate": 4.282310899344848e-05,
+      "loss": 0.4386,
+      "step": 429
+    },
+    {
+      "epoch": 0.6907630522088354,
+      "grad_norm": 0.28508990279511975,
+      "learning_rate": 4.279332936271591e-05,
+      "loss": 0.4608,
+      "step": 430
+    },
+    {
+      "epoch": 0.6923694779116466,
+      "grad_norm": 0.2673675554092482,
+      "learning_rate": 4.276354973198333e-05,
+      "loss": 0.452,
+      "step": 431
+    },
+    {
+      "epoch": 0.6939759036144578,
+      "grad_norm": 0.25167957382435874,
+      "learning_rate": 4.2733770101250746e-05,
+      "loss": 0.4448,
+      "step": 432
+    },
+    {
+      "epoch": 0.695582329317269,
+      "grad_norm": 0.2748905077940865,
+      "learning_rate": 4.2703990470518165e-05,
+      "loss": 0.4328,
+      "step": 433
+    },
+    {
+      "epoch": 0.6971887550200804,
+      "grad_norm": 0.25694191753243817,
+      "learning_rate": 4.267421083978559e-05,
+      "loss": 0.4433,
+      "step": 434
+    },
+    {
+      "epoch": 0.6987951807228916,
+      "grad_norm": 0.2953083002140047,
+      "learning_rate": 4.264443120905301e-05,
+      "loss": 0.4346,
+      "step": 435
+    },
+    {
+      "epoch": 0.7004016064257028,
+      "grad_norm": 0.31346385697940604,
+      "learning_rate": 4.261465157832043e-05,
+      "loss": 0.4585,
+      "step": 436
+    },
+    {
+      "epoch": 0.702008032128514,
+      "grad_norm": 0.2943130181110173,
+      "learning_rate": 4.2584871947587854e-05,
+      "loss": 0.4301,
+      "step": 437
+    },
+    {
+      "epoch": 0.7036144578313253,
+      "grad_norm": 0.2970743817353812,
+      "learning_rate": 4.255509231685527e-05,
+      "loss": 0.4532,
+      "step": 438
+    },
+    {
+      "epoch": 0.7052208835341366,
+      "grad_norm": 0.2831138548453156,
+      "learning_rate": 4.252531268612269e-05,
+      "loss": 0.4466,
+      "step": 439
+    },
+    {
+      "epoch": 0.7068273092369478,
+      "grad_norm": 0.26374093187932474,
+      "learning_rate": 4.249553305539011e-05,
+      "loss": 0.4274,
+      "step": 440
+    },
+    {
+      "epoch": 0.708433734939759,
+      "grad_norm": 0.24426239862559207,
+      "learning_rate": 4.2465753424657536e-05,
+      "loss": 0.4424,
+      "step": 441
+    },
+    {
+      "epoch": 0.7100401606425703,
+      "grad_norm": 0.2641064913637875,
+      "learning_rate": 4.2435973793924955e-05,
+      "loss": 0.4408,
+      "step": 442
+    },
+    {
+      "epoch": 0.7116465863453815,
+      "grad_norm": 0.2235304256466447,
+      "learning_rate": 4.240619416319238e-05,
+      "loss": 0.4275,
+      "step": 443
+    },
+    {
+      "epoch": 0.7132530120481928,
+      "grad_norm": 0.24784504525361167,
+      "learning_rate": 4.23764145324598e-05,
+      "loss": 0.4381,
+      "step": 444
+    },
+    {
+      "epoch": 0.714859437751004,
+      "grad_norm": 0.2370545453905028,
+      "learning_rate": 4.2346634901727225e-05,
+      "loss": 0.4468,
+      "step": 445
+    },
+    {
+      "epoch": 0.7164658634538152,
+      "grad_norm": 0.2562860337291831,
+      "learning_rate": 4.2316855270994644e-05,
+      "loss": 0.4705,
+      "step": 446
+    },
+    {
+      "epoch": 0.7180722891566265,
+      "grad_norm": 0.23651519929818382,
+      "learning_rate": 4.228707564026206e-05,
+      "loss": 0.4372,
+      "step": 447
+    },
+    {
+      "epoch": 0.7196787148594378,
+      "grad_norm": 0.28608450172484184,
+      "learning_rate": 4.225729600952948e-05,
+      "loss": 0.4408,
+      "step": 448
+    },
+    {
+      "epoch": 0.721285140562249,
+      "grad_norm": 0.2378382301528455,
+      "learning_rate": 4.222751637879691e-05,
+      "loss": 0.4695,
+      "step": 449
+    },
+    {
+      "epoch": 0.7228915662650602,
+      "grad_norm": 0.30330808848040486,
+      "learning_rate": 4.2197736748064326e-05,
+      "loss": 0.4597,
+      "step": 450
+    },
+    {
+      "epoch": 0.7244979919678715,
+      "grad_norm": 0.26876726977866233,
+      "learning_rate": 4.2167957117331745e-05,
+      "loss": 0.4658,
+      "step": 451
+    },
+    {
+      "epoch": 0.7261044176706827,
+      "grad_norm": 0.2844637152748802,
+      "learning_rate": 4.213817748659917e-05,
+      "loss": 0.4413,
+      "step": 452
+    },
+    {
+      "epoch": 0.727710843373494,
+      "grad_norm": 0.2615741396788308,
+      "learning_rate": 4.210839785586659e-05,
+      "loss": 0.4299,
+      "step": 453
+    },
+    {
+      "epoch": 0.7293172690763052,
+      "grad_norm": 0.27257653372835033,
+      "learning_rate": 4.207861822513401e-05,
+      "loss": 0.4481,
+      "step": 454
+    },
+    {
+      "epoch": 0.7309236947791165,
+      "grad_norm": 0.25639819683469073,
+      "learning_rate": 4.204883859440143e-05,
+      "loss": 0.4324,
+      "step": 455
+    },
+    {
+      "epoch": 0.7325301204819277,
+      "grad_norm": 0.23672287570546763,
+      "learning_rate": 4.201905896366885e-05,
+      "loss": 0.4341,
+      "step": 456
+    },
+    {
+      "epoch": 0.734136546184739,
+      "grad_norm": 0.23858902027069817,
+      "learning_rate": 4.198927933293627e-05,
+      "loss": 0.4454,
+      "step": 457
+    },
+    {
+      "epoch": 0.7357429718875502,
+      "grad_norm": 0.26312986415658596,
+      "learning_rate": 4.19594997022037e-05,
+      "loss": 0.4503,
+      "step": 458
+    },
+    {
+      "epoch": 0.7373493975903614,
+      "grad_norm": 0.22359659568823306,
+      "learning_rate": 4.1929720071471116e-05,
+      "loss": 0.4464,
+      "step": 459
+    },
+    {
+      "epoch": 0.7389558232931727,
+      "grad_norm": 0.2771729748099005,
+      "learning_rate": 4.189994044073854e-05,
+      "loss": 0.4714,
+      "step": 460
+    },
+    {
+      "epoch": 0.7405622489959839,
+      "grad_norm": 0.21993164913808572,
+      "learning_rate": 4.187016081000596e-05,
+      "loss": 0.4325,
+      "step": 461
+    },
+    {
+      "epoch": 0.7421686746987952,
+      "grad_norm": 0.2889751049620364,
+      "learning_rate": 4.184038117927338e-05,
+      "loss": 0.4449,
+      "step": 462
+    },
+    {
+      "epoch": 0.7437751004016064,
+      "grad_norm": 0.2949855484027405,
+      "learning_rate": 4.18106015485408e-05,
+      "loss": 0.4262,
+      "step": 463
+    },
+    {
+      "epoch": 0.7453815261044177,
+      "grad_norm": 0.26377183894449047,
+      "learning_rate": 4.1780821917808224e-05,
+      "loss": 0.4545,
+      "step": 464
+    },
+    {
+      "epoch": 0.7469879518072289,
+      "grad_norm": 0.30726865720505625,
+      "learning_rate": 4.175104228707564e-05,
+      "loss": 0.4382,
+      "step": 465
+    },
+    {
+      "epoch": 0.7485943775100402,
+      "grad_norm": 0.23122382644642359,
+      "learning_rate": 4.172126265634306e-05,
+      "loss": 0.4406,
+      "step": 466
+    },
+    {
+      "epoch": 0.7502008032128514,
+      "grad_norm": 0.276949750828653,
+      "learning_rate": 4.169148302561048e-05,
+      "loss": 0.4418,
+      "step": 467
+    },
+    {
+      "epoch": 0.7518072289156627,
+      "grad_norm": 0.27713071319003524,
+      "learning_rate": 4.1661703394877906e-05,
+      "loss": 0.4537,
+      "step": 468
+    },
+    {
+      "epoch": 0.7534136546184739,
+      "grad_norm": 0.2526256098750081,
+      "learning_rate": 4.1631923764145325e-05,
+      "loss": 0.4228,
+      "step": 469
+    },
+    {
+      "epoch": 0.7550200803212851,
+      "grad_norm": 0.2638363919651469,
+      "learning_rate": 4.1602144133412744e-05,
+      "loss": 0.4285,
+      "step": 470
+    },
+    {
+      "epoch": 0.7566265060240964,
+      "grad_norm": 0.258408639046892,
+      "learning_rate": 4.157236450268017e-05,
+      "loss": 0.4783,
+      "step": 471
+    },
+    {
+      "epoch": 0.7582329317269076,
+      "grad_norm": 0.27786588034932264,
+      "learning_rate": 4.154258487194759e-05,
+      "loss": 0.4233,
+      "step": 472
+    },
+    {
+      "epoch": 0.7598393574297189,
+      "grad_norm": 0.2755128011878773,
+      "learning_rate": 4.1512805241215014e-05,
+      "loss": 0.4381,
+      "step": 473
+    },
+    {
+      "epoch": 0.7614457831325301,
+      "grad_norm": 0.2413178291750358,
+      "learning_rate": 4.148302561048243e-05,
+      "loss": 0.4426,
+      "step": 474
+    },
+    {
+      "epoch": 0.7630522088353414,
+      "grad_norm": 0.2625618384300778,
+      "learning_rate": 4.145324597974985e-05,
+      "loss": 0.4254,
+      "step": 475
+    },
+    {
+      "epoch": 0.7646586345381526,
+      "grad_norm": 0.24765395667571197,
+      "learning_rate": 4.142346634901728e-05,
+      "loss": 0.4085,
+      "step": 476
+    },
+    {
+      "epoch": 0.7662650602409639,
+      "grad_norm": 0.24032530369165062,
+      "learning_rate": 4.1393686718284696e-05,
+      "loss": 0.4362,
+      "step": 477
+    },
+    {
+      "epoch": 0.7678714859437751,
+      "grad_norm": 0.29944650953941004,
+      "learning_rate": 4.1363907087552115e-05,
+      "loss": 0.4418,
+      "step": 478
+    },
+    {
+      "epoch": 0.7694779116465863,
+      "grad_norm": 0.23902526885476016,
+      "learning_rate": 4.133412745681954e-05,
+      "loss": 0.4428,
+      "step": 479
+    },
+    {
+      "epoch": 0.7710843373493976,
+      "grad_norm": 0.2765681820302345,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 0.4303,
+      "step": 480
+    },
+    {
+      "epoch": 0.7726907630522089,
+      "grad_norm": 0.2842130912447409,
+      "learning_rate": 4.127456819535438e-05,
+      "loss": 0.4475,
+      "step": 481
+    },
+    {
+      "epoch": 0.7742971887550201,
+      "grad_norm": 0.2438936958185551,
+      "learning_rate": 4.12447885646218e-05,
+      "loss": 0.4347,
+      "step": 482
+    },
+    {
+      "epoch": 0.7759036144578313,
+      "grad_norm": 0.3168618858927945,
+      "learning_rate": 4.121500893388922e-05,
+      "loss": 0.4269,
+      "step": 483
+    },
+    {
+      "epoch": 0.7775100401606426,
+      "grad_norm": 0.2758789903313452,
+      "learning_rate": 4.118522930315664e-05,
+      "loss": 0.4459,
+      "step": 484
+    },
+    {
+      "epoch": 0.7791164658634538,
+      "grad_norm": 0.2493181686533524,
+      "learning_rate": 4.115544967242406e-05,
+      "loss": 0.4148,
+      "step": 485
+    },
+    {
+      "epoch": 0.7807228915662651,
+      "grad_norm": 0.308879700403648,
+      "learning_rate": 4.1125670041691486e-05,
+      "loss": 0.4234,
+      "step": 486
+    },
+    {
+      "epoch": 0.7823293172690763,
+      "grad_norm": 0.25111177333337475,
+      "learning_rate": 4.1095890410958905e-05,
+      "loss": 0.4326,
+      "step": 487
+    },
+    {
+      "epoch": 0.7839357429718875,
+      "grad_norm": 0.27869616318648066,
+      "learning_rate": 4.1066110780226324e-05,
+      "loss": 0.4636,
+      "step": 488
+    },
+    {
+      "epoch": 0.7855421686746988,
+      "grad_norm": 0.26843185289016686,
+      "learning_rate": 4.103633114949375e-05,
+      "loss": 0.4489,
+      "step": 489
+    },
+    {
+      "epoch": 0.7871485943775101,
+      "grad_norm": 0.2515875459811497,
+      "learning_rate": 4.100655151876117e-05,
+      "loss": 0.4365,
+      "step": 490
+    },
+    {
+      "epoch": 0.7887550200803213,
+      "grad_norm": 0.3043990948350012,
+      "learning_rate": 4.0976771888028594e-05,
+      "loss": 0.4461,
+      "step": 491
+    },
+    {
+      "epoch": 0.7903614457831325,
+      "grad_norm": 0.2574069687114595,
+      "learning_rate": 4.094699225729601e-05,
+      "loss": 0.457,
+      "step": 492
+    },
+    {
+      "epoch": 0.7919678714859437,
+      "grad_norm": 0.27116949616970154,
+      "learning_rate": 4.091721262656343e-05,
+      "loss": 0.4409,
+      "step": 493
+    },
+    {
+      "epoch": 0.793574297188755,
+      "grad_norm": 0.25968831554328287,
+      "learning_rate": 4.088743299583086e-05,
+      "loss": 0.4468,
+      "step": 494
+    },
+    {
+      "epoch": 0.7951807228915663,
+      "grad_norm": 0.26335068526437494,
+      "learning_rate": 4.0857653365098276e-05,
+      "loss": 0.4327,
+      "step": 495
+    },
+    {
+      "epoch": 0.7967871485943775,
+      "grad_norm": 0.24080104395558735,
+      "learning_rate": 4.0827873734365695e-05,
+      "loss": 0.4634,
+      "step": 496
+    },
+    {
+      "epoch": 0.7983935742971887,
+      "grad_norm": 0.26294966531248926,
+      "learning_rate": 4.0798094103633114e-05,
+      "loss": 0.4391,
+      "step": 497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25051730546706547,
+      "learning_rate": 4.076831447290054e-05,
+      "loss": 0.4325,
+      "step": 498
+    },
+    {
+      "epoch": 0.8016064257028113,
+      "grad_norm": 0.2427733402448281,
+      "learning_rate": 4.073853484216796e-05,
+      "loss": 0.4374,
+      "step": 499
+    },
+    {
+      "epoch": 0.8032128514056225,
+      "grad_norm": 0.24425221729505334,
+      "learning_rate": 4.070875521143538e-05,
+      "loss": 0.4473,
+      "step": 500
+    },
+    {
+      "epoch": 0.8048192771084337,
+      "grad_norm": 0.2683791345440276,
+      "learning_rate": 4.0678975580702796e-05,
+      "loss": 0.4337,
+      "step": 501
+    },
+    {
+      "epoch": 0.8064257028112449,
+      "grad_norm": 0.24711519506263632,
+      "learning_rate": 4.064919594997022e-05,
+      "loss": 0.4282,
+      "step": 502
+    },
+    {
+      "epoch": 0.8080321285140563,
+      "grad_norm": 0.28860665208927944,
+      "learning_rate": 4.061941631923764e-05,
+      "loss": 0.4279,
+      "step": 503
+    },
+    {
+      "epoch": 0.8096385542168675,
+      "grad_norm": 0.26118713217431333,
+      "learning_rate": 4.0589636688505067e-05,
+      "loss": 0.4426,
+      "step": 504
+    },
+    {
+      "epoch": 0.8112449799196787,
+      "grad_norm": 0.2730270735024781,
+      "learning_rate": 4.0559857057772485e-05,
+      "loss": 0.4335,
+      "step": 505
+    },
+    {
+      "epoch": 0.8128514056224899,
+      "grad_norm": 0.26209320634654937,
+      "learning_rate": 4.053007742703991e-05,
+      "loss": 0.4197,
+      "step": 506
+    },
+    {
+      "epoch": 0.8144578313253013,
+      "grad_norm": 0.2805274090853642,
+      "learning_rate": 4.050029779630733e-05,
+      "loss": 0.4378,
+      "step": 507
+    },
+    {
+      "epoch": 0.8160642570281125,
+      "grad_norm": 0.2667961879265187,
+      "learning_rate": 4.047051816557475e-05,
+      "loss": 0.424,
+      "step": 508
+    },
+    {
+      "epoch": 0.8176706827309237,
+      "grad_norm": 0.25928083584015393,
+      "learning_rate": 4.044073853484217e-05,
+      "loss": 0.4422,
+      "step": 509
+    },
+    {
+      "epoch": 0.8192771084337349,
+      "grad_norm": 0.2901803923450964,
+      "learning_rate": 4.041095890410959e-05,
+      "loss": 0.4414,
+      "step": 510
+    },
+    {
+      "epoch": 0.8208835341365461,
+      "grad_norm": 0.3193674709588745,
+      "learning_rate": 4.038117927337701e-05,
+      "loss": 0.4534,
+      "step": 511
+    },
+    {
+      "epoch": 0.8224899598393575,
+      "grad_norm": 0.24227403826420443,
+      "learning_rate": 4.035139964264443e-05,
+      "loss": 0.4587,
+      "step": 512
+    },
+    {
+      "epoch": 0.8240963855421687,
+      "grad_norm": 0.2964629851317735,
+      "learning_rate": 4.0321620011911857e-05,
+      "loss": 0.4323,
+      "step": 513
+    },
+    {
+      "epoch": 0.8257028112449799,
+      "grad_norm": 0.23139253151927322,
+      "learning_rate": 4.0291840381179275e-05,
+      "loss": 0.4428,
+      "step": 514
+    },
+    {
+      "epoch": 0.8273092369477911,
+      "grad_norm": 0.2766551845263243,
+      "learning_rate": 4.0262060750446694e-05,
+      "loss": 0.4396,
+      "step": 515
+    },
+    {
+      "epoch": 0.8289156626506025,
+      "grad_norm": 0.2558099712232151,
+      "learning_rate": 4.023228111971411e-05,
+      "loss": 0.4531,
+      "step": 516
+    },
+    {
+      "epoch": 0.8305220883534137,
+      "grad_norm": 0.248132214894419,
+      "learning_rate": 4.020250148898154e-05,
+      "loss": 0.4343,
+      "step": 517
+    },
+    {
+      "epoch": 0.8321285140562249,
+      "grad_norm": 0.218481604590402,
+      "learning_rate": 4.017272185824896e-05,
+      "loss": 0.4422,
+      "step": 518
+    },
+    {
+      "epoch": 0.8337349397590361,
+      "grad_norm": 0.24735349079682337,
+      "learning_rate": 4.0142942227516376e-05,
+      "loss": 0.448,
+      "step": 519
+    },
+    {
+      "epoch": 0.8353413654618473,
+      "grad_norm": 0.24763740642710605,
+      "learning_rate": 4.01131625967838e-05,
+      "loss": 0.4681,
+      "step": 520
+    },
+    {
+      "epoch": 0.8369477911646587,
+      "grad_norm": 0.2318003082200106,
+      "learning_rate": 4.008338296605123e-05,
+      "loss": 0.4266,
+      "step": 521
+    },
+    {
+      "epoch": 0.8385542168674699,
+      "grad_norm": 0.2938561150927858,
+      "learning_rate": 4.005360333531865e-05,
+      "loss": 0.4539,
+      "step": 522
+    },
+    {
+      "epoch": 0.8401606425702811,
+      "grad_norm": 0.21988408888482563,
+      "learning_rate": 4.0023823704586065e-05,
+      "loss": 0.4212,
+      "step": 523
+    },
+    {
+      "epoch": 0.8417670682730923,
+      "grad_norm": 0.26776989445692106,
+      "learning_rate": 3.9994044073853484e-05,
+      "loss": 0.419,
+      "step": 524
+    },
+    {
+      "epoch": 0.8433734939759037,
+      "grad_norm": 0.27196304375993613,
+      "learning_rate": 3.996426444312091e-05,
+      "loss": 0.4393,
+      "step": 525
+    },
+    {
+      "epoch": 0.8449799196787149,
+      "grad_norm": 0.23459828989262607,
+      "learning_rate": 3.993448481238833e-05,
+      "loss": 0.4506,
+      "step": 526
+    },
+    {
+      "epoch": 0.8465863453815261,
+      "grad_norm": 0.2867482982512224,
+      "learning_rate": 3.990470518165575e-05,
+      "loss": 0.4156,
+      "step": 527
+    },
+    {
+      "epoch": 0.8481927710843373,
+      "grad_norm": 0.2379855499866802,
+      "learning_rate": 3.987492555092317e-05,
+      "loss": 0.445,
+      "step": 528
+    },
+    {
+      "epoch": 0.8497991967871485,
+      "grad_norm": 0.26332013956044786,
+      "learning_rate": 3.984514592019059e-05,
+      "loss": 0.4326,
+      "step": 529
+    },
+    {
+      "epoch": 0.8514056224899599,
+      "grad_norm": 0.26813329833183946,
+      "learning_rate": 3.981536628945801e-05,
+      "loss": 0.43,
+      "step": 530
+    },
+    {
+      "epoch": 0.8530120481927711,
+      "grad_norm": 0.2389778853354646,
+      "learning_rate": 3.978558665872543e-05,
+      "loss": 0.422,
+      "step": 531
+    },
+    {
+      "epoch": 0.8546184738955823,
+      "grad_norm": 0.25166138672793514,
+      "learning_rate": 3.9755807027992856e-05,
+      "loss": 0.4398,
+      "step": 532
+    },
+    {
+      "epoch": 0.8562248995983935,
+      "grad_norm": 0.30458405837769154,
+      "learning_rate": 3.9726027397260274e-05,
+      "loss": 0.4363,
+      "step": 533
+    },
+    {
+      "epoch": 0.8578313253012049,
+      "grad_norm": 0.2561268269052372,
+      "learning_rate": 3.969624776652769e-05,
+      "loss": 0.4318,
+      "step": 534
+    },
+    {
+      "epoch": 0.8594377510040161,
+      "grad_norm": 0.31382863625958424,
+      "learning_rate": 3.966646813579512e-05,
+      "loss": 0.4519,
+      "step": 535
+    },
+    {
+      "epoch": 0.8610441767068273,
+      "grad_norm": 0.27399140852336135,
+      "learning_rate": 3.9636688505062545e-05,
+      "loss": 0.4493,
+      "step": 536
+    },
+    {
+      "epoch": 0.8626506024096385,
+      "grad_norm": 0.2708572339865821,
+      "learning_rate": 3.960690887432996e-05,
+      "loss": 0.4157,
+      "step": 537
+    },
+    {
+      "epoch": 0.8642570281124498,
+      "grad_norm": 0.3013531430351654,
+      "learning_rate": 3.957712924359738e-05,
+      "loss": 0.428,
+      "step": 538
+    },
+    {
+      "epoch": 0.8658634538152611,
+      "grad_norm": 0.2154447354851553,
+      "learning_rate": 3.95473496128648e-05,
+      "loss": 0.4423,
+      "step": 539
+    },
+    {
+      "epoch": 0.8674698795180723,
+      "grad_norm": 0.27209553350557136,
+      "learning_rate": 3.951756998213223e-05,
+      "loss": 0.4493,
+      "step": 540
+    },
+    {
+      "epoch": 0.8690763052208835,
+      "grad_norm": 0.25010423326607467,
+      "learning_rate": 3.9487790351399646e-05,
+      "loss": 0.4415,
+      "step": 541
+    },
+    {
+      "epoch": 0.8706827309236947,
+      "grad_norm": 0.3204511427575982,
+      "learning_rate": 3.9458010720667064e-05,
+      "loss": 0.4375,
+      "step": 542
+    },
+    {
+      "epoch": 0.8722891566265061,
+      "grad_norm": 0.24205733838152174,
+      "learning_rate": 3.942823108993448e-05,
+      "loss": 0.4444,
+      "step": 543
+    },
+    {
+      "epoch": 0.8738955823293173,
+      "grad_norm": 0.2294213191593051,
+      "learning_rate": 3.939845145920191e-05,
+      "loss": 0.4267,
+      "step": 544
+    },
+    {
+      "epoch": 0.8755020080321285,
+      "grad_norm": 0.2257426838434749,
+      "learning_rate": 3.936867182846933e-05,
+      "loss": 0.4221,
+      "step": 545
+    },
+    {
+      "epoch": 0.8771084337349397,
+      "grad_norm": 0.2706901866438114,
+      "learning_rate": 3.933889219773675e-05,
+      "loss": 0.4269,
+      "step": 546
+    },
+    {
+      "epoch": 0.878714859437751,
+      "grad_norm": 0.2225358386339718,
+      "learning_rate": 3.930911256700417e-05,
+      "loss": 0.4252,
+      "step": 547
+    },
+    {
+      "epoch": 0.8803212851405623,
+      "grad_norm": 0.237633405654431,
+      "learning_rate": 3.927933293627159e-05,
+      "loss": 0.4472,
+      "step": 548
+    },
+    {
+      "epoch": 0.8819277108433735,
+      "grad_norm": 0.25486484276588334,
+      "learning_rate": 3.924955330553901e-05,
+      "loss": 0.4509,
+      "step": 549
+    },
+    {
+      "epoch": 0.8835341365461847,
+      "grad_norm": 0.2959254315333727,
+      "learning_rate": 3.9219773674806436e-05,
+      "loss": 0.4357,
+      "step": 550
+    },
+    {
+      "epoch": 0.885140562248996,
+      "grad_norm": 0.2517468143857465,
+      "learning_rate": 3.9189994044073855e-05,
+      "loss": 0.4346,
+      "step": 551
+    },
+    {
+      "epoch": 0.8867469879518072,
+      "grad_norm": 0.24459158559461538,
+      "learning_rate": 3.916021441334128e-05,
+      "loss": 0.4457,
+      "step": 552
+    },
+    {
+      "epoch": 0.8883534136546185,
+      "grad_norm": 0.29549737944364596,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.4382,
+      "step": 553
+    },
+    {
+      "epoch": 0.8899598393574297,
+      "grad_norm": 0.26996542381353816,
+      "learning_rate": 3.910065515187612e-05,
+      "loss": 0.459,
+      "step": 554
+    },
+    {
+      "epoch": 0.891566265060241,
+      "grad_norm": 0.23705682208176873,
+      "learning_rate": 3.9070875521143543e-05,
+      "loss": 0.4128,
+      "step": 555
+    },
+    {
+      "epoch": 0.8931726907630522,
+      "grad_norm": 0.31888298370501633,
+      "learning_rate": 3.904109589041096e-05,
+      "loss": 0.4352,
+      "step": 556
+    },
+    {
+      "epoch": 0.8947791164658635,
+      "grad_norm": 0.25562028810153453,
+      "learning_rate": 3.901131625967838e-05,
+      "loss": 0.4206,
+      "step": 557
+    },
+    {
+      "epoch": 0.8963855421686747,
+      "grad_norm": 0.2990297387911476,
+      "learning_rate": 3.89815366289458e-05,
+      "loss": 0.4325,
+      "step": 558
+    },
+    {
+      "epoch": 0.8979919678714859,
+      "grad_norm": 0.28378737269849236,
+      "learning_rate": 3.8951756998213226e-05,
+      "loss": 0.4305,
+      "step": 559
+    },
+    {
+      "epoch": 0.8995983935742972,
+      "grad_norm": 0.24043681590576332,
+      "learning_rate": 3.8921977367480645e-05,
+      "loss": 0.4289,
+      "step": 560
+    },
+    {
+      "epoch": 0.9012048192771084,
+      "grad_norm": 0.27719811894493096,
+      "learning_rate": 3.8892197736748063e-05,
+      "loss": 0.427,
+      "step": 561
+    },
+    {
+      "epoch": 0.9028112449799197,
+      "grad_norm": 0.2419776006882972,
+      "learning_rate": 3.886241810601549e-05,
+      "loss": 0.4124,
+      "step": 562
+    },
+    {
+      "epoch": 0.9044176706827309,
+      "grad_norm": 0.24970037471349055,
+      "learning_rate": 3.883263847528291e-05,
+      "loss": 0.4319,
+      "step": 563
+    },
+    {
+      "epoch": 0.9060240963855422,
+      "grad_norm": 0.2906891588270049,
+      "learning_rate": 3.880285884455033e-05,
+      "loss": 0.456,
+      "step": 564
+    },
+    {
+      "epoch": 0.9076305220883534,
+      "grad_norm": 0.23010103423549969,
+      "learning_rate": 3.8773079213817746e-05,
+      "loss": 0.4387,
+      "step": 565
+    },
+    {
+      "epoch": 0.9092369477911647,
+      "grad_norm": 0.25700570128801525,
+      "learning_rate": 3.874329958308517e-05,
+      "loss": 0.4476,
+      "step": 566
+    },
+    {
+      "epoch": 0.9108433734939759,
+      "grad_norm": 0.24990675488537026,
+      "learning_rate": 3.87135199523526e-05,
+      "loss": 0.4552,
+      "step": 567
+    },
+    {
+      "epoch": 0.9124497991967871,
+      "grad_norm": 0.2307282093649637,
+      "learning_rate": 3.8683740321620016e-05,
+      "loss": 0.4311,
+      "step": 568
+    },
+    {
+      "epoch": 0.9140562248995984,
+      "grad_norm": 0.27738696314612776,
+      "learning_rate": 3.8653960690887435e-05,
+      "loss": 0.4444,
+      "step": 569
+    },
+    {
+      "epoch": 0.9156626506024096,
+      "grad_norm": 0.2552012268400254,
+      "learning_rate": 3.862418106015486e-05,
+      "loss": 0.4373,
+      "step": 570
+    },
+    {
+      "epoch": 0.9172690763052209,
+      "grad_norm": 0.24376104605564367,
+      "learning_rate": 3.859440142942228e-05,
+      "loss": 0.4437,
+      "step": 571
+    },
+    {
+      "epoch": 0.9188755020080321,
+      "grad_norm": 0.24493252905116827,
+      "learning_rate": 3.85646217986897e-05,
+      "loss": 0.4288,
+      "step": 572
+    },
+    {
+      "epoch": 0.9204819277108434,
+      "grad_norm": 0.26709579469582134,
+      "learning_rate": 3.853484216795712e-05,
+      "loss": 0.4506,
+      "step": 573
+    },
+    {
+      "epoch": 0.9220883534136546,
+      "grad_norm": 0.22939166264690128,
+      "learning_rate": 3.850506253722454e-05,
+      "loss": 0.4201,
+      "step": 574
+    },
+    {
+      "epoch": 0.9236947791164659,
+      "grad_norm": 0.2760775596448828,
+      "learning_rate": 3.847528290649196e-05,
+      "loss": 0.4302,
+      "step": 575
+    },
+    {
+      "epoch": 0.9253012048192771,
+      "grad_norm": 0.2341167029172588,
+      "learning_rate": 3.844550327575938e-05,
+      "loss": 0.4206,
+      "step": 576
+    },
+    {
+      "epoch": 0.9269076305220884,
+      "grad_norm": 0.28845596603073687,
+      "learning_rate": 3.84157236450268e-05,
+      "loss": 0.4303,
+      "step": 577
+    },
+    {
+      "epoch": 0.9285140562248996,
+      "grad_norm": 0.26016271659068463,
+      "learning_rate": 3.8385944014294225e-05,
+      "loss": 0.4277,
+      "step": 578
+    },
+    {
+      "epoch": 0.9301204819277108,
+      "grad_norm": 0.2573057822086778,
+      "learning_rate": 3.8356164383561644e-05,
+      "loss": 0.4343,
+      "step": 579
+    },
+    {
+      "epoch": 0.9317269076305221,
+      "grad_norm": 0.2600638208053094,
+      "learning_rate": 3.832638475282906e-05,
+      "loss": 0.4426,
+      "step": 580
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 0.23909074619126763,
+      "learning_rate": 3.829660512209649e-05,
+      "loss": 0.438,
+      "step": 581
+    },
+    {
+      "epoch": 0.9349397590361446,
+      "grad_norm": 0.2779421249279328,
+      "learning_rate": 3.8266825491363914e-05,
+      "loss": 0.4413,
+      "step": 582
+    },
+    {
+      "epoch": 0.9365461847389558,
+      "grad_norm": 0.23613930700473348,
+      "learning_rate": 3.823704586063133e-05,
+      "loss": 0.4524,
+      "step": 583
+    },
+    {
+      "epoch": 0.9381526104417671,
+      "grad_norm": 0.24231621353205904,
+      "learning_rate": 3.820726622989875e-05,
+      "loss": 0.4172,
+      "step": 584
+    },
+    {
+      "epoch": 0.9397590361445783,
+      "grad_norm": 0.27548795148960975,
+      "learning_rate": 3.817748659916617e-05,
+      "loss": 0.4234,
+      "step": 585
+    },
+    {
+      "epoch": 0.9413654618473896,
+      "grad_norm": 0.25527044160278795,
+      "learning_rate": 3.8147706968433596e-05,
+      "loss": 0.4382,
+      "step": 586
+    },
+    {
+      "epoch": 0.9429718875502008,
+      "grad_norm": 0.2349394226856958,
+      "learning_rate": 3.8117927337701015e-05,
+      "loss": 0.4349,
+      "step": 587
+    },
+    {
+      "epoch": 0.944578313253012,
+      "grad_norm": 0.22907020884629825,
+      "learning_rate": 3.8088147706968434e-05,
+      "loss": 0.4228,
+      "step": 588
+    },
+    {
+      "epoch": 0.9461847389558233,
+      "grad_norm": 0.28681246806321475,
+      "learning_rate": 3.805836807623586e-05,
+      "loss": 0.4303,
+      "step": 589
+    },
+    {
+      "epoch": 0.9477911646586346,
+      "grad_norm": 0.23123430367180658,
+      "learning_rate": 3.802858844550328e-05,
+      "loss": 0.451,
+      "step": 590
+    },
+    {
+      "epoch": 0.9493975903614458,
+      "grad_norm": 0.25837739673979626,
+      "learning_rate": 3.79988088147707e-05,
+      "loss": 0.4246,
+      "step": 591
+    },
+    {
+      "epoch": 0.951004016064257,
+      "grad_norm": 0.26932761173674435,
+      "learning_rate": 3.7969029184038116e-05,
+      "loss": 0.4302,
+      "step": 592
+    },
+    {
+      "epoch": 0.9526104417670683,
+      "grad_norm": 0.2689813929288827,
+      "learning_rate": 3.793924955330554e-05,
+      "loss": 0.4196,
+      "step": 593
+    },
+    {
+      "epoch": 0.9542168674698795,
+      "grad_norm": 0.24482217188234856,
+      "learning_rate": 3.790946992257296e-05,
+      "loss": 0.4383,
+      "step": 594
+    },
+    {
+      "epoch": 0.9558232931726908,
+      "grad_norm": 0.29914093028504873,
+      "learning_rate": 3.787969029184038e-05,
+      "loss": 0.4522,
+      "step": 595
+    },
+    {
+      "epoch": 0.957429718875502,
+      "grad_norm": 0.2359952335036009,
+      "learning_rate": 3.78499106611078e-05,
+      "loss": 0.4465,
+      "step": 596
+    },
+    {
+      "epoch": 0.9590361445783132,
+      "grad_norm": 0.2532581617564896,
+      "learning_rate": 3.782013103037523e-05,
+      "loss": 0.4335,
+      "step": 597
+    },
+    {
+      "epoch": 0.9606425702811245,
+      "grad_norm": 0.2514263045025721,
+      "learning_rate": 3.779035139964265e-05,
+      "loss": 0.4325,
+      "step": 598
+    },
+    {
+      "epoch": 0.9622489959839358,
+      "grad_norm": 0.23570835352148856,
+      "learning_rate": 3.776057176891007e-05,
+      "loss": 0.4314,
+      "step": 599
+    },
+    {
+      "epoch": 0.963855421686747,
+      "grad_norm": 0.23976696550069423,
+      "learning_rate": 3.773079213817749e-05,
+      "loss": 0.4427,
+      "step": 600
+    },
+    {
+      "epoch": 0.9654618473895582,
+      "grad_norm": 0.22055205479897635,
+      "learning_rate": 3.770101250744491e-05,
+      "loss": 0.4221,
+      "step": 601
+    },
+    {
+      "epoch": 0.9670682730923694,
+      "grad_norm": 0.24343117535869208,
+      "learning_rate": 3.767123287671233e-05,
+      "loss": 0.4421,
+      "step": 602
+    },
+    {
+      "epoch": 0.9686746987951808,
+      "grad_norm": 0.24390631127271806,
+      "learning_rate": 3.764145324597975e-05,
+      "loss": 0.4313,
+      "step": 603
+    },
+    {
+      "epoch": 0.970281124497992,
+      "grad_norm": 0.23440381138018293,
+      "learning_rate": 3.7611673615247176e-05,
+      "loss": 0.4335,
+      "step": 604
+    },
+    {
+      "epoch": 0.9718875502008032,
+      "grad_norm": 0.23653143643840432,
+      "learning_rate": 3.7581893984514595e-05,
+      "loss": 0.4501,
+      "step": 605
+    },
+    {
+      "epoch": 0.9734939759036144,
+      "grad_norm": 0.25580575112258935,
+      "learning_rate": 3.7552114353782014e-05,
+      "loss": 0.4318,
+      "step": 606
+    },
+    {
+      "epoch": 0.9751004016064257,
+      "grad_norm": 0.26355197800261765,
+      "learning_rate": 3.752233472304943e-05,
+      "loss": 0.4197,
+      "step": 607
+    },
+    {
+      "epoch": 0.976706827309237,
+      "grad_norm": 0.24291180092148457,
+      "learning_rate": 3.749255509231686e-05,
+      "loss": 0.4234,
+      "step": 608
+    },
+    {
+      "epoch": 0.9783132530120482,
+      "grad_norm": 0.2847500740226297,
+      "learning_rate": 3.746277546158428e-05,
+      "loss": 0.4443,
+      "step": 609
+    },
+    {
+      "epoch": 0.9799196787148594,
+      "grad_norm": 0.2610758749435212,
+      "learning_rate": 3.7432995830851696e-05,
+      "loss": 0.4289,
+      "step": 610
+    },
+    {
+      "epoch": 0.9815261044176706,
+      "grad_norm": 0.24968845981927007,
+      "learning_rate": 3.7403216200119115e-05,
+      "loss": 0.4276,
+      "step": 611
+    },
+    {
+      "epoch": 0.983132530120482,
+      "grad_norm": 0.3323334996216219,
+      "learning_rate": 3.737343656938655e-05,
+      "loss": 0.4574,
+      "step": 612
+    },
+    {
+      "epoch": 0.9847389558232932,
+      "grad_norm": 0.26279163924567633,
+      "learning_rate": 3.7343656938653966e-05,
+      "loss": 0.4336,
+      "step": 613
+    },
+    {
+      "epoch": 0.9863453815261044,
+      "grad_norm": 0.24348461047320755,
+      "learning_rate": 3.7313877307921385e-05,
+      "loss": 0.434,
+      "step": 614
+    },
+    {
+      "epoch": 0.9879518072289156,
+      "grad_norm": 0.24839350837891416,
+      "learning_rate": 3.7284097677188804e-05,
+      "loss": 0.4136,
+      "step": 615
+    },
+    {
+      "epoch": 0.989558232931727,
+      "grad_norm": 0.26281972473580856,
+      "learning_rate": 3.725431804645623e-05,
+      "loss": 0.4431,
+      "step": 616
+    },
+    {
+      "epoch": 0.9911646586345382,
+      "grad_norm": 0.2313225564914852,
+      "learning_rate": 3.722453841572365e-05,
+      "loss": 0.4277,
+      "step": 617
+    },
+    {
+      "epoch": 0.9927710843373494,
+      "grad_norm": 0.2615224862640514,
+      "learning_rate": 3.719475878499107e-05,
+      "loss": 0.4286,
+      "step": 618
+    },
+    {
+      "epoch": 0.9943775100401606,
+      "grad_norm": 0.2520240607539348,
+      "learning_rate": 3.7164979154258486e-05,
+      "loss": 0.4286,
+      "step": 619
+    },
+    {
+      "epoch": 0.9959839357429718,
+      "grad_norm": 0.25262720498315444,
+      "learning_rate": 3.713519952352591e-05,
+      "loss": 0.4311,
+      "step": 620
+    },
+    {
+      "epoch": 0.9975903614457832,
+      "grad_norm": 0.22632961503761864,
+      "learning_rate": 3.710541989279333e-05,
+      "loss": 0.4415,
+      "step": 621
+    },
+    {
+      "epoch": 0.9991967871485944,
+      "grad_norm": 0.2371805967007009,
+      "learning_rate": 3.707564026206075e-05,
+      "loss": 0.4015,
+      "step": 622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2371805967007009,
+      "learning_rate": 3.7045860631328175e-05,
+      "loss": 0.4294,
+      "step": 623
+    },
+    {
+      "epoch": 1.0016064257028112,
+      "grad_norm": 0.37932341441984047,
+      "learning_rate": 3.7016081000595594e-05,
+      "loss": 0.3495,
+      "step": 624
+    },
+    {
+      "epoch": 1.0032128514056224,
+      "grad_norm": 0.2759554804160266,
+      "learning_rate": 3.698630136986301e-05,
+      "loss": 0.3745,
+      "step": 625
+    },
+    {
+      "epoch": 1.0048192771084337,
+      "grad_norm": 0.2546676949628782,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 0.3681,
+      "step": 626
+    },
+    {
+      "epoch": 1.0064257028112449,
+      "grad_norm": 0.2935521485778336,
+      "learning_rate": 3.692674210839786e-05,
+      "loss": 0.3734,
+      "step": 627
+    },
+    {
+      "epoch": 1.0080321285140563,
+      "grad_norm": 0.29990889534005793,
+      "learning_rate": 3.689696247766528e-05,
+      "loss": 0.3801,
+      "step": 628
+    },
+    {
+      "epoch": 1.0096385542168675,
+      "grad_norm": 0.26748247562403804,
+      "learning_rate": 3.68671828469327e-05,
+      "loss": 0.3743,
+      "step": 629
+    },
+    {
+      "epoch": 1.0112449799196788,
+      "grad_norm": 0.26734512122421833,
+      "learning_rate": 3.683740321620012e-05,
+      "loss": 0.3478,
+      "step": 630
+    },
+    {
+      "epoch": 1.01285140562249,
+      "grad_norm": 0.2847423153327123,
+      "learning_rate": 3.6807623585467546e-05,
+      "loss": 0.3699,
+      "step": 631
+    },
+    {
+      "epoch": 1.0144578313253012,
+      "grad_norm": 0.2650366814150442,
+      "learning_rate": 3.6777843954734965e-05,
+      "loss": 0.3488,
+      "step": 632
+    },
+    {
+      "epoch": 1.0160642570281124,
+      "grad_norm": 0.2596723308109665,
+      "learning_rate": 3.6748064324002384e-05,
+      "loss": 0.3839,
+      "step": 633
+    },
+    {
+      "epoch": 1.0176706827309236,
+      "grad_norm": 0.2775391732767592,
+      "learning_rate": 3.67182846932698e-05,
+      "loss": 0.3634,
+      "step": 634
+    },
+    {
+      "epoch": 1.0192771084337349,
+      "grad_norm": 0.3039539665914044,
+      "learning_rate": 3.668850506253723e-05,
+      "loss": 0.3689,
+      "step": 635
+    },
+    {
+      "epoch": 1.020883534136546,
+      "grad_norm": 0.28485132467442625,
+      "learning_rate": 3.665872543180465e-05,
+      "loss": 0.3769,
+      "step": 636
+    },
+    {
+      "epoch": 1.0224899598393575,
+      "grad_norm": 0.2890348401840446,
+      "learning_rate": 3.6628945801072066e-05,
+      "loss": 0.3741,
+      "step": 637
+    },
+    {
+      "epoch": 1.0240963855421688,
+      "grad_norm": 0.2865555236023417,
+      "learning_rate": 3.659916617033949e-05,
+      "loss": 0.3675,
+      "step": 638
+    },
+    {
+      "epoch": 1.02570281124498,
+      "grad_norm": 0.2675353012405362,
+      "learning_rate": 3.656938653960691e-05,
+      "loss": 0.3748,
+      "step": 639
+    },
+    {
+      "epoch": 1.0273092369477912,
+      "grad_norm": 0.2596693456604531,
+      "learning_rate": 3.653960690887433e-05,
+      "loss": 0.3594,
+      "step": 640
+    },
+    {
+      "epoch": 1.0289156626506024,
+      "grad_norm": 0.270414602391874,
+      "learning_rate": 3.650982727814175e-05,
+      "loss": 0.3633,
+      "step": 641
+    },
+    {
+      "epoch": 1.0305220883534136,
+      "grad_norm": 0.22619014147747027,
+      "learning_rate": 3.6480047647409174e-05,
+      "loss": 0.3776,
+      "step": 642
+    },
+    {
+      "epoch": 1.0321285140562249,
+      "grad_norm": 0.29226927393529,
+      "learning_rate": 3.64502680166766e-05,
+      "loss": 0.3646,
+      "step": 643
+    },
+    {
+      "epoch": 1.033734939759036,
+      "grad_norm": 0.2395639919856441,
+      "learning_rate": 3.642048838594402e-05,
+      "loss": 0.3654,
+      "step": 644
+    },
+    {
+      "epoch": 1.0353413654618473,
+      "grad_norm": 0.25348584398956553,
+      "learning_rate": 3.639070875521144e-05,
+      "loss": 0.3924,
+      "step": 645
+    },
+    {
+      "epoch": 1.0369477911646587,
+      "grad_norm": 0.2672275246998648,
+      "learning_rate": 3.636092912447886e-05,
+      "loss": 0.3538,
+      "step": 646
+    },
+    {
+      "epoch": 1.03855421686747,
+      "grad_norm": 0.2346022042725061,
+      "learning_rate": 3.633114949374628e-05,
+      "loss": 0.367,
+      "step": 647
+    },
+    {
+      "epoch": 1.0401606425702812,
+      "grad_norm": 0.2525831094063313,
+      "learning_rate": 3.63013698630137e-05,
+      "loss": 0.3665,
+      "step": 648
+    },
+    {
+      "epoch": 1.0417670682730924,
+      "grad_norm": 0.23157749157994945,
+      "learning_rate": 3.627159023228112e-05,
+      "loss": 0.3746,
+      "step": 649
+    },
+    {
+      "epoch": 1.0433734939759036,
+      "grad_norm": 0.23185354633237557,
+      "learning_rate": 3.6241810601548545e-05,
+      "loss": 0.3868,
+      "step": 650
+    },
+    {
+      "epoch": 1.0449799196787148,
+      "grad_norm": 0.2432844999424583,
+      "learning_rate": 3.6212030970815964e-05,
+      "loss": 0.3719,
+      "step": 651
+    },
+    {
+      "epoch": 1.046586345381526,
+      "grad_norm": 0.2192600088467529,
+      "learning_rate": 3.618225134008338e-05,
+      "loss": 0.3644,
+      "step": 652
+    },
+    {
+      "epoch": 1.0481927710843373,
+      "grad_norm": 0.25117602242785403,
+      "learning_rate": 3.61524717093508e-05,
+      "loss": 0.3516,
+      "step": 653
+    },
+    {
+      "epoch": 1.0497991967871485,
+      "grad_norm": 0.24915531821305087,
+      "learning_rate": 3.612269207861823e-05,
+      "loss": 0.3514,
+      "step": 654
+    },
+    {
+      "epoch": 1.05140562248996,
+      "grad_norm": 0.22692662484237394,
+      "learning_rate": 3.6092912447885646e-05,
+      "loss": 0.3804,
+      "step": 655
+    },
+    {
+      "epoch": 1.0530120481927712,
+      "grad_norm": 0.25324103379881957,
+      "learning_rate": 3.6063132817153065e-05,
+      "loss": 0.3615,
+      "step": 656
+    },
+    {
+      "epoch": 1.0546184738955824,
+      "grad_norm": 0.2429896691080188,
+      "learning_rate": 3.603335318642049e-05,
+      "loss": 0.3781,
+      "step": 657
+    },
+    {
+      "epoch": 1.0562248995983936,
+      "grad_norm": 0.23497370537491616,
+      "learning_rate": 3.6003573555687916e-05,
+      "loss": 0.3857,
+      "step": 658
+    },
+    {
+      "epoch": 1.0578313253012048,
+      "grad_norm": 0.24625075829341786,
+      "learning_rate": 3.5973793924955335e-05,
+      "loss": 0.3752,
+      "step": 659
+    },
+    {
+      "epoch": 1.059437751004016,
+      "grad_norm": 0.25709099687928116,
+      "learning_rate": 3.5944014294222754e-05,
+      "loss": 0.3596,
+      "step": 660
+    },
+    {
+      "epoch": 1.0610441767068273,
+      "grad_norm": 0.23056836241774717,
+      "learning_rate": 3.591423466349017e-05,
+      "loss": 0.3483,
+      "step": 661
+    },
+    {
+      "epoch": 1.0626506024096385,
+      "grad_norm": 0.24580094444062736,
+      "learning_rate": 3.58844550327576e-05,
+      "loss": 0.3603,
+      "step": 662
+    },
+    {
+      "epoch": 1.0642570281124497,
+      "grad_norm": 0.27979904138127654,
+      "learning_rate": 3.585467540202502e-05,
+      "loss": 0.3821,
+      "step": 663
+    },
+    {
+      "epoch": 1.0658634538152612,
+      "grad_norm": 0.221287404622989,
+      "learning_rate": 3.5824895771292436e-05,
+      "loss": 0.3512,
+      "step": 664
+    },
+    {
+      "epoch": 1.0674698795180724,
+      "grad_norm": 0.26457549896181337,
+      "learning_rate": 3.579511614055986e-05,
+      "loss": 0.3723,
+      "step": 665
+    },
+    {
+      "epoch": 1.0690763052208836,
+      "grad_norm": 0.22452705568993894,
+      "learning_rate": 3.576533650982728e-05,
+      "loss": 0.3708,
+      "step": 666
+    },
+    {
+      "epoch": 1.0706827309236948,
+      "grad_norm": 0.2360253339758698,
+      "learning_rate": 3.57355568790947e-05,
+      "loss": 0.3539,
+      "step": 667
+    },
+    {
+      "epoch": 1.072289156626506,
+      "grad_norm": 0.23874769004827587,
+      "learning_rate": 3.570577724836212e-05,
+      "loss": 0.3567,
+      "step": 668
+    },
+    {
+      "epoch": 1.0738955823293173,
+      "grad_norm": 0.2477102370108865,
+      "learning_rate": 3.5675997617629544e-05,
+      "loss": 0.3655,
+      "step": 669
+    },
+    {
+      "epoch": 1.0755020080321285,
+      "grad_norm": 0.20742644891736414,
+      "learning_rate": 3.564621798689696e-05,
+      "loss": 0.3618,
+      "step": 670
+    },
+    {
+      "epoch": 1.0771084337349397,
+      "grad_norm": 0.25364731028659276,
+      "learning_rate": 3.561643835616438e-05,
+      "loss": 0.3633,
+      "step": 671
+    },
+    {
+      "epoch": 1.078714859437751,
+      "grad_norm": 0.21206121186535307,
+      "learning_rate": 3.55866587254318e-05,
+      "loss": 0.3829,
+      "step": 672
+    },
+    {
+      "epoch": 1.0803212851405624,
+      "grad_norm": 0.23809685517326412,
+      "learning_rate": 3.5556879094699226e-05,
+      "loss": 0.3771,
+      "step": 673
+    },
+    {
+      "epoch": 1.0819277108433736,
+      "grad_norm": 0.25148288245072126,
+      "learning_rate": 3.552709946396665e-05,
+      "loss": 0.3573,
+      "step": 674
+    },
+    {
+      "epoch": 1.0835341365461848,
+      "grad_norm": 0.24148870358153765,
+      "learning_rate": 3.549731983323407e-05,
+      "loss": 0.3677,
+      "step": 675
+    },
+    {
+      "epoch": 1.085140562248996,
+      "grad_norm": 0.23469672202622077,
+      "learning_rate": 3.546754020250149e-05,
+      "loss": 0.3621,
+      "step": 676
+    },
+    {
+      "epoch": 1.0867469879518072,
+      "grad_norm": 0.2184344550196849,
+      "learning_rate": 3.5437760571768915e-05,
+      "loss": 0.3489,
+      "step": 677
+    },
+    {
+      "epoch": 1.0883534136546185,
+      "grad_norm": 0.24328481848849765,
+      "learning_rate": 3.5407980941036334e-05,
+      "loss": 0.3684,
+      "step": 678
+    },
+    {
+      "epoch": 1.0899598393574297,
+      "grad_norm": 0.24479406299704576,
+      "learning_rate": 3.537820131030375e-05,
+      "loss": 0.3907,
+      "step": 679
+    },
+    {
+      "epoch": 1.091566265060241,
+      "grad_norm": 0.22146902705233215,
+      "learning_rate": 3.534842167957118e-05,
+      "loss": 0.3802,
+      "step": 680
+    },
+    {
+      "epoch": 1.0931726907630521,
+      "grad_norm": 0.22443432974493835,
+      "learning_rate": 3.53186420488386e-05,
+      "loss": 0.3693,
+      "step": 681
+    },
+    {
+      "epoch": 1.0947791164658636,
+      "grad_norm": 0.2142069300586743,
+      "learning_rate": 3.5288862418106016e-05,
+      "loss": 0.3597,
+      "step": 682
+    },
+    {
+      "epoch": 1.0963855421686748,
+      "grad_norm": 0.21276658389057349,
+      "learning_rate": 3.5259082787373435e-05,
+      "loss": 0.3791,
+      "step": 683
+    },
+    {
+      "epoch": 1.097991967871486,
+      "grad_norm": 0.22001139189062285,
+      "learning_rate": 3.522930315664086e-05,
+      "loss": 0.359,
+      "step": 684
+    },
+    {
+      "epoch": 1.0995983935742972,
+      "grad_norm": 0.27030583552920734,
+      "learning_rate": 3.519952352590828e-05,
+      "loss": 0.3639,
+      "step": 685
+    },
+    {
+      "epoch": 1.1012048192771084,
+      "grad_norm": 0.21645081863138144,
+      "learning_rate": 3.51697438951757e-05,
+      "loss": 0.3845,
+      "step": 686
+    },
+    {
+      "epoch": 1.1028112449799197,
+      "grad_norm": 0.23213039417581863,
+      "learning_rate": 3.513996426444312e-05,
+      "loss": 0.3695,
+      "step": 687
+    },
+    {
+      "epoch": 1.104417670682731,
+      "grad_norm": 0.2473426640084354,
+      "learning_rate": 3.511018463371054e-05,
+      "loss": 0.353,
+      "step": 688
+    },
+    {
+      "epoch": 1.106024096385542,
+      "grad_norm": 0.2389357656774047,
+      "learning_rate": 3.508040500297797e-05,
+      "loss": 0.3749,
+      "step": 689
+    },
+    {
+      "epoch": 1.1076305220883533,
+      "grad_norm": 0.2552532077149984,
+      "learning_rate": 3.505062537224539e-05,
+      "loss": 0.3653,
+      "step": 690
+    },
+    {
+      "epoch": 1.1092369477911648,
+      "grad_norm": 0.25188268389731727,
+      "learning_rate": 3.5020845741512806e-05,
+      "loss": 0.3715,
+      "step": 691
+    },
+    {
+      "epoch": 1.110843373493976,
+      "grad_norm": 0.25344476363181,
+      "learning_rate": 3.499106611078023e-05,
+      "loss": 0.3752,
+      "step": 692
+    },
+    {
+      "epoch": 1.1124497991967872,
+      "grad_norm": 0.22356543631446663,
+      "learning_rate": 3.496128648004765e-05,
+      "loss": 0.3608,
+      "step": 693
+    },
+    {
+      "epoch": 1.1140562248995984,
+      "grad_norm": 0.22701786012495878,
+      "learning_rate": 3.493150684931507e-05,
+      "loss": 0.3657,
+      "step": 694
+    },
+    {
+      "epoch": 1.1156626506024097,
+      "grad_norm": 0.25341083733534187,
+      "learning_rate": 3.490172721858249e-05,
+      "loss": 0.3767,
+      "step": 695
+    },
+    {
+      "epoch": 1.1172690763052209,
+      "grad_norm": 0.23595910429128436,
+      "learning_rate": 3.4871947587849914e-05,
+      "loss": 0.3487,
+      "step": 696
+    },
+    {
+      "epoch": 1.118875502008032,
+      "grad_norm": 0.23068513649779648,
+      "learning_rate": 3.484216795711733e-05,
+      "loss": 0.3622,
+      "step": 697
+    },
+    {
+      "epoch": 1.1204819277108433,
+      "grad_norm": 0.23256189957817827,
+      "learning_rate": 3.481238832638475e-05,
+      "loss": 0.3406,
+      "step": 698
+    },
+    {
+      "epoch": 1.1220883534136545,
+      "grad_norm": 0.25966731002559584,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.3446,
+      "step": 699
+    },
+    {
+      "epoch": 1.123694779116466,
+      "grad_norm": 0.20834092240579738,
+      "learning_rate": 3.4752829064919597e-05,
+      "loss": 0.3594,
+      "step": 700
+    },
+    {
+      "epoch": 1.1253012048192772,
+      "grad_norm": 0.2464905111090441,
+      "learning_rate": 3.4723049434187015e-05,
+      "loss": 0.3509,
+      "step": 701
+    },
+    {
+      "epoch": 1.1269076305220884,
+      "grad_norm": 0.25837388163460817,
+      "learning_rate": 3.4693269803454434e-05,
+      "loss": 0.3846,
+      "step": 702
+    },
+    {
+      "epoch": 1.1285140562248996,
+      "grad_norm": 0.24333941968302056,
+      "learning_rate": 3.466349017272186e-05,
+      "loss": 0.3896,
+      "step": 703
+    },
+    {
+      "epoch": 1.1301204819277109,
+      "grad_norm": 0.26919093911243963,
+      "learning_rate": 3.463371054198928e-05,
+      "loss": 0.3811,
+      "step": 704
+    },
+    {
+      "epoch": 1.131726907630522,
+      "grad_norm": 0.2791697185654261,
+      "learning_rate": 3.4603930911256704e-05,
+      "loss": 0.3628,
+      "step": 705
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 0.2904893441742124,
+      "learning_rate": 3.457415128052412e-05,
+      "loss": 0.3643,
+      "step": 706
+    },
+    {
+      "epoch": 1.1349397590361445,
+      "grad_norm": 0.25399716724167515,
+      "learning_rate": 3.454437164979155e-05,
+      "loss": 0.3606,
+      "step": 707
+    },
+    {
+      "epoch": 1.1365461847389557,
+      "grad_norm": 0.27569474381536324,
+      "learning_rate": 3.451459201905897e-05,
+      "loss": 0.3774,
+      "step": 708
+    },
+    {
+      "epoch": 1.1381526104417672,
+      "grad_norm": 0.29988319701980504,
+      "learning_rate": 3.4484812388326387e-05,
+      "loss": 0.3536,
+      "step": 709
+    },
+    {
+      "epoch": 1.1397590361445784,
+      "grad_norm": 0.23099364928570554,
+      "learning_rate": 3.4455032757593805e-05,
+      "loss": 0.3611,
+      "step": 710
+    },
+    {
+      "epoch": 1.1413654618473896,
+      "grad_norm": 0.2805499124103727,
+      "learning_rate": 3.442525312686123e-05,
+      "loss": 0.3572,
+      "step": 711
+    },
+    {
+      "epoch": 1.1429718875502008,
+      "grad_norm": 0.26449450906138705,
+      "learning_rate": 3.439547349612865e-05,
+      "loss": 0.3439,
+      "step": 712
+    },
+    {
+      "epoch": 1.144578313253012,
+      "grad_norm": 0.288733039004294,
+      "learning_rate": 3.436569386539607e-05,
+      "loss": 0.3709,
+      "step": 713
+    },
+    {
+      "epoch": 1.1461847389558233,
+      "grad_norm": 0.3036364491451544,
+      "learning_rate": 3.433591423466349e-05,
+      "loss": 0.367,
+      "step": 714
+    },
+    {
+      "epoch": 1.1477911646586345,
+      "grad_norm": 0.29675889950840706,
+      "learning_rate": 3.430613460393091e-05,
+      "loss": 0.3775,
+      "step": 715
+    },
+    {
+      "epoch": 1.1493975903614457,
+      "grad_norm": 0.3259907065820817,
+      "learning_rate": 3.427635497319833e-05,
+      "loss": 0.3543,
+      "step": 716
+    },
+    {
+      "epoch": 1.151004016064257,
+      "grad_norm": 0.28452983500030143,
+      "learning_rate": 3.424657534246575e-05,
+      "loss": 0.3758,
+      "step": 717
+    },
+    {
+      "epoch": 1.1526104417670684,
+      "grad_norm": 0.2965433380474179,
+      "learning_rate": 3.421679571173318e-05,
+      "loss": 0.3595,
+      "step": 718
+    },
+    {
+      "epoch": 1.1542168674698796,
+      "grad_norm": 0.32133186307611517,
+      "learning_rate": 3.4187016081000595e-05,
+      "loss": 0.3589,
+      "step": 719
+    },
+    {
+      "epoch": 1.1558232931726908,
+      "grad_norm": 0.24143132234827158,
+      "learning_rate": 3.415723645026802e-05,
+      "loss": 0.3657,
+      "step": 720
+    },
+    {
+      "epoch": 1.157429718875502,
+      "grad_norm": 0.28034999472823113,
+      "learning_rate": 3.412745681953544e-05,
+      "loss": 0.3732,
+      "step": 721
+    },
+    {
+      "epoch": 1.1590361445783133,
+      "grad_norm": 0.3567620337310111,
+      "learning_rate": 3.4097677188802866e-05,
+      "loss": 0.3765,
+      "step": 722
+    },
+    {
+      "epoch": 1.1606425702811245,
+      "grad_norm": 0.2734839295396049,
+      "learning_rate": 3.4067897558070284e-05,
+      "loss": 0.349,
+      "step": 723
+    },
+    {
+      "epoch": 1.1622489959839357,
+      "grad_norm": 0.302098968412399,
+      "learning_rate": 3.40381179273377e-05,
+      "loss": 0.3703,
+      "step": 724
+    },
+    {
+      "epoch": 1.163855421686747,
+      "grad_norm": 0.2770751634090622,
+      "learning_rate": 3.400833829660512e-05,
+      "loss": 0.3525,
+      "step": 725
+    },
+    {
+      "epoch": 1.1654618473895582,
+      "grad_norm": 0.2338401666046532,
+      "learning_rate": 3.397855866587255e-05,
+      "loss": 0.3652,
+      "step": 726
+    },
+    {
+      "epoch": 1.1670682730923696,
+      "grad_norm": 0.2881681866550499,
+      "learning_rate": 3.394877903513997e-05,
+      "loss": 0.3586,
+      "step": 727
+    },
+    {
+      "epoch": 1.1686746987951806,
+      "grad_norm": 0.22750968897316734,
+      "learning_rate": 3.3918999404407386e-05,
+      "loss": 0.3681,
+      "step": 728
+    },
+    {
+      "epoch": 1.170281124497992,
+      "grad_norm": 0.3040420516406043,
+      "learning_rate": 3.3889219773674804e-05,
+      "loss": 0.3672,
+      "step": 729
+    },
+    {
+      "epoch": 1.1718875502008033,
+      "grad_norm": 0.2330663790560336,
+      "learning_rate": 3.385944014294223e-05,
+      "loss": 0.358,
+      "step": 730
+    },
+    {
+      "epoch": 1.1734939759036145,
+      "grad_norm": 0.2636617763473861,
+      "learning_rate": 3.382966051220965e-05,
+      "loss": 0.3557,
+      "step": 731
+    },
+    {
+      "epoch": 1.1751004016064257,
+      "grad_norm": 0.2712954225906756,
+      "learning_rate": 3.379988088147707e-05,
+      "loss": 0.3658,
+      "step": 732
+    },
+    {
+      "epoch": 1.176706827309237,
+      "grad_norm": 0.25644409732677503,
+      "learning_rate": 3.377010125074449e-05,
+      "loss": 0.3772,
+      "step": 733
+    },
+    {
+      "epoch": 1.1783132530120481,
+      "grad_norm": 0.2736176276124936,
+      "learning_rate": 3.374032162001191e-05,
+      "loss": 0.3523,
+      "step": 734
+    },
+    {
+      "epoch": 1.1799196787148594,
+      "grad_norm": 0.24856888153925333,
+      "learning_rate": 3.371054198927934e-05,
+      "loss": 0.3644,
+      "step": 735
+    },
+    {
+      "epoch": 1.1815261044176706,
+      "grad_norm": 0.23373015060576086,
+      "learning_rate": 3.368076235854676e-05,
+      "loss": 0.3685,
+      "step": 736
+    },
+    {
+      "epoch": 1.1831325301204818,
+      "grad_norm": 0.23172878048190607,
+      "learning_rate": 3.3650982727814176e-05,
+      "loss": 0.3765,
+      "step": 737
+    },
+    {
+      "epoch": 1.1847389558232932,
+      "grad_norm": 0.25808246859075906,
+      "learning_rate": 3.36212030970816e-05,
+      "loss": 0.379,
+      "step": 738
+    },
+    {
+      "epoch": 1.1863453815261045,
+      "grad_norm": 0.2166593271905805,
+      "learning_rate": 3.359142346634902e-05,
+      "loss": 0.3593,
+      "step": 739
+    },
+    {
+      "epoch": 1.1879518072289157,
+      "grad_norm": 0.21395631375601631,
+      "learning_rate": 3.356164383561644e-05,
+      "loss": 0.3452,
+      "step": 740
+    },
+    {
+      "epoch": 1.189558232931727,
+      "grad_norm": 0.22667954531445253,
+      "learning_rate": 3.3531864204883865e-05,
+      "loss": 0.3602,
+      "step": 741
+    },
+    {
+      "epoch": 1.1911646586345381,
+      "grad_norm": 0.21873569788689487,
+      "learning_rate": 3.3502084574151283e-05,
+      "loss": 0.3334,
+      "step": 742
+    },
+    {
+      "epoch": 1.1927710843373494,
+      "grad_norm": 0.1920911060477098,
+      "learning_rate": 3.34723049434187e-05,
+      "loss": 0.3633,
+      "step": 743
+    },
+    {
+      "epoch": 1.1943775100401606,
+      "grad_norm": 0.2574898749410916,
+      "learning_rate": 3.344252531268612e-05,
+      "loss": 0.3503,
+      "step": 744
+    },
+    {
+      "epoch": 1.1959839357429718,
+      "grad_norm": 0.22614786106563914,
+      "learning_rate": 3.341274568195355e-05,
+      "loss": 0.3623,
+      "step": 745
+    },
+    {
+      "epoch": 1.197590361445783,
+      "grad_norm": 0.26822964876930067,
+      "learning_rate": 3.3382966051220966e-05,
+      "loss": 0.3735,
+      "step": 746
+    },
+    {
+      "epoch": 1.1991967871485945,
+      "grad_norm": 0.22224303726380826,
+      "learning_rate": 3.3353186420488385e-05,
+      "loss": 0.372,
+      "step": 747
+    },
+    {
+      "epoch": 1.2008032128514057,
+      "grad_norm": 0.22078850305949052,
+      "learning_rate": 3.33234067897558e-05,
+      "loss": 0.3676,
+      "step": 748
+    },
+    {
+      "epoch": 1.202409638554217,
+      "grad_norm": 0.26633556835778005,
+      "learning_rate": 3.329362715902323e-05,
+      "loss": 0.371,
+      "step": 749
+    },
+    {
+      "epoch": 1.2040160642570281,
+      "grad_norm": 0.2788503075716381,
+      "learning_rate": 3.326384752829065e-05,
+      "loss": 0.3653,
+      "step": 750
+    },
+    {
+      "epoch": 1.2056224899598393,
+      "grad_norm": 0.23849352579520705,
+      "learning_rate": 3.3234067897558073e-05,
+      "loss": 0.3663,
+      "step": 751
+    },
+    {
+      "epoch": 1.2072289156626506,
+      "grad_norm": 0.24013045056887264,
+      "learning_rate": 3.320428826682549e-05,
+      "loss": 0.353,
+      "step": 752
+    },
+    {
+      "epoch": 1.2088353413654618,
+      "grad_norm": 0.20855339337194004,
+      "learning_rate": 3.317450863609292e-05,
+      "loss": 0.3774,
+      "step": 753
+    },
+    {
+      "epoch": 1.210441767068273,
+      "grad_norm": 0.2424426225359671,
+      "learning_rate": 3.314472900536034e-05,
+      "loss": 0.3663,
+      "step": 754
+    },
+    {
+      "epoch": 1.2120481927710842,
+      "grad_norm": 0.21880338853845513,
+      "learning_rate": 3.3114949374627756e-05,
+      "loss": 0.3702,
+      "step": 755
+    },
+    {
+      "epoch": 1.2136546184738957,
+      "grad_norm": 0.2153739479920725,
+      "learning_rate": 3.308516974389518e-05,
+      "loss": 0.3497,
+      "step": 756
+    },
+    {
+      "epoch": 1.2152610441767069,
+      "grad_norm": 0.23494454996837524,
+      "learning_rate": 3.30553901131626e-05,
+      "loss": 0.3559,
+      "step": 757
+    },
+    {
+      "epoch": 1.216867469879518,
+      "grad_norm": 0.21465927488059922,
+      "learning_rate": 3.302561048243002e-05,
+      "loss": 0.3676,
+      "step": 758
+    },
+    {
+      "epoch": 1.2184738955823293,
+      "grad_norm": 0.29985885980431837,
+      "learning_rate": 3.299583085169744e-05,
+      "loss": 0.3597,
+      "step": 759
+    },
+    {
+      "epoch": 1.2200803212851405,
+      "grad_norm": 0.22961190904343967,
+      "learning_rate": 3.2966051220964864e-05,
+      "loss": 0.3739,
+      "step": 760
+    },
+    {
+      "epoch": 1.2216867469879518,
+      "grad_norm": 0.24411542482113738,
+      "learning_rate": 3.293627159023228e-05,
+      "loss": 0.3695,
+      "step": 761
+    },
+    {
+      "epoch": 1.223293172690763,
+      "grad_norm": 0.22666726531935338,
+      "learning_rate": 3.29064919594997e-05,
+      "loss": 0.3663,
+      "step": 762
+    },
+    {
+      "epoch": 1.2248995983935742,
+      "grad_norm": 0.24629498678132544,
+      "learning_rate": 3.287671232876712e-05,
+      "loss": 0.3586,
+      "step": 763
+    },
+    {
+      "epoch": 1.2265060240963854,
+      "grad_norm": 0.21406505000866682,
+      "learning_rate": 3.2846932698034546e-05,
+      "loss": 0.3776,
+      "step": 764
+    },
+    {
+      "epoch": 1.2281124497991969,
+      "grad_norm": 0.24598567814668207,
+      "learning_rate": 3.2817153067301965e-05,
+      "loss": 0.3755,
+      "step": 765
+    },
+    {
+      "epoch": 1.229718875502008,
+      "grad_norm": 0.22021354268684526,
+      "learning_rate": 3.278737343656939e-05,
+      "loss": 0.3956,
+      "step": 766
+    },
+    {
+      "epoch": 1.2313253012048193,
+      "grad_norm": 0.2179847387106146,
+      "learning_rate": 3.275759380583681e-05,
+      "loss": 0.3699,
+      "step": 767
+    },
+    {
+      "epoch": 1.2329317269076305,
+      "grad_norm": 0.25186296667067626,
+      "learning_rate": 3.2727814175104235e-05,
+      "loss": 0.3541,
+      "step": 768
+    },
+    {
+      "epoch": 1.2345381526104418,
+      "grad_norm": 0.21321707223657882,
+      "learning_rate": 3.2698034544371654e-05,
+      "loss": 0.3511,
+      "step": 769
+    },
+    {
+      "epoch": 1.236144578313253,
+      "grad_norm": 0.21028796507808997,
+      "learning_rate": 3.266825491363907e-05,
+      "loss": 0.3713,
+      "step": 770
+    },
+    {
+      "epoch": 1.2377510040160642,
+      "grad_norm": 0.22137129389730711,
+      "learning_rate": 3.263847528290649e-05,
+      "loss": 0.3615,
+      "step": 771
+    },
+    {
+      "epoch": 1.2393574297188754,
+      "grad_norm": 0.21424357254598164,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.3646,
+      "step": 772
+    },
+    {
+      "epoch": 1.2409638554216866,
+      "grad_norm": 0.21445314064635118,
+      "learning_rate": 3.2578916021441336e-05,
+      "loss": 0.3725,
+      "step": 773
+    },
+    {
+      "epoch": 1.242570281124498,
+      "grad_norm": 0.22051212791632466,
+      "learning_rate": 3.2549136390708755e-05,
+      "loss": 0.3661,
+      "step": 774
+    },
+    {
+      "epoch": 1.2441767068273093,
+      "grad_norm": 0.2344032506467021,
+      "learning_rate": 3.251935675997618e-05,
+      "loss": 0.3542,
+      "step": 775
+    },
+    {
+      "epoch": 1.2457831325301205,
+      "grad_norm": 0.227403160247532,
+      "learning_rate": 3.24895771292436e-05,
+      "loss": 0.3582,
+      "step": 776
+    },
+    {
+      "epoch": 1.2473895582329317,
+      "grad_norm": 0.21636326335669,
+      "learning_rate": 3.245979749851102e-05,
+      "loss": 0.3505,
+      "step": 777
+    },
+    {
+      "epoch": 1.248995983935743,
+      "grad_norm": 0.2281562730752796,
+      "learning_rate": 3.243001786777844e-05,
+      "loss": 0.364,
+      "step": 778
+    },
+    {
+      "epoch": 1.2506024096385542,
+      "grad_norm": 0.2292249330966563,
+      "learning_rate": 3.240023823704586e-05,
+      "loss": 0.3722,
+      "step": 779
+    },
+    {
+      "epoch": 1.2522088353413654,
+      "grad_norm": 0.24790847991703668,
+      "learning_rate": 3.237045860631328e-05,
+      "loss": 0.37,
+      "step": 780
+    },
+    {
+      "epoch": 1.2538152610441768,
+      "grad_norm": 0.23734414857881575,
+      "learning_rate": 3.23406789755807e-05,
+      "loss": 0.3851,
+      "step": 781
+    },
+    {
+      "epoch": 1.2554216867469878,
+      "grad_norm": 0.23966985858047482,
+      "learning_rate": 3.2310899344848126e-05,
+      "loss": 0.3513,
+      "step": 782
+    },
+    {
+      "epoch": 1.2570281124497993,
+      "grad_norm": 0.26861252599311647,
+      "learning_rate": 3.228111971411555e-05,
+      "loss": 0.3742,
+      "step": 783
+    },
+    {
+      "epoch": 1.2586345381526105,
+      "grad_norm": 0.2405381223420924,
+      "learning_rate": 3.225134008338297e-05,
+      "loss": 0.3675,
+      "step": 784
+    },
+    {
+      "epoch": 1.2602409638554217,
+      "grad_norm": 0.25133074956933993,
+      "learning_rate": 3.222156045265039e-05,
+      "loss": 0.3476,
+      "step": 785
+    },
+    {
+      "epoch": 1.261847389558233,
+      "grad_norm": 0.22002417112468556,
+      "learning_rate": 3.219178082191781e-05,
+      "loss": 0.3613,
+      "step": 786
+    },
+    {
+      "epoch": 1.2634538152610442,
+      "grad_norm": 0.23643945898649082,
+      "learning_rate": 3.2162001191185234e-05,
+      "loss": 0.3805,
+      "step": 787
+    },
+    {
+      "epoch": 1.2650602409638554,
+      "grad_norm": 0.2340836057538758,
+      "learning_rate": 3.213222156045265e-05,
+      "loss": 0.3598,
+      "step": 788
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.24621197268298853,
+      "learning_rate": 3.210244192972007e-05,
+      "loss": 0.3667,
+      "step": 789
+    },
+    {
+      "epoch": 1.268273092369478,
+      "grad_norm": 0.21790940067218326,
+      "learning_rate": 3.207266229898749e-05,
+      "loss": 0.3901,
+      "step": 790
+    },
+    {
+      "epoch": 1.269879518072289,
+      "grad_norm": 0.27818463747836164,
+      "learning_rate": 3.2042882668254916e-05,
+      "loss": 0.36,
+      "step": 791
+    },
+    {
+      "epoch": 1.2714859437751005,
+      "grad_norm": 0.23485828790810853,
+      "learning_rate": 3.2013103037522335e-05,
+      "loss": 0.379,
+      "step": 792
+    },
+    {
+      "epoch": 1.2730923694779117,
+      "grad_norm": 0.2428969954061332,
+      "learning_rate": 3.1983323406789754e-05,
+      "loss": 0.3839,
+      "step": 793
+    },
+    {
+      "epoch": 1.274698795180723,
+      "grad_norm": 0.25201173026955526,
+      "learning_rate": 3.195354377605718e-05,
+      "loss": 0.3578,
+      "step": 794
+    },
+    {
+      "epoch": 1.2763052208835342,
+      "grad_norm": 0.2543167963872766,
+      "learning_rate": 3.19237641453246e-05,
+      "loss": 0.3735,
+      "step": 795
+    },
+    {
+      "epoch": 1.2779116465863454,
+      "grad_norm": 0.21114445186703706,
+      "learning_rate": 3.189398451459202e-05,
+      "loss": 0.357,
+      "step": 796
+    },
+    {
+      "epoch": 1.2795180722891566,
+      "grad_norm": 0.2734430777164395,
+      "learning_rate": 3.186420488385944e-05,
+      "loss": 0.3593,
+      "step": 797
+    },
+    {
+      "epoch": 1.2811244979919678,
+      "grad_norm": 0.21392607525142876,
+      "learning_rate": 3.183442525312687e-05,
+      "loss": 0.3496,
+      "step": 798
+    },
+    {
+      "epoch": 1.282730923694779,
+      "grad_norm": 0.21427758438676245,
+      "learning_rate": 3.180464562239429e-05,
+      "loss": 0.3695,
+      "step": 799
+    },
+    {
+      "epoch": 1.2843373493975903,
+      "grad_norm": 0.23448772916727945,
+      "learning_rate": 3.1774865991661706e-05,
+      "loss": 0.3715,
+      "step": 800
+    },
+    {
+      "epoch": 1.2859437751004017,
+      "grad_norm": 0.23891654697924247,
+      "learning_rate": 3.1745086360929125e-05,
+      "loss": 0.3794,
+      "step": 801
+    },
+    {
+      "epoch": 1.287550200803213,
+      "grad_norm": 0.25040550089891145,
+      "learning_rate": 3.171530673019655e-05,
+      "loss": 0.3802,
+      "step": 802
+    },
+    {
+      "epoch": 1.2891566265060241,
+      "grad_norm": 0.22959015673231148,
+      "learning_rate": 3.168552709946397e-05,
+      "loss": 0.3855,
+      "step": 803
+    },
+    {
+      "epoch": 1.2907630522088354,
+      "grad_norm": 0.21504738444833543,
+      "learning_rate": 3.165574746873139e-05,
+      "loss": 0.3706,
+      "step": 804
+    },
+    {
+      "epoch": 1.2923694779116466,
+      "grad_norm": 0.22916959554798327,
+      "learning_rate": 3.162596783799881e-05,
+      "loss": 0.3734,
+      "step": 805
+    },
+    {
+      "epoch": 1.2939759036144578,
+      "grad_norm": 0.2261734860287735,
+      "learning_rate": 3.159618820726623e-05,
+      "loss": 0.3796,
+      "step": 806
+    },
+    {
+      "epoch": 1.295582329317269,
+      "grad_norm": 0.22653415939680627,
+      "learning_rate": 3.156640857653365e-05,
+      "loss": 0.3686,
+      "step": 807
+    },
+    {
+      "epoch": 1.2971887550200802,
+      "grad_norm": 0.23392504669301062,
+      "learning_rate": 3.153662894580107e-05,
+      "loss": 0.3802,
+      "step": 808
+    },
+    {
+      "epoch": 1.2987951807228915,
+      "grad_norm": 0.22088590455041548,
+      "learning_rate": 3.1506849315068496e-05,
+      "loss": 0.367,
+      "step": 809
+    },
+    {
+      "epoch": 1.300401606425703,
+      "grad_norm": 0.23792429321888736,
+      "learning_rate": 3.1477069684335915e-05,
+      "loss": 0.3563,
+      "step": 810
+    },
+    {
+      "epoch": 1.3020080321285141,
+      "grad_norm": 0.20161084796105802,
+      "learning_rate": 3.1447290053603334e-05,
+      "loss": 0.3676,
+      "step": 811
+    },
+    {
+      "epoch": 1.3036144578313253,
+      "grad_norm": 0.2277725701014789,
+      "learning_rate": 3.141751042287076e-05,
+      "loss": 0.3529,
+      "step": 812
+    },
+    {
+      "epoch": 1.3052208835341366,
+      "grad_norm": 0.23045297881934104,
+      "learning_rate": 3.138773079213818e-05,
+      "loss": 0.3847,
+      "step": 813
+    },
+    {
+      "epoch": 1.3068273092369478,
+      "grad_norm": 0.22210517944789068,
+      "learning_rate": 3.1357951161405604e-05,
+      "loss": 0.3586,
+      "step": 814
+    },
+    {
+      "epoch": 1.308433734939759,
+      "grad_norm": 0.2072831680450955,
+      "learning_rate": 3.132817153067302e-05,
+      "loss": 0.3573,
+      "step": 815
+    },
+    {
+      "epoch": 1.3100401606425702,
+      "grad_norm": 0.2398466310783302,
+      "learning_rate": 3.129839189994044e-05,
+      "loss": 0.3615,
+      "step": 816
+    },
+    {
+      "epoch": 1.3116465863453814,
+      "grad_norm": 0.19600369300484732,
+      "learning_rate": 3.126861226920787e-05,
+      "loss": 0.3497,
+      "step": 817
+    },
+    {
+      "epoch": 1.3132530120481927,
+      "grad_norm": 0.2189372363986866,
+      "learning_rate": 3.1238832638475286e-05,
+      "loss": 0.3571,
+      "step": 818
+    },
+    {
+      "epoch": 1.3148594377510041,
+      "grad_norm": 0.24132969320550957,
+      "learning_rate": 3.1209053007742705e-05,
+      "loss": 0.3717,
+      "step": 819
+    },
+    {
+      "epoch": 1.3164658634538153,
+      "grad_norm": 0.2220779044504832,
+      "learning_rate": 3.1179273377010124e-05,
+      "loss": 0.3668,
+      "step": 820
+    },
+    {
+      "epoch": 1.3180722891566266,
+      "grad_norm": 0.21951027884242172,
+      "learning_rate": 3.114949374627755e-05,
+      "loss": 0.3666,
+      "step": 821
+    },
+    {
+      "epoch": 1.3196787148594378,
+      "grad_norm": 0.23139773130369037,
+      "learning_rate": 3.111971411554497e-05,
+      "loss": 0.3819,
+      "step": 822
+    },
+    {
+      "epoch": 1.321285140562249,
+      "grad_norm": 0.2436895797315047,
+      "learning_rate": 3.108993448481239e-05,
+      "loss": 0.3756,
+      "step": 823
+    },
+    {
+      "epoch": 1.3228915662650602,
+      "grad_norm": 0.24398206222736174,
+      "learning_rate": 3.1060154854079806e-05,
+      "loss": 0.3784,
+      "step": 824
+    },
+    {
+      "epoch": 1.3244979919678714,
+      "grad_norm": 0.22574883293170708,
+      "learning_rate": 3.103037522334723e-05,
+      "loss": 0.351,
+      "step": 825
+    },
+    {
+      "epoch": 1.3261044176706827,
+      "grad_norm": 0.23562300733545966,
+      "learning_rate": 3.100059559261465e-05,
+      "loss": 0.3525,
+      "step": 826
+    },
+    {
+      "epoch": 1.3277108433734939,
+      "grad_norm": 0.21956446777264724,
+      "learning_rate": 3.097081596188207e-05,
+      "loss": 0.365,
+      "step": 827
+    },
+    {
+      "epoch": 1.3293172690763053,
+      "grad_norm": 0.22065986302717655,
+      "learning_rate": 3.0941036331149495e-05,
+      "loss": 0.3596,
+      "step": 828
+    },
+    {
+      "epoch": 1.3309236947791165,
+      "grad_norm": 0.21384764922460403,
+      "learning_rate": 3.091125670041692e-05,
+      "loss": 0.3425,
+      "step": 829
+    },
+    {
+      "epoch": 1.3325301204819278,
+      "grad_norm": 0.21762415293360973,
+      "learning_rate": 3.088147706968434e-05,
+      "loss": 0.3682,
+      "step": 830
+    },
+    {
+      "epoch": 1.334136546184739,
+      "grad_norm": 0.221693082369684,
+      "learning_rate": 3.085169743895176e-05,
+      "loss": 0.3386,
+      "step": 831
+    },
+    {
+      "epoch": 1.3357429718875502,
+      "grad_norm": 0.22070044364874203,
+      "learning_rate": 3.082191780821918e-05,
+      "loss": 0.3803,
+      "step": 832
+    },
+    {
+      "epoch": 1.3373493975903614,
+      "grad_norm": 0.24697590279902848,
+      "learning_rate": 3.07921381774866e-05,
+      "loss": 0.3609,
+      "step": 833
+    },
+    {
+      "epoch": 1.3389558232931726,
+      "grad_norm": 0.20975548827859883,
+      "learning_rate": 3.076235854675402e-05,
+      "loss": 0.3568,
+      "step": 834
+    },
+    {
+      "epoch": 1.3405622489959839,
+      "grad_norm": 0.24626094973921908,
+      "learning_rate": 3.073257891602144e-05,
+      "loss": 0.3821,
+      "step": 835
+    },
+    {
+      "epoch": 1.342168674698795,
+      "grad_norm": 0.21878005364172165,
+      "learning_rate": 3.0702799285288866e-05,
+      "loss": 0.3541,
+      "step": 836
+    },
+    {
+      "epoch": 1.3437751004016065,
+      "grad_norm": 0.2091599497342082,
+      "learning_rate": 3.0673019654556285e-05,
+      "loss": 0.3777,
+      "step": 837
+    },
+    {
+      "epoch": 1.3453815261044177,
+      "grad_norm": 0.23698156981964014,
+      "learning_rate": 3.0643240023823704e-05,
+      "loss": 0.3577,
+      "step": 838
+    },
+    {
+      "epoch": 1.346987951807229,
+      "grad_norm": 0.19976369485296863,
+      "learning_rate": 3.061346039309112e-05,
+      "loss": 0.3492,
+      "step": 839
+    },
+    {
+      "epoch": 1.3485943775100402,
+      "grad_norm": 0.25041720546760476,
+      "learning_rate": 3.058368076235855e-05,
+      "loss": 0.3788,
+      "step": 840
+    },
+    {
+      "epoch": 1.3502008032128514,
+      "grad_norm": 0.22628458952334474,
+      "learning_rate": 3.055390113162597e-05,
+      "loss": 0.3777,
+      "step": 841
+    },
+    {
+      "epoch": 1.3518072289156626,
+      "grad_norm": 0.22779377389101385,
+      "learning_rate": 3.0524121500893386e-05,
+      "loss": 0.3465,
+      "step": 842
+    },
+    {
+      "epoch": 1.3534136546184738,
+      "grad_norm": 0.2028780371997436,
+      "learning_rate": 3.0494341870160815e-05,
+      "loss": 0.3584,
+      "step": 843
+    },
+    {
+      "epoch": 1.355020080321285,
+      "grad_norm": 0.2258560159139495,
+      "learning_rate": 3.0464562239428234e-05,
+      "loss": 0.3809,
+      "step": 844
+    },
+    {
+      "epoch": 1.3566265060240963,
+      "grad_norm": 0.22782391594390167,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.3741,
+      "step": 845
+    },
+    {
+      "epoch": 1.3582329317269077,
+      "grad_norm": 0.21440199029792373,
+      "learning_rate": 3.0405002977963075e-05,
+      "loss": 0.3584,
+      "step": 846
+    },
+    {
+      "epoch": 1.359839357429719,
+      "grad_norm": 0.21844067462947783,
+      "learning_rate": 3.0375223347230497e-05,
+      "loss": 0.3921,
+      "step": 847
+    },
+    {
+      "epoch": 1.3614457831325302,
+      "grad_norm": 0.23535442333385953,
+      "learning_rate": 3.0345443716497916e-05,
+      "loss": 0.3811,
+      "step": 848
+    },
+    {
+      "epoch": 1.3630522088353414,
+      "grad_norm": 0.2428521987029711,
+      "learning_rate": 3.031566408576534e-05,
+      "loss": 0.3566,
+      "step": 849
+    },
+    {
+      "epoch": 1.3646586345381526,
+      "grad_norm": 0.2354500934566656,
+      "learning_rate": 3.028588445503276e-05,
+      "loss": 0.3474,
+      "step": 850
+    },
+    {
+      "epoch": 1.3662650602409638,
+      "grad_norm": 0.2380376127583646,
+      "learning_rate": 3.025610482430018e-05,
+      "loss": 0.3645,
+      "step": 851
+    },
+    {
+      "epoch": 1.367871485943775,
+      "grad_norm": 0.24030027929429704,
+      "learning_rate": 3.0226325193567602e-05,
+      "loss": 0.3644,
+      "step": 852
+    },
+    {
+      "epoch": 1.3694779116465863,
+      "grad_norm": 0.22628103813644257,
+      "learning_rate": 3.019654556283502e-05,
+      "loss": 0.3616,
+      "step": 853
+    },
+    {
+      "epoch": 1.3710843373493975,
+      "grad_norm": 0.2481381906601932,
+      "learning_rate": 3.0166765932102443e-05,
+      "loss": 0.3686,
+      "step": 854
+    },
+    {
+      "epoch": 1.372690763052209,
+      "grad_norm": 0.22932123919593422,
+      "learning_rate": 3.0136986301369862e-05,
+      "loss": 0.3513,
+      "step": 855
+    },
+    {
+      "epoch": 1.3742971887550202,
+      "grad_norm": 0.232607475833928,
+      "learning_rate": 3.0107206670637284e-05,
+      "loss": 0.356,
+      "step": 856
+    },
+    {
+      "epoch": 1.3759036144578314,
+      "grad_norm": 0.23919609411368228,
+      "learning_rate": 3.0077427039904703e-05,
+      "loss": 0.3628,
+      "step": 857
+    },
+    {
+      "epoch": 1.3775100401606426,
+      "grad_norm": 0.24735662052608867,
+      "learning_rate": 3.0047647409172125e-05,
+      "loss": 0.38,
+      "step": 858
+    },
+    {
+      "epoch": 1.3791164658634538,
+      "grad_norm": 0.24312498440342986,
+      "learning_rate": 3.001786777843955e-05,
+      "loss": 0.3824,
+      "step": 859
+    },
+    {
+      "epoch": 1.380722891566265,
+      "grad_norm": 0.24614164137683403,
+      "learning_rate": 2.9988088147706973e-05,
+      "loss": 0.3669,
+      "step": 860
+    },
+    {
+      "epoch": 1.3823293172690763,
+      "grad_norm": 0.23007440185768782,
+      "learning_rate": 2.9958308516974392e-05,
+      "loss": 0.3823,
+      "step": 861
+    },
+    {
+      "epoch": 1.3839357429718875,
+      "grad_norm": 0.23503403725119434,
+      "learning_rate": 2.9928528886241814e-05,
+      "loss": 0.3547,
+      "step": 862
+    },
+    {
+      "epoch": 1.3855421686746987,
+      "grad_norm": 0.26213313448474795,
+      "learning_rate": 2.9898749255509233e-05,
+      "loss": 0.3902,
+      "step": 863
+    },
+    {
+      "epoch": 1.3871485943775101,
+      "grad_norm": 0.27248878472797816,
+      "learning_rate": 2.9868969624776655e-05,
+      "loss": 0.3638,
+      "step": 864
+    },
+    {
+      "epoch": 1.3887550200803214,
+      "grad_norm": 0.21904082404496375,
+      "learning_rate": 2.9839189994044074e-05,
+      "loss": 0.3781,
+      "step": 865
+    },
+    {
+      "epoch": 1.3903614457831326,
+      "grad_norm": 0.23427228859116567,
+      "learning_rate": 2.9809410363311496e-05,
+      "loss": 0.3681,
+      "step": 866
+    },
+    {
+      "epoch": 1.3919678714859438,
+      "grad_norm": 0.25377813900510793,
+      "learning_rate": 2.977963073257892e-05,
+      "loss": 0.3519,
+      "step": 867
+    },
+    {
+      "epoch": 1.393574297188755,
+      "grad_norm": 0.20959477667017692,
+      "learning_rate": 2.9749851101846337e-05,
+      "loss": 0.3822,
+      "step": 868
+    },
+    {
+      "epoch": 1.3951807228915662,
+      "grad_norm": 0.22810492747487612,
+      "learning_rate": 2.972007147111376e-05,
+      "loss": 0.3637,
+      "step": 869
+    },
+    {
+      "epoch": 1.3967871485943775,
+      "grad_norm": 0.2474576219838295,
+      "learning_rate": 2.969029184038118e-05,
+      "loss": 0.381,
+      "step": 870
+    },
+    {
+      "epoch": 1.3983935742971887,
+      "grad_norm": 0.21986525317161276,
+      "learning_rate": 2.96605122096486e-05,
+      "loss": 0.363,
+      "step": 871
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21908538162021823,
+      "learning_rate": 2.963073257891602e-05,
+      "loss": 0.3608,
+      "step": 872
+    },
+    {
+      "epoch": 1.4016064257028114,
+      "grad_norm": 0.23107734597072727,
+      "learning_rate": 2.9600952948183442e-05,
+      "loss": 0.3856,
+      "step": 873
+    },
+    {
+      "epoch": 1.4032128514056224,
+      "grad_norm": 0.2162366677993556,
+      "learning_rate": 2.9571173317450868e-05,
+      "loss": 0.3601,
+      "step": 874
+    },
+    {
+      "epoch": 1.4048192771084338,
+      "grad_norm": 0.23032507693748092,
+      "learning_rate": 2.954139368671829e-05,
+      "loss": 0.3668,
+      "step": 875
+    },
+    {
+      "epoch": 1.406425702811245,
+      "grad_norm": 0.23005753209891203,
+      "learning_rate": 2.951161405598571e-05,
+      "loss": 0.3871,
+      "step": 876
+    },
+    {
+      "epoch": 1.4080321285140562,
+      "grad_norm": 0.23665682010297576,
+      "learning_rate": 2.948183442525313e-05,
+      "loss": 0.3671,
+      "step": 877
+    },
+    {
+      "epoch": 1.4096385542168675,
+      "grad_norm": 0.2123172712814544,
+      "learning_rate": 2.945205479452055e-05,
+      "loss": 0.3657,
+      "step": 878
+    },
+    {
+      "epoch": 1.4112449799196787,
+      "grad_norm": 0.23159086396276327,
+      "learning_rate": 2.9422275163787972e-05,
+      "loss": 0.3569,
+      "step": 879
+    },
+    {
+      "epoch": 1.41285140562249,
+      "grad_norm": 0.20610842351742073,
+      "learning_rate": 2.939249553305539e-05,
+      "loss": 0.3667,
+      "step": 880
+    },
+    {
+      "epoch": 1.4144578313253011,
+      "grad_norm": 0.2095943517768558,
+      "learning_rate": 2.9362715902322813e-05,
+      "loss": 0.3594,
+      "step": 881
+    },
+    {
+      "epoch": 1.4160642570281126,
+      "grad_norm": 0.20383729710674806,
+      "learning_rate": 2.9332936271590232e-05,
+      "loss": 0.3812,
+      "step": 882
+    },
+    {
+      "epoch": 1.4176706827309236,
+      "grad_norm": 0.2148605870055888,
+      "learning_rate": 2.9303156640857654e-05,
+      "loss": 0.3574,
+      "step": 883
+    },
+    {
+      "epoch": 1.419277108433735,
+      "grad_norm": 0.20538045682489592,
+      "learning_rate": 2.9273377010125076e-05,
+      "loss": 0.3593,
+      "step": 884
+    },
+    {
+      "epoch": 1.4208835341365462,
+      "grad_norm": 0.22659131989469658,
+      "learning_rate": 2.9243597379392495e-05,
+      "loss": 0.3691,
+      "step": 885
+    },
+    {
+      "epoch": 1.4224899598393574,
+      "grad_norm": 0.2369038481320581,
+      "learning_rate": 2.9213817748659918e-05,
+      "loss": 0.3549,
+      "step": 886
+    },
+    {
+      "epoch": 1.4240963855421687,
+      "grad_norm": 0.21487405668840318,
+      "learning_rate": 2.9184038117927336e-05,
+      "loss": 0.3672,
+      "step": 887
+    },
+    {
+      "epoch": 1.4257028112449799,
+      "grad_norm": 0.21882296224297848,
+      "learning_rate": 2.915425848719476e-05,
+      "loss": 0.3819,
+      "step": 888
+    },
+    {
+      "epoch": 1.427309236947791,
+      "grad_norm": 0.2122041283250008,
+      "learning_rate": 2.9124478856462184e-05,
+      "loss": 0.3684,
+      "step": 889
+    },
+    {
+      "epoch": 1.4289156626506023,
+      "grad_norm": 0.2100792740049444,
+      "learning_rate": 2.9094699225729603e-05,
+      "loss": 0.3691,
+      "step": 890
+    },
+    {
+      "epoch": 1.4305220883534138,
+      "grad_norm": 0.21515277117579934,
+      "learning_rate": 2.9064919594997025e-05,
+      "loss": 0.3558,
+      "step": 891
+    },
+    {
+      "epoch": 1.4321285140562248,
+      "grad_norm": 0.19445546811223982,
+      "learning_rate": 2.9035139964264448e-05,
+      "loss": 0.3556,
+      "step": 892
+    },
+    {
+      "epoch": 1.4337349397590362,
+      "grad_norm": 0.22756377734983396,
+      "learning_rate": 2.9005360333531867e-05,
+      "loss": 0.3652,
+      "step": 893
+    },
+    {
+      "epoch": 1.4353413654618474,
+      "grad_norm": 0.2149240696923472,
+      "learning_rate": 2.897558070279929e-05,
+      "loss": 0.3656,
+      "step": 894
+    },
+    {
+      "epoch": 1.4369477911646586,
+      "grad_norm": 0.20526738516022955,
+      "learning_rate": 2.8945801072066708e-05,
+      "loss": 0.3529,
+      "step": 895
+    },
+    {
+      "epoch": 1.4385542168674699,
+      "grad_norm": 0.2429605718067559,
+      "learning_rate": 2.891602144133413e-05,
+      "loss": 0.3689,
+      "step": 896
+    },
+    {
+      "epoch": 1.440160642570281,
+      "grad_norm": 0.2149812497343841,
+      "learning_rate": 2.888624181060155e-05,
+      "loss": 0.3979,
+      "step": 897
+    },
+    {
+      "epoch": 1.4417670682730923,
+      "grad_norm": 0.23192601126786705,
+      "learning_rate": 2.885646217986897e-05,
+      "loss": 0.3912,
+      "step": 898
+    },
+    {
+      "epoch": 1.4433734939759035,
+      "grad_norm": 0.22698204921354354,
+      "learning_rate": 2.882668254913639e-05,
+      "loss": 0.3546,
+      "step": 899
+    },
+    {
+      "epoch": 1.444979919678715,
+      "grad_norm": 0.22205606795192698,
+      "learning_rate": 2.8796902918403812e-05,
+      "loss": 0.3692,
+      "step": 900
+    },
+    {
+      "epoch": 1.446586345381526,
+      "grad_norm": 0.2111212056897702,
+      "learning_rate": 2.8767123287671234e-05,
+      "loss": 0.3787,
+      "step": 901
+    },
+    {
+      "epoch": 1.4481927710843374,
+      "grad_norm": 0.23052767402474444,
+      "learning_rate": 2.8737343656938653e-05,
+      "loss": 0.372,
+      "step": 902
+    },
+    {
+      "epoch": 1.4497991967871486,
+      "grad_norm": 0.2424302535235894,
+      "learning_rate": 2.8707564026206075e-05,
+      "loss": 0.3748,
+      "step": 903
+    },
+    {
+      "epoch": 1.4514056224899599,
+      "grad_norm": 0.20197438070774545,
+      "learning_rate": 2.8677784395473494e-05,
+      "loss": 0.3635,
+      "step": 904
+    },
+    {
+      "epoch": 1.453012048192771,
+      "grad_norm": 0.2084681253188935,
+      "learning_rate": 2.864800476474092e-05,
+      "loss": 0.3925,
+      "step": 905
+    },
+    {
+      "epoch": 1.4546184738955823,
+      "grad_norm": 0.2072242386279762,
+      "learning_rate": 2.8618225134008342e-05,
+      "loss": 0.3672,
+      "step": 906
+    },
+    {
+      "epoch": 1.4562248995983935,
+      "grad_norm": 0.21284756895769136,
+      "learning_rate": 2.858844550327576e-05,
+      "loss": 0.3685,
+      "step": 907
+    },
+    {
+      "epoch": 1.4578313253012047,
+      "grad_norm": 0.20227667670803087,
+      "learning_rate": 2.8558665872543183e-05,
+      "loss": 0.3597,
+      "step": 908
+    },
+    {
+      "epoch": 1.4594377510040162,
+      "grad_norm": 0.21127860676616467,
+      "learning_rate": 2.8528886241810606e-05,
+      "loss": 0.3488,
+      "step": 909
+    },
+    {
+      "epoch": 1.4610441767068272,
+      "grad_norm": 0.19786787955222504,
+      "learning_rate": 2.8499106611078024e-05,
+      "loss": 0.3725,
+      "step": 910
+    },
+    {
+      "epoch": 1.4626506024096386,
+      "grad_norm": 0.21027102248989119,
+      "learning_rate": 2.8469326980345447e-05,
+      "loss": 0.3727,
+      "step": 911
+    },
+    {
+      "epoch": 1.4642570281124498,
+      "grad_norm": 0.2219598070879816,
+      "learning_rate": 2.8439547349612866e-05,
+      "loss": 0.385,
+      "step": 912
+    },
+    {
+      "epoch": 1.465863453815261,
+      "grad_norm": 0.21792884213216784,
+      "learning_rate": 2.8409767718880288e-05,
+      "loss": 0.3659,
+      "step": 913
+    },
+    {
+      "epoch": 1.4674698795180723,
+      "grad_norm": 0.23692472629136208,
+      "learning_rate": 2.8379988088147707e-05,
+      "loss": 0.3716,
+      "step": 914
+    },
+    {
+      "epoch": 1.4690763052208835,
+      "grad_norm": 0.22574261052333702,
+      "learning_rate": 2.835020845741513e-05,
+      "loss": 0.3884,
+      "step": 915
+    },
+    {
+      "epoch": 1.4706827309236947,
+      "grad_norm": 0.26972786984809455,
+      "learning_rate": 2.8320428826682548e-05,
+      "loss": 0.3477,
+      "step": 916
+    },
+    {
+      "epoch": 1.472289156626506,
+      "grad_norm": 0.20622807458145592,
+      "learning_rate": 2.829064919594997e-05,
+      "loss": 0.373,
+      "step": 917
+    },
+    {
+      "epoch": 1.4738955823293174,
+      "grad_norm": 0.22586468226366713,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 0.3612,
+      "step": 918
+    },
+    {
+      "epoch": 1.4755020080321284,
+      "grad_norm": 0.22600368188790596,
+      "learning_rate": 2.823108993448481e-05,
+      "loss": 0.3624,
+      "step": 919
+    },
+    {
+      "epoch": 1.4771084337349398,
+      "grad_norm": 0.20939787587771433,
+      "learning_rate": 2.8201310303752237e-05,
+      "loss": 0.3628,
+      "step": 920
+    },
+    {
+      "epoch": 1.478714859437751,
+      "grad_norm": 0.21476583410976668,
+      "learning_rate": 2.817153067301966e-05,
+      "loss": 0.3607,
+      "step": 921
+    },
+    {
+      "epoch": 1.4803212851405623,
+      "grad_norm": 0.21734844652117358,
+      "learning_rate": 2.8141751042287078e-05,
+      "loss": 0.3635,
+      "step": 922
+    },
+    {
+      "epoch": 1.4819277108433735,
+      "grad_norm": 0.2144356999561757,
+      "learning_rate": 2.81119714115545e-05,
+      "loss": 0.3555,
+      "step": 923
+    },
+    {
+      "epoch": 1.4835341365461847,
+      "grad_norm": 0.234862318388648,
+      "learning_rate": 2.808219178082192e-05,
+      "loss": 0.3663,
+      "step": 924
+    },
+    {
+      "epoch": 1.485140562248996,
+      "grad_norm": 0.23031757949966822,
+      "learning_rate": 2.805241215008934e-05,
+      "loss": 0.3713,
+      "step": 925
+    },
+    {
+      "epoch": 1.4867469879518072,
+      "grad_norm": 0.2157553880030866,
+      "learning_rate": 2.8022632519356763e-05,
+      "loss": 0.3756,
+      "step": 926
+    },
+    {
+      "epoch": 1.4883534136546186,
+      "grad_norm": 0.2276004175355454,
+      "learning_rate": 2.7992852888624182e-05,
+      "loss": 0.3695,
+      "step": 927
+    },
+    {
+      "epoch": 1.4899598393574296,
+      "grad_norm": 0.218014308267801,
+      "learning_rate": 2.7963073257891605e-05,
+      "loss": 0.343,
+      "step": 928
+    },
+    {
+      "epoch": 1.491566265060241,
+      "grad_norm": 0.21269986535144367,
+      "learning_rate": 2.7933293627159023e-05,
+      "loss": 0.3783,
+      "step": 929
+    },
+    {
+      "epoch": 1.4931726907630523,
+      "grad_norm": 0.20789162882212137,
+      "learning_rate": 2.7903513996426446e-05,
+      "loss": 0.3755,
+      "step": 930
+    },
+    {
+      "epoch": 1.4947791164658635,
+      "grad_norm": 0.2166785940860413,
+      "learning_rate": 2.7873734365693864e-05,
+      "loss": 0.351,
+      "step": 931
+    },
+    {
+      "epoch": 1.4963855421686747,
+      "grad_norm": 0.21274777459162647,
+      "learning_rate": 2.7843954734961287e-05,
+      "loss": 0.3676,
+      "step": 932
+    },
+    {
+      "epoch": 1.497991967871486,
+      "grad_norm": 0.2219183771458503,
+      "learning_rate": 2.7814175104228706e-05,
+      "loss": 0.377,
+      "step": 933
+    },
+    {
+      "epoch": 1.4995983935742971,
+      "grad_norm": 0.22762225908211084,
+      "learning_rate": 2.7784395473496128e-05,
+      "loss": 0.3715,
+      "step": 934
+    },
+    {
+      "epoch": 1.5012048192771084,
+      "grad_norm": 0.2388083493092833,
+      "learning_rate": 2.7754615842763547e-05,
+      "loss": 0.3473,
+      "step": 935
+    },
+    {
+      "epoch": 1.5028112449799198,
+      "grad_norm": 0.20429686311494033,
+      "learning_rate": 2.7724836212030976e-05,
+      "loss": 0.3548,
+      "step": 936
+    },
+    {
+      "epoch": 1.5044176706827308,
+      "grad_norm": 0.23481779789206692,
+      "learning_rate": 2.7695056581298395e-05,
+      "loss": 0.3702,
+      "step": 937
+    },
+    {
+      "epoch": 1.5060240963855422,
+      "grad_norm": 0.25347931950970687,
+      "learning_rate": 2.7665276950565817e-05,
+      "loss": 0.3549,
+      "step": 938
+    },
+    {
+      "epoch": 1.5076305220883535,
+      "grad_norm": 0.2530466881443676,
+      "learning_rate": 2.7635497319833236e-05,
+      "loss": 0.3652,
+      "step": 939
+    },
+    {
+      "epoch": 1.5092369477911647,
+      "grad_norm": 0.2099481634380821,
+      "learning_rate": 2.7605717689100658e-05,
+      "loss": 0.3698,
+      "step": 940
+    },
+    {
+      "epoch": 1.510843373493976,
+      "grad_norm": 0.22383646269194873,
+      "learning_rate": 2.7575938058368077e-05,
+      "loss": 0.3473,
+      "step": 941
+    },
+    {
+      "epoch": 1.5124497991967871,
+      "grad_norm": 0.23510008704046406,
+      "learning_rate": 2.75461584276355e-05,
+      "loss": 0.3642,
+      "step": 942
+    },
+    {
+      "epoch": 1.5140562248995983,
+      "grad_norm": 0.2365960136406408,
+      "learning_rate": 2.751637879690292e-05,
+      "loss": 0.3587,
+      "step": 943
+    },
+    {
+      "epoch": 1.5156626506024096,
+      "grad_norm": 0.20194874258388565,
+      "learning_rate": 2.748659916617034e-05,
+      "loss": 0.3421,
+      "step": 944
+    },
+    {
+      "epoch": 1.517269076305221,
+      "grad_norm": 0.21419410413526951,
+      "learning_rate": 2.7456819535437762e-05,
+      "loss": 0.3642,
+      "step": 945
+    },
+    {
+      "epoch": 1.518875502008032,
+      "grad_norm": 0.23503233240382296,
+      "learning_rate": 2.742703990470518e-05,
+      "loss": 0.3639,
+      "step": 946
+    },
+    {
+      "epoch": 1.5204819277108435,
+      "grad_norm": 0.19989894199559718,
+      "learning_rate": 2.7397260273972603e-05,
+      "loss": 0.3547,
+      "step": 947
+    },
+    {
+      "epoch": 1.5220883534136547,
+      "grad_norm": 0.22996814320536904,
+      "learning_rate": 2.7367480643240022e-05,
+      "loss": 0.393,
+      "step": 948
+    },
+    {
+      "epoch": 1.523694779116466,
+      "grad_norm": 0.23563323680885603,
+      "learning_rate": 2.7337701012507445e-05,
+      "loss": 0.3789,
+      "step": 949
+    },
+    {
+      "epoch": 1.5253012048192771,
+      "grad_norm": 0.22544780319410296,
+      "learning_rate": 2.7307921381774863e-05,
+      "loss": 0.346,
+      "step": 950
+    },
+    {
+      "epoch": 1.5269076305220883,
+      "grad_norm": 0.21358695063820807,
+      "learning_rate": 2.7278141751042292e-05,
+      "loss": 0.3591,
+      "step": 951
+    },
+    {
+      "epoch": 1.5285140562248996,
+      "grad_norm": 0.23704356976175486,
+      "learning_rate": 2.724836212030971e-05,
+      "loss": 0.3839,
+      "step": 952
+    },
+    {
+      "epoch": 1.5301204819277108,
+      "grad_norm": 0.27475630912830123,
+      "learning_rate": 2.7218582489577134e-05,
+      "loss": 0.3553,
+      "step": 953
+    },
+    {
+      "epoch": 1.5317269076305222,
+      "grad_norm": 0.2015377422723893,
+      "learning_rate": 2.7188802858844552e-05,
+      "loss": 0.3523,
+      "step": 954
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.21028984207219548,
+      "learning_rate": 2.7159023228111975e-05,
+      "loss": 0.3636,
+      "step": 955
+    },
+    {
+      "epoch": 1.5349397590361447,
+      "grad_norm": 0.2761749899383388,
+      "learning_rate": 2.7129243597379394e-05,
+      "loss": 0.3714,
+      "step": 956
+    },
+    {
+      "epoch": 1.5365461847389559,
+      "grad_norm": 0.24211596663596374,
+      "learning_rate": 2.7099463966646816e-05,
+      "loss": 0.3663,
+      "step": 957
+    },
+    {
+      "epoch": 1.538152610441767,
+      "grad_norm": 0.22792483992791712,
+      "learning_rate": 2.7069684335914235e-05,
+      "loss": 0.364,
+      "step": 958
+    },
+    {
+      "epoch": 1.5397590361445783,
+      "grad_norm": 0.24171640195202226,
+      "learning_rate": 2.7039904705181657e-05,
+      "loss": 0.3732,
+      "step": 959
+    },
+    {
+      "epoch": 1.5413654618473895,
+      "grad_norm": 0.24911482394201456,
+      "learning_rate": 2.701012507444908e-05,
+      "loss": 0.355,
+      "step": 960
+    },
+    {
+      "epoch": 1.5429718875502008,
+      "grad_norm": 0.2319688639910931,
+      "learning_rate": 2.6980345443716498e-05,
+      "loss": 0.365,
+      "step": 961
+    },
+    {
+      "epoch": 1.544578313253012,
+      "grad_norm": 0.23455520769533375,
+      "learning_rate": 2.695056581298392e-05,
+      "loss": 0.3874,
+      "step": 962
+    },
+    {
+      "epoch": 1.5461847389558234,
+      "grad_norm": 0.23995257002178408,
+      "learning_rate": 2.692078618225134e-05,
+      "loss": 0.3595,
+      "step": 963
+    },
+    {
+      "epoch": 1.5477911646586344,
+      "grad_norm": 0.23535956707436564,
+      "learning_rate": 2.689100655151876e-05,
+      "loss": 0.3456,
+      "step": 964
+    },
+    {
+      "epoch": 1.5493975903614459,
+      "grad_norm": 0.21283048878661445,
+      "learning_rate": 2.686122692078618e-05,
+      "loss": 0.3794,
+      "step": 965
+    },
+    {
+      "epoch": 1.5510040160642569,
+      "grad_norm": 0.27485947364881846,
+      "learning_rate": 2.6831447290053606e-05,
+      "loss": 0.3886,
+      "step": 966
+    },
+    {
+      "epoch": 1.5526104417670683,
+      "grad_norm": 0.25782677358973055,
+      "learning_rate": 2.6801667659321028e-05,
+      "loss": 0.3842,
+      "step": 967
+    },
+    {
+      "epoch": 1.5542168674698795,
+      "grad_norm": 0.22391534287768072,
+      "learning_rate": 2.677188802858845e-05,
+      "loss": 0.364,
+      "step": 968
+    },
+    {
+      "epoch": 1.5558232931726907,
+      "grad_norm": 0.20344190008680924,
+      "learning_rate": 2.674210839785587e-05,
+      "loss": 0.3648,
+      "step": 969
+    },
+    {
+      "epoch": 1.557429718875502,
+      "grad_norm": 0.2632931728366601,
+      "learning_rate": 2.671232876712329e-05,
+      "loss": 0.345,
+      "step": 970
+    },
+    {
+      "epoch": 1.5590361445783132,
+      "grad_norm": 0.22477219142520344,
+      "learning_rate": 2.668254913639071e-05,
+      "loss": 0.3616,
+      "step": 971
+    },
+    {
+      "epoch": 1.5606425702811246,
+      "grad_norm": 0.20847250954768615,
+      "learning_rate": 2.6652769505658133e-05,
+      "loss": 0.3869,
+      "step": 972
+    },
+    {
+      "epoch": 1.5622489959839356,
+      "grad_norm": 0.2105481288432062,
+      "learning_rate": 2.662298987492555e-05,
+      "loss": 0.3641,
+      "step": 973
+    },
+    {
+      "epoch": 1.563855421686747,
+      "grad_norm": 0.22889309732013546,
+      "learning_rate": 2.6593210244192974e-05,
+      "loss": 0.3535,
+      "step": 974
+    },
+    {
+      "epoch": 1.565461847389558,
+      "grad_norm": 0.20431010393481647,
+      "learning_rate": 2.6563430613460393e-05,
+      "loss": 0.3637,
+      "step": 975
+    },
+    {
+      "epoch": 1.5670682730923695,
+      "grad_norm": 0.21790437848384872,
+      "learning_rate": 2.6533650982727815e-05,
+      "loss": 0.3682,
+      "step": 976
+    },
+    {
+      "epoch": 1.5686746987951807,
+      "grad_norm": 0.20127235332587692,
+      "learning_rate": 2.6503871351995234e-05,
+      "loss": 0.3668,
+      "step": 977
+    },
+    {
+      "epoch": 1.570281124497992,
+      "grad_norm": 0.209472785322983,
+      "learning_rate": 2.6474091721262656e-05,
+      "loss": 0.3483,
+      "step": 978
+    },
+    {
+      "epoch": 1.5718875502008032,
+      "grad_norm": 0.2076162814357429,
+      "learning_rate": 2.6444312090530078e-05,
+      "loss": 0.3795,
+      "step": 979
+    },
+    {
+      "epoch": 1.5734939759036144,
+      "grad_norm": 0.21586409436702833,
+      "learning_rate": 2.6414532459797497e-05,
+      "loss": 0.3928,
+      "step": 980
+    },
+    {
+      "epoch": 1.5751004016064258,
+      "grad_norm": 0.22452727262006167,
+      "learning_rate": 2.638475282906492e-05,
+      "loss": 0.3734,
+      "step": 981
+    },
+    {
+      "epoch": 1.5767068273092368,
+      "grad_norm": 0.21841183925586227,
+      "learning_rate": 2.6354973198332345e-05,
+      "loss": 0.349,
+      "step": 982
+    },
+    {
+      "epoch": 1.5783132530120483,
+      "grad_norm": 0.21497213489470451,
+      "learning_rate": 2.6325193567599764e-05,
+      "loss": 0.3579,
+      "step": 983
+    },
+    {
+      "epoch": 1.5799196787148593,
+      "grad_norm": 0.21853135828760742,
+      "learning_rate": 2.6295413936867186e-05,
+      "loss": 0.3757,
+      "step": 984
+    },
+    {
+      "epoch": 1.5815261044176707,
+      "grad_norm": 0.22156628370310655,
+      "learning_rate": 2.6265634306134608e-05,
+      "loss": 0.3758,
+      "step": 985
+    },
+    {
+      "epoch": 1.583132530120482,
+      "grad_norm": 0.25664803260355046,
+      "learning_rate": 2.6235854675402027e-05,
+      "loss": 0.3524,
+      "step": 986
+    },
+    {
+      "epoch": 1.5847389558232932,
+      "grad_norm": 0.21833937858097816,
+      "learning_rate": 2.620607504466945e-05,
+      "loss": 0.3668,
+      "step": 987
+    },
+    {
+      "epoch": 1.5863453815261044,
+      "grad_norm": 0.2174828652080452,
+      "learning_rate": 2.6176295413936868e-05,
+      "loss": 0.3613,
+      "step": 988
+    },
+    {
+      "epoch": 1.5879518072289156,
+      "grad_norm": 0.2353983024978749,
+      "learning_rate": 2.614651578320429e-05,
+      "loss": 0.3589,
+      "step": 989
+    },
+    {
+      "epoch": 1.589558232931727,
+      "grad_norm": 0.2174023484648924,
+      "learning_rate": 2.611673615247171e-05,
+      "loss": 0.3724,
+      "step": 990
+    },
+    {
+      "epoch": 1.591164658634538,
+      "grad_norm": 0.23618451910653723,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.3671,
+      "step": 991
+    },
+    {
+      "epoch": 1.5927710843373495,
+      "grad_norm": 0.20739430564186828,
+      "learning_rate": 2.605717689100655e-05,
+      "loss": 0.3625,
+      "step": 992
+    },
+    {
+      "epoch": 1.5943775100401605,
+      "grad_norm": 0.21228651394370981,
+      "learning_rate": 2.6027397260273973e-05,
+      "loss": 0.35,
+      "step": 993
+    },
+    {
+      "epoch": 1.595983935742972,
+      "grad_norm": 0.2085036082140482,
+      "learning_rate": 2.599761762954139e-05,
+      "loss": 0.371,
+      "step": 994
+    },
+    {
+      "epoch": 1.5975903614457831,
+      "grad_norm": 0.23282178824648542,
+      "learning_rate": 2.5967837998808814e-05,
+      "loss": 0.3612,
+      "step": 995
+    },
+    {
+      "epoch": 1.5991967871485944,
+      "grad_norm": 0.2290510109947705,
+      "learning_rate": 2.5938058368076236e-05,
+      "loss": 0.3748,
+      "step": 996
+    },
+    {
+      "epoch": 1.6008032128514056,
+      "grad_norm": 0.22200176495038842,
+      "learning_rate": 2.590827873734366e-05,
+      "loss": 0.3825,
+      "step": 997
+    },
+    {
+      "epoch": 1.6024096385542168,
+      "grad_norm": 0.22439874377419233,
+      "learning_rate": 2.587849910661108e-05,
+      "loss": 0.3539,
+      "step": 998
+    },
+    {
+      "epoch": 1.6040160642570283,
+      "grad_norm": 0.23407001058299182,
+      "learning_rate": 2.5848719475878503e-05,
+      "loss": 0.3781,
+      "step": 999
+    },
+    {
+      "epoch": 1.6056224899598392,
+      "grad_norm": 0.21886484223211836,
+      "learning_rate": 2.581893984514592e-05,
+      "loss": 0.3739,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6072289156626507,
+      "grad_norm": 0.2530388270263431,
+      "learning_rate": 2.5789160214413344e-05,
+      "loss": 0.3684,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6088353413654617,
+      "grad_norm": 0.2036275976702559,
+      "learning_rate": 2.5759380583680766e-05,
+      "loss": 0.3607,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6104417670682731,
+      "grad_norm": 0.23121711876497358,
+      "learning_rate": 2.5729600952948185e-05,
+      "loss": 0.351,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6120481927710844,
+      "grad_norm": 0.21462527213179255,
+      "learning_rate": 2.5699821322215607e-05,
+      "loss": 0.3575,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6136546184738956,
+      "grad_norm": 0.20052158615851726,
+      "learning_rate": 2.5670041691483026e-05,
+      "loss": 0.3754,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6152610441767068,
+      "grad_norm": 0.22616020302746706,
+      "learning_rate": 2.5640262060750448e-05,
+      "loss": 0.3677,
+      "step": 1006
+    },
+    {
+      "epoch": 1.616867469879518,
+      "grad_norm": 0.2226715559894779,
+      "learning_rate": 2.5610482430017867e-05,
+      "loss": 0.3617,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6184738955823295,
+      "grad_norm": 0.20836554253125414,
+      "learning_rate": 2.558070279928529e-05,
+      "loss": 0.3724,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6200803212851405,
+      "grad_norm": 0.22463188594728506,
+      "learning_rate": 2.5550923168552708e-05,
+      "loss": 0.3634,
+      "step": 1009
+    },
+    {
+      "epoch": 1.621686746987952,
+      "grad_norm": 0.21187153182745774,
+      "learning_rate": 2.552114353782013e-05,
+      "loss": 0.3709,
+      "step": 1010
+    },
+    {
+      "epoch": 1.623293172690763,
+      "grad_norm": 0.20305345286193022,
+      "learning_rate": 2.549136390708755e-05,
+      "loss": 0.3547,
+      "step": 1011
+    },
+    {
+      "epoch": 1.6248995983935743,
+      "grad_norm": 0.18531268063010506,
+      "learning_rate": 2.546158427635497e-05,
+      "loss": 0.3707,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6265060240963856,
+      "grad_norm": 0.2041856506255866,
+      "learning_rate": 2.5431804645622397e-05,
+      "loss": 0.3715,
+      "step": 1013
+    },
+    {
+      "epoch": 1.6281124497991968,
+      "grad_norm": 0.2134984077840653,
+      "learning_rate": 2.540202501488982e-05,
+      "loss": 0.3497,
+      "step": 1014
+    },
+    {
+      "epoch": 1.629718875502008,
+      "grad_norm": 0.1952331371500388,
+      "learning_rate": 2.537224538415724e-05,
+      "loss": 0.3727,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6313253012048192,
+      "grad_norm": 0.22469385469458966,
+      "learning_rate": 2.534246575342466e-05,
+      "loss": 0.3511,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6329317269076307,
+      "grad_norm": 0.19794746408070624,
+      "learning_rate": 2.531268612269208e-05,
+      "loss": 0.3779,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6345381526104417,
+      "grad_norm": 0.19967397274850174,
+      "learning_rate": 2.5282906491959502e-05,
+      "loss": 0.3747,
+      "step": 1018
+    },
+    {
+      "epoch": 1.636144578313253,
+      "grad_norm": 0.22538749173025088,
+      "learning_rate": 2.5253126861226924e-05,
+      "loss": 0.352,
+      "step": 1019
+    },
+    {
+      "epoch": 1.637751004016064,
+      "grad_norm": 0.22398508134124584,
+      "learning_rate": 2.5223347230494343e-05,
+      "loss": 0.3625,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6393574297188755,
+      "grad_norm": 0.1883127548632391,
+      "learning_rate": 2.5193567599761765e-05,
+      "loss": 0.3322,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6409638554216868,
+      "grad_norm": 0.20492467820901458,
+      "learning_rate": 2.5163787969029184e-05,
+      "loss": 0.3416,
+      "step": 1022
+    },
+    {
+      "epoch": 1.642570281124498,
+      "grad_norm": 0.19950205181583336,
+      "learning_rate": 2.5134008338296606e-05,
+      "loss": 0.356,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6441767068273092,
+      "grad_norm": 0.20706482851814978,
+      "learning_rate": 2.5104228707564025e-05,
+      "loss": 0.3681,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6457831325301204,
+      "grad_norm": 0.1960488866576575,
+      "learning_rate": 2.5074449076831447e-05,
+      "loss": 0.3611,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6473895582329319,
+      "grad_norm": 0.21928337797665387,
+      "learning_rate": 2.5044669446098866e-05,
+      "loss": 0.354,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6489959839357429,
+      "grad_norm": 0.2049749443964654,
+      "learning_rate": 2.501488981536629e-05,
+      "loss": 0.3483,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6506024096385543,
+      "grad_norm": 0.21975575711160597,
+      "learning_rate": 2.498511018463371e-05,
+      "loss": 0.3617,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6522088353413653,
+      "grad_norm": 0.25435722039840947,
+      "learning_rate": 2.4955330553901133e-05,
+      "loss": 0.3595,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6538152610441768,
+      "grad_norm": 0.21053271933640544,
+      "learning_rate": 2.4925550923168552e-05,
+      "loss": 0.3538,
+      "step": 1030
+    },
+    {
+      "epoch": 1.655421686746988,
+      "grad_norm": 0.21092960698173877,
+      "learning_rate": 2.4895771292435974e-05,
+      "loss": 0.3708,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6570281124497992,
+      "grad_norm": 0.25037893658218124,
+      "learning_rate": 2.4865991661703396e-05,
+      "loss": 0.3572,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6586345381526104,
+      "grad_norm": 0.21614435230428244,
+      "learning_rate": 2.483621203097082e-05,
+      "loss": 0.3616,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6602409638554216,
+      "grad_norm": 0.21762495218791192,
+      "learning_rate": 2.4806432400238237e-05,
+      "loss": 0.3541,
+      "step": 1034
+    },
+    {
+      "epoch": 1.661847389558233,
+      "grad_norm": 0.21227742351106402,
+      "learning_rate": 2.477665276950566e-05,
+      "loss": 0.3489,
+      "step": 1035
+    },
+    {
+      "epoch": 1.663453815261044,
+      "grad_norm": 0.22051322792778752,
+      "learning_rate": 2.474687313877308e-05,
+      "loss": 0.3788,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6650602409638555,
+      "grad_norm": 0.1967781801087776,
+      "learning_rate": 2.47170935080405e-05,
+      "loss": 0.345,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.21567566845974007,
+      "learning_rate": 2.4687313877307923e-05,
+      "loss": 0.3706,
+      "step": 1038
+    },
+    {
+      "epoch": 1.668273092369478,
+      "grad_norm": 0.2056270298713853,
+      "learning_rate": 2.4657534246575342e-05,
+      "loss": 0.3504,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6698795180722892,
+      "grad_norm": 0.21004166292669,
+      "learning_rate": 2.4627754615842767e-05,
+      "loss": 0.3633,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6714859437751004,
+      "grad_norm": 0.21043296463452388,
+      "learning_rate": 2.4597974985110186e-05,
+      "loss": 0.3668,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6730923694779116,
+      "grad_norm": 0.21326051661342343,
+      "learning_rate": 2.456819535437761e-05,
+      "loss": 0.3598,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6746987951807228,
+      "grad_norm": 0.21795643765282252,
+      "learning_rate": 2.4538415723645027e-05,
+      "loss": 0.3798,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6763052208835343,
+      "grad_norm": 0.18662735508569567,
+      "learning_rate": 2.450863609291245e-05,
+      "loss": 0.344,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6779116465863453,
+      "grad_norm": 0.18983173307642734,
+      "learning_rate": 2.447885646217987e-05,
+      "loss": 0.3537,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6795180722891567,
+      "grad_norm": 0.20726848993601318,
+      "learning_rate": 2.444907683144729e-05,
+      "loss": 0.3826,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6811244979919677,
+      "grad_norm": 0.20680344779356397,
+      "learning_rate": 2.4419297200714713e-05,
+      "loss": 0.3468,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6827309236947792,
+      "grad_norm": 0.21615630603434186,
+      "learning_rate": 2.4389517569982135e-05,
+      "loss": 0.3592,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6843373493975904,
+      "grad_norm": 0.1952368775804799,
+      "learning_rate": 2.4359737939249554e-05,
+      "loss": 0.3604,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6859437751004016,
+      "grad_norm": 0.2100215577635232,
+      "learning_rate": 2.4329958308516976e-05,
+      "loss": 0.3589,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6875502008032128,
+      "grad_norm": 0.21355439868810364,
+      "learning_rate": 2.4300178677784395e-05,
+      "loss": 0.3501,
+      "step": 1051
+    },
+    {
+      "epoch": 1.689156626506024,
+      "grad_norm": 0.23457242221702107,
+      "learning_rate": 2.4270399047051817e-05,
+      "loss": 0.3713,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6907630522088355,
+      "grad_norm": 0.1965153502918042,
+      "learning_rate": 2.4240619416319236e-05,
+      "loss": 0.3668,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6923694779116465,
+      "grad_norm": 0.22811037493361058,
+      "learning_rate": 2.421083978558666e-05,
+      "loss": 0.3588,
+      "step": 1054
+    },
+    {
+      "epoch": 1.693975903614458,
+      "grad_norm": 0.19865555229011633,
+      "learning_rate": 2.418106015485408e-05,
+      "loss": 0.3601,
+      "step": 1055
+    },
+    {
+      "epoch": 1.695582329317269,
+      "grad_norm": 0.22658627459573066,
+      "learning_rate": 2.4151280524121503e-05,
+      "loss": 0.3523,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6971887550200804,
+      "grad_norm": 0.23275848437999672,
+      "learning_rate": 2.4121500893388925e-05,
+      "loss": 0.3554,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6987951807228916,
+      "grad_norm": 0.22113612954878686,
+      "learning_rate": 2.4091721262656344e-05,
+      "loss": 0.3725,
+      "step": 1058
+    },
+    {
+      "epoch": 1.7004016064257028,
+      "grad_norm": 0.224473754437892,
+      "learning_rate": 2.4061941631923766e-05,
+      "loss": 0.3776,
+      "step": 1059
+    },
+    {
+      "epoch": 1.702008032128514,
+      "grad_norm": 0.2295222836812493,
+      "learning_rate": 2.4032162001191185e-05,
+      "loss": 0.3581,
+      "step": 1060
+    },
+    {
+      "epoch": 1.7036144578313253,
+      "grad_norm": 0.24314433625472318,
+      "learning_rate": 2.4002382370458608e-05,
+      "loss": 0.3568,
+      "step": 1061
+    },
+    {
+      "epoch": 1.7052208835341367,
+      "grad_norm": 0.34175260472943086,
+      "learning_rate": 2.3972602739726026e-05,
+      "loss": 0.3416,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7068273092369477,
+      "grad_norm": 0.2514460521674827,
+      "learning_rate": 2.3942823108993452e-05,
+      "loss": 0.3724,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7084337349397591,
+      "grad_norm": 0.2276723687182961,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 0.3665,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7100401606425701,
+      "grad_norm": 0.21923249265172579,
+      "learning_rate": 2.3883263847528293e-05,
+      "loss": 0.356,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7116465863453816,
+      "grad_norm": 0.20314212509455468,
+      "learning_rate": 2.3853484216795712e-05,
+      "loss": 0.3619,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7132530120481928,
+      "grad_norm": 0.22203074761564148,
+      "learning_rate": 2.3823704586063134e-05,
+      "loss": 0.3657,
+      "step": 1067
+    },
+    {
+      "epoch": 1.714859437751004,
+      "grad_norm": 0.20207033700363516,
+      "learning_rate": 2.3793924955330553e-05,
+      "loss": 0.3589,
+      "step": 1068
+    },
+    {
+      "epoch": 1.7164658634538152,
+      "grad_norm": 0.2209021600211984,
+      "learning_rate": 2.3764145324597975e-05,
+      "loss": 0.369,
+      "step": 1069
+    },
+    {
+      "epoch": 1.7180722891566265,
+      "grad_norm": 0.20768196432692904,
+      "learning_rate": 2.3734365693865398e-05,
+      "loss": 0.3643,
+      "step": 1070
+    },
+    {
+      "epoch": 1.719678714859438,
+      "grad_norm": 0.2098995668547178,
+      "learning_rate": 2.370458606313282e-05,
+      "loss": 0.3578,
+      "step": 1071
+    },
+    {
+      "epoch": 1.721285140562249,
+      "grad_norm": 0.20500363490217155,
+      "learning_rate": 2.367480643240024e-05,
+      "loss": 0.3455,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7228915662650603,
+      "grad_norm": 0.20815589579222898,
+      "learning_rate": 2.364502680166766e-05,
+      "loss": 0.3551,
+      "step": 1073
+    },
+    {
+      "epoch": 1.7244979919678713,
+      "grad_norm": 0.218238069485204,
+      "learning_rate": 2.361524717093508e-05,
+      "loss": 0.3572,
+      "step": 1074
+    },
+    {
+      "epoch": 1.7261044176706828,
+      "grad_norm": 0.21261831858723657,
+      "learning_rate": 2.3585467540202502e-05,
+      "loss": 0.3611,
+      "step": 1075
+    },
+    {
+      "epoch": 1.727710843373494,
+      "grad_norm": 0.218199623953151,
+      "learning_rate": 2.3555687909469924e-05,
+      "loss": 0.3732,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7293172690763052,
+      "grad_norm": 0.21077091020641125,
+      "learning_rate": 2.3525908278737343e-05,
+      "loss": 0.3589,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7309236947791165,
+      "grad_norm": 0.20811140385555477,
+      "learning_rate": 2.349612864800477e-05,
+      "loss": 0.3499,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7325301204819277,
+      "grad_norm": 0.224004592451593,
+      "learning_rate": 2.3466349017272188e-05,
+      "loss": 0.3656,
+      "step": 1079
+    },
+    {
+      "epoch": 1.7341365461847391,
+      "grad_norm": 0.20937722059078773,
+      "learning_rate": 2.343656938653961e-05,
+      "loss": 0.3875,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7357429718875501,
+      "grad_norm": 0.19918778480602867,
+      "learning_rate": 2.340678975580703e-05,
+      "loss": 0.3566,
+      "step": 1081
+    },
+    {
+      "epoch": 1.7373493975903616,
+      "grad_norm": 0.21098302310101957,
+      "learning_rate": 2.337701012507445e-05,
+      "loss": 0.3716,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7389558232931726,
+      "grad_norm": 0.2093261199980322,
+      "learning_rate": 2.334723049434187e-05,
+      "loss": 0.3576,
+      "step": 1083
+    },
+    {
+      "epoch": 1.740562248995984,
+      "grad_norm": 0.1970208286024692,
+      "learning_rate": 2.3317450863609292e-05,
+      "loss": 0.3714,
+      "step": 1084
+    },
+    {
+      "epoch": 1.7421686746987952,
+      "grad_norm": 0.2016557441886884,
+      "learning_rate": 2.328767123287671e-05,
+      "loss": 0.3655,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7437751004016064,
+      "grad_norm": 0.21169338027090684,
+      "learning_rate": 2.3257891602144137e-05,
+      "loss": 0.3483,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7453815261044177,
+      "grad_norm": 0.20200737567041832,
+      "learning_rate": 2.3228111971411555e-05,
+      "loss": 0.3745,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7469879518072289,
+      "grad_norm": 0.22578155521015078,
+      "learning_rate": 2.3198332340678978e-05,
+      "loss": 0.3745,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7485943775100403,
+      "grad_norm": 0.19093185238889984,
+      "learning_rate": 2.3168552709946397e-05,
+      "loss": 0.3627,
+      "step": 1089
+    },
+    {
+      "epoch": 1.7502008032128513,
+      "grad_norm": 0.25205916180047044,
+      "learning_rate": 2.313877307921382e-05,
+      "loss": 0.3631,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7518072289156628,
+      "grad_norm": 0.20141539644570866,
+      "learning_rate": 2.3108993448481238e-05,
+      "loss": 0.3684,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7534136546184738,
+      "grad_norm": 0.21474090300526755,
+      "learning_rate": 2.307921381774866e-05,
+      "loss": 0.3624,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7550200803212852,
+      "grad_norm": 0.2054655600297508,
+      "learning_rate": 2.3049434187016082e-05,
+      "loss": 0.3553,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7566265060240964,
+      "grad_norm": 0.20955065103071374,
+      "learning_rate": 2.3019654556283504e-05,
+      "loss": 0.3633,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7582329317269076,
+      "grad_norm": 0.19589875551410302,
+      "learning_rate": 2.2989874925550923e-05,
+      "loss": 0.3289,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7598393574297189,
+      "grad_norm": 0.2361099076001025,
+      "learning_rate": 2.2960095294818345e-05,
+      "loss": 0.353,
+      "step": 1096
+    },
+    {
+      "epoch": 1.76144578313253,
+      "grad_norm": 0.21498956022370994,
+      "learning_rate": 2.2930315664085768e-05,
+      "loss": 0.3828,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7630522088353415,
+      "grad_norm": 0.21650382135153187,
+      "learning_rate": 2.2900536033353187e-05,
+      "loss": 0.368,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7646586345381525,
+      "grad_norm": 0.20893736406148616,
+      "learning_rate": 2.287075640262061e-05,
+      "loss": 0.3539,
+      "step": 1099
+    },
+    {
+      "epoch": 1.766265060240964,
+      "grad_norm": 0.23082011015119872,
+      "learning_rate": 2.2840976771888028e-05,
+      "loss": 0.3494,
+      "step": 1100
+    },
+    {
+      "epoch": 1.767871485943775,
+      "grad_norm": 0.19694576672664207,
+      "learning_rate": 2.2811197141155453e-05,
+      "loss": 0.3629,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7694779116465864,
+      "grad_norm": 0.20193978071730798,
+      "learning_rate": 2.2781417510422872e-05,
+      "loss": 0.3859,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7710843373493976,
+      "grad_norm": 0.23092561419314273,
+      "learning_rate": 2.2751637879690294e-05,
+      "loss": 0.3516,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7726907630522089,
+      "grad_norm": 0.20832654839220274,
+      "learning_rate": 2.2721858248957713e-05,
+      "loss": 0.3478,
+      "step": 1104
+    },
+    {
+      "epoch": 1.77429718875502,
+      "grad_norm": 0.22535285467644003,
+      "learning_rate": 2.2692078618225136e-05,
+      "loss": 0.3649,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7759036144578313,
+      "grad_norm": 0.20726821679440688,
+      "learning_rate": 2.2662298987492554e-05,
+      "loss": 0.3571,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7775100401606427,
+      "grad_norm": 0.22444597601238636,
+      "learning_rate": 2.2632519356759977e-05,
+      "loss": 0.3645,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7791164658634537,
+      "grad_norm": 0.2009713091668314,
+      "learning_rate": 2.2602739726027396e-05,
+      "loss": 0.3519,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7807228915662652,
+      "grad_norm": 0.21711474490245392,
+      "learning_rate": 2.257296009529482e-05,
+      "loss": 0.3583,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7823293172690762,
+      "grad_norm": 0.22516044847286923,
+      "learning_rate": 2.254318046456224e-05,
+      "loss": 0.3799,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7839357429718876,
+      "grad_norm": 0.2073918693034101,
+      "learning_rate": 2.2513400833829662e-05,
+      "loss": 0.363,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7855421686746988,
+      "grad_norm": 0.22634597043012206,
+      "learning_rate": 2.248362120309708e-05,
+      "loss": 0.3535,
+      "step": 1112
+    },
+    {
+      "epoch": 1.78714859437751,
+      "grad_norm": 0.20760649450417196,
+      "learning_rate": 2.2453841572364503e-05,
+      "loss": 0.3762,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7887550200803213,
+      "grad_norm": 0.24727436296116567,
+      "learning_rate": 2.2424061941631926e-05,
+      "loss": 0.3698,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7903614457831325,
+      "grad_norm": 0.19266851010173805,
+      "learning_rate": 2.2394282310899344e-05,
+      "loss": 0.3444,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7919678714859437,
+      "grad_norm": 0.2211464369383969,
+      "learning_rate": 2.2364502680166767e-05,
+      "loss": 0.3663,
+      "step": 1116
+    },
+    {
+      "epoch": 1.793574297188755,
+      "grad_norm": 0.20631058168368754,
+      "learning_rate": 2.233472304943419e-05,
+      "loss": 0.3519,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7951807228915664,
+      "grad_norm": 0.19331297422580224,
+      "learning_rate": 2.230494341870161e-05,
+      "loss": 0.3652,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7967871485943774,
+      "grad_norm": 0.2200641847902077,
+      "learning_rate": 2.227516378796903e-05,
+      "loss": 0.3621,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7983935742971888,
+      "grad_norm": 0.20279217114814665,
+      "learning_rate": 2.2245384157236452e-05,
+      "loss": 0.3801,
+      "step": 1120
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.21870788059156782,
+      "learning_rate": 2.221560452650387e-05,
+      "loss": 0.3457,
+      "step": 1121
+    },
+    {
+      "epoch": 1.8016064257028113,
+      "grad_norm": 0.21189789954147503,
+      "learning_rate": 2.2185824895771293e-05,
+      "loss": 0.3564,
+      "step": 1122
+    },
+    {
+      "epoch": 1.8032128514056225,
+      "grad_norm": 0.20086799106080055,
+      "learning_rate": 2.2156045265038712e-05,
+      "loss": 0.3733,
+      "step": 1123
+    },
+    {
+      "epoch": 1.8048192771084337,
+      "grad_norm": 0.1986048299225642,
+      "learning_rate": 2.2126265634306138e-05,
+      "loss": 0.3671,
+      "step": 1124
+    },
+    {
+      "epoch": 1.806425702811245,
+      "grad_norm": 0.2083000604689775,
+      "learning_rate": 2.2096486003573557e-05,
+      "loss": 0.3455,
+      "step": 1125
+    },
+    {
+      "epoch": 1.8080321285140561,
+      "grad_norm": 0.19057526404306752,
+      "learning_rate": 2.206670637284098e-05,
+      "loss": 0.3665,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8096385542168676,
+      "grad_norm": 0.21925743385948493,
+      "learning_rate": 2.2036926742108398e-05,
+      "loss": 0.3673,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8112449799196786,
+      "grad_norm": 0.2068789882250303,
+      "learning_rate": 2.200714711137582e-05,
+      "loss": 0.3459,
+      "step": 1128
+    },
+    {
+      "epoch": 1.81285140562249,
+      "grad_norm": 0.19327520335944015,
+      "learning_rate": 2.197736748064324e-05,
+      "loss": 0.3659,
+      "step": 1129
+    },
+    {
+      "epoch": 1.8144578313253013,
+      "grad_norm": 0.2078167143831951,
+      "learning_rate": 2.194758784991066e-05,
+      "loss": 0.3727,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8160642570281125,
+      "grad_norm": 0.18858502029684826,
+      "learning_rate": 2.1917808219178083e-05,
+      "loss": 0.3556,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8176706827309237,
+      "grad_norm": 0.21278063382666992,
+      "learning_rate": 2.1888028588445506e-05,
+      "loss": 0.3568,
+      "step": 1132
+    },
+    {
+      "epoch": 1.819277108433735,
+      "grad_norm": 0.2180291628817106,
+      "learning_rate": 2.1858248957712925e-05,
+      "loss": 0.3643,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8208835341365461,
+      "grad_norm": 0.18948521028304013,
+      "learning_rate": 2.1828469326980347e-05,
+      "loss": 0.3528,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8224899598393574,
+      "grad_norm": 0.20853568994549157,
+      "learning_rate": 2.179868969624777e-05,
+      "loss": 0.3467,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8240963855421688,
+      "grad_norm": 0.19566677865636536,
+      "learning_rate": 2.1768910065515188e-05,
+      "loss": 0.3732,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8257028112449798,
+      "grad_norm": 0.19445461615137893,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.3488,
+      "step": 1137
+    },
+    {
+      "epoch": 1.8273092369477912,
+      "grad_norm": 0.19624375302208613,
+      "learning_rate": 2.170935080405003e-05,
+      "loss": 0.3624,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8289156626506025,
+      "grad_norm": 0.2040335887516193,
+      "learning_rate": 2.167957117331745e-05,
+      "loss": 0.3729,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8305220883534137,
+      "grad_norm": 0.20102650556518836,
+      "learning_rate": 2.1649791542584874e-05,
+      "loss": 0.3466,
+      "step": 1140
+    },
+    {
+      "epoch": 1.832128514056225,
+      "grad_norm": 0.21005309955955231,
+      "learning_rate": 2.1620011911852296e-05,
+      "loss": 0.3632,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8337349397590361,
+      "grad_norm": 0.20305833408054538,
+      "learning_rate": 2.1590232281119715e-05,
+      "loss": 0.3534,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8353413654618473,
+      "grad_norm": 0.20124327635765527,
+      "learning_rate": 2.1560452650387137e-05,
+      "loss": 0.3466,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8369477911646586,
+      "grad_norm": 0.1832866917039283,
+      "learning_rate": 2.1530673019654556e-05,
+      "loss": 0.3716,
+      "step": 1144
+    },
+    {
+      "epoch": 1.83855421686747,
+      "grad_norm": 0.20990146281914185,
+      "learning_rate": 2.1500893388921978e-05,
+      "loss": 0.3616,
+      "step": 1145
+    },
+    {
+      "epoch": 1.840160642570281,
+      "grad_norm": 0.20672268930674145,
+      "learning_rate": 2.1471113758189397e-05,
+      "loss": 0.3608,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8417670682730924,
+      "grad_norm": 0.2107260397722923,
+      "learning_rate": 2.1441334127456822e-05,
+      "loss": 0.3706,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8433734939759037,
+      "grad_norm": 0.19672011468893483,
+      "learning_rate": 2.141155449672424e-05,
+      "loss": 0.3596,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8449799196787149,
+      "grad_norm": 0.1926451723956918,
+      "learning_rate": 2.1381774865991664e-05,
+      "loss": 0.3772,
+      "step": 1149
+    },
+    {
+      "epoch": 1.846586345381526,
+      "grad_norm": 0.21932150933370484,
+      "learning_rate": 2.1351995235259082e-05,
+      "loss": 0.3532,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8481927710843373,
+      "grad_norm": 0.23147096635393158,
+      "learning_rate": 2.1322215604526505e-05,
+      "loss": 0.3664,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8497991967871485,
+      "grad_norm": 0.19934971134196694,
+      "learning_rate": 2.1292435973793927e-05,
+      "loss": 0.3711,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8514056224899598,
+      "grad_norm": 0.2272191026129902,
+      "learning_rate": 2.1262656343061346e-05,
+      "loss": 0.3504,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8530120481927712,
+      "grad_norm": 0.21777434254774575,
+      "learning_rate": 2.1232876712328768e-05,
+      "loss": 0.3498,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8546184738955822,
+      "grad_norm": 0.2124750716341292,
+      "learning_rate": 2.120309708159619e-05,
+      "loss": 0.3522,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8562248995983937,
+      "grad_norm": 0.22228255865646168,
+      "learning_rate": 2.1173317450863613e-05,
+      "loss": 0.3578,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8578313253012049,
+      "grad_norm": 0.22231705030783963,
+      "learning_rate": 2.114353782013103e-05,
+      "loss": 0.3394,
+      "step": 1157
+    },
+    {
+      "epoch": 1.859437751004016,
+      "grad_norm": 0.22285029234412623,
+      "learning_rate": 2.1113758189398454e-05,
+      "loss": 0.3584,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8610441767068273,
+      "grad_norm": 0.20482179387652538,
+      "learning_rate": 2.1083978558665873e-05,
+      "loss": 0.3525,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8626506024096385,
+      "grad_norm": 0.24264556011157912,
+      "learning_rate": 2.1054198927933295e-05,
+      "loss": 0.3688,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8642570281124498,
+      "grad_norm": 0.1957672647540786,
+      "learning_rate": 2.1024419297200714e-05,
+      "loss": 0.3363,
+      "step": 1161
+    },
+    {
+      "epoch": 1.865863453815261,
+      "grad_norm": 0.2108725681692963,
+      "learning_rate": 2.0994639666468136e-05,
+      "loss": 0.3664,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8674698795180724,
+      "grad_norm": 0.21125049983033603,
+      "learning_rate": 2.0964860035735558e-05,
+      "loss": 0.3728,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8690763052208834,
+      "grad_norm": 0.21069255508441262,
+      "learning_rate": 2.093508040500298e-05,
+      "loss": 0.3624,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8706827309236949,
+      "grad_norm": 0.20365274655393906,
+      "learning_rate": 2.09053007742704e-05,
+      "loss": 0.3452,
+      "step": 1165
+    },
+    {
+      "epoch": 1.872289156626506,
+      "grad_norm": 0.2195954631278454,
+      "learning_rate": 2.087552114353782e-05,
+      "loss": 0.3615,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8738955823293173,
+      "grad_norm": 0.21517553233692346,
+      "learning_rate": 2.084574151280524e-05,
+      "loss": 0.3608,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8755020080321285,
+      "grad_norm": 0.20756181473622315,
+      "learning_rate": 2.0815961882072663e-05,
+      "loss": 0.3512,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8771084337349397,
+      "grad_norm": 0.22271333786027447,
+      "learning_rate": 2.0786182251340085e-05,
+      "loss": 0.3497,
+      "step": 1169
+    },
+    {
+      "epoch": 1.878714859437751,
+      "grad_norm": 0.21853129021119494,
+      "learning_rate": 2.0756402620607507e-05,
+      "loss": 0.3574,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8803212851405622,
+      "grad_norm": 0.21585643665601728,
+      "learning_rate": 2.0726622989874926e-05,
+      "loss": 0.3594,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8819277108433736,
+      "grad_norm": 0.20331060293257677,
+      "learning_rate": 2.0696843359142348e-05,
+      "loss": 0.3666,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8835341365461846,
+      "grad_norm": 0.2565749022088396,
+      "learning_rate": 2.066706372840977e-05,
+      "loss": 0.3526,
+      "step": 1173
+    },
+    {
+      "epoch": 1.885140562248996,
+      "grad_norm": 0.20758070527927017,
+      "learning_rate": 2.063728409767719e-05,
+      "loss": 0.3827,
+      "step": 1174
+    },
+    {
+      "epoch": 1.886746987951807,
+      "grad_norm": 0.22210523230297302,
+      "learning_rate": 2.060750446694461e-05,
+      "loss": 0.3588,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8883534136546185,
+      "grad_norm": 0.21091506966574514,
+      "learning_rate": 2.057772483621203e-05,
+      "loss": 0.3531,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8899598393574297,
+      "grad_norm": 0.20149717266780776,
+      "learning_rate": 2.0547945205479453e-05,
+      "loss": 0.3576,
+      "step": 1177
+    },
+    {
+      "epoch": 1.891566265060241,
+      "grad_norm": 0.22621131534500763,
+      "learning_rate": 2.0518165574746875e-05,
+      "loss": 0.3751,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8931726907630522,
+      "grad_norm": 0.20613955089295916,
+      "learning_rate": 2.0488385944014297e-05,
+      "loss": 0.3589,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8947791164658634,
+      "grad_norm": 0.19083016481113835,
+      "learning_rate": 2.0458606313281716e-05,
+      "loss": 0.3426,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8963855421686748,
+      "grad_norm": 0.22126022453936145,
+      "learning_rate": 2.0428826682549138e-05,
+      "loss": 0.3505,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8979919678714858,
+      "grad_norm": 0.20323408257329473,
+      "learning_rate": 2.0399047051816557e-05,
+      "loss": 0.3469,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8995983935742973,
+      "grad_norm": 0.20039417903514725,
+      "learning_rate": 2.036926742108398e-05,
+      "loss": 0.3618,
+      "step": 1183
+    },
+    {
+      "epoch": 1.9012048192771083,
+      "grad_norm": 0.21296230999792612,
+      "learning_rate": 2.0339487790351398e-05,
+      "loss": 0.3583,
+      "step": 1184
+    },
+    {
+      "epoch": 1.9028112449799197,
+      "grad_norm": 0.20265904210834976,
+      "learning_rate": 2.030970815961882e-05,
+      "loss": 0.3613,
+      "step": 1185
+    },
+    {
+      "epoch": 1.904417670682731,
+      "grad_norm": 0.21826476667287842,
+      "learning_rate": 2.0279928528886243e-05,
+      "loss": 0.3663,
+      "step": 1186
+    },
+    {
+      "epoch": 1.9060240963855422,
+      "grad_norm": 0.22207635068754514,
+      "learning_rate": 2.0250148898153665e-05,
+      "loss": 0.3482,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9076305220883534,
+      "grad_norm": 0.22381259846748205,
+      "learning_rate": 2.0220369267421084e-05,
+      "loss": 0.3667,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9092369477911646,
+      "grad_norm": 0.19895299553105997,
+      "learning_rate": 2.0190589636688506e-05,
+      "loss": 0.3443,
+      "step": 1189
+    },
+    {
+      "epoch": 1.910843373493976,
+      "grad_norm": 0.21122074307230462,
+      "learning_rate": 2.0160810005955928e-05,
+      "loss": 0.366,
+      "step": 1190
+    },
+    {
+      "epoch": 1.912449799196787,
+      "grad_norm": 0.21887960654596514,
+      "learning_rate": 2.0131030375223347e-05,
+      "loss": 0.3553,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9140562248995985,
+      "grad_norm": 0.21709717023702313,
+      "learning_rate": 2.010125074449077e-05,
+      "loss": 0.3556,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9156626506024095,
+      "grad_norm": 0.2170193651130133,
+      "learning_rate": 2.0071471113758188e-05,
+      "loss": 0.3674,
+      "step": 1193
+    },
+    {
+      "epoch": 1.917269076305221,
+      "grad_norm": 0.2029081551011821,
+      "learning_rate": 2.0041691483025614e-05,
+      "loss": 0.3549,
+      "step": 1194
+    },
+    {
+      "epoch": 1.9188755020080321,
+      "grad_norm": 0.19745275159480208,
+      "learning_rate": 2.0011911852293033e-05,
+      "loss": 0.359,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9204819277108434,
+      "grad_norm": 0.19719600657200462,
+      "learning_rate": 1.9982132221560455e-05,
+      "loss": 0.3481,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9220883534136546,
+      "grad_norm": 0.20721127133510045,
+      "learning_rate": 1.9952352590827874e-05,
+      "loss": 0.3832,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9236947791164658,
+      "grad_norm": 0.1956761467782788,
+      "learning_rate": 1.9922572960095296e-05,
+      "loss": 0.3506,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9253012048192772,
+      "grad_norm": 0.19780280846955045,
+      "learning_rate": 1.9892793329362715e-05,
+      "loss": 0.3578,
+      "step": 1199
+    },
+    {
+      "epoch": 1.9269076305220882,
+      "grad_norm": 0.20566585919858663,
+      "learning_rate": 1.9863013698630137e-05,
+      "loss": 0.364,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9285140562248997,
+      "grad_norm": 0.1991383339452239,
+      "learning_rate": 1.983323406789756e-05,
+      "loss": 0.3541,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9301204819277107,
+      "grad_norm": 0.2155869647709782,
+      "learning_rate": 1.980345443716498e-05,
+      "loss": 0.3441,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9317269076305221,
+      "grad_norm": 0.2047880536730749,
+      "learning_rate": 1.97736748064324e-05,
+      "loss": 0.3601,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 0.20606661706770807,
+      "learning_rate": 1.9743895175699823e-05,
+      "loss": 0.3643,
+      "step": 1204
+    },
+    {
+      "epoch": 1.9349397590361446,
+      "grad_norm": 0.19908190373445847,
+      "learning_rate": 1.971411554496724e-05,
+      "loss": 0.3736,
+      "step": 1205
+    },
+    {
+      "epoch": 1.9365461847389558,
+      "grad_norm": 0.2308674494926228,
+      "learning_rate": 1.9684335914234664e-05,
+      "loss": 0.3631,
+      "step": 1206
+    },
+    {
+      "epoch": 1.938152610441767,
+      "grad_norm": 0.2066133059057739,
+      "learning_rate": 1.9654556283502086e-05,
+      "loss": 0.356,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9397590361445785,
+      "grad_norm": 0.20384597819015435,
+      "learning_rate": 1.9624776652769505e-05,
+      "loss": 0.3571,
+      "step": 1208
+    },
+    {
+      "epoch": 1.9413654618473895,
+      "grad_norm": 0.2347153369190148,
+      "learning_rate": 1.9594997022036927e-05,
+      "loss": 0.35,
+      "step": 1209
+    },
+    {
+      "epoch": 1.942971887550201,
+      "grad_norm": 0.20581681587401848,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.346,
+      "step": 1210
+    },
+    {
+      "epoch": 1.944578313253012,
+      "grad_norm": 0.21019457597339639,
+      "learning_rate": 1.9535437760571772e-05,
+      "loss": 0.3546,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9461847389558233,
+      "grad_norm": 0.20152245518746795,
+      "learning_rate": 1.950565812983919e-05,
+      "loss": 0.3314,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9477911646586346,
+      "grad_norm": 0.22313459443419506,
+      "learning_rate": 1.9475878499106613e-05,
+      "loss": 0.3433,
+      "step": 1213
+    },
+    {
+      "epoch": 1.9493975903614458,
+      "grad_norm": 0.20031556726002392,
+      "learning_rate": 1.9446098868374032e-05,
+      "loss": 0.3699,
+      "step": 1214
+    },
+    {
+      "epoch": 1.951004016064257,
+      "grad_norm": 0.21068240359744894,
+      "learning_rate": 1.9416319237641454e-05,
+      "loss": 0.3717,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9526104417670682,
+      "grad_norm": 0.18049575736825463,
+      "learning_rate": 1.9386539606908873e-05,
+      "loss": 0.3355,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9542168674698797,
+      "grad_norm": 0.19194566457192386,
+      "learning_rate": 1.93567599761763e-05,
+      "loss": 0.3523,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9558232931726907,
+      "grad_norm": 0.21603061394583895,
+      "learning_rate": 1.9326980345443717e-05,
+      "loss": 0.3438,
+      "step": 1218
+    },
+    {
+      "epoch": 1.957429718875502,
+      "grad_norm": 0.20014961841231446,
+      "learning_rate": 1.929720071471114e-05,
+      "loss": 0.3375,
+      "step": 1219
+    },
+    {
+      "epoch": 1.959036144578313,
+      "grad_norm": 0.17999062215530368,
+      "learning_rate": 1.926742108397856e-05,
+      "loss": 0.3636,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9606425702811245,
+      "grad_norm": 0.20451266412122704,
+      "learning_rate": 1.923764145324598e-05,
+      "loss": 0.3569,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9622489959839358,
+      "grad_norm": 0.204885341355471,
+      "learning_rate": 1.92078618225134e-05,
+      "loss": 0.3664,
+      "step": 1222
+    },
+    {
+      "epoch": 1.963855421686747,
+      "grad_norm": 0.20325761883578033,
+      "learning_rate": 1.9178082191780822e-05,
+      "loss": 0.3653,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9654618473895582,
+      "grad_norm": 0.19011063209861206,
+      "learning_rate": 1.9148302561048244e-05,
+      "loss": 0.3452,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9670682730923694,
+      "grad_norm": 0.19018854040935476,
+      "learning_rate": 1.9118522930315666e-05,
+      "loss": 0.3619,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9686746987951809,
+      "grad_norm": 0.20318943026957367,
+      "learning_rate": 1.9088743299583085e-05,
+      "loss": 0.3503,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9702811244979919,
+      "grad_norm": 0.2036147004934244,
+      "learning_rate": 1.9058963668850507e-05,
+      "loss": 0.3768,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9718875502008033,
+      "grad_norm": 0.2031461574425529,
+      "learning_rate": 1.902918403811793e-05,
+      "loss": 0.3607,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9734939759036143,
+      "grad_norm": 0.20194319934341554,
+      "learning_rate": 1.899940440738535e-05,
+      "loss": 0.3552,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9751004016064257,
+      "grad_norm": 0.2159941288822771,
+      "learning_rate": 1.896962477665277e-05,
+      "loss": 0.3401,
+      "step": 1230
+    },
+    {
+      "epoch": 1.976706827309237,
+      "grad_norm": 0.19857621324762853,
+      "learning_rate": 1.893984514592019e-05,
+      "loss": 0.3759,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9783132530120482,
+      "grad_norm": 0.21210683481011638,
+      "learning_rate": 1.8910065515187615e-05,
+      "loss": 0.3661,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9799196787148594,
+      "grad_norm": 0.21464726089214603,
+      "learning_rate": 1.8880285884455034e-05,
+      "loss": 0.3749,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9815261044176706,
+      "grad_norm": 0.19853232290078882,
+      "learning_rate": 1.8850506253722456e-05,
+      "loss": 0.334,
+      "step": 1234
+    },
+    {
+      "epoch": 1.983132530120482,
+      "grad_norm": 0.20279986647636403,
+      "learning_rate": 1.8820726622989875e-05,
+      "loss": 0.3449,
+      "step": 1235
+    },
+    {
+      "epoch": 1.984738955823293,
+      "grad_norm": 0.2068529080085828,
+      "learning_rate": 1.8790946992257297e-05,
+      "loss": 0.3469,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9863453815261045,
+      "grad_norm": 0.19233002925201434,
+      "learning_rate": 1.8761167361524716e-05,
+      "loss": 0.3386,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9879518072289155,
+      "grad_norm": 0.20185567313273858,
+      "learning_rate": 1.873138773079214e-05,
+      "loss": 0.3432,
+      "step": 1238
+    },
+    {
+      "epoch": 1.989558232931727,
+      "grad_norm": 0.21079326889696912,
+      "learning_rate": 1.8701608100059557e-05,
+      "loss": 0.3734,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9911646586345382,
+      "grad_norm": 0.23227970805229742,
+      "learning_rate": 1.8671828469326983e-05,
+      "loss": 0.3526,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9927710843373494,
+      "grad_norm": 0.18597898236125074,
+      "learning_rate": 1.8642048838594402e-05,
+      "loss": 0.3483,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9943775100401606,
+      "grad_norm": 0.21314270191104023,
+      "learning_rate": 1.8612269207861824e-05,
+      "loss": 0.3784,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9959839357429718,
+      "grad_norm": 0.23543703444673417,
+      "learning_rate": 1.8582489577129243e-05,
+      "loss": 0.3406,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9975903614457833,
+      "grad_norm": 0.17715702141617948,
+      "learning_rate": 1.8552709946396665e-05,
+      "loss": 0.3423,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9991967871485943,
+      "grad_norm": 0.19746388647741975,
+      "learning_rate": 1.8522930315664087e-05,
+      "loss": 0.3691,
+      "step": 1245
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3174062639755838,
+      "learning_rate": 1.8493150684931506e-05,
+      "loss": 0.2983,
+      "step": 1246
+    },
+    {
+      "epoch": 2.0016064257028114,
+      "grad_norm": 0.23522152993607706,
+      "learning_rate": 1.846337105419893e-05,
+      "loss": 0.2727,
+      "step": 1247
+    },
+    {
+      "epoch": 2.0032128514056224,
+      "grad_norm": 0.20737054864646431,
+      "learning_rate": 1.843359142346635e-05,
+      "loss": 0.2767,
+      "step": 1248
+    },
+    {
+      "epoch": 2.004819277108434,
+      "grad_norm": 0.2400288684257797,
+      "learning_rate": 1.8403811792733773e-05,
+      "loss": 0.2795,
+      "step": 1249
+    },
+    {
+      "epoch": 2.006425702811245,
+      "grad_norm": 0.2630539207789394,
+      "learning_rate": 1.8374032162001192e-05,
+      "loss": 0.2916,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0080321285140563,
+      "grad_norm": 0.2201641599575612,
+      "learning_rate": 1.8344252531268614e-05,
+      "loss": 0.2813,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0096385542168673,
+      "grad_norm": 0.220685005990676,
+      "learning_rate": 1.8314472900536033e-05,
+      "loss": 0.2824,
+      "step": 1252
+    },
+    {
+      "epoch": 2.0112449799196788,
+      "grad_norm": 0.26390181703132115,
+      "learning_rate": 1.8284693269803455e-05,
+      "loss": 0.2861,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0128514056224898,
+      "grad_norm": 0.23610994889531656,
+      "learning_rate": 1.8254913639070874e-05,
+      "loss": 0.301,
+      "step": 1254
+    },
+    {
+      "epoch": 2.014457831325301,
+      "grad_norm": 0.2383546717776889,
+      "learning_rate": 1.82251340083383e-05,
+      "loss": 0.2908,
+      "step": 1255
+    },
+    {
+      "epoch": 2.0160642570281126,
+      "grad_norm": 1.7848307825699223,
+      "learning_rate": 1.819535437760572e-05,
+      "loss": 0.3335,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0176706827309236,
+      "grad_norm": 0.23696784503136437,
+      "learning_rate": 1.816557474687314e-05,
+      "loss": 0.2669,
+      "step": 1257
+    },
+    {
+      "epoch": 2.019277108433735,
+      "grad_norm": 0.21888481866636475,
+      "learning_rate": 1.813579511614056e-05,
+      "loss": 0.2909,
+      "step": 1258
+    },
+    {
+      "epoch": 2.020883534136546,
+      "grad_norm": 0.20332382254985962,
+      "learning_rate": 1.8106015485407982e-05,
+      "loss": 0.2782,
+      "step": 1259
+    },
+    {
+      "epoch": 2.0224899598393575,
+      "grad_norm": 0.20809561245158395,
+      "learning_rate": 1.80762358546754e-05,
+      "loss": 0.2951,
+      "step": 1260
+    },
+    {
+      "epoch": 2.0240963855421685,
+      "grad_norm": 0.21814457062033693,
+      "learning_rate": 1.8046456223942823e-05,
+      "loss": 0.2784,
+      "step": 1261
+    },
+    {
+      "epoch": 2.02570281124498,
+      "grad_norm": 0.21872054164235094,
+      "learning_rate": 1.8016676593210245e-05,
+      "loss": 0.2836,
+      "step": 1262
+    },
+    {
+      "epoch": 2.027309236947791,
+      "grad_norm": 0.19842979874754635,
+      "learning_rate": 1.7986896962477668e-05,
+      "loss": 0.2803,
+      "step": 1263
+    },
+    {
+      "epoch": 2.0289156626506024,
+      "grad_norm": 0.21881186415450113,
+      "learning_rate": 1.7957117331745086e-05,
+      "loss": 0.2881,
+      "step": 1264
+    },
+    {
+      "epoch": 2.030522088353414,
+      "grad_norm": 0.22443328960794542,
+      "learning_rate": 1.792733770101251e-05,
+      "loss": 0.2941,
+      "step": 1265
+    },
+    {
+      "epoch": 2.032128514056225,
+      "grad_norm": 0.21587260600025693,
+      "learning_rate": 1.789755807027993e-05,
+      "loss": 0.2757,
+      "step": 1266
+    },
+    {
+      "epoch": 2.0337349397590363,
+      "grad_norm": 0.20115770456663162,
+      "learning_rate": 1.786777843954735e-05,
+      "loss": 0.2787,
+      "step": 1267
+    },
+    {
+      "epoch": 2.0353413654618473,
+      "grad_norm": 0.2287670251960821,
+      "learning_rate": 1.7837998808814772e-05,
+      "loss": 0.2681,
+      "step": 1268
+    },
+    {
+      "epoch": 2.0369477911646587,
+      "grad_norm": 0.22109238751154517,
+      "learning_rate": 1.780821917808219e-05,
+      "loss": 0.2833,
+      "step": 1269
+    },
+    {
+      "epoch": 2.0385542168674697,
+      "grad_norm": 0.20007219850099878,
+      "learning_rate": 1.7778439547349613e-05,
+      "loss": 0.2862,
+      "step": 1270
+    },
+    {
+      "epoch": 2.040160642570281,
+      "grad_norm": 0.2161847201043952,
+      "learning_rate": 1.7748659916617035e-05,
+      "loss": 0.2616,
+      "step": 1271
+    },
+    {
+      "epoch": 2.041767068273092,
+      "grad_norm": 0.19586347077594904,
+      "learning_rate": 1.7718880285884458e-05,
+      "loss": 0.2744,
+      "step": 1272
+    },
+    {
+      "epoch": 2.0433734939759036,
+      "grad_norm": 0.1977339318614,
+      "learning_rate": 1.7689100655151877e-05,
+      "loss": 0.2709,
+      "step": 1273
+    },
+    {
+      "epoch": 2.044979919678715,
+      "grad_norm": 0.21405570213223185,
+      "learning_rate": 1.76593210244193e-05,
+      "loss": 0.2684,
+      "step": 1274
+    },
+    {
+      "epoch": 2.046586345381526,
+      "grad_norm": 0.19886670005088752,
+      "learning_rate": 1.7629541393686718e-05,
+      "loss": 0.2812,
+      "step": 1275
+    },
+    {
+      "epoch": 2.0481927710843375,
+      "grad_norm": 0.20121731499762072,
+      "learning_rate": 1.759976176295414e-05,
+      "loss": 0.2583,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0497991967871485,
+      "grad_norm": 0.22436613750052545,
+      "learning_rate": 1.756998213222156e-05,
+      "loss": 0.292,
+      "step": 1277
+    },
+    {
+      "epoch": 2.05140562248996,
+      "grad_norm": 0.2104802982109658,
+      "learning_rate": 1.7540202501488984e-05,
+      "loss": 0.2886,
+      "step": 1278
+    },
+    {
+      "epoch": 2.053012048192771,
+      "grad_norm": 0.1856838445294004,
+      "learning_rate": 1.7510422870756403e-05,
+      "loss": 0.2853,
+      "step": 1279
+    },
+    {
+      "epoch": 2.0546184738955824,
+      "grad_norm": 0.23394464779780705,
+      "learning_rate": 1.7480643240023825e-05,
+      "loss": 0.2841,
+      "step": 1280
+    },
+    {
+      "epoch": 2.0562248995983934,
+      "grad_norm": 0.21119273123023982,
+      "learning_rate": 1.7450863609291244e-05,
+      "loss": 0.2775,
+      "step": 1281
+    },
+    {
+      "epoch": 2.057831325301205,
+      "grad_norm": 0.19989610493561433,
+      "learning_rate": 1.7421083978558667e-05,
+      "loss": 0.2734,
+      "step": 1282
+    },
+    {
+      "epoch": 2.0594377510040163,
+      "grad_norm": 0.2049461925085804,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.2793,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0610441767068273,
+      "grad_norm": 0.21027456687662827,
+      "learning_rate": 1.7361524717093508e-05,
+      "loss": 0.284,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0626506024096387,
+      "grad_norm": 0.20592602304519445,
+      "learning_rate": 1.733174508636093e-05,
+      "loss": 0.2876,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0642570281124497,
+      "grad_norm": 0.20045695701272662,
+      "learning_rate": 1.7301965455628352e-05,
+      "loss": 0.2821,
+      "step": 1286
+    },
+    {
+      "epoch": 2.065863453815261,
+      "grad_norm": 0.2030737729176968,
+      "learning_rate": 1.7272185824895774e-05,
+      "loss": 0.2751,
+      "step": 1287
+    },
+    {
+      "epoch": 2.067469879518072,
+      "grad_norm": 0.20606854023821006,
+      "learning_rate": 1.7242406194163193e-05,
+      "loss": 0.2786,
+      "step": 1288
+    },
+    {
+      "epoch": 2.0690763052208836,
+      "grad_norm": 0.19050518539243924,
+      "learning_rate": 1.7212626563430616e-05,
+      "loss": 0.2806,
+      "step": 1289
+    },
+    {
+      "epoch": 2.0706827309236946,
+      "grad_norm": 0.2016324872075261,
+      "learning_rate": 1.7182846932698034e-05,
+      "loss": 0.2912,
+      "step": 1290
+    },
+    {
+      "epoch": 2.072289156626506,
+      "grad_norm": 0.19937464734252766,
+      "learning_rate": 1.7153067301965457e-05,
+      "loss": 0.292,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0738955823293175,
+      "grad_norm": 0.2030184865559162,
+      "learning_rate": 1.7123287671232875e-05,
+      "loss": 0.2782,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0755020080321285,
+      "grad_norm": 0.19153969529826628,
+      "learning_rate": 1.7093508040500298e-05,
+      "loss": 0.281,
+      "step": 1293
+    },
+    {
+      "epoch": 2.07710843373494,
+      "grad_norm": 0.20342913607352534,
+      "learning_rate": 1.706372840976772e-05,
+      "loss": 0.2819,
+      "step": 1294
+    },
+    {
+      "epoch": 2.078714859437751,
+      "grad_norm": 0.19249096532827192,
+      "learning_rate": 1.7033948779035142e-05,
+      "loss": 0.2815,
+      "step": 1295
+    },
+    {
+      "epoch": 2.0803212851405624,
+      "grad_norm": 0.19416016554818794,
+      "learning_rate": 1.700416914830256e-05,
+      "loss": 0.2837,
+      "step": 1296
+    },
+    {
+      "epoch": 2.0819277108433734,
+      "grad_norm": 0.20169308055930382,
+      "learning_rate": 1.6974389517569983e-05,
+      "loss": 0.2735,
+      "step": 1297
+    },
+    {
+      "epoch": 2.083534136546185,
+      "grad_norm": 0.20000708163336223,
+      "learning_rate": 1.6944609886837402e-05,
+      "loss": 0.2771,
+      "step": 1298
+    },
+    {
+      "epoch": 2.085140562248996,
+      "grad_norm": 0.22245446601955285,
+      "learning_rate": 1.6914830256104824e-05,
+      "loss": 0.2744,
+      "step": 1299
+    },
+    {
+      "epoch": 2.0867469879518072,
+      "grad_norm": 0.20367878601775039,
+      "learning_rate": 1.6885050625372247e-05,
+      "loss": 0.2703,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0883534136546187,
+      "grad_norm": 0.21379513521178495,
+      "learning_rate": 1.685527099463967e-05,
+      "loss": 0.2854,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0899598393574297,
+      "grad_norm": 0.2124703802627289,
+      "learning_rate": 1.6825491363907088e-05,
+      "loss": 0.2791,
+      "step": 1302
+    },
+    {
+      "epoch": 2.091566265060241,
+      "grad_norm": 0.19744090340380965,
+      "learning_rate": 1.679571173317451e-05,
+      "loss": 0.2831,
+      "step": 1303
+    },
+    {
+      "epoch": 2.093172690763052,
+      "grad_norm": 0.2030029634127148,
+      "learning_rate": 1.6765932102441932e-05,
+      "loss": 0.2677,
+      "step": 1304
+    },
+    {
+      "epoch": 2.0947791164658636,
+      "grad_norm": 0.2180252902147147,
+      "learning_rate": 1.673615247170935e-05,
+      "loss": 0.2821,
+      "step": 1305
+    },
+    {
+      "epoch": 2.0963855421686746,
+      "grad_norm": 0.19691447044091656,
+      "learning_rate": 1.6706372840976773e-05,
+      "loss": 0.2864,
+      "step": 1306
+    },
+    {
+      "epoch": 2.097991967871486,
+      "grad_norm": 0.19265504935281613,
+      "learning_rate": 1.6676593210244192e-05,
+      "loss": 0.2927,
+      "step": 1307
+    },
+    {
+      "epoch": 2.099598393574297,
+      "grad_norm": 0.19508824514066675,
+      "learning_rate": 1.6646813579511615e-05,
+      "loss": 0.283,
+      "step": 1308
+    },
+    {
+      "epoch": 2.1012048192771084,
+      "grad_norm": 0.20194415821250655,
+      "learning_rate": 1.6617033948779037e-05,
+      "loss": 0.2768,
+      "step": 1309
+    },
+    {
+      "epoch": 2.10281124497992,
+      "grad_norm": 0.19516202353557202,
+      "learning_rate": 1.658725431804646e-05,
+      "loss": 0.2684,
+      "step": 1310
+    },
+    {
+      "epoch": 2.104417670682731,
+      "grad_norm": 0.19986830870005645,
+      "learning_rate": 1.6557474687313878e-05,
+      "loss": 0.2767,
+      "step": 1311
+    },
+    {
+      "epoch": 2.1060240963855423,
+      "grad_norm": 0.2055813489906199,
+      "learning_rate": 1.65276950565813e-05,
+      "loss": 0.2884,
+      "step": 1312
+    },
+    {
+      "epoch": 2.1076305220883533,
+      "grad_norm": 0.20050715479446987,
+      "learning_rate": 1.649791542584872e-05,
+      "loss": 0.2758,
+      "step": 1313
+    },
+    {
+      "epoch": 2.1092369477911648,
+      "grad_norm": 0.19236285983507476,
+      "learning_rate": 1.646813579511614e-05,
+      "loss": 0.2719,
+      "step": 1314
+    },
+    {
+      "epoch": 2.1108433734939758,
+      "grad_norm": 0.18775008327406353,
+      "learning_rate": 1.643835616438356e-05,
+      "loss": 0.2769,
+      "step": 1315
+    },
+    {
+      "epoch": 2.112449799196787,
+      "grad_norm": 0.21571708856539357,
+      "learning_rate": 1.6408576533650982e-05,
+      "loss": 0.2925,
+      "step": 1316
+    },
+    {
+      "epoch": 2.114056224899598,
+      "grad_norm": 0.1936950743747594,
+      "learning_rate": 1.6378796902918405e-05,
+      "loss": 0.28,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1156626506024097,
+      "grad_norm": 0.2199167999394077,
+      "learning_rate": 1.6349017272185827e-05,
+      "loss": 0.293,
+      "step": 1318
+    },
+    {
+      "epoch": 2.117269076305221,
+      "grad_norm": 0.195144728617892,
+      "learning_rate": 1.6319237641453246e-05,
+      "loss": 0.2745,
+      "step": 1319
+    },
+    {
+      "epoch": 2.118875502008032,
+      "grad_norm": 0.2080366011261388,
+      "learning_rate": 1.6289458010720668e-05,
+      "loss": 0.2958,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1204819277108435,
+      "grad_norm": 0.20834567278969648,
+      "learning_rate": 1.625967837998809e-05,
+      "loss": 0.2919,
+      "step": 1321
+    },
+    {
+      "epoch": 2.1220883534136545,
+      "grad_norm": 0.20330161112032538,
+      "learning_rate": 1.622989874925551e-05,
+      "loss": 0.295,
+      "step": 1322
+    },
+    {
+      "epoch": 2.123694779116466,
+      "grad_norm": 0.19423024126446417,
+      "learning_rate": 1.620011911852293e-05,
+      "loss": 0.2705,
+      "step": 1323
+    },
+    {
+      "epoch": 2.125301204819277,
+      "grad_norm": 0.2074284814583773,
+      "learning_rate": 1.617033948779035e-05,
+      "loss": 0.2942,
+      "step": 1324
+    },
+    {
+      "epoch": 2.1269076305220884,
+      "grad_norm": 0.2224622605312662,
+      "learning_rate": 1.6140559857057776e-05,
+      "loss": 0.2948,
+      "step": 1325
+    },
+    {
+      "epoch": 2.1285140562248994,
+      "grad_norm": 0.18843797510856755,
+      "learning_rate": 1.6110780226325195e-05,
+      "loss": 0.2855,
+      "step": 1326
+    },
+    {
+      "epoch": 2.130120481927711,
+      "grad_norm": 0.3675079769449738,
+      "learning_rate": 1.6081000595592617e-05,
+      "loss": 0.2731,
+      "step": 1327
+    },
+    {
+      "epoch": 2.1317269076305223,
+      "grad_norm": 0.22536574097578618,
+      "learning_rate": 1.6051220964860036e-05,
+      "loss": 0.2885,
+      "step": 1328
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.20765939969070893,
+      "learning_rate": 1.6021441334127458e-05,
+      "loss": 0.2737,
+      "step": 1329
+    },
+    {
+      "epoch": 2.1349397590361447,
+      "grad_norm": 0.1917959472732135,
+      "learning_rate": 1.5991661703394877e-05,
+      "loss": 0.2698,
+      "step": 1330
+    },
+    {
+      "epoch": 2.1365461847389557,
+      "grad_norm": 0.20833088178715453,
+      "learning_rate": 1.59618820726623e-05,
+      "loss": 0.2812,
+      "step": 1331
+    },
+    {
+      "epoch": 2.138152610441767,
+      "grad_norm": 0.2266944112068008,
+      "learning_rate": 1.593210244192972e-05,
+      "loss": 0.2925,
+      "step": 1332
+    },
+    {
+      "epoch": 2.139759036144578,
+      "grad_norm": 0.19873930251384897,
+      "learning_rate": 1.5902322811197144e-05,
+      "loss": 0.2832,
+      "step": 1333
+    },
+    {
+      "epoch": 2.1413654618473896,
+      "grad_norm": 0.211257087356807,
+      "learning_rate": 1.5872543180464562e-05,
+      "loss": 0.2927,
+      "step": 1334
+    },
+    {
+      "epoch": 2.1429718875502006,
+      "grad_norm": 0.22087096040863327,
+      "learning_rate": 1.5842763549731985e-05,
+      "loss": 0.2749,
+      "step": 1335
+    },
+    {
+      "epoch": 2.144578313253012,
+      "grad_norm": 0.2137852392339819,
+      "learning_rate": 1.5812983918999404e-05,
+      "loss": 0.2746,
+      "step": 1336
+    },
+    {
+      "epoch": 2.1461847389558235,
+      "grad_norm": 0.21811352320657051,
+      "learning_rate": 1.5783204288266826e-05,
+      "loss": 0.2873,
+      "step": 1337
+    },
+    {
+      "epoch": 2.1477911646586345,
+      "grad_norm": 0.20957211886331442,
+      "learning_rate": 1.5753424657534248e-05,
+      "loss": 0.276,
+      "step": 1338
+    },
+    {
+      "epoch": 2.149397590361446,
+      "grad_norm": 0.22560776369291363,
+      "learning_rate": 1.5723645026801667e-05,
+      "loss": 0.3123,
+      "step": 1339
+    },
+    {
+      "epoch": 2.151004016064257,
+      "grad_norm": 0.22339377515881612,
+      "learning_rate": 1.569386539606909e-05,
+      "loss": 0.285,
+      "step": 1340
+    },
+    {
+      "epoch": 2.1526104417670684,
+      "grad_norm": 0.20990067995837738,
+      "learning_rate": 1.566408576533651e-05,
+      "loss": 0.2857,
+      "step": 1341
+    },
+    {
+      "epoch": 2.1542168674698794,
+      "grad_norm": 0.19663492374905836,
+      "learning_rate": 1.5634306134603934e-05,
+      "loss": 0.2718,
+      "step": 1342
+    },
+    {
+      "epoch": 2.155823293172691,
+      "grad_norm": 0.21682519090898078,
+      "learning_rate": 1.5604526503871352e-05,
+      "loss": 0.2743,
+      "step": 1343
+    },
+    {
+      "epoch": 2.157429718875502,
+      "grad_norm": 0.2171805234137457,
+      "learning_rate": 1.5574746873138775e-05,
+      "loss": 0.2804,
+      "step": 1344
+    },
+    {
+      "epoch": 2.1590361445783133,
+      "grad_norm": 0.2140861366836785,
+      "learning_rate": 1.5544967242406194e-05,
+      "loss": 0.2808,
+      "step": 1345
+    },
+    {
+      "epoch": 2.1606425702811247,
+      "grad_norm": 0.21635984908185377,
+      "learning_rate": 1.5515187611673616e-05,
+      "loss": 0.2799,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1622489959839357,
+      "grad_norm": 0.20798428133406957,
+      "learning_rate": 1.5485407980941035e-05,
+      "loss": 0.2671,
+      "step": 1347
+    },
+    {
+      "epoch": 2.163855421686747,
+      "grad_norm": 0.22593474257136462,
+      "learning_rate": 1.545562835020846e-05,
+      "loss": 0.2809,
+      "step": 1348
+    },
+    {
+      "epoch": 2.165461847389558,
+      "grad_norm": 0.22565128373678292,
+      "learning_rate": 1.542584871947588e-05,
+      "loss": 0.2898,
+      "step": 1349
+    },
+    {
+      "epoch": 2.1670682730923696,
+      "grad_norm": 0.21453394498328968,
+      "learning_rate": 1.53960690887433e-05,
+      "loss": 0.2669,
+      "step": 1350
+    },
+    {
+      "epoch": 2.1686746987951806,
+      "grad_norm": 0.21565741702247793,
+      "learning_rate": 1.536628945801072e-05,
+      "loss": 0.2888,
+      "step": 1351
+    },
+    {
+      "epoch": 2.170281124497992,
+      "grad_norm": 0.20920208759066525,
+      "learning_rate": 1.5336509827278143e-05,
+      "loss": 0.2793,
+      "step": 1352
+    },
+    {
+      "epoch": 2.171887550200803,
+      "grad_norm": 0.2139532549748069,
+      "learning_rate": 1.530673019654556e-05,
+      "loss": 0.2816,
+      "step": 1353
+    },
+    {
+      "epoch": 2.1734939759036145,
+      "grad_norm": 0.2170410666396211,
+      "learning_rate": 1.5276950565812984e-05,
+      "loss": 0.2877,
+      "step": 1354
+    },
+    {
+      "epoch": 2.175100401606426,
+      "grad_norm": 0.19884229667013362,
+      "learning_rate": 1.5247170935080408e-05,
+      "loss": 0.2842,
+      "step": 1355
+    },
+    {
+      "epoch": 2.176706827309237,
+      "grad_norm": 0.19604489046763632,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.2833,
+      "step": 1356
+    },
+    {
+      "epoch": 2.1783132530120484,
+      "grad_norm": 0.2143224822362246,
+      "learning_rate": 1.5187611673615249e-05,
+      "loss": 0.2787,
+      "step": 1357
+    },
+    {
+      "epoch": 2.1799196787148594,
+      "grad_norm": 0.19359948352798684,
+      "learning_rate": 1.515783204288267e-05,
+      "loss": 0.2734,
+      "step": 1358
+    },
+    {
+      "epoch": 2.181526104417671,
+      "grad_norm": 0.19451280888655836,
+      "learning_rate": 1.512805241215009e-05,
+      "loss": 0.2826,
+      "step": 1359
+    },
+    {
+      "epoch": 2.183132530120482,
+      "grad_norm": 0.20343592326016344,
+      "learning_rate": 1.509827278141751e-05,
+      "loss": 0.2725,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1847389558232932,
+      "grad_norm": 0.20372631119579326,
+      "learning_rate": 1.5068493150684931e-05,
+      "loss": 0.2804,
+      "step": 1361
+    },
+    {
+      "epoch": 2.1863453815261042,
+      "grad_norm": 0.19976055476333732,
+      "learning_rate": 1.5038713519952351e-05,
+      "loss": 0.2847,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1879518072289157,
+      "grad_norm": 0.2126072933668119,
+      "learning_rate": 1.5008933889219775e-05,
+      "loss": 0.2801,
+      "step": 1363
+    },
+    {
+      "epoch": 2.189558232931727,
+      "grad_norm": 0.21783118111222857,
+      "learning_rate": 1.4979154258487196e-05,
+      "loss": 0.2762,
+      "step": 1364
+    },
+    {
+      "epoch": 2.191164658634538,
+      "grad_norm": 0.22424590850184806,
+      "learning_rate": 1.4949374627754617e-05,
+      "loss": 0.2695,
+      "step": 1365
+    },
+    {
+      "epoch": 2.1927710843373496,
+      "grad_norm": 0.20898807650299586,
+      "learning_rate": 1.4919594997022037e-05,
+      "loss": 0.2866,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1943775100401606,
+      "grad_norm": 0.2024609480649088,
+      "learning_rate": 1.488981536628946e-05,
+      "loss": 0.2593,
+      "step": 1367
+    },
+    {
+      "epoch": 2.195983935742972,
+      "grad_norm": 0.2318390578922345,
+      "learning_rate": 1.486003573555688e-05,
+      "loss": 0.2852,
+      "step": 1368
+    },
+    {
+      "epoch": 2.197590361445783,
+      "grad_norm": 0.2234576266891764,
+      "learning_rate": 1.48302561048243e-05,
+      "loss": 0.2979,
+      "step": 1369
+    },
+    {
+      "epoch": 2.1991967871485945,
+      "grad_norm": 0.25087746453484566,
+      "learning_rate": 1.4800476474091721e-05,
+      "loss": 0.2952,
+      "step": 1370
+    },
+    {
+      "epoch": 2.2008032128514055,
+      "grad_norm": 0.20779299268249266,
+      "learning_rate": 1.4770696843359145e-05,
+      "loss": 0.291,
+      "step": 1371
+    },
+    {
+      "epoch": 2.202409638554217,
+      "grad_norm": 0.19503797552144309,
+      "learning_rate": 1.4740917212626565e-05,
+      "loss": 0.2889,
+      "step": 1372
+    },
+    {
+      "epoch": 2.2040160642570283,
+      "grad_norm": 0.22183786011491854,
+      "learning_rate": 1.4711137581893986e-05,
+      "loss": 0.2876,
+      "step": 1373
+    },
+    {
+      "epoch": 2.2056224899598393,
+      "grad_norm": 0.20644315848684305,
+      "learning_rate": 1.4681357951161407e-05,
+      "loss": 0.2823,
+      "step": 1374
+    },
+    {
+      "epoch": 2.207228915662651,
+      "grad_norm": 0.18929674842471658,
+      "learning_rate": 1.4651578320428827e-05,
+      "loss": 0.2783,
+      "step": 1375
+    },
+    {
+      "epoch": 2.208835341365462,
+      "grad_norm": 0.21308959207082992,
+      "learning_rate": 1.4621798689696248e-05,
+      "loss": 0.288,
+      "step": 1376
+    },
+    {
+      "epoch": 2.2104417670682732,
+      "grad_norm": 0.20666635961218438,
+      "learning_rate": 1.4592019058963668e-05,
+      "loss": 0.2738,
+      "step": 1377
+    },
+    {
+      "epoch": 2.212048192771084,
+      "grad_norm": 0.2304141465628837,
+      "learning_rate": 1.4562239428231092e-05,
+      "loss": 0.2898,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2136546184738957,
+      "grad_norm": 0.20788763343300304,
+      "learning_rate": 1.4532459797498513e-05,
+      "loss": 0.2878,
+      "step": 1379
+    },
+    {
+      "epoch": 2.2152610441767067,
+      "grad_norm": 0.20588499147123118,
+      "learning_rate": 1.4502680166765933e-05,
+      "loss": 0.2657,
+      "step": 1380
+    },
+    {
+      "epoch": 2.216867469879518,
+      "grad_norm": 0.20418894178510408,
+      "learning_rate": 1.4472900536033354e-05,
+      "loss": 0.2779,
+      "step": 1381
+    },
+    {
+      "epoch": 2.2184738955823295,
+      "grad_norm": 0.20491189247655886,
+      "learning_rate": 1.4443120905300774e-05,
+      "loss": 0.2753,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2200803212851405,
+      "grad_norm": 0.19753400988709927,
+      "learning_rate": 1.4413341274568195e-05,
+      "loss": 0.2829,
+      "step": 1383
+    },
+    {
+      "epoch": 2.221686746987952,
+      "grad_norm": 0.44309112407887496,
+      "learning_rate": 1.4383561643835617e-05,
+      "loss": 0.2795,
+      "step": 1384
+    },
+    {
+      "epoch": 2.223293172690763,
+      "grad_norm": 0.195193110652276,
+      "learning_rate": 1.4353782013103038e-05,
+      "loss": 0.2857,
+      "step": 1385
+    },
+    {
+      "epoch": 2.2248995983935744,
+      "grad_norm": 0.1821677494511547,
+      "learning_rate": 1.432400238237046e-05,
+      "loss": 0.2772,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2265060240963854,
+      "grad_norm": 0.20054081495677123,
+      "learning_rate": 1.429422275163788e-05,
+      "loss": 0.2915,
+      "step": 1387
+    },
+    {
+      "epoch": 2.228112449799197,
+      "grad_norm": 0.1912911850679274,
+      "learning_rate": 1.4264443120905303e-05,
+      "loss": 0.2752,
+      "step": 1388
+    },
+    {
+      "epoch": 2.229718875502008,
+      "grad_norm": 0.203267085985269,
+      "learning_rate": 1.4234663490172723e-05,
+      "loss": 0.2828,
+      "step": 1389
+    },
+    {
+      "epoch": 2.2313253012048193,
+      "grad_norm": 0.1905952075992503,
+      "learning_rate": 1.4204883859440144e-05,
+      "loss": 0.286,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2329317269076308,
+      "grad_norm": 0.195695539557944,
+      "learning_rate": 1.4175104228707564e-05,
+      "loss": 0.2835,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2345381526104418,
+      "grad_norm": 0.18676759927959077,
+      "learning_rate": 1.4145324597974985e-05,
+      "loss": 0.2796,
+      "step": 1392
+    },
+    {
+      "epoch": 2.236144578313253,
+      "grad_norm": 0.20852779760290824,
+      "learning_rate": 1.4115544967242406e-05,
+      "loss": 0.2945,
+      "step": 1393
+    },
+    {
+      "epoch": 2.237751004016064,
+      "grad_norm": 0.1948314886127861,
+      "learning_rate": 1.408576533650983e-05,
+      "loss": 0.287,
+      "step": 1394
+    },
+    {
+      "epoch": 2.2393574297188756,
+      "grad_norm": 0.19534317098927167,
+      "learning_rate": 1.405598570577725e-05,
+      "loss": 0.2907,
+      "step": 1395
+    },
+    {
+      "epoch": 2.2409638554216866,
+      "grad_norm": 0.1851818001811194,
+      "learning_rate": 1.402620607504467e-05,
+      "loss": 0.2705,
+      "step": 1396
+    },
+    {
+      "epoch": 2.242570281124498,
+      "grad_norm": 0.19855454405040163,
+      "learning_rate": 1.3996426444312091e-05,
+      "loss": 0.2743,
+      "step": 1397
+    },
+    {
+      "epoch": 2.244176706827309,
+      "grad_norm": 0.18856697121755286,
+      "learning_rate": 1.3966646813579512e-05,
+      "loss": 0.2742,
+      "step": 1398
+    },
+    {
+      "epoch": 2.2457831325301205,
+      "grad_norm": 0.20949440819860526,
+      "learning_rate": 1.3936867182846932e-05,
+      "loss": 0.2983,
+      "step": 1399
+    },
+    {
+      "epoch": 2.247389558232932,
+      "grad_norm": 0.18799229527821282,
+      "learning_rate": 1.3907087552114353e-05,
+      "loss": 0.2777,
+      "step": 1400
+    },
+    {
+      "epoch": 2.248995983935743,
+      "grad_norm": 0.20817802926054224,
+      "learning_rate": 1.3877307921381773e-05,
+      "loss": 0.2748,
+      "step": 1401
+    },
+    {
+      "epoch": 2.2506024096385544,
+      "grad_norm": 0.19971412125106477,
+      "learning_rate": 1.3847528290649197e-05,
+      "loss": 0.2697,
+      "step": 1402
+    },
+    {
+      "epoch": 2.2522088353413654,
+      "grad_norm": 0.20260285826421598,
+      "learning_rate": 1.3817748659916618e-05,
+      "loss": 0.2824,
+      "step": 1403
+    },
+    {
+      "epoch": 2.253815261044177,
+      "grad_norm": 0.195358413178943,
+      "learning_rate": 1.3787969029184038e-05,
+      "loss": 0.2685,
+      "step": 1404
+    },
+    {
+      "epoch": 2.255421686746988,
+      "grad_norm": 0.20138950773431835,
+      "learning_rate": 1.375818939845146e-05,
+      "loss": 0.2883,
+      "step": 1405
+    },
+    {
+      "epoch": 2.2570281124497993,
+      "grad_norm": 0.20198637318846824,
+      "learning_rate": 1.3728409767718881e-05,
+      "loss": 0.2917,
+      "step": 1406
+    },
+    {
+      "epoch": 2.2586345381526103,
+      "grad_norm": 0.19903212810034507,
+      "learning_rate": 1.3698630136986302e-05,
+      "loss": 0.2727,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2602409638554217,
+      "grad_norm": 0.2014873051634944,
+      "learning_rate": 1.3668850506253722e-05,
+      "loss": 0.27,
+      "step": 1408
+    },
+    {
+      "epoch": 2.261847389558233,
+      "grad_norm": 0.20301125119497848,
+      "learning_rate": 1.3639070875521146e-05,
+      "loss": 0.2888,
+      "step": 1409
+    },
+    {
+      "epoch": 2.263453815261044,
+      "grad_norm": 0.18947958537365203,
+      "learning_rate": 1.3609291244788567e-05,
+      "loss": 0.2851,
+      "step": 1410
+    },
+    {
+      "epoch": 2.2650602409638556,
+      "grad_norm": 0.19886368646745592,
+      "learning_rate": 1.3579511614055987e-05,
+      "loss": 0.2887,
+      "step": 1411
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 0.19329147766946875,
+      "learning_rate": 1.3549731983323408e-05,
+      "loss": 0.2923,
+      "step": 1412
+    },
+    {
+      "epoch": 2.268273092369478,
+      "grad_norm": 0.1888562779617201,
+      "learning_rate": 1.3519952352590828e-05,
+      "loss": 0.292,
+      "step": 1413
+    },
+    {
+      "epoch": 2.269879518072289,
+      "grad_norm": 0.19788625383031824,
+      "learning_rate": 1.3490172721858249e-05,
+      "loss": 0.286,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2714859437751005,
+      "grad_norm": 0.21236176682280908,
+      "learning_rate": 1.346039309112567e-05,
+      "loss": 0.2908,
+      "step": 1415
+    },
+    {
+      "epoch": 2.2730923694779115,
+      "grad_norm": 0.1876516240958428,
+      "learning_rate": 1.343061346039309e-05,
+      "loss": 0.2804,
+      "step": 1416
+    },
+    {
+      "epoch": 2.274698795180723,
+      "grad_norm": 0.1831457191251782,
+      "learning_rate": 1.3400833829660514e-05,
+      "loss": 0.2762,
+      "step": 1417
+    },
+    {
+      "epoch": 2.2763052208835344,
+      "grad_norm": 0.20772669907762084,
+      "learning_rate": 1.3371054198927935e-05,
+      "loss": 0.2893,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2779116465863454,
+      "grad_norm": 0.20745250121591247,
+      "learning_rate": 1.3341274568195355e-05,
+      "loss": 0.2813,
+      "step": 1419
+    },
+    {
+      "epoch": 2.279518072289157,
+      "grad_norm": 0.19787394808531827,
+      "learning_rate": 1.3311494937462776e-05,
+      "loss": 0.2786,
+      "step": 1420
+    },
+    {
+      "epoch": 2.281124497991968,
+      "grad_norm": 0.19693061759333913,
+      "learning_rate": 1.3281715306730196e-05,
+      "loss": 0.2731,
+      "step": 1421
+    },
+    {
+      "epoch": 2.2827309236947793,
+      "grad_norm": 0.200179553066319,
+      "learning_rate": 1.3251935675997617e-05,
+      "loss": 0.2712,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2843373493975903,
+      "grad_norm": 0.20538459255470326,
+      "learning_rate": 1.3222156045265039e-05,
+      "loss": 0.2937,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2859437751004017,
+      "grad_norm": 0.20540469526629412,
+      "learning_rate": 1.319237641453246e-05,
+      "loss": 0.2867,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2875502008032127,
+      "grad_norm": 0.19424062293228397,
+      "learning_rate": 1.3162596783799882e-05,
+      "loss": 0.2848,
+      "step": 1425
+    },
+    {
+      "epoch": 2.289156626506024,
+      "grad_norm": 0.19772370771483927,
+      "learning_rate": 1.3132817153067304e-05,
+      "loss": 0.2761,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2907630522088356,
+      "grad_norm": 0.19070855889734556,
+      "learning_rate": 1.3103037522334725e-05,
+      "loss": 0.287,
+      "step": 1427
+    },
+    {
+      "epoch": 2.2923694779116466,
+      "grad_norm": 0.19207092404570347,
+      "learning_rate": 1.3073257891602145e-05,
+      "loss": 0.2766,
+      "step": 1428
+    },
+    {
+      "epoch": 2.293975903614458,
+      "grad_norm": 0.18790104378715738,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.2707,
+      "step": 1429
+    },
+    {
+      "epoch": 2.295582329317269,
+      "grad_norm": 0.1990292808311974,
+      "learning_rate": 1.3013698630136986e-05,
+      "loss": 0.2849,
+      "step": 1430
+    },
+    {
+      "epoch": 2.2971887550200805,
+      "grad_norm": 0.22742348746658536,
+      "learning_rate": 1.2983918999404407e-05,
+      "loss": 0.2966,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2987951807228915,
+      "grad_norm": 0.19765174567815721,
+      "learning_rate": 1.295413936867183e-05,
+      "loss": 0.2851,
+      "step": 1432
+    },
+    {
+      "epoch": 2.300401606425703,
+      "grad_norm": 0.18088850697034875,
+      "learning_rate": 1.2924359737939251e-05,
+      "loss": 0.2657,
+      "step": 1433
+    },
+    {
+      "epoch": 2.302008032128514,
+      "grad_norm": 0.19373049223838326,
+      "learning_rate": 1.2894580107206672e-05,
+      "loss": 0.2839,
+      "step": 1434
+    },
+    {
+      "epoch": 2.3036144578313253,
+      "grad_norm": 0.20654452155770917,
+      "learning_rate": 1.2864800476474092e-05,
+      "loss": 0.2896,
+      "step": 1435
+    },
+    {
+      "epoch": 2.305220883534137,
+      "grad_norm": 0.20375004192296206,
+      "learning_rate": 1.2835020845741513e-05,
+      "loss": 0.2899,
+      "step": 1436
+    },
+    {
+      "epoch": 2.306827309236948,
+      "grad_norm": 0.2057218215028387,
+      "learning_rate": 1.2805241215008934e-05,
+      "loss": 0.2871,
+      "step": 1437
+    },
+    {
+      "epoch": 2.3084337349397592,
+      "grad_norm": 0.1955208741700446,
+      "learning_rate": 1.2775461584276354e-05,
+      "loss": 0.2726,
+      "step": 1438
+    },
+    {
+      "epoch": 2.3100401606425702,
+      "grad_norm": 0.21049631645666206,
+      "learning_rate": 1.2745681953543775e-05,
+      "loss": 0.2948,
+      "step": 1439
+    },
+    {
+      "epoch": 2.3116465863453817,
+      "grad_norm": 0.20512260944309865,
+      "learning_rate": 1.2715902322811199e-05,
+      "loss": 0.2776,
+      "step": 1440
+    },
+    {
+      "epoch": 2.3132530120481927,
+      "grad_norm": 0.18114258763003166,
+      "learning_rate": 1.268612269207862e-05,
+      "loss": 0.2732,
+      "step": 1441
+    },
+    {
+      "epoch": 2.314859437751004,
+      "grad_norm": 0.1959388399118363,
+      "learning_rate": 1.265634306134604e-05,
+      "loss": 0.2812,
+      "step": 1442
+    },
+    {
+      "epoch": 2.316465863453815,
+      "grad_norm": 0.18349205225454254,
+      "learning_rate": 1.2626563430613462e-05,
+      "loss": 0.2612,
+      "step": 1443
+    },
+    {
+      "epoch": 2.3180722891566266,
+      "grad_norm": 0.1849928114980666,
+      "learning_rate": 1.2596783799880883e-05,
+      "loss": 0.2763,
+      "step": 1444
+    },
+    {
+      "epoch": 2.319678714859438,
+      "grad_norm": 0.1839742763610793,
+      "learning_rate": 1.2567004169148303e-05,
+      "loss": 0.2741,
+      "step": 1445
+    },
+    {
+      "epoch": 2.321285140562249,
+      "grad_norm": 0.1936421692979358,
+      "learning_rate": 1.2537224538415724e-05,
+      "loss": 0.2878,
+      "step": 1446
+    },
+    {
+      "epoch": 2.32289156626506,
+      "grad_norm": 0.19257864798083466,
+      "learning_rate": 1.2507444907683144e-05,
+      "loss": 0.2827,
+      "step": 1447
+    },
+    {
+      "epoch": 2.3244979919678714,
+      "grad_norm": 0.1989288231161012,
+      "learning_rate": 1.2477665276950566e-05,
+      "loss": 0.2785,
+      "step": 1448
+    },
+    {
+      "epoch": 2.326104417670683,
+      "grad_norm": 0.19337325518819184,
+      "learning_rate": 1.2447885646217987e-05,
+      "loss": 0.2779,
+      "step": 1449
+    },
+    {
+      "epoch": 2.327710843373494,
+      "grad_norm": 0.18727054478458935,
+      "learning_rate": 1.241810601548541e-05,
+      "loss": 0.2787,
+      "step": 1450
+    },
+    {
+      "epoch": 2.3293172690763053,
+      "grad_norm": 0.1967745024997371,
+      "learning_rate": 1.238832638475283e-05,
+      "loss": 0.2886,
+      "step": 1451
+    },
+    {
+      "epoch": 2.3309236947791163,
+      "grad_norm": 0.20060793013447148,
+      "learning_rate": 1.235854675402025e-05,
+      "loss": 0.2695,
+      "step": 1452
+    },
+    {
+      "epoch": 2.3325301204819278,
+      "grad_norm": 0.2013438687767352,
+      "learning_rate": 1.2328767123287671e-05,
+      "loss": 0.2658,
+      "step": 1453
+    },
+    {
+      "epoch": 2.334136546184739,
+      "grad_norm": 0.18529352907025987,
+      "learning_rate": 1.2298987492555093e-05,
+      "loss": 0.2772,
+      "step": 1454
+    },
+    {
+      "epoch": 2.33574297188755,
+      "grad_norm": 0.19042540749718512,
+      "learning_rate": 1.2269207861822514e-05,
+      "loss": 0.2908,
+      "step": 1455
+    },
+    {
+      "epoch": 2.337349397590361,
+      "grad_norm": 0.18922713550520592,
+      "learning_rate": 1.2239428231089934e-05,
+      "loss": 0.2882,
+      "step": 1456
+    },
+    {
+      "epoch": 2.3389558232931726,
+      "grad_norm": 0.20939876918201103,
+      "learning_rate": 1.2209648600357356e-05,
+      "loss": 0.2794,
+      "step": 1457
+    },
+    {
+      "epoch": 2.340562248995984,
+      "grad_norm": 0.18838061426751868,
+      "learning_rate": 1.2179868969624777e-05,
+      "loss": 0.2883,
+      "step": 1458
+    },
+    {
+      "epoch": 2.342168674698795,
+      "grad_norm": 0.18809567266436786,
+      "learning_rate": 1.2150089338892198e-05,
+      "loss": 0.2868,
+      "step": 1459
+    },
+    {
+      "epoch": 2.3437751004016065,
+      "grad_norm": 0.18724157835249255,
+      "learning_rate": 1.2120309708159618e-05,
+      "loss": 0.3008,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3453815261044175,
+      "grad_norm": 0.18753869589680505,
+      "learning_rate": 1.209053007742704e-05,
+      "loss": 0.2907,
+      "step": 1461
+    },
+    {
+      "epoch": 2.346987951807229,
+      "grad_norm": 0.18196888532868866,
+      "learning_rate": 1.2060750446694463e-05,
+      "loss": 0.2741,
+      "step": 1462
+    },
+    {
+      "epoch": 2.3485943775100404,
+      "grad_norm": 0.18421309904049457,
+      "learning_rate": 1.2030970815961883e-05,
+      "loss": 0.2814,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3502008032128514,
+      "grad_norm": 0.18040962820252937,
+      "learning_rate": 1.2001191185229304e-05,
+      "loss": 0.2851,
+      "step": 1464
+    },
+    {
+      "epoch": 2.3518072289156624,
+      "grad_norm": 0.19206627034468482,
+      "learning_rate": 1.1971411554496726e-05,
+      "loss": 0.2771,
+      "step": 1465
+    },
+    {
+      "epoch": 2.353413654618474,
+      "grad_norm": 0.1892992889850213,
+      "learning_rate": 1.1941631923764147e-05,
+      "loss": 0.278,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3550200803212853,
+      "grad_norm": 0.18386969475637321,
+      "learning_rate": 1.1911852293031567e-05,
+      "loss": 0.2849,
+      "step": 1467
+    },
+    {
+      "epoch": 2.3566265060240963,
+      "grad_norm": 0.18682928678552546,
+      "learning_rate": 1.1882072662298988e-05,
+      "loss": 0.2849,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3582329317269077,
+      "grad_norm": 0.194393928066698,
+      "learning_rate": 1.185229303156641e-05,
+      "loss": 0.3004,
+      "step": 1469
+    },
+    {
+      "epoch": 2.3598393574297187,
+      "grad_norm": 0.18579247490094822,
+      "learning_rate": 1.182251340083383e-05,
+      "loss": 0.2809,
+      "step": 1470
+    },
+    {
+      "epoch": 2.36144578313253,
+      "grad_norm": 0.18682075099005277,
+      "learning_rate": 1.1792733770101251e-05,
+      "loss": 0.2924,
+      "step": 1471
+    },
+    {
+      "epoch": 2.363052208835341,
+      "grad_norm": 0.18642046974686882,
+      "learning_rate": 1.1762954139368672e-05,
+      "loss": 0.2806,
+      "step": 1472
+    },
+    {
+      "epoch": 2.3646586345381526,
+      "grad_norm": 0.18866799498275091,
+      "learning_rate": 1.1733174508636094e-05,
+      "loss": 0.2892,
+      "step": 1473
+    },
+    {
+      "epoch": 2.3662650602409636,
+      "grad_norm": 0.1890942207554411,
+      "learning_rate": 1.1703394877903514e-05,
+      "loss": 0.2942,
+      "step": 1474
+    },
+    {
+      "epoch": 2.367871485943775,
+      "grad_norm": 0.18048434103136618,
+      "learning_rate": 1.1673615247170935e-05,
+      "loss": 0.2833,
+      "step": 1475
+    },
+    {
+      "epoch": 2.3694779116465865,
+      "grad_norm": 0.19632406720450757,
+      "learning_rate": 1.1643835616438355e-05,
+      "loss": 0.2848,
+      "step": 1476
+    },
+    {
+      "epoch": 2.3710843373493975,
+      "grad_norm": 0.18719345956190125,
+      "learning_rate": 1.1614055985705778e-05,
+      "loss": 0.2874,
+      "step": 1477
+    },
+    {
+      "epoch": 2.372690763052209,
+      "grad_norm": 0.19841712109494833,
+      "learning_rate": 1.1584276354973198e-05,
+      "loss": 0.2839,
+      "step": 1478
+    },
+    {
+      "epoch": 2.37429718875502,
+      "grad_norm": 0.18642681858163637,
+      "learning_rate": 1.1554496724240619e-05,
+      "loss": 0.2792,
+      "step": 1479
+    },
+    {
+      "epoch": 2.3759036144578314,
+      "grad_norm": 0.1792405548472351,
+      "learning_rate": 1.1524717093508041e-05,
+      "loss": 0.2688,
+      "step": 1480
+    },
+    {
+      "epoch": 2.3775100401606424,
+      "grad_norm": 0.19477960567033956,
+      "learning_rate": 1.1494937462775462e-05,
+      "loss": 0.2808,
+      "step": 1481
+    },
+    {
+      "epoch": 2.379116465863454,
+      "grad_norm": 0.19812997269342178,
+      "learning_rate": 1.1465157832042884e-05,
+      "loss": 0.2983,
+      "step": 1482
+    },
+    {
+      "epoch": 2.380722891566265,
+      "grad_norm": 0.19755348401845638,
+      "learning_rate": 1.1435378201310304e-05,
+      "loss": 0.2949,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3823293172690763,
+      "grad_norm": 0.19900905996115784,
+      "learning_rate": 1.1405598570577727e-05,
+      "loss": 0.2938,
+      "step": 1484
+    },
+    {
+      "epoch": 2.3839357429718877,
+      "grad_norm": 0.19677337928790387,
+      "learning_rate": 1.1375818939845147e-05,
+      "loss": 0.2967,
+      "step": 1485
+    },
+    {
+      "epoch": 2.3855421686746987,
+      "grad_norm": 0.19364024075681263,
+      "learning_rate": 1.1346039309112568e-05,
+      "loss": 0.2815,
+      "step": 1486
+    },
+    {
+      "epoch": 2.38714859437751,
+      "grad_norm": 0.19161486254413096,
+      "learning_rate": 1.1316259678379988e-05,
+      "loss": 0.2808,
+      "step": 1487
+    },
+    {
+      "epoch": 2.388755020080321,
+      "grad_norm": 0.20890719417104178,
+      "learning_rate": 1.128648004764741e-05,
+      "loss": 0.2819,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3903614457831326,
+      "grad_norm": 0.18146426376706648,
+      "learning_rate": 1.1256700416914831e-05,
+      "loss": 0.2693,
+      "step": 1489
+    },
+    {
+      "epoch": 2.3919678714859436,
+      "grad_norm": 0.1921463250218022,
+      "learning_rate": 1.1226920786182252e-05,
+      "loss": 0.2933,
+      "step": 1490
+    },
+    {
+      "epoch": 2.393574297188755,
+      "grad_norm": 0.18715864298714197,
+      "learning_rate": 1.1197141155449672e-05,
+      "loss": 0.2667,
+      "step": 1491
+    },
+    {
+      "epoch": 2.395180722891566,
+      "grad_norm": 0.18764049048200124,
+      "learning_rate": 1.1167361524717094e-05,
+      "loss": 0.2841,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3967871485943775,
+      "grad_norm": 0.18633386696731774,
+      "learning_rate": 1.1137581893984515e-05,
+      "loss": 0.283,
+      "step": 1493
+    },
+    {
+      "epoch": 2.398393574297189,
+      "grad_norm": 0.18102587136896592,
+      "learning_rate": 1.1107802263251936e-05,
+      "loss": 0.2775,
+      "step": 1494
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.18499734311638247,
+      "learning_rate": 1.1078022632519356e-05,
+      "loss": 0.2815,
+      "step": 1495
+    },
+    {
+      "epoch": 2.4016064257028114,
+      "grad_norm": 0.18726230604101202,
+      "learning_rate": 1.1048243001786778e-05,
+      "loss": 0.2755,
+      "step": 1496
+    },
+    {
+      "epoch": 2.4032128514056224,
+      "grad_norm": 0.19396755026779727,
+      "learning_rate": 1.1018463371054199e-05,
+      "loss": 0.2746,
+      "step": 1497
+    },
+    {
+      "epoch": 2.404819277108434,
+      "grad_norm": 0.19963469527589364,
+      "learning_rate": 1.098868374032162e-05,
+      "loss": 0.2784,
+      "step": 1498
+    },
+    {
+      "epoch": 2.406425702811245,
+      "grad_norm": 0.19452591890113338,
+      "learning_rate": 1.0958904109589042e-05,
+      "loss": 0.278,
+      "step": 1499
+    },
+    {
+      "epoch": 2.4080321285140562,
+      "grad_norm": 0.18801090615789337,
+      "learning_rate": 1.0929124478856462e-05,
+      "loss": 0.2758,
+      "step": 1500
+    },
+    {
+      "epoch": 2.4096385542168672,
+      "grad_norm": 0.20654038553918738,
+      "learning_rate": 1.0899344848123885e-05,
+      "loss": 0.2949,
+      "step": 1501
+    },
+    {
+      "epoch": 2.4112449799196787,
+      "grad_norm": 0.18834258722356456,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.271,
+      "step": 1502
+    },
+    {
+      "epoch": 2.41285140562249,
+      "grad_norm": 0.19237206886647512,
+      "learning_rate": 1.0839785586658726e-05,
+      "loss": 0.2702,
+      "step": 1503
+    },
+    {
+      "epoch": 2.414457831325301,
+      "grad_norm": 0.18993808278326732,
+      "learning_rate": 1.0810005955926148e-05,
+      "loss": 0.2784,
+      "step": 1504
+    },
+    {
+      "epoch": 2.4160642570281126,
+      "grad_norm": 0.20480903011469814,
+      "learning_rate": 1.0780226325193568e-05,
+      "loss": 0.2995,
+      "step": 1505
+    },
+    {
+      "epoch": 2.4176706827309236,
+      "grad_norm": 0.208136543604581,
+      "learning_rate": 1.0750446694460989e-05,
+      "loss": 0.3076,
+      "step": 1506
+    },
+    {
+      "epoch": 2.419277108433735,
+      "grad_norm": 0.18244357017251567,
+      "learning_rate": 1.0720667063728411e-05,
+      "loss": 0.2682,
+      "step": 1507
+    },
+    {
+      "epoch": 2.420883534136546,
+      "grad_norm": 0.16971261958781741,
+      "learning_rate": 1.0690887432995832e-05,
+      "loss": 0.2579,
+      "step": 1508
+    },
+    {
+      "epoch": 2.4224899598393574,
+      "grad_norm": 0.18274700243771452,
+      "learning_rate": 1.0661107802263252e-05,
+      "loss": 0.2746,
+      "step": 1509
+    },
+    {
+      "epoch": 2.4240963855421684,
+      "grad_norm": 0.19118813600451662,
+      "learning_rate": 1.0631328171530673e-05,
+      "loss": 0.2797,
+      "step": 1510
+    },
+    {
+      "epoch": 2.42570281124498,
+      "grad_norm": 0.18303626521117353,
+      "learning_rate": 1.0601548540798095e-05,
+      "loss": 0.2896,
+      "step": 1511
+    },
+    {
+      "epoch": 2.4273092369477913,
+      "grad_norm": 0.18305566120895966,
+      "learning_rate": 1.0571768910065516e-05,
+      "loss": 0.2675,
+      "step": 1512
+    },
+    {
+      "epoch": 2.4289156626506023,
+      "grad_norm": 0.1893884338871087,
+      "learning_rate": 1.0541989279332936e-05,
+      "loss": 0.2714,
+      "step": 1513
+    },
+    {
+      "epoch": 2.4305220883534138,
+      "grad_norm": 0.18491595866432944,
+      "learning_rate": 1.0512209648600357e-05,
+      "loss": 0.2917,
+      "step": 1514
+    },
+    {
+      "epoch": 2.4321285140562248,
+      "grad_norm": 0.18918777323863992,
+      "learning_rate": 1.0482430017867779e-05,
+      "loss": 0.2888,
+      "step": 1515
+    },
+    {
+      "epoch": 2.433734939759036,
+      "grad_norm": 0.1850411963758118,
+      "learning_rate": 1.04526503871352e-05,
+      "loss": 0.286,
+      "step": 1516
+    },
+    {
+      "epoch": 2.435341365461847,
+      "grad_norm": 0.1848750327219421,
+      "learning_rate": 1.042287075640262e-05,
+      "loss": 0.274,
+      "step": 1517
+    },
+    {
+      "epoch": 2.4369477911646586,
+      "grad_norm": 0.2004762828981936,
+      "learning_rate": 1.0393091125670042e-05,
+      "loss": 0.2819,
+      "step": 1518
+    },
+    {
+      "epoch": 2.4385542168674696,
+      "grad_norm": 0.18728371181745848,
+      "learning_rate": 1.0363311494937463e-05,
+      "loss": 0.2845,
+      "step": 1519
+    },
+    {
+      "epoch": 2.440160642570281,
+      "grad_norm": 0.18628971274859396,
+      "learning_rate": 1.0333531864204885e-05,
+      "loss": 0.291,
+      "step": 1520
+    },
+    {
+      "epoch": 2.4417670682730925,
+      "grad_norm": 0.1897329118843444,
+      "learning_rate": 1.0303752233472306e-05,
+      "loss": 0.3044,
+      "step": 1521
+    },
+    {
+      "epoch": 2.4433734939759035,
+      "grad_norm": 0.18472072022984715,
+      "learning_rate": 1.0273972602739726e-05,
+      "loss": 0.2896,
+      "step": 1522
+    },
+    {
+      "epoch": 2.444979919678715,
+      "grad_norm": 0.1956552598842094,
+      "learning_rate": 1.0244192972007149e-05,
+      "loss": 0.2855,
+      "step": 1523
+    },
+    {
+      "epoch": 2.446586345381526,
+      "grad_norm": 0.1958651207777498,
+      "learning_rate": 1.0214413341274569e-05,
+      "loss": 0.2951,
+      "step": 1524
+    },
+    {
+      "epoch": 2.4481927710843374,
+      "grad_norm": 0.19318712685154302,
+      "learning_rate": 1.018463371054199e-05,
+      "loss": 0.281,
+      "step": 1525
+    },
+    {
+      "epoch": 2.4497991967871484,
+      "grad_norm": 0.17614917193183735,
+      "learning_rate": 1.015485407980941e-05,
+      "loss": 0.268,
+      "step": 1526
+    },
+    {
+      "epoch": 2.45140562248996,
+      "grad_norm": 0.1955853333253292,
+      "learning_rate": 1.0125074449076832e-05,
+      "loss": 0.2775,
+      "step": 1527
+    },
+    {
+      "epoch": 2.453012048192771,
+      "grad_norm": 0.19608910484702638,
+      "learning_rate": 1.0095294818344253e-05,
+      "loss": 0.2811,
+      "step": 1528
+    },
+    {
+      "epoch": 2.4546184738955823,
+      "grad_norm": 0.20332125720916808,
+      "learning_rate": 1.0065515187611674e-05,
+      "loss": 0.2936,
+      "step": 1529
+    },
+    {
+      "epoch": 2.4562248995983937,
+      "grad_norm": 0.19947230185866924,
+      "learning_rate": 1.0035735556879094e-05,
+      "loss": 0.2694,
+      "step": 1530
+    },
+    {
+      "epoch": 2.4578313253012047,
+      "grad_norm": 0.1879744700611198,
+      "learning_rate": 1.0005955926146516e-05,
+      "loss": 0.2822,
+      "step": 1531
+    },
+    {
+      "epoch": 2.459437751004016,
+      "grad_norm": 0.19580220039910437,
+      "learning_rate": 9.976176295413937e-06,
+      "loss": 0.2879,
+      "step": 1532
+    },
+    {
+      "epoch": 2.461044176706827,
+      "grad_norm": 0.19902210658108496,
+      "learning_rate": 9.946396664681357e-06,
+      "loss": 0.2744,
+      "step": 1533
+    },
+    {
+      "epoch": 2.4626506024096386,
+      "grad_norm": 0.19067665389503385,
+      "learning_rate": 9.91661703394878e-06,
+      "loss": 0.2764,
+      "step": 1534
+    },
+    {
+      "epoch": 2.4642570281124496,
+      "grad_norm": 0.1872003159605713,
+      "learning_rate": 9.8868374032162e-06,
+      "loss": 0.2758,
+      "step": 1535
+    },
+    {
+      "epoch": 2.465863453815261,
+      "grad_norm": 0.18018037288858052,
+      "learning_rate": 9.85705777248362e-06,
+      "loss": 0.2716,
+      "step": 1536
+    },
+    {
+      "epoch": 2.467469879518072,
+      "grad_norm": 0.20592796772981525,
+      "learning_rate": 9.827278141751043e-06,
+      "loss": 0.2901,
+      "step": 1537
+    },
+    {
+      "epoch": 2.4690763052208835,
+      "grad_norm": 0.19330384391872885,
+      "learning_rate": 9.797498511018464e-06,
+      "loss": 0.2878,
+      "step": 1538
+    },
+    {
+      "epoch": 2.470682730923695,
+      "grad_norm": 0.18564767070391436,
+      "learning_rate": 9.767718880285886e-06,
+      "loss": 0.2743,
+      "step": 1539
+    },
+    {
+      "epoch": 2.472289156626506,
+      "grad_norm": 0.1948652825643195,
+      "learning_rate": 9.737939249553306e-06,
+      "loss": 0.2895,
+      "step": 1540
+    },
+    {
+      "epoch": 2.4738955823293174,
+      "grad_norm": 0.19144249683042172,
+      "learning_rate": 9.708159618820727e-06,
+      "loss": 0.2806,
+      "step": 1541
+    },
+    {
+      "epoch": 2.4755020080321284,
+      "grad_norm": 0.2029879781027749,
+      "learning_rate": 9.67837998808815e-06,
+      "loss": 0.2903,
+      "step": 1542
+    },
+    {
+      "epoch": 2.47710843373494,
+      "grad_norm": 0.18251041716955976,
+      "learning_rate": 9.64860035735557e-06,
+      "loss": 0.2733,
+      "step": 1543
+    },
+    {
+      "epoch": 2.478714859437751,
+      "grad_norm": 0.18821726631570496,
+      "learning_rate": 9.61882072662299e-06,
+      "loss": 0.2811,
+      "step": 1544
+    },
+    {
+      "epoch": 2.4803212851405623,
+      "grad_norm": 0.1890188257297131,
+      "learning_rate": 9.589041095890411e-06,
+      "loss": 0.2848,
+      "step": 1545
+    },
+    {
+      "epoch": 2.4819277108433733,
+      "grad_norm": 0.18495983065658275,
+      "learning_rate": 9.559261465157833e-06,
+      "loss": 0.2776,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4835341365461847,
+      "grad_norm": 0.18760223528588468,
+      "learning_rate": 9.529481834425254e-06,
+      "loss": 0.288,
+      "step": 1547
+    },
+    {
+      "epoch": 2.485140562248996,
+      "grad_norm": 0.18121608702361014,
+      "learning_rate": 9.499702203692674e-06,
+      "loss": 0.279,
+      "step": 1548
+    },
+    {
+      "epoch": 2.486746987951807,
+      "grad_norm": 0.1851818964587047,
+      "learning_rate": 9.469922572960095e-06,
+      "loss": 0.2834,
+      "step": 1549
+    },
+    {
+      "epoch": 2.4883534136546186,
+      "grad_norm": 0.20877119041333475,
+      "learning_rate": 9.440142942227517e-06,
+      "loss": 0.2799,
+      "step": 1550
+    },
+    {
+      "epoch": 2.4899598393574296,
+      "grad_norm": 0.18641603386738315,
+      "learning_rate": 9.410363311494938e-06,
+      "loss": 0.2857,
+      "step": 1551
+    },
+    {
+      "epoch": 2.491566265060241,
+      "grad_norm": 0.17930127671625415,
+      "learning_rate": 9.380583680762358e-06,
+      "loss": 0.2809,
+      "step": 1552
+    },
+    {
+      "epoch": 2.493172690763052,
+      "grad_norm": 0.19095375958840477,
+      "learning_rate": 9.350804050029779e-06,
+      "loss": 0.2892,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4947791164658635,
+      "grad_norm": 0.2208696008778948,
+      "learning_rate": 9.321024419297201e-06,
+      "loss": 0.29,
+      "step": 1554
+    },
+    {
+      "epoch": 2.4963855421686745,
+      "grad_norm": 0.17527580643972474,
+      "learning_rate": 9.291244788564621e-06,
+      "loss": 0.267,
+      "step": 1555
+    },
+    {
+      "epoch": 2.497991967871486,
+      "grad_norm": 0.17976690381454716,
+      "learning_rate": 9.261465157832044e-06,
+      "loss": 0.2735,
+      "step": 1556
+    },
+    {
+      "epoch": 2.4995983935742974,
+      "grad_norm": 0.20219422951098998,
+      "learning_rate": 9.231685527099464e-06,
+      "loss": 0.2809,
+      "step": 1557
+    },
+    {
+      "epoch": 2.5012048192771084,
+      "grad_norm": 0.18319632216230614,
+      "learning_rate": 9.201905896366887e-06,
+      "loss": 0.2847,
+      "step": 1558
+    },
+    {
+      "epoch": 2.50281124497992,
+      "grad_norm": 0.19027327482072806,
+      "learning_rate": 9.172126265634307e-06,
+      "loss": 0.2798,
+      "step": 1559
+    },
+    {
+      "epoch": 2.504417670682731,
+      "grad_norm": 0.17500345115805288,
+      "learning_rate": 9.142346634901728e-06,
+      "loss": 0.2822,
+      "step": 1560
+    },
+    {
+      "epoch": 2.5060240963855422,
+      "grad_norm": 0.19553868356390833,
+      "learning_rate": 9.11256700416915e-06,
+      "loss": 0.2881,
+      "step": 1561
+    },
+    {
+      "epoch": 2.5076305220883537,
+      "grad_norm": 0.1982913757337272,
+      "learning_rate": 9.08278737343657e-06,
+      "loss": 0.2689,
+      "step": 1562
+    },
+    {
+      "epoch": 2.5092369477911647,
+      "grad_norm": 0.17241622331403286,
+      "learning_rate": 9.053007742703991e-06,
+      "loss": 0.265,
+      "step": 1563
+    },
+    {
+      "epoch": 2.5108433734939757,
+      "grad_norm": 0.20400008665717287,
+      "learning_rate": 9.023228111971412e-06,
+      "loss": 0.2861,
+      "step": 1564
+    },
+    {
+      "epoch": 2.512449799196787,
+      "grad_norm": 0.18947077833730427,
+      "learning_rate": 8.993448481238834e-06,
+      "loss": 0.29,
+      "step": 1565
+    },
+    {
+      "epoch": 2.5140562248995986,
+      "grad_norm": 0.1931313038797478,
+      "learning_rate": 8.963668850506254e-06,
+      "loss": 0.2825,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5156626506024096,
+      "grad_norm": 0.1924361737065017,
+      "learning_rate": 8.933889219773675e-06,
+      "loss": 0.2914,
+      "step": 1567
+    },
+    {
+      "epoch": 2.517269076305221,
+      "grad_norm": 0.20161485896708153,
+      "learning_rate": 8.904109589041095e-06,
+      "loss": 0.2844,
+      "step": 1568
+    },
+    {
+      "epoch": 2.518875502008032,
+      "grad_norm": 0.1938947469685224,
+      "learning_rate": 8.874329958308518e-06,
+      "loss": 0.2809,
+      "step": 1569
+    },
+    {
+      "epoch": 2.5204819277108435,
+      "grad_norm": 0.1981657020185443,
+      "learning_rate": 8.844550327575938e-06,
+      "loss": 0.2776,
+      "step": 1570
+    },
+    {
+      "epoch": 2.522088353413655,
+      "grad_norm": 0.1852544079685128,
+      "learning_rate": 8.814770696843359e-06,
+      "loss": 0.283,
+      "step": 1571
+    },
+    {
+      "epoch": 2.523694779116466,
+      "grad_norm": 0.18257009414972816,
+      "learning_rate": 8.78499106611078e-06,
+      "loss": 0.2777,
+      "step": 1572
+    },
+    {
+      "epoch": 2.525301204819277,
+      "grad_norm": 0.20960854561551556,
+      "learning_rate": 8.755211435378202e-06,
+      "loss": 0.2774,
+      "step": 1573
+    },
+    {
+      "epoch": 2.5269076305220883,
+      "grad_norm": 0.21428861794421236,
+      "learning_rate": 8.725431804645622e-06,
+      "loss": 0.2798,
+      "step": 1574
+    },
+    {
+      "epoch": 2.5285140562248998,
+      "grad_norm": 0.2148320921752406,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.2922,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5301204819277108,
+      "grad_norm": 0.18357213594521285,
+      "learning_rate": 8.665872543180465e-06,
+      "loss": 0.2868,
+      "step": 1576
+    },
+    {
+      "epoch": 2.531726907630522,
+      "grad_norm": 0.19535424046278613,
+      "learning_rate": 8.636092912447887e-06,
+      "loss": 0.2743,
+      "step": 1577
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 0.20462130886223842,
+      "learning_rate": 8.606313281715308e-06,
+      "loss": 0.2815,
+      "step": 1578
+    },
+    {
+      "epoch": 2.5349397590361447,
+      "grad_norm": 0.19449775734044208,
+      "learning_rate": 8.576533650982728e-06,
+      "loss": 0.2885,
+      "step": 1579
+    },
+    {
+      "epoch": 2.536546184738956,
+      "grad_norm": 0.19910861376895686,
+      "learning_rate": 8.546754020250149e-06,
+      "loss": 0.2866,
+      "step": 1580
+    },
+    {
+      "epoch": 2.538152610441767,
+      "grad_norm": 0.19417866975855358,
+      "learning_rate": 8.516974389517571e-06,
+      "loss": 0.2781,
+      "step": 1581
+    },
+    {
+      "epoch": 2.539759036144578,
+      "grad_norm": 0.18478437068715128,
+      "learning_rate": 8.487194758784992e-06,
+      "loss": 0.2826,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5413654618473895,
+      "grad_norm": 0.19700676917349086,
+      "learning_rate": 8.457415128052412e-06,
+      "loss": 0.2783,
+      "step": 1583
+    },
+    {
+      "epoch": 2.542971887550201,
+      "grad_norm": 0.19512337491230494,
+      "learning_rate": 8.427635497319834e-06,
+      "loss": 0.2708,
+      "step": 1584
+    },
+    {
+      "epoch": 2.544578313253012,
+      "grad_norm": 0.1875799539925133,
+      "learning_rate": 8.397855866587255e-06,
+      "loss": 0.2773,
+      "step": 1585
+    },
+    {
+      "epoch": 2.5461847389558234,
+      "grad_norm": 0.18076364858709523,
+      "learning_rate": 8.368076235854676e-06,
+      "loss": 0.288,
+      "step": 1586
+    },
+    {
+      "epoch": 2.5477911646586344,
+      "grad_norm": 0.18428627299969846,
+      "learning_rate": 8.338296605122096e-06,
+      "loss": 0.2973,
+      "step": 1587
+    },
+    {
+      "epoch": 2.549397590361446,
+      "grad_norm": 0.18854751285290158,
+      "learning_rate": 8.308516974389518e-06,
+      "loss": 0.2815,
+      "step": 1588
+    },
+    {
+      "epoch": 2.551004016064257,
+      "grad_norm": 0.18930031139579048,
+      "learning_rate": 8.278737343656939e-06,
+      "loss": 0.2879,
+      "step": 1589
+    },
+    {
+      "epoch": 2.5526104417670683,
+      "grad_norm": 0.18859917440773347,
+      "learning_rate": 8.24895771292436e-06,
+      "loss": 0.282,
+      "step": 1590
+    },
+    {
+      "epoch": 2.5542168674698793,
+      "grad_norm": 0.18547151485453509,
+      "learning_rate": 8.21917808219178e-06,
+      "loss": 0.2921,
+      "step": 1591
+    },
+    {
+      "epoch": 2.5558232931726907,
+      "grad_norm": 0.18825341883711819,
+      "learning_rate": 8.189398451459202e-06,
+      "loss": 0.2694,
+      "step": 1592
+    },
+    {
+      "epoch": 2.557429718875502,
+      "grad_norm": 0.1804570932899742,
+      "learning_rate": 8.159618820726623e-06,
+      "loss": 0.2758,
+      "step": 1593
+    },
+    {
+      "epoch": 2.559036144578313,
+      "grad_norm": 0.19094162633691125,
+      "learning_rate": 8.129839189994045e-06,
+      "loss": 0.2676,
+      "step": 1594
+    },
+    {
+      "epoch": 2.5606425702811246,
+      "grad_norm": 0.19021680588033746,
+      "learning_rate": 8.100059559261466e-06,
+      "loss": 0.2768,
+      "step": 1595
+    },
+    {
+      "epoch": 2.5622489959839356,
+      "grad_norm": 0.21083832609587919,
+      "learning_rate": 8.070279928528888e-06,
+      "loss": 0.2983,
+      "step": 1596
+    },
+    {
+      "epoch": 2.563855421686747,
+      "grad_norm": 0.1830825332075518,
+      "learning_rate": 8.040500297796308e-06,
+      "loss": 0.2692,
+      "step": 1597
+    },
+    {
+      "epoch": 2.565461847389558,
+      "grad_norm": 0.18269990559890056,
+      "learning_rate": 8.010720667063729e-06,
+      "loss": 0.2925,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5670682730923695,
+      "grad_norm": 0.19413648799393457,
+      "learning_rate": 7.98094103633115e-06,
+      "loss": 0.2847,
+      "step": 1599
+    },
+    {
+      "epoch": 2.5686746987951805,
+      "grad_norm": 0.19049420255904484,
+      "learning_rate": 7.951161405598572e-06,
+      "loss": 0.2784,
+      "step": 1600
+    },
+    {
+      "epoch": 2.570281124497992,
+      "grad_norm": 0.18580716065517364,
+      "learning_rate": 7.921381774865992e-06,
+      "loss": 0.2775,
+      "step": 1601
+    },
+    {
+      "epoch": 2.5718875502008034,
+      "grad_norm": 0.18486961647225106,
+      "learning_rate": 7.891602144133413e-06,
+      "loss": 0.2781,
+      "step": 1602
+    },
+    {
+      "epoch": 2.5734939759036144,
+      "grad_norm": 0.18132379562904336,
+      "learning_rate": 7.861822513400833e-06,
+      "loss": 0.2792,
+      "step": 1603
+    },
+    {
+      "epoch": 2.575100401606426,
+      "grad_norm": 0.18415948796334825,
+      "learning_rate": 7.832042882668256e-06,
+      "loss": 0.2769,
+      "step": 1604
+    },
+    {
+      "epoch": 2.576706827309237,
+      "grad_norm": 0.19136470915187226,
+      "learning_rate": 7.802263251935676e-06,
+      "loss": 0.2783,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5783132530120483,
+      "grad_norm": 0.20649143995929636,
+      "learning_rate": 7.772483621203097e-06,
+      "loss": 0.2873,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5799196787148593,
+      "grad_norm": 0.19068669901482754,
+      "learning_rate": 7.742703990470517e-06,
+      "loss": 0.2809,
+      "step": 1607
+    },
+    {
+      "epoch": 2.5815261044176707,
+      "grad_norm": 0.18357700581539366,
+      "learning_rate": 7.71292435973794e-06,
+      "loss": 0.2902,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5831325301204817,
+      "grad_norm": 0.18669030669393102,
+      "learning_rate": 7.68314472900536e-06,
+      "loss": 0.292,
+      "step": 1609
+    },
+    {
+      "epoch": 2.584738955823293,
+      "grad_norm": 0.22347819315091055,
+      "learning_rate": 7.65336509827278e-06,
+      "loss": 0.2946,
+      "step": 1610
+    },
+    {
+      "epoch": 2.5863453815261046,
+      "grad_norm": 0.18607762756921192,
+      "learning_rate": 7.623585467540204e-06,
+      "loss": 0.2705,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5879518072289156,
+      "grad_norm": 0.1958821500543021,
+      "learning_rate": 7.593805836807624e-06,
+      "loss": 0.2966,
+      "step": 1612
+    },
+    {
+      "epoch": 2.589558232931727,
+      "grad_norm": 0.1696731750041806,
+      "learning_rate": 7.564026206075045e-06,
+      "loss": 0.2552,
+      "step": 1613
+    },
+    {
+      "epoch": 2.591164658634538,
+      "grad_norm": 0.18225337475718914,
+      "learning_rate": 7.5342465753424655e-06,
+      "loss": 0.2722,
+      "step": 1614
+    },
+    {
+      "epoch": 2.5927710843373495,
+      "grad_norm": 0.19685522891076973,
+      "learning_rate": 7.504466944609888e-06,
+      "loss": 0.2832,
+      "step": 1615
+    },
+    {
+      "epoch": 2.5943775100401605,
+      "grad_norm": 0.17949581001712991,
+      "learning_rate": 7.474687313877308e-06,
+      "loss": 0.2733,
+      "step": 1616
+    },
+    {
+      "epoch": 2.595983935742972,
+      "grad_norm": 0.18660771363916165,
+      "learning_rate": 7.44490768314473e-06,
+      "loss": 0.2745,
+      "step": 1617
+    },
+    {
+      "epoch": 2.597590361445783,
+      "grad_norm": 0.18638889562090943,
+      "learning_rate": 7.41512805241215e-06,
+      "loss": 0.2689,
+      "step": 1618
+    },
+    {
+      "epoch": 2.5991967871485944,
+      "grad_norm": 0.18704008147122786,
+      "learning_rate": 7.3853484216795725e-06,
+      "loss": 0.2759,
+      "step": 1619
+    },
+    {
+      "epoch": 2.600803212851406,
+      "grad_norm": 0.18443469365017764,
+      "learning_rate": 7.355568790946993e-06,
+      "loss": 0.2658,
+      "step": 1620
+    },
+    {
+      "epoch": 2.602409638554217,
+      "grad_norm": 0.19062879189349496,
+      "learning_rate": 7.3257891602144136e-06,
+      "loss": 0.2788,
+      "step": 1621
+    },
+    {
+      "epoch": 2.6040160642570283,
+      "grad_norm": 0.1899150420248041,
+      "learning_rate": 7.296009529481834e-06,
+      "loss": 0.278,
+      "step": 1622
+    },
+    {
+      "epoch": 2.6056224899598392,
+      "grad_norm": 0.1949125724322077,
+      "learning_rate": 7.266229898749256e-06,
+      "loss": 0.2899,
+      "step": 1623
+    },
+    {
+      "epoch": 2.6072289156626507,
+      "grad_norm": 0.1699098074378404,
+      "learning_rate": 7.236450268016677e-06,
+      "loss": 0.2794,
+      "step": 1624
+    },
+    {
+      "epoch": 2.6088353413654617,
+      "grad_norm": 0.1915104653537304,
+      "learning_rate": 7.2066706372840975e-06,
+      "loss": 0.2787,
+      "step": 1625
+    },
+    {
+      "epoch": 2.610441767068273,
+      "grad_norm": 0.197531787181883,
+      "learning_rate": 7.176891006551519e-06,
+      "loss": 0.277,
+      "step": 1626
+    },
+    {
+      "epoch": 2.612048192771084,
+      "grad_norm": 0.1791632253109545,
+      "learning_rate": 7.14711137581894e-06,
+      "loss": 0.284,
+      "step": 1627
+    },
+    {
+      "epoch": 2.6136546184738956,
+      "grad_norm": 0.1882752270203241,
+      "learning_rate": 7.117331745086362e-06,
+      "loss": 0.2852,
+      "step": 1628
+    },
+    {
+      "epoch": 2.615261044176707,
+      "grad_norm": 0.17840672102386784,
+      "learning_rate": 7.087552114353782e-06,
+      "loss": 0.2613,
+      "step": 1629
+    },
+    {
+      "epoch": 2.616867469879518,
+      "grad_norm": 0.18391588960980307,
+      "learning_rate": 7.057772483621203e-06,
+      "loss": 0.2797,
+      "step": 1630
+    },
+    {
+      "epoch": 2.6184738955823295,
+      "grad_norm": 0.18819826058200362,
+      "learning_rate": 7.027992852888625e-06,
+      "loss": 0.2874,
+      "step": 1631
+    },
+    {
+      "epoch": 2.6200803212851405,
+      "grad_norm": 0.18523122078571658,
+      "learning_rate": 6.9982132221560456e-06,
+      "loss": 0.282,
+      "step": 1632
+    },
+    {
+      "epoch": 2.621686746987952,
+      "grad_norm": 0.17922978175589946,
+      "learning_rate": 6.968433591423466e-06,
+      "loss": 0.2735,
+      "step": 1633
+    },
+    {
+      "epoch": 2.623293172690763,
+      "grad_norm": 0.18372374618188894,
+      "learning_rate": 6.938653960690887e-06,
+      "loss": 0.2808,
+      "step": 1634
+    },
+    {
+      "epoch": 2.6248995983935743,
+      "grad_norm": 0.1990798931781952,
+      "learning_rate": 6.908874329958309e-06,
+      "loss": 0.2864,
+      "step": 1635
+    },
+    {
+      "epoch": 2.6265060240963853,
+      "grad_norm": 0.18455806479532721,
+      "learning_rate": 6.87909469922573e-06,
+      "loss": 0.2774,
+      "step": 1636
+    },
+    {
+      "epoch": 2.628112449799197,
+      "grad_norm": 0.1807836176874832,
+      "learning_rate": 6.849315068493151e-06,
+      "loss": 0.2879,
+      "step": 1637
+    },
+    {
+      "epoch": 2.6297188755020082,
+      "grad_norm": 0.19383858435079449,
+      "learning_rate": 6.819535437760573e-06,
+      "loss": 0.2858,
+      "step": 1638
+    },
+    {
+      "epoch": 2.6313253012048192,
+      "grad_norm": 0.19245116237163992,
+      "learning_rate": 6.789755807027994e-06,
+      "loss": 0.2831,
+      "step": 1639
+    },
+    {
+      "epoch": 2.6329317269076307,
+      "grad_norm": 0.17990454268365383,
+      "learning_rate": 6.759976176295414e-06,
+      "loss": 0.2803,
+      "step": 1640
+    },
+    {
+      "epoch": 2.6345381526104417,
+      "grad_norm": 0.17767290721583814,
+      "learning_rate": 6.730196545562835e-06,
+      "loss": 0.281,
+      "step": 1641
+    },
+    {
+      "epoch": 2.636144578313253,
+      "grad_norm": 0.19991589531881362,
+      "learning_rate": 6.700416914830257e-06,
+      "loss": 0.2835,
+      "step": 1642
+    },
+    {
+      "epoch": 2.637751004016064,
+      "grad_norm": 0.18699080009874672,
+      "learning_rate": 6.670637284097678e-06,
+      "loss": 0.284,
+      "step": 1643
+    },
+    {
+      "epoch": 2.6393574297188755,
+      "grad_norm": 0.17930928117500708,
+      "learning_rate": 6.640857653365098e-06,
+      "loss": 0.2786,
+      "step": 1644
+    },
+    {
+      "epoch": 2.6409638554216865,
+      "grad_norm": 0.19453857949440811,
+      "learning_rate": 6.6110780226325195e-06,
+      "loss": 0.2734,
+      "step": 1645
+    },
+    {
+      "epoch": 2.642570281124498,
+      "grad_norm": 0.17684358944960415,
+      "learning_rate": 6.581298391899941e-06,
+      "loss": 0.2666,
+      "step": 1646
+    },
+    {
+      "epoch": 2.6441767068273094,
+      "grad_norm": 0.17025227550260216,
+      "learning_rate": 6.551518761167362e-06,
+      "loss": 0.2845,
+      "step": 1647
+    },
+    {
+      "epoch": 2.6457831325301204,
+      "grad_norm": 0.17834204727479083,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.2673,
+      "step": 1648
+    },
+    {
+      "epoch": 2.647389558232932,
+      "grad_norm": 0.18912658642714128,
+      "learning_rate": 6.4919594997022034e-06,
+      "loss": 0.2819,
+      "step": 1649
+    },
+    {
+      "epoch": 2.648995983935743,
+      "grad_norm": 0.1737681616774698,
+      "learning_rate": 6.462179868969626e-06,
+      "loss": 0.2811,
+      "step": 1650
+    },
+    {
+      "epoch": 2.6506024096385543,
+      "grad_norm": 0.17286827101324814,
+      "learning_rate": 6.432400238237046e-06,
+      "loss": 0.2758,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6522088353413653,
+      "grad_norm": 0.19212478696833551,
+      "learning_rate": 6.402620607504467e-06,
+      "loss": 0.2969,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6538152610441768,
+      "grad_norm": 0.17906924074942082,
+      "learning_rate": 6.372840976771887e-06,
+      "loss": 0.2749,
+      "step": 1653
+    },
+    {
+      "epoch": 2.6554216867469878,
+      "grad_norm": 0.18947641365558332,
+      "learning_rate": 6.34306134603931e-06,
+      "loss": 0.2822,
+      "step": 1654
+    },
+    {
+      "epoch": 2.657028112449799,
+      "grad_norm": 0.17356368064329025,
+      "learning_rate": 6.313281715306731e-06,
+      "loss": 0.2825,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6586345381526106,
+      "grad_norm": 0.18057458726583847,
+      "learning_rate": 6.2835020845741515e-06,
+      "loss": 0.2825,
+      "step": 1656
+    },
+    {
+      "epoch": 2.6602409638554216,
+      "grad_norm": 0.1877542933350748,
+      "learning_rate": 6.253722453841572e-06,
+      "loss": 0.2914,
+      "step": 1657
+    },
+    {
+      "epoch": 2.661847389558233,
+      "grad_norm": 0.17969877942256923,
+      "learning_rate": 6.2239428231089935e-06,
+      "loss": 0.2829,
+      "step": 1658
+    },
+    {
+      "epoch": 2.663453815261044,
+      "grad_norm": 0.17870162644623386,
+      "learning_rate": 6.194163192376415e-06,
+      "loss": 0.2639,
+      "step": 1659
+    },
+    {
+      "epoch": 2.6650602409638555,
+      "grad_norm": 0.17385453495340514,
+      "learning_rate": 6.1643835616438354e-06,
+      "loss": 0.2822,
+      "step": 1660
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.17484147276070983,
+      "learning_rate": 6.134603930911257e-06,
+      "loss": 0.268,
+      "step": 1661
+    },
+    {
+      "epoch": 2.668273092369478,
+      "grad_norm": 0.1825438106082261,
+      "learning_rate": 6.104824300178678e-06,
+      "loss": 0.2854,
+      "step": 1662
+    },
+    {
+      "epoch": 2.669879518072289,
+      "grad_norm": 0.1836659740160302,
+      "learning_rate": 6.075044669446099e-06,
+      "loss": 0.2762,
+      "step": 1663
+    },
+    {
+      "epoch": 2.6714859437751004,
+      "grad_norm": 0.19695563144999662,
+      "learning_rate": 6.04526503871352e-06,
+      "loss": 0.2793,
+      "step": 1664
+    },
+    {
+      "epoch": 2.673092369477912,
+      "grad_norm": 0.1791319681791505,
+      "learning_rate": 6.015485407980942e-06,
+      "loss": 0.2773,
+      "step": 1665
+    },
+    {
+      "epoch": 2.674698795180723,
+      "grad_norm": 0.18117849182348353,
+      "learning_rate": 5.985705777248363e-06,
+      "loss": 0.2782,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6763052208835343,
+      "grad_norm": 0.18394861678586363,
+      "learning_rate": 5.9559261465157836e-06,
+      "loss": 0.2724,
+      "step": 1667
+    },
+    {
+      "epoch": 2.6779116465863453,
+      "grad_norm": 0.17558775366985108,
+      "learning_rate": 5.926146515783205e-06,
+      "loss": 0.2763,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6795180722891567,
+      "grad_norm": 0.1817180623536754,
+      "learning_rate": 5.8963668850506255e-06,
+      "loss": 0.2812,
+      "step": 1669
+    },
+    {
+      "epoch": 2.6811244979919677,
+      "grad_norm": 0.1794961501282779,
+      "learning_rate": 5.866587254318047e-06,
+      "loss": 0.2808,
+      "step": 1670
+    },
+    {
+      "epoch": 2.682730923694779,
+      "grad_norm": 0.17953298862907435,
+      "learning_rate": 5.8368076235854675e-06,
+      "loss": 0.2897,
+      "step": 1671
+    },
+    {
+      "epoch": 2.68433734939759,
+      "grad_norm": 0.1781021623495441,
+      "learning_rate": 5.807027992852889e-06,
+      "loss": 0.2774,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6859437751004016,
+      "grad_norm": 0.19580053060440614,
+      "learning_rate": 5.777248362120309e-06,
+      "loss": 0.2828,
+      "step": 1673
+    },
+    {
+      "epoch": 2.687550200803213,
+      "grad_norm": 0.18017972515436526,
+      "learning_rate": 5.747468731387731e-06,
+      "loss": 0.282,
+      "step": 1674
+    },
+    {
+      "epoch": 2.689156626506024,
+      "grad_norm": 0.17822514747203339,
+      "learning_rate": 5.717689100655152e-06,
+      "loss": 0.2773,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6907630522088355,
+      "grad_norm": 0.18518154759023206,
+      "learning_rate": 5.687909469922574e-06,
+      "loss": 0.283,
+      "step": 1676
+    },
+    {
+      "epoch": 2.6923694779116465,
+      "grad_norm": 0.19759436455167156,
+      "learning_rate": 5.658129839189994e-06,
+      "loss": 0.2893,
+      "step": 1677
+    },
+    {
+      "epoch": 2.693975903614458,
+      "grad_norm": 0.17831832185569604,
+      "learning_rate": 5.6283502084574156e-06,
+      "loss": 0.2788,
+      "step": 1678
+    },
+    {
+      "epoch": 2.695582329317269,
+      "grad_norm": 0.1822729335393522,
+      "learning_rate": 5.598570577724836e-06,
+      "loss": 0.2815,
+      "step": 1679
+    },
+    {
+      "epoch": 2.6971887550200804,
+      "grad_norm": 0.18637284782160773,
+      "learning_rate": 5.5687909469922575e-06,
+      "loss": 0.2754,
+      "step": 1680
+    },
+    {
+      "epoch": 2.6987951807228914,
+      "grad_norm": 0.21541418445055555,
+      "learning_rate": 5.539011316259678e-06,
+      "loss": 0.2699,
+      "step": 1681
+    },
+    {
+      "epoch": 2.700401606425703,
+      "grad_norm": 0.184036491711288,
+      "learning_rate": 5.5092316855270995e-06,
+      "loss": 0.2719,
+      "step": 1682
+    },
+    {
+      "epoch": 2.7020080321285143,
+      "grad_norm": 0.18589778378254496,
+      "learning_rate": 5.479452054794521e-06,
+      "loss": 0.2707,
+      "step": 1683
+    },
+    {
+      "epoch": 2.7036144578313253,
+      "grad_norm": 0.19434906672779567,
+      "learning_rate": 5.449672424061942e-06,
+      "loss": 0.3102,
+      "step": 1684
+    },
+    {
+      "epoch": 2.7052208835341367,
+      "grad_norm": 0.19696554447531195,
+      "learning_rate": 5.419892793329363e-06,
+      "loss": 0.2913,
+      "step": 1685
+    },
+    {
+      "epoch": 2.7068273092369477,
+      "grad_norm": 0.1872426169089665,
+      "learning_rate": 5.390113162596784e-06,
+      "loss": 0.2812,
+      "step": 1686
+    },
+    {
+      "epoch": 2.708433734939759,
+      "grad_norm": 0.18592737019395528,
+      "learning_rate": 5.360333531864206e-06,
+      "loss": 0.2693,
+      "step": 1687
+    },
+    {
+      "epoch": 2.71004016064257,
+      "grad_norm": 0.19133953951267615,
+      "learning_rate": 5.330553901131626e-06,
+      "loss": 0.2874,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7116465863453816,
+      "grad_norm": 0.1910105265401887,
+      "learning_rate": 5.3007742703990476e-06,
+      "loss": 0.2839,
+      "step": 1689
+    },
+    {
+      "epoch": 2.7132530120481926,
+      "grad_norm": 0.20107523425093335,
+      "learning_rate": 5.270994639666468e-06,
+      "loss": 0.2848,
+      "step": 1690
+    },
+    {
+      "epoch": 2.714859437751004,
+      "grad_norm": 0.18131315076806906,
+      "learning_rate": 5.2412150089338895e-06,
+      "loss": 0.2936,
+      "step": 1691
+    },
+    {
+      "epoch": 2.7164658634538155,
+      "grad_norm": 0.17998179499764488,
+      "learning_rate": 5.21143537820131e-06,
+      "loss": 0.2726,
+      "step": 1692
+    },
+    {
+      "epoch": 2.7180722891566265,
+      "grad_norm": 0.17958592327392905,
+      "learning_rate": 5.1816557474687315e-06,
+      "loss": 0.2852,
+      "step": 1693
+    },
+    {
+      "epoch": 2.719678714859438,
+      "grad_norm": 0.183308753969173,
+      "learning_rate": 5.151876116736153e-06,
+      "loss": 0.2844,
+      "step": 1694
+    },
+    {
+      "epoch": 2.721285140562249,
+      "grad_norm": 0.19652975978403583,
+      "learning_rate": 5.122096486003574e-06,
+      "loss": 0.276,
+      "step": 1695
+    },
+    {
+      "epoch": 2.7228915662650603,
+      "grad_norm": 0.1862766900571352,
+      "learning_rate": 5.092316855270995e-06,
+      "loss": 0.291,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7244979919678713,
+      "grad_norm": 0.1789975019779833,
+      "learning_rate": 5.062537224538416e-06,
+      "loss": 0.2678,
+      "step": 1697
+    },
+    {
+      "epoch": 2.726104417670683,
+      "grad_norm": 0.1804473631478777,
+      "learning_rate": 5.032757593805837e-06,
+      "loss": 0.2768,
+      "step": 1698
+    },
+    {
+      "epoch": 2.727710843373494,
+      "grad_norm": 0.17484370027170906,
+      "learning_rate": 5.002977963073258e-06,
+      "loss": 0.2745,
+      "step": 1699
+    },
+    {
+      "epoch": 2.7293172690763052,
+      "grad_norm": 0.18840033620754235,
+      "learning_rate": 4.973198332340679e-06,
+      "loss": 0.2935,
+      "step": 1700
+    },
+    {
+      "epoch": 2.7309236947791167,
+      "grad_norm": 0.1774236454463638,
+      "learning_rate": 4.9434187016081e-06,
+      "loss": 0.2861,
+      "step": 1701
+    },
+    {
+      "epoch": 2.7325301204819277,
+      "grad_norm": 0.174204170118579,
+      "learning_rate": 4.9136390708755215e-06,
+      "loss": 0.2874,
+      "step": 1702
+    },
+    {
+      "epoch": 2.734136546184739,
+      "grad_norm": 0.18746916237309122,
+      "learning_rate": 4.883859440142943e-06,
+      "loss": 0.2988,
+      "step": 1703
+    },
+    {
+      "epoch": 2.73574297188755,
+      "grad_norm": 0.18760654275484376,
+      "learning_rate": 4.8540798094103635e-06,
+      "loss": 0.3052,
+      "step": 1704
+    },
+    {
+      "epoch": 2.7373493975903616,
+      "grad_norm": 0.18517560114201603,
+      "learning_rate": 4.824300178677785e-06,
+      "loss": 0.2854,
+      "step": 1705
+    },
+    {
+      "epoch": 2.7389558232931726,
+      "grad_norm": 0.1876520550357728,
+      "learning_rate": 4.7945205479452054e-06,
+      "loss": 0.2932,
+      "step": 1706
+    },
+    {
+      "epoch": 2.740562248995984,
+      "grad_norm": 0.17966388420192841,
+      "learning_rate": 4.764740917212627e-06,
+      "loss": 0.28,
+      "step": 1707
+    },
+    {
+      "epoch": 2.742168674698795,
+      "grad_norm": 0.179020115591498,
+      "learning_rate": 4.734961286480047e-06,
+      "loss": 0.2864,
+      "step": 1708
+    },
+    {
+      "epoch": 2.7437751004016064,
+      "grad_norm": 0.1817568004265401,
+      "learning_rate": 4.705181655747469e-06,
+      "loss": 0.2923,
+      "step": 1709
+    },
+    {
+      "epoch": 2.745381526104418,
+      "grad_norm": 0.177238610054722,
+      "learning_rate": 4.675402025014889e-06,
+      "loss": 0.2762,
+      "step": 1710
+    },
+    {
+      "epoch": 2.746987951807229,
+      "grad_norm": 0.1896915852118844,
+      "learning_rate": 4.645622394282311e-06,
+      "loss": 0.2784,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7485943775100403,
+      "grad_norm": 0.1942534377604623,
+      "learning_rate": 4.615842763549732e-06,
+      "loss": 0.2856,
+      "step": 1712
+    },
+    {
+      "epoch": 2.7502008032128513,
+      "grad_norm": 0.18421086652363167,
+      "learning_rate": 4.5860631328171535e-06,
+      "loss": 0.2779,
+      "step": 1713
+    },
+    {
+      "epoch": 2.7518072289156628,
+      "grad_norm": 0.17539043641917704,
+      "learning_rate": 4.556283502084575e-06,
+      "loss": 0.2826,
+      "step": 1714
+    },
+    {
+      "epoch": 2.7534136546184738,
+      "grad_norm": 0.1744055331592788,
+      "learning_rate": 4.5265038713519955e-06,
+      "loss": 0.2731,
+      "step": 1715
+    },
+    {
+      "epoch": 2.755020080321285,
+      "grad_norm": 0.1724485729429514,
+      "learning_rate": 4.496724240619417e-06,
+      "loss": 0.2882,
+      "step": 1716
+    },
+    {
+      "epoch": 2.756626506024096,
+      "grad_norm": 0.18273987004893796,
+      "learning_rate": 4.4669446098868375e-06,
+      "loss": 0.2851,
+      "step": 1717
+    },
+    {
+      "epoch": 2.7582329317269076,
+      "grad_norm": 0.182839997686776,
+      "learning_rate": 4.437164979154259e-06,
+      "loss": 0.2718,
+      "step": 1718
+    },
+    {
+      "epoch": 2.759839357429719,
+      "grad_norm": 0.18031944958924873,
+      "learning_rate": 4.407385348421679e-06,
+      "loss": 0.2852,
+      "step": 1719
+    },
+    {
+      "epoch": 2.76144578313253,
+      "grad_norm": 0.18304611053583184,
+      "learning_rate": 4.377605717689101e-06,
+      "loss": 0.2753,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7630522088353415,
+      "grad_norm": 0.18653131767935718,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.2912,
+      "step": 1721
+    },
+    {
+      "epoch": 2.7646586345381525,
+      "grad_norm": 0.18280709251927277,
+      "learning_rate": 4.318046456223944e-06,
+      "loss": 0.2814,
+      "step": 1722
+    },
+    {
+      "epoch": 2.766265060240964,
+      "grad_norm": 0.1730903954482395,
+      "learning_rate": 4.288266825491364e-06,
+      "loss": 0.2782,
+      "step": 1723
+    },
+    {
+      "epoch": 2.767871485943775,
+      "grad_norm": 0.1860699548269679,
+      "learning_rate": 4.2584871947587856e-06,
+      "loss": 0.273,
+      "step": 1724
+    },
+    {
+      "epoch": 2.7694779116465864,
+      "grad_norm": 0.1781795487696371,
+      "learning_rate": 4.228707564026206e-06,
+      "loss": 0.2807,
+      "step": 1725
+    },
+    {
+      "epoch": 2.7710843373493974,
+      "grad_norm": 0.18497222231295918,
+      "learning_rate": 4.1989279332936275e-06,
+      "loss": 0.2904,
+      "step": 1726
+    },
+    {
+      "epoch": 2.772690763052209,
+      "grad_norm": 0.20870406694142798,
+      "learning_rate": 4.169148302561048e-06,
+      "loss": 0.2924,
+      "step": 1727
+    },
+    {
+      "epoch": 2.7742971887550203,
+      "grad_norm": 0.18032970445124402,
+      "learning_rate": 4.1393686718284695e-06,
+      "loss": 0.2752,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7759036144578313,
+      "grad_norm": 0.17508172159384913,
+      "learning_rate": 4.10958904109589e-06,
+      "loss": 0.2726,
+      "step": 1729
+    },
+    {
+      "epoch": 2.7775100401606427,
+      "grad_norm": 0.17555459903201376,
+      "learning_rate": 4.079809410363311e-06,
+      "loss": 0.2855,
+      "step": 1730
+    },
+    {
+      "epoch": 2.7791164658634537,
+      "grad_norm": 0.17205832766295667,
+      "learning_rate": 4.050029779630733e-06,
+      "loss": 0.2657,
+      "step": 1731
+    },
+    {
+      "epoch": 2.780722891566265,
+      "grad_norm": 0.1829609417989265,
+      "learning_rate": 4.020250148898154e-06,
+      "loss": 0.2855,
+      "step": 1732
+    },
+    {
+      "epoch": 2.782329317269076,
+      "grad_norm": 0.17878692408711844,
+      "learning_rate": 3.990470518165575e-06,
+      "loss": 0.2793,
+      "step": 1733
+    },
+    {
+      "epoch": 2.7839357429718876,
+      "grad_norm": 0.19508318768702912,
+      "learning_rate": 3.960690887432996e-06,
+      "loss": 0.2842,
+      "step": 1734
+    },
+    {
+      "epoch": 2.7855421686746986,
+      "grad_norm": 0.17987705205522847,
+      "learning_rate": 3.930911256700417e-06,
+      "loss": 0.2763,
+      "step": 1735
+    },
+    {
+      "epoch": 2.78714859437751,
+      "grad_norm": 0.17185264560334,
+      "learning_rate": 3.901131625967838e-06,
+      "loss": 0.2793,
+      "step": 1736
+    },
+    {
+      "epoch": 2.7887550200803215,
+      "grad_norm": 0.1798915631146029,
+      "learning_rate": 3.871351995235259e-06,
+      "loss": 0.2823,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7903614457831325,
+      "grad_norm": 0.17936783365240172,
+      "learning_rate": 3.84157236450268e-06,
+      "loss": 0.2923,
+      "step": 1738
+    },
+    {
+      "epoch": 2.7919678714859435,
+      "grad_norm": 0.17881359287676532,
+      "learning_rate": 3.811792733770102e-06,
+      "loss": 0.283,
+      "step": 1739
+    },
+    {
+      "epoch": 2.793574297188755,
+      "grad_norm": 0.17750412281273836,
+      "learning_rate": 3.7820131030375225e-06,
+      "loss": 0.2774,
+      "step": 1740
+    },
+    {
+      "epoch": 2.7951807228915664,
+      "grad_norm": 0.17978912748152973,
+      "learning_rate": 3.752233472304944e-06,
+      "loss": 0.3089,
+      "step": 1741
+    },
+    {
+      "epoch": 2.7967871485943774,
+      "grad_norm": 0.1826655687491039,
+      "learning_rate": 3.722453841572365e-06,
+      "loss": 0.2709,
+      "step": 1742
+    },
+    {
+      "epoch": 2.798393574297189,
+      "grad_norm": 0.18001177870589455,
+      "learning_rate": 3.6926742108397862e-06,
+      "loss": 0.2737,
+      "step": 1743
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.17713047219916325,
+      "learning_rate": 3.6628945801072068e-06,
+      "loss": 0.2752,
+      "step": 1744
+    },
+    {
+      "epoch": 2.8016064257028113,
+      "grad_norm": 0.18298092591553738,
+      "learning_rate": 3.633114949374628e-06,
+      "loss": 0.2873,
+      "step": 1745
+    },
+    {
+      "epoch": 2.8032128514056227,
+      "grad_norm": 0.17757847821734857,
+      "learning_rate": 3.6033353186420487e-06,
+      "loss": 0.2846,
+      "step": 1746
+    },
+    {
+      "epoch": 2.8048192771084337,
+      "grad_norm": 0.18958642415753324,
+      "learning_rate": 3.57355568790947e-06,
+      "loss": 0.2926,
+      "step": 1747
+    },
+    {
+      "epoch": 2.8064257028112447,
+      "grad_norm": 0.18146180673271486,
+      "learning_rate": 3.543776057176891e-06,
+      "loss": 0.2791,
+      "step": 1748
+    },
+    {
+      "epoch": 2.808032128514056,
+      "grad_norm": 0.17453241970228875,
+      "learning_rate": 3.5139964264443125e-06,
+      "loss": 0.2724,
+      "step": 1749
+    },
+    {
+      "epoch": 2.8096385542168676,
+      "grad_norm": 0.17966834865117448,
+      "learning_rate": 3.484216795711733e-06,
+      "loss": 0.2787,
+      "step": 1750
+    },
+    {
+      "epoch": 2.8112449799196786,
+      "grad_norm": 0.1784681684891437,
+      "learning_rate": 3.4544371649791545e-06,
+      "loss": 0.2764,
+      "step": 1751
+    },
+    {
+      "epoch": 2.81285140562249,
+      "grad_norm": 0.18572317206096217,
+      "learning_rate": 3.4246575342465754e-06,
+      "loss": 0.2709,
+      "step": 1752
+    },
+    {
+      "epoch": 2.814457831325301,
+      "grad_norm": 0.1759488290957744,
+      "learning_rate": 3.394877903513997e-06,
+      "loss": 0.2815,
+      "step": 1753
+    },
+    {
+      "epoch": 2.8160642570281125,
+      "grad_norm": 0.19336096846491638,
+      "learning_rate": 3.3650982727814174e-06,
+      "loss": 0.2744,
+      "step": 1754
+    },
+    {
+      "epoch": 2.817670682730924,
+      "grad_norm": 0.18283745315959676,
+      "learning_rate": 3.335318642048839e-06,
+      "loss": 0.2921,
+      "step": 1755
+    },
+    {
+      "epoch": 2.819277108433735,
+      "grad_norm": 0.17472330034447986,
+      "learning_rate": 3.3055390113162598e-06,
+      "loss": 0.2756,
+      "step": 1756
+    },
+    {
+      "epoch": 2.820883534136546,
+      "grad_norm": 0.1800313759664309,
+      "learning_rate": 3.275759380583681e-06,
+      "loss": 0.2812,
+      "step": 1757
+    },
+    {
+      "epoch": 2.8224899598393574,
+      "grad_norm": 0.1756785630336977,
+      "learning_rate": 3.2459797498511017e-06,
+      "loss": 0.2587,
+      "step": 1758
+    },
+    {
+      "epoch": 2.824096385542169,
+      "grad_norm": 0.1932923197044465,
+      "learning_rate": 3.216200119118523e-06,
+      "loss": 0.2707,
+      "step": 1759
+    },
+    {
+      "epoch": 2.82570281124498,
+      "grad_norm": 0.1840245526187382,
+      "learning_rate": 3.1864204883859437e-06,
+      "loss": 0.2837,
+      "step": 1760
+    },
+    {
+      "epoch": 2.8273092369477912,
+      "grad_norm": 0.1766729183541744,
+      "learning_rate": 3.1566408576533655e-06,
+      "loss": 0.2845,
+      "step": 1761
+    },
+    {
+      "epoch": 2.8289156626506022,
+      "grad_norm": 0.1841631130200818,
+      "learning_rate": 3.126861226920786e-06,
+      "loss": 0.2809,
+      "step": 1762
+    },
+    {
+      "epoch": 2.8305220883534137,
+      "grad_norm": 0.17038920533561577,
+      "learning_rate": 3.0970815961882074e-06,
+      "loss": 0.2872,
+      "step": 1763
+    },
+    {
+      "epoch": 2.832128514056225,
+      "grad_norm": 0.18490513608282008,
+      "learning_rate": 3.0673019654556284e-06,
+      "loss": 0.2827,
+      "step": 1764
+    },
+    {
+      "epoch": 2.833734939759036,
+      "grad_norm": 0.1793068608455851,
+      "learning_rate": 3.0375223347230494e-06,
+      "loss": 0.2915,
+      "step": 1765
+    },
+    {
+      "epoch": 2.835341365461847,
+      "grad_norm": 0.18910291135979634,
+      "learning_rate": 3.007742703990471e-06,
+      "loss": 0.2954,
+      "step": 1766
+    },
+    {
+      "epoch": 2.8369477911646586,
+      "grad_norm": 0.1794649230471941,
+      "learning_rate": 2.9779630732578918e-06,
+      "loss": 0.282,
+      "step": 1767
+    },
+    {
+      "epoch": 2.83855421686747,
+      "grad_norm": 0.17996525722863402,
+      "learning_rate": 2.9481834425253128e-06,
+      "loss": 0.2905,
+      "step": 1768
+    },
+    {
+      "epoch": 2.840160642570281,
+      "grad_norm": 0.16999154285816043,
+      "learning_rate": 2.9184038117927337e-06,
+      "loss": 0.2645,
+      "step": 1769
+    },
+    {
+      "epoch": 2.8417670682730924,
+      "grad_norm": 0.1739829332617308,
+      "learning_rate": 2.8886241810601547e-06,
+      "loss": 0.2778,
+      "step": 1770
+    },
+    {
+      "epoch": 2.8433734939759034,
+      "grad_norm": 0.1862630294313089,
+      "learning_rate": 2.858844550327576e-06,
+      "loss": 0.2734,
+      "step": 1771
+    },
+    {
+      "epoch": 2.844979919678715,
+      "grad_norm": 0.1833390723721438,
+      "learning_rate": 2.829064919594997e-06,
+      "loss": 0.3016,
+      "step": 1772
+    },
+    {
+      "epoch": 2.8465863453815263,
+      "grad_norm": 0.17222764961420253,
+      "learning_rate": 2.799285288862418e-06,
+      "loss": 0.2616,
+      "step": 1773
+    },
+    {
+      "epoch": 2.8481927710843373,
+      "grad_norm": 0.17723986699960997,
+      "learning_rate": 2.769505658129839e-06,
+      "loss": 0.2816,
+      "step": 1774
+    },
+    {
+      "epoch": 2.8497991967871483,
+      "grad_norm": 0.17550492152209196,
+      "learning_rate": 2.7397260273972604e-06,
+      "loss": 0.2999,
+      "step": 1775
+    },
+    {
+      "epoch": 2.8514056224899598,
+      "grad_norm": 0.17567847530277153,
+      "learning_rate": 2.7099463966646814e-06,
+      "loss": 0.2903,
+      "step": 1776
+    },
+    {
+      "epoch": 2.853012048192771,
+      "grad_norm": 0.1825234418627062,
+      "learning_rate": 2.680166765932103e-06,
+      "loss": 0.2873,
+      "step": 1777
+    },
+    {
+      "epoch": 2.854618473895582,
+      "grad_norm": 0.17878324566094148,
+      "learning_rate": 2.6503871351995238e-06,
+      "loss": 0.2885,
+      "step": 1778
+    },
+    {
+      "epoch": 2.8562248995983937,
+      "grad_norm": 0.18249488651514392,
+      "learning_rate": 2.6206075044669448e-06,
+      "loss": 0.2871,
+      "step": 1779
+    },
+    {
+      "epoch": 2.8578313253012047,
+      "grad_norm": 0.17760052228984852,
+      "learning_rate": 2.5908278737343657e-06,
+      "loss": 0.2717,
+      "step": 1780
+    },
+    {
+      "epoch": 2.859437751004016,
+      "grad_norm": 0.18808062189817043,
+      "learning_rate": 2.561048243001787e-06,
+      "loss": 0.2876,
+      "step": 1781
+    },
+    {
+      "epoch": 2.8610441767068275,
+      "grad_norm": 0.1823780795570633,
+      "learning_rate": 2.531268612269208e-06,
+      "loss": 0.2867,
+      "step": 1782
+    },
+    {
+      "epoch": 2.8626506024096385,
+      "grad_norm": 0.1885706803388553,
+      "learning_rate": 2.501488981536629e-06,
+      "loss": 0.2897,
+      "step": 1783
+    },
+    {
+      "epoch": 2.8642570281124495,
+      "grad_norm": 0.17633385450538436,
+      "learning_rate": 2.47170935080405e-06,
+      "loss": 0.2786,
+      "step": 1784
+    },
+    {
+      "epoch": 2.865863453815261,
+      "grad_norm": 0.18185187408978562,
+      "learning_rate": 2.4419297200714715e-06,
+      "loss": 0.2918,
+      "step": 1785
+    },
+    {
+      "epoch": 2.8674698795180724,
+      "grad_norm": 0.18061017522143138,
+      "learning_rate": 2.4121500893388924e-06,
+      "loss": 0.282,
+      "step": 1786
+    },
+    {
+      "epoch": 2.8690763052208834,
+      "grad_norm": 0.1703671192800184,
+      "learning_rate": 2.3823704586063134e-06,
+      "loss": 0.2802,
+      "step": 1787
+    },
+    {
+      "epoch": 2.870682730923695,
+      "grad_norm": 0.18089410922349705,
+      "learning_rate": 2.3525908278737344e-06,
+      "loss": 0.2762,
+      "step": 1788
+    },
+    {
+      "epoch": 2.872289156626506,
+      "grad_norm": 0.18344826515624643,
+      "learning_rate": 2.3228111971411554e-06,
+      "loss": 0.2847,
+      "step": 1789
+    },
+    {
+      "epoch": 2.8738955823293173,
+      "grad_norm": 0.18790690808141583,
+      "learning_rate": 2.2930315664085768e-06,
+      "loss": 0.2832,
+      "step": 1790
+    },
+    {
+      "epoch": 2.8755020080321287,
+      "grad_norm": 0.19268300685339843,
+      "learning_rate": 2.2632519356759978e-06,
+      "loss": 0.2924,
+      "step": 1791
+    },
+    {
+      "epoch": 2.8771084337349397,
+      "grad_norm": 0.18108361744769685,
+      "learning_rate": 2.2334723049434187e-06,
+      "loss": 0.2666,
+      "step": 1792
+    },
+    {
+      "epoch": 2.8787148594377507,
+      "grad_norm": 0.1964453513606187,
+      "learning_rate": 2.2036926742108397e-06,
+      "loss": 0.2834,
+      "step": 1793
+    },
+    {
+      "epoch": 2.880321285140562,
+      "grad_norm": 0.1788510567768747,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.2884,
+      "step": 1794
+    },
+    {
+      "epoch": 2.8819277108433736,
+      "grad_norm": 0.17295138712409047,
+      "learning_rate": 2.144133412745682e-06,
+      "loss": 0.2711,
+      "step": 1795
+    },
+    {
+      "epoch": 2.8835341365461846,
+      "grad_norm": 0.17567031763593471,
+      "learning_rate": 2.114353782013103e-06,
+      "loss": 0.2717,
+      "step": 1796
+    },
+    {
+      "epoch": 2.885140562248996,
+      "grad_norm": 0.16678222696781173,
+      "learning_rate": 2.084574151280524e-06,
+      "loss": 0.2706,
+      "step": 1797
+    },
+    {
+      "epoch": 2.886746987951807,
+      "grad_norm": 0.17470142569680275,
+      "learning_rate": 2.054794520547945e-06,
+      "loss": 0.2843,
+      "step": 1798
+    },
+    {
+      "epoch": 2.8883534136546185,
+      "grad_norm": 0.1829261532971885,
+      "learning_rate": 2.0250148898153664e-06,
+      "loss": 0.2782,
+      "step": 1799
+    },
+    {
+      "epoch": 2.88995983935743,
+      "grad_norm": 0.1722757025350079,
+      "learning_rate": 1.9952352590827874e-06,
+      "loss": 0.2774,
+      "step": 1800
+    },
+    {
+      "epoch": 2.891566265060241,
+      "grad_norm": 0.18349308042009962,
+      "learning_rate": 1.9654556283502084e-06,
+      "loss": 0.2832,
+      "step": 1801
+    },
+    {
+      "epoch": 2.893172690763052,
+      "grad_norm": 0.1754837052446264,
+      "learning_rate": 1.9356759976176293e-06,
+      "loss": 0.2822,
+      "step": 1802
+    },
+    {
+      "epoch": 2.8947791164658634,
+      "grad_norm": 0.1792463120774883,
+      "learning_rate": 1.905896366885051e-06,
+      "loss": 0.295,
+      "step": 1803
+    },
+    {
+      "epoch": 2.896385542168675,
+      "grad_norm": 0.17685710971676047,
+      "learning_rate": 1.876116736152472e-06,
+      "loss": 0.2888,
+      "step": 1804
+    },
+    {
+      "epoch": 2.897991967871486,
+      "grad_norm": 0.17535217611636308,
+      "learning_rate": 1.8463371054198931e-06,
+      "loss": 0.2783,
+      "step": 1805
+    },
+    {
+      "epoch": 2.8995983935742973,
+      "grad_norm": 0.18039479385992338,
+      "learning_rate": 1.816557474687314e-06,
+      "loss": 0.272,
+      "step": 1806
+    },
+    {
+      "epoch": 2.9012048192771083,
+      "grad_norm": 0.17519948862975526,
+      "learning_rate": 1.786777843954735e-06,
+      "loss": 0.2683,
+      "step": 1807
+    },
+    {
+      "epoch": 2.9028112449799197,
+      "grad_norm": 0.17420513363314138,
+      "learning_rate": 1.7569982132221563e-06,
+      "loss": 0.2727,
+      "step": 1808
+    },
+    {
+      "epoch": 2.904417670682731,
+      "grad_norm": 0.1744433206516152,
+      "learning_rate": 1.7272185824895772e-06,
+      "loss": 0.2783,
+      "step": 1809
+    },
+    {
+      "epoch": 2.906024096385542,
+      "grad_norm": 0.1680639127482644,
+      "learning_rate": 1.6974389517569984e-06,
+      "loss": 0.2862,
+      "step": 1810
+    },
+    {
+      "epoch": 2.907630522088353,
+      "grad_norm": 0.17920150463037057,
+      "learning_rate": 1.6676593210244194e-06,
+      "loss": 0.2869,
+      "step": 1811
+    },
+    {
+      "epoch": 2.9092369477911646,
+      "grad_norm": 0.17136726709539002,
+      "learning_rate": 1.6378796902918406e-06,
+      "loss": 0.2847,
+      "step": 1812
+    },
+    {
+      "epoch": 2.910843373493976,
+      "grad_norm": 0.17658479204211133,
+      "learning_rate": 1.6081000595592616e-06,
+      "loss": 0.2788,
+      "step": 1813
+    },
+    {
+      "epoch": 2.912449799196787,
+      "grad_norm": 0.17962465801415045,
+      "learning_rate": 1.5783204288266827e-06,
+      "loss": 0.277,
+      "step": 1814
+    },
+    {
+      "epoch": 2.9140562248995985,
+      "grad_norm": 0.17585752402402058,
+      "learning_rate": 1.5485407980941037e-06,
+      "loss": 0.2822,
+      "step": 1815
+    },
+    {
+      "epoch": 2.9156626506024095,
+      "grad_norm": 0.20014161600135602,
+      "learning_rate": 1.5187611673615247e-06,
+      "loss": 0.27,
+      "step": 1816
+    },
+    {
+      "epoch": 2.917269076305221,
+      "grad_norm": 0.18031244115450326,
+      "learning_rate": 1.4889815366289459e-06,
+      "loss": 0.2776,
+      "step": 1817
+    },
+    {
+      "epoch": 2.9188755020080324,
+      "grad_norm": 0.17203231141150208,
+      "learning_rate": 1.4592019058963669e-06,
+      "loss": 0.2709,
+      "step": 1818
+    },
+    {
+      "epoch": 2.9204819277108434,
+      "grad_norm": 0.17632883689576834,
+      "learning_rate": 1.429422275163788e-06,
+      "loss": 0.2862,
+      "step": 1819
+    },
+    {
+      "epoch": 2.9220883534136544,
+      "grad_norm": 0.17158335200283492,
+      "learning_rate": 1.399642644431209e-06,
+      "loss": 0.2805,
+      "step": 1820
+    },
+    {
+      "epoch": 2.923694779116466,
+      "grad_norm": 0.1727754387032493,
+      "learning_rate": 1.3698630136986302e-06,
+      "loss": 0.2862,
+      "step": 1821
+    },
+    {
+      "epoch": 2.9253012048192772,
+      "grad_norm": 0.17308901484014633,
+      "learning_rate": 1.3400833829660514e-06,
+      "loss": 0.2757,
+      "step": 1822
+    },
+    {
+      "epoch": 2.9269076305220882,
+      "grad_norm": 0.17421930479432718,
+      "learning_rate": 1.3103037522334724e-06,
+      "loss": 0.2834,
+      "step": 1823
+    },
+    {
+      "epoch": 2.9285140562248997,
+      "grad_norm": 0.17882219625729937,
+      "learning_rate": 1.2805241215008936e-06,
+      "loss": 0.2643,
+      "step": 1824
+    },
+    {
+      "epoch": 2.9301204819277107,
+      "grad_norm": 0.1698144392762095,
+      "learning_rate": 1.2507444907683145e-06,
+      "loss": 0.2791,
+      "step": 1825
+    },
+    {
+      "epoch": 2.931726907630522,
+      "grad_norm": 0.176433464798481,
+      "learning_rate": 1.2209648600357357e-06,
+      "loss": 0.2885,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 0.17291091781017842,
+      "learning_rate": 1.1911852293031567e-06,
+      "loss": 0.2876,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9349397590361446,
+      "grad_norm": 0.1745399160440178,
+      "learning_rate": 1.1614055985705777e-06,
+      "loss": 0.2804,
+      "step": 1828
+    },
+    {
+      "epoch": 2.9365461847389556,
+      "grad_norm": 0.17469125289305582,
+      "learning_rate": 1.1316259678379989e-06,
+      "loss": 0.2886,
+      "step": 1829
+    },
+    {
+      "epoch": 2.938152610441767,
+      "grad_norm": 0.17188888134373598,
+      "learning_rate": 1.1018463371054199e-06,
+      "loss": 0.2747,
+      "step": 1830
+    },
+    {
+      "epoch": 2.9397590361445785,
+      "grad_norm": 0.17315914977583619,
+      "learning_rate": 1.072066706372841e-06,
+      "loss": 0.2718,
+      "step": 1831
+    },
+    {
+      "epoch": 2.9413654618473895,
+      "grad_norm": 0.18424998554829677,
+      "learning_rate": 1.042287075640262e-06,
+      "loss": 0.2927,
+      "step": 1832
+    },
+    {
+      "epoch": 2.942971887550201,
+      "grad_norm": 0.17405613312679283,
+      "learning_rate": 1.0125074449076832e-06,
+      "loss": 0.2801,
+      "step": 1833
+    },
+    {
+      "epoch": 2.944578313253012,
+      "grad_norm": 0.17392389279259265,
+      "learning_rate": 9.827278141751042e-07,
+      "loss": 0.2672,
+      "step": 1834
+    },
+    {
+      "epoch": 2.9461847389558233,
+      "grad_norm": 0.1750803761886037,
+      "learning_rate": 9.529481834425255e-07,
+      "loss": 0.2753,
+      "step": 1835
+    },
+    {
+      "epoch": 2.9477911646586348,
+      "grad_norm": 0.17516743336571347,
+      "learning_rate": 9.231685527099466e-07,
+      "loss": 0.2898,
+      "step": 1836
+    },
+    {
+      "epoch": 2.9493975903614458,
+      "grad_norm": 0.17815036030448045,
+      "learning_rate": 8.933889219773675e-07,
+      "loss": 0.2843,
+      "step": 1837
+    },
+    {
+      "epoch": 2.9510040160642568,
+      "grad_norm": 0.17553406497898083,
+      "learning_rate": 8.636092912447886e-07,
+      "loss": 0.286,
+      "step": 1838
+    },
+    {
+      "epoch": 2.952610441767068,
+      "grad_norm": 0.17395921166402753,
+      "learning_rate": 8.338296605122097e-07,
+      "loss": 0.28,
+      "step": 1839
+    },
+    {
+      "epoch": 2.9542168674698797,
+      "grad_norm": 0.18520083979603033,
+      "learning_rate": 8.040500297796308e-07,
+      "loss": 0.3018,
+      "step": 1840
+    },
+    {
+      "epoch": 2.9558232931726907,
+      "grad_norm": 0.17913590766346535,
+      "learning_rate": 7.742703990470519e-07,
+      "loss": 0.2882,
+      "step": 1841
+    },
+    {
+      "epoch": 2.957429718875502,
+      "grad_norm": 0.177731379147974,
+      "learning_rate": 7.444907683144729e-07,
+      "loss": 0.291,
+      "step": 1842
+    },
+    {
+      "epoch": 2.959036144578313,
+      "grad_norm": 0.17932734702480063,
+      "learning_rate": 7.14711137581894e-07,
+      "loss": 0.2796,
+      "step": 1843
+    },
+    {
+      "epoch": 2.9606425702811245,
+      "grad_norm": 0.17871304503420032,
+      "learning_rate": 6.849315068493151e-07,
+      "loss": 0.2838,
+      "step": 1844
+    },
+    {
+      "epoch": 2.962248995983936,
+      "grad_norm": 0.17164070830208178,
+      "learning_rate": 6.551518761167362e-07,
+      "loss": 0.2827,
+      "step": 1845
+    },
+    {
+      "epoch": 2.963855421686747,
+      "grad_norm": 0.16835947713847804,
+      "learning_rate": 6.253722453841573e-07,
+      "loss": 0.2802,
+      "step": 1846
+    },
+    {
+      "epoch": 2.965461847389558,
+      "grad_norm": 0.1693631561465372,
+      "learning_rate": 5.955926146515784e-07,
+      "loss": 0.2768,
+      "step": 1847
+    },
+    {
+      "epoch": 2.9670682730923694,
+      "grad_norm": 0.17243093099494994,
+      "learning_rate": 5.658129839189994e-07,
+      "loss": 0.2757,
+      "step": 1848
+    },
+    {
+      "epoch": 2.968674698795181,
+      "grad_norm": 0.17605861774049406,
+      "learning_rate": 5.360333531864205e-07,
+      "loss": 0.2673,
+      "step": 1849
+    },
+    {
+      "epoch": 2.970281124497992,
+      "grad_norm": 0.1726383252083538,
+      "learning_rate": 5.062537224538416e-07,
+      "loss": 0.2873,
+      "step": 1850
+    },
+    {
+      "epoch": 2.9718875502008033,
+      "grad_norm": 0.1726403649864526,
+      "learning_rate": 4.7647409172126274e-07,
+      "loss": 0.2686,
+      "step": 1851
+    },
+    {
+      "epoch": 2.9734939759036143,
+      "grad_norm": 0.1809245764248369,
+      "learning_rate": 4.4669446098868377e-07,
+      "loss": 0.2838,
+      "step": 1852
+    },
+    {
+      "epoch": 2.9751004016064257,
+      "grad_norm": 0.17092605908352648,
+      "learning_rate": 4.1691483025610485e-07,
+      "loss": 0.2716,
+      "step": 1853
+    },
+    {
+      "epoch": 2.976706827309237,
+      "grad_norm": 0.17322511760116535,
+      "learning_rate": 3.8713519952352593e-07,
+      "loss": 0.2777,
+      "step": 1854
+    },
+    {
+      "epoch": 2.978313253012048,
+      "grad_norm": 0.18375085828597734,
+      "learning_rate": 3.57355568790947e-07,
+      "loss": 0.2794,
+      "step": 1855
+    },
+    {
+      "epoch": 2.979919678714859,
+      "grad_norm": 0.179038088679026,
+      "learning_rate": 3.275759380583681e-07,
+      "loss": 0.279,
+      "step": 1856
+    },
+    {
+      "epoch": 2.9815261044176706,
+      "grad_norm": 0.17767349005638372,
+      "learning_rate": 2.977963073257892e-07,
+      "loss": 0.2852,
+      "step": 1857
+    },
+    {
+      "epoch": 2.983132530120482,
+      "grad_norm": 0.18020745119137233,
+      "learning_rate": 2.6801667659321026e-07,
+      "loss": 0.3029,
+      "step": 1858
+    },
+    {
+      "epoch": 2.984738955823293,
+      "grad_norm": 0.17012300307211053,
+      "learning_rate": 2.3823704586063137e-07,
+      "loss": 0.2752,
+      "step": 1859
+    },
+    {
+      "epoch": 2.9863453815261045,
+      "grad_norm": 0.17381821055232685,
+      "learning_rate": 2.0845741512805242e-07,
+      "loss": 0.2859,
+      "step": 1860
+    },
+    {
+      "epoch": 2.9879518072289155,
+      "grad_norm": 0.17160169606492132,
+      "learning_rate": 1.786777843954735e-07,
+      "loss": 0.2707,
+      "step": 1861
+    },
+    {
+      "epoch": 2.989558232931727,
+      "grad_norm": 0.17490271990116033,
+      "learning_rate": 1.488981536628946e-07,
+      "loss": 0.289,
+      "step": 1862
+    },
+    {
+      "epoch": 2.9911646586345384,
+      "grad_norm": 0.17694062193658128,
+      "learning_rate": 1.1911852293031568e-07,
+      "loss": 0.2818,
+      "step": 1863
+    },
+    {
+      "epoch": 2.9927710843373494,
+      "grad_norm": 0.17511488153591193,
+      "learning_rate": 8.933889219773675e-08,
+      "loss": 0.2849,
+      "step": 1864
+    },
+    {
+      "epoch": 2.9943775100401604,
+      "grad_norm": 0.18167443886988274,
+      "learning_rate": 5.955926146515784e-08,
+      "loss": 0.2883,
+      "step": 1865
+    },
+    {
+      "epoch": 2.995983935742972,
+      "grad_norm": 0.1638730430540001,
+      "learning_rate": 2.977963073257892e-08,
+      "loss": 0.271,
+      "step": 1866
+    },
+    {
+      "epoch": 2.995983935742972,
+      "step": 1866,
+      "total_flos": 2.074186974392379e+19,
+      "train_loss": 0.4476111942836326,
+      "train_runtime": 54710.37,
+      "train_samples_per_second": 0.546,
+      "train_steps_per_second": 0.034
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1866,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.074186974392379e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}