diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14091 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2007,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014947683109118087,
+      "grad_norm": 52.97605206736214,
+      "learning_rate": 2.4875621890547267e-07,
+      "loss": 11.2314,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029895366218236174,
+      "grad_norm": 53.20346264564822,
+      "learning_rate": 4.975124378109453e-07,
+      "loss": 11.2027,
+      "step": 2
+    },
+    {
+      "epoch": 0.004484304932735426,
+      "grad_norm": 52.32262680258478,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 11.2714,
+      "step": 3
+    },
+    {
+      "epoch": 0.005979073243647235,
+      "grad_norm": 53.67489631830734,
+      "learning_rate": 9.950248756218907e-07,
+      "loss": 11.1742,
+      "step": 4
+    },
+    {
+      "epoch": 0.007473841554559043,
+      "grad_norm": 54.361671382803394,
+      "learning_rate": 1.2437810945273632e-06,
+      "loss": 11.1189,
+      "step": 5
+    },
+    {
+      "epoch": 0.008968609865470852,
+      "grad_norm": 53.8581220548159,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 11.1285,
+      "step": 6
+    },
+    {
+      "epoch": 0.01046337817638266,
+      "grad_norm": 54.33983519849067,
+      "learning_rate": 1.7412935323383086e-06,
+      "loss": 11.2061,
+      "step": 7
+    },
+    {
+      "epoch": 0.01195814648729447,
+      "grad_norm": 56.79667121530081,
+      "learning_rate": 1.9900497512437813e-06,
+      "loss": 10.8985,
+      "step": 8
+    },
+    {
+      "epoch": 0.013452914798206279,
+      "grad_norm": 57.54644493464062,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 10.8386,
+      "step": 9
+    },
+    {
+      "epoch": 0.014947683109118086,
+      "grad_norm": 80.67522415176944,
+      "learning_rate": 2.4875621890547264e-06,
+      "loss": 9.6661,
+      "step": 10
+    },
+    {
+      "epoch": 0.016442451420029897,
+      "grad_norm": 83.04310068292683,
+      "learning_rate": 2.736318407960199e-06,
+      "loss": 9.6005,
+      "step": 11
+    },
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 100.28897124121201,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 8.8483,
+      "step": 12
+    },
+    {
+      "epoch": 0.01943198804185351,
+      "grad_norm": 73.38668385695874,
+      "learning_rate": 3.233830845771144e-06,
+      "loss": 4.137,
+      "step": 13
+    },
+    {
+      "epoch": 0.02092675635276532,
+      "grad_norm": 60.90440912823211,
+      "learning_rate": 3.482587064676617e-06,
+      "loss": 3.6773,
+      "step": 14
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 49.97847113607188,
+      "learning_rate": 3.7313432835820893e-06,
+      "loss": 3.2139,
+      "step": 15
+    },
+    {
+      "epoch": 0.02391629297458894,
+      "grad_norm": 37.966063438895745,
+      "learning_rate": 3.980099502487563e-06,
+      "loss": 2.7046,
+      "step": 16
+    },
+    {
+      "epoch": 0.025411061285500747,
+      "grad_norm": 13.914507213864729,
+      "learning_rate": 4.228855721393035e-06,
+      "loss": 1.8255,
+      "step": 17
+    },
+    {
+      "epoch": 0.026905829596412557,
+      "grad_norm": 6.105398239613559,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 1.4353,
+      "step": 18
+    },
+    {
+      "epoch": 0.028400597907324365,
+      "grad_norm": 5.02484844884908,
+      "learning_rate": 4.72636815920398e-06,
+      "loss": 1.3793,
+      "step": 19
+    },
+    {
+      "epoch": 0.029895366218236172,
+      "grad_norm": 4.070946798935725,
+      "learning_rate": 4.975124378109453e-06,
+      "loss": 1.3733,
+      "step": 20
+    },
+    {
+      "epoch": 0.03139013452914798,
+      "grad_norm": 2.9308075429659852,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 1.2583,
+      "step": 21
+    },
+    {
+      "epoch": 0.03288490284005979,
+      "grad_norm": 2.414753481035493,
+      "learning_rate": 5.472636815920398e-06,
+      "loss": 1.2214,
+      "step": 22
+    },
+    {
+      "epoch": 0.0343796711509716,
+      "grad_norm": 1.9109986031208839,
+      "learning_rate": 5.7213930348258714e-06,
+      "loss": 1.1769,
+      "step": 23
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 7.4315385075775895,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 1.0463,
+      "step": 24
+    },
+    {
+      "epoch": 0.03736920777279522,
+      "grad_norm": 21.676397817905844,
+      "learning_rate": 6.2189054726368165e-06,
+      "loss": 1.0013,
+      "step": 25
+    },
+    {
+      "epoch": 0.03886397608370702,
+      "grad_norm": 1.9685801894369204,
+      "learning_rate": 6.467661691542288e-06,
+      "loss": 0.9601,
+      "step": 26
+    },
+    {
+      "epoch": 0.04035874439461883,
+      "grad_norm": 1.2885852106521303,
+      "learning_rate": 6.716417910447762e-06,
+      "loss": 0.9682,
+      "step": 27
+    },
+    {
+      "epoch": 0.04185351270553064,
+      "grad_norm": 1.0614985591914423,
+      "learning_rate": 6.965174129353234e-06,
+      "loss": 0.9696,
+      "step": 28
+    },
+    {
+      "epoch": 0.043348281016442454,
+      "grad_norm": 0.916474552137227,
+      "learning_rate": 7.213930348258708e-06,
+      "loss": 0.9619,
+      "step": 29
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 0.8609957840049264,
+      "learning_rate": 7.4626865671641785e-06,
+      "loss": 0.8916,
+      "step": 30
+    },
+    {
+      "epoch": 0.04633781763826607,
+      "grad_norm": 0.811747444746741,
+      "learning_rate": 7.711442786069652e-06,
+      "loss": 0.8831,
+      "step": 31
+    },
+    {
+      "epoch": 0.04783258594917788,
+      "grad_norm": 0.7405639773094553,
+      "learning_rate": 7.960199004975125e-06,
+      "loss": 0.8552,
+      "step": 32
+    },
+    {
+      "epoch": 0.04932735426008968,
+      "grad_norm": 0.6511770752034179,
+      "learning_rate": 8.208955223880597e-06,
+      "loss": 0.8719,
+      "step": 33
+    },
+    {
+      "epoch": 0.05082212257100149,
+      "grad_norm": 0.6811193896058715,
+      "learning_rate": 8.45771144278607e-06,
+      "loss": 0.8602,
+      "step": 34
+    },
+    {
+      "epoch": 0.052316890881913304,
+      "grad_norm": 0.655453322024125,
+      "learning_rate": 8.706467661691544e-06,
+      "loss": 0.8121,
+      "step": 35
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 0.7081849807944613,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.8223,
+      "step": 36
+    },
+    {
+      "epoch": 0.05530642750373692,
+      "grad_norm": 0.6518185475908654,
+      "learning_rate": 9.203980099502487e-06,
+      "loss": 0.7809,
+      "step": 37
+    },
+    {
+      "epoch": 0.05680119581464873,
+      "grad_norm": 0.5516265544693141,
+      "learning_rate": 9.45273631840796e-06,
+      "loss": 0.8079,
+      "step": 38
+    },
+    {
+      "epoch": 0.05829596412556054,
+      "grad_norm": 0.5162578549951131,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 0.8059,
+      "step": 39
+    },
+    {
+      "epoch": 0.059790732436472344,
+      "grad_norm": 0.5275735763047418,
+      "learning_rate": 9.950248756218906e-06,
+      "loss": 0.7685,
+      "step": 40
+    },
+    {
+      "epoch": 0.061285500747384154,
+      "grad_norm": 0.5691142572032146,
+      "learning_rate": 1.0199004975124378e-05,
+      "loss": 0.7536,
+      "step": 41
+    },
+    {
+      "epoch": 0.06278026905829596,
+      "grad_norm": 0.6176915802531199,
+      "learning_rate": 1.0447761194029851e-05,
+      "loss": 0.7608,
+      "step": 42
+    },
+    {
+      "epoch": 0.06427503736920777,
+      "grad_norm": 0.43847475950575654,
+      "learning_rate": 1.0696517412935324e-05,
+      "loss": 0.7306,
+      "step": 43
+    },
+    {
+      "epoch": 0.06576980568011959,
+      "grad_norm": 0.408842616367869,
+      "learning_rate": 1.0945273631840796e-05,
+      "loss": 0.7191,
+      "step": 44
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 0.42321463616177823,
+      "learning_rate": 1.119402985074627e-05,
+      "loss": 0.7127,
+      "step": 45
+    },
+    {
+      "epoch": 0.0687593423019432,
+      "grad_norm": 0.5449730455456911,
+      "learning_rate": 1.1442786069651743e-05,
+      "loss": 0.7563,
+      "step": 46
+    },
+    {
+      "epoch": 0.07025411061285501,
+      "grad_norm": 0.47557196050344996,
+      "learning_rate": 1.1691542288557215e-05,
+      "loss": 0.7089,
+      "step": 47
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 0.3659276118805365,
+      "learning_rate": 1.1940298507462686e-05,
+      "loss": 0.7233,
+      "step": 48
+    },
+    {
+      "epoch": 0.07324364723467862,
+      "grad_norm": 0.362093758437109,
+      "learning_rate": 1.218905472636816e-05,
+      "loss": 0.7245,
+      "step": 49
+    },
+    {
+      "epoch": 0.07473841554559044,
+      "grad_norm": 0.4458678158631584,
+      "learning_rate": 1.2437810945273633e-05,
+      "loss": 0.7196,
+      "step": 50
+    },
+    {
+      "epoch": 0.07623318385650224,
+      "grad_norm": 0.3955261494328727,
+      "learning_rate": 1.2686567164179105e-05,
+      "loss": 0.7105,
+      "step": 51
+    },
+    {
+      "epoch": 0.07772795216741404,
+      "grad_norm": 0.3939645993583809,
+      "learning_rate": 1.2935323383084577e-05,
+      "loss": 0.7089,
+      "step": 52
+    },
+    {
+      "epoch": 0.07922272047832586,
+      "grad_norm": 0.3304062343488697,
+      "learning_rate": 1.3184079601990052e-05,
+      "loss": 0.6851,
+      "step": 53
+    },
+    {
+      "epoch": 0.08071748878923767,
+      "grad_norm": 0.45083568522530126,
+      "learning_rate": 1.3432835820895523e-05,
+      "loss": 0.7,
+      "step": 54
+    },
+    {
+      "epoch": 0.08221225710014948,
+      "grad_norm": 0.38532253241304415,
+      "learning_rate": 1.3681592039800995e-05,
+      "loss": 0.6807,
+      "step": 55
+    },
+    {
+      "epoch": 0.08370702541106129,
+      "grad_norm": 0.35325264002433726,
+      "learning_rate": 1.3930348258706468e-05,
+      "loss": 0.7005,
+      "step": 56
+    },
+    {
+      "epoch": 0.08520179372197309,
+      "grad_norm": 0.3257648860591999,
+      "learning_rate": 1.417910447761194e-05,
+      "loss": 0.6889,
+      "step": 57
+    },
+    {
+      "epoch": 0.08669656203288491,
+      "grad_norm": 0.31759129566026123,
+      "learning_rate": 1.4427860696517415e-05,
+      "loss": 0.6859,
+      "step": 58
+    },
+    {
+      "epoch": 0.08819133034379671,
+      "grad_norm": 0.3090776860180793,
+      "learning_rate": 1.4676616915422885e-05,
+      "loss": 0.6504,
+      "step": 59
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 0.31771147822650486,
+      "learning_rate": 1.4925373134328357e-05,
+      "loss": 0.6556,
+      "step": 60
+    },
+    {
+      "epoch": 0.09118086696562033,
+      "grad_norm": 0.30641417983675256,
+      "learning_rate": 1.5174129353233832e-05,
+      "loss": 0.6821,
+      "step": 61
+    },
+    {
+      "epoch": 0.09267563527653214,
+      "grad_norm": 0.33713288403630953,
+      "learning_rate": 1.5422885572139304e-05,
+      "loss": 0.6679,
+      "step": 62
+    },
+    {
+      "epoch": 0.09417040358744394,
+      "grad_norm": 0.344511792009641,
+      "learning_rate": 1.5671641791044777e-05,
+      "loss": 0.6629,
+      "step": 63
+    },
+    {
+      "epoch": 0.09566517189835576,
+      "grad_norm": 0.3072328509633548,
+      "learning_rate": 1.592039800995025e-05,
+      "loss": 0.638,
+      "step": 64
+    },
+    {
+      "epoch": 0.09715994020926756,
+      "grad_norm": 0.322606871552221,
+      "learning_rate": 1.6169154228855724e-05,
+      "loss": 0.6454,
+      "step": 65
+    },
+    {
+      "epoch": 0.09865470852017937,
+      "grad_norm": 0.32296233013357595,
+      "learning_rate": 1.6417910447761194e-05,
+      "loss": 0.6662,
+      "step": 66
+    },
+    {
+      "epoch": 0.10014947683109118,
+      "grad_norm": 0.32436335009848727,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6401,
+      "step": 67
+    },
+    {
+      "epoch": 0.10164424514200299,
+      "grad_norm": 0.2690412665707551,
+      "learning_rate": 1.691542288557214e-05,
+      "loss": 0.6662,
+      "step": 68
+    },
+    {
+      "epoch": 0.1031390134529148,
+      "grad_norm": 0.3205365165269462,
+      "learning_rate": 1.716417910447761e-05,
+      "loss": 0.6581,
+      "step": 69
+    },
+    {
+      "epoch": 0.10463378176382661,
+      "grad_norm": 0.3099372088840112,
+      "learning_rate": 1.7412935323383088e-05,
+      "loss": 0.6356,
+      "step": 70
+    },
+    {
+      "epoch": 0.10612855007473841,
+      "grad_norm": 0.27771855076834717,
+      "learning_rate": 1.7661691542288558e-05,
+      "loss": 0.6294,
+      "step": 71
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 0.2818680959299575,
+      "learning_rate": 1.791044776119403e-05,
+      "loss": 0.6387,
+      "step": 72
+    },
+    {
+      "epoch": 0.10911808669656203,
+      "grad_norm": 0.30091652869874297,
+      "learning_rate": 1.8159203980099505e-05,
+      "loss": 0.6632,
+      "step": 73
+    },
+    {
+      "epoch": 0.11061285500747384,
+      "grad_norm": 0.3057793531876186,
+      "learning_rate": 1.8407960199004975e-05,
+      "loss": 0.6153,
+      "step": 74
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 0.2747323330695542,
+      "learning_rate": 1.865671641791045e-05,
+      "loss": 0.6347,
+      "step": 75
+    },
+    {
+      "epoch": 0.11360239162929746,
+      "grad_norm": 0.3404013968811091,
+      "learning_rate": 1.890547263681592e-05,
+      "loss": 0.6388,
+      "step": 76
+    },
+    {
+      "epoch": 0.11509715994020926,
+      "grad_norm": 0.31480886991149687,
+      "learning_rate": 1.915422885572139e-05,
+      "loss": 0.6228,
+      "step": 77
+    },
+    {
+      "epoch": 0.11659192825112108,
+      "grad_norm": 0.290029997703859,
+      "learning_rate": 1.9402985074626868e-05,
+      "loss": 0.6372,
+      "step": 78
+    },
+    {
+      "epoch": 0.11808669656203288,
+      "grad_norm": 0.33623322026536195,
+      "learning_rate": 1.9651741293532338e-05,
+      "loss": 0.6483,
+      "step": 79
+    },
+    {
+      "epoch": 0.11958146487294469,
+      "grad_norm": 0.3602405204489083,
+      "learning_rate": 1.990049751243781e-05,
+      "loss": 0.6366,
+      "step": 80
+    },
+    {
+      "epoch": 0.1210762331838565,
+      "grad_norm": 0.30450347674713135,
+      "learning_rate": 2.0149253731343285e-05,
+      "loss": 0.5974,
+      "step": 81
+    },
+    {
+      "epoch": 0.12257100149476831,
+      "grad_norm": 0.40466074816502096,
+      "learning_rate": 2.0398009950248755e-05,
+      "loss": 0.6766,
+      "step": 82
+    },
+    {
+      "epoch": 0.12406576980568013,
+      "grad_norm": 0.3685435155451652,
+      "learning_rate": 2.0646766169154232e-05,
+      "loss": 0.6326,
+      "step": 83
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 0.3644523158111927,
+      "learning_rate": 2.0895522388059702e-05,
+      "loss": 0.6245,
+      "step": 84
+    },
+    {
+      "epoch": 0.12705530642750373,
+      "grad_norm": 0.33808390943643307,
+      "learning_rate": 2.1144278606965175e-05,
+      "loss": 0.6371,
+      "step": 85
+    },
+    {
+      "epoch": 0.12855007473841554,
+      "grad_norm": 0.3607700151040439,
+      "learning_rate": 2.139303482587065e-05,
+      "loss": 0.6174,
+      "step": 86
+    },
+    {
+      "epoch": 0.13004484304932734,
+      "grad_norm": 0.3879716661659854,
+      "learning_rate": 2.164179104477612e-05,
+      "loss": 0.5959,
+      "step": 87
+    },
+    {
+      "epoch": 0.13153961136023917,
+      "grad_norm": 0.3191058571134585,
+      "learning_rate": 2.1890547263681592e-05,
+      "loss": 0.6196,
+      "step": 88
+    },
+    {
+      "epoch": 0.13303437967115098,
+      "grad_norm": 0.36346378044422506,
+      "learning_rate": 2.2139303482587065e-05,
+      "loss": 0.6193,
+      "step": 89
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 0.319655757191322,
+      "learning_rate": 2.238805970149254e-05,
+      "loss": 0.6066,
+      "step": 90
+    },
+    {
+      "epoch": 0.13602391629297458,
+      "grad_norm": 0.34477410213774246,
+      "learning_rate": 2.2636815920398012e-05,
+      "loss": 0.621,
+      "step": 91
+    },
+    {
+      "epoch": 0.1375186846038864,
+      "grad_norm": 0.34058386776081856,
+      "learning_rate": 2.2885572139303486e-05,
+      "loss": 0.6308,
+      "step": 92
+    },
+    {
+      "epoch": 0.13901345291479822,
+      "grad_norm": 0.32386683073233313,
+      "learning_rate": 2.3134328358208956e-05,
+      "loss": 0.592,
+      "step": 93
+    },
+    {
+      "epoch": 0.14050822122571002,
+      "grad_norm": 0.3231317140601968,
+      "learning_rate": 2.338308457711443e-05,
+      "loss": 0.6229,
+      "step": 94
+    },
+    {
+      "epoch": 0.14200298953662183,
+      "grad_norm": 0.3048873678393069,
+      "learning_rate": 2.3631840796019903e-05,
+      "loss": 0.6221,
+      "step": 95
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.283886474509331,
+      "learning_rate": 2.3880597014925373e-05,
+      "loss": 0.6272,
+      "step": 96
+    },
+    {
+      "epoch": 0.14499252615844543,
+      "grad_norm": 0.3216515822305372,
+      "learning_rate": 2.412935323383085e-05,
+      "loss": 0.626,
+      "step": 97
+    },
+    {
+      "epoch": 0.14648729446935724,
+      "grad_norm": 0.309301546629261,
+      "learning_rate": 2.437810945273632e-05,
+      "loss": 0.6239,
+      "step": 98
+    },
+    {
+      "epoch": 0.14798206278026907,
+      "grad_norm": 0.29307479412185894,
+      "learning_rate": 2.4626865671641793e-05,
+      "loss": 0.6025,
+      "step": 99
+    },
+    {
+      "epoch": 0.14947683109118087,
+      "grad_norm": 0.3642321759981609,
+      "learning_rate": 2.4875621890547266e-05,
+      "loss": 0.6124,
+      "step": 100
+    },
+    {
+      "epoch": 0.15097159940209268,
+      "grad_norm": 0.35121637493507746,
+      "learning_rate": 2.512437810945274e-05,
+      "loss": 0.5987,
+      "step": 101
+    },
+    {
+      "epoch": 0.15246636771300448,
+      "grad_norm": 0.357267860233352,
+      "learning_rate": 2.537313432835821e-05,
+      "loss": 0.6068,
+      "step": 102
+    },
+    {
+      "epoch": 0.15396113602391628,
+      "grad_norm": 0.39197403891612,
+      "learning_rate": 2.5621890547263683e-05,
+      "loss": 0.6224,
+      "step": 103
+    },
+    {
+      "epoch": 0.1554559043348281,
+      "grad_norm": 0.3772686433580899,
+      "learning_rate": 2.5870646766169153e-05,
+      "loss": 0.6164,
+      "step": 104
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 0.3544153893836608,
+      "learning_rate": 2.6119402985074626e-05,
+      "loss": 0.6004,
+      "step": 105
+    },
+    {
+      "epoch": 0.15844544095665172,
+      "grad_norm": 0.35655489835419585,
+      "learning_rate": 2.6368159203980103e-05,
+      "loss": 0.5978,
+      "step": 106
+    },
+    {
+      "epoch": 0.15994020926756353,
+      "grad_norm": 0.3389703706269328,
+      "learning_rate": 2.6616915422885573e-05,
+      "loss": 0.577,
+      "step": 107
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.37776328351077276,
+      "learning_rate": 2.6865671641791047e-05,
+      "loss": 0.5982,
+      "step": 108
+    },
+    {
+      "epoch": 0.16292974588938713,
+      "grad_norm": 0.37580639414506484,
+      "learning_rate": 2.7114427860696517e-05,
+      "loss": 0.5879,
+      "step": 109
+    },
+    {
+      "epoch": 0.16442451420029897,
+      "grad_norm": 0.41583445520417034,
+      "learning_rate": 2.736318407960199e-05,
+      "loss": 0.6101,
+      "step": 110
+    },
+    {
+      "epoch": 0.16591928251121077,
+      "grad_norm": 0.42607509950322203,
+      "learning_rate": 2.7611940298507467e-05,
+      "loss": 0.6066,
+      "step": 111
+    },
+    {
+      "epoch": 0.16741405082212257,
+      "grad_norm": 0.3867589959062827,
+      "learning_rate": 2.7860696517412937e-05,
+      "loss": 0.619,
+      "step": 112
+    },
+    {
+      "epoch": 0.16890881913303438,
+      "grad_norm": 0.4458680267053343,
+      "learning_rate": 2.810945273631841e-05,
+      "loss": 0.6133,
+      "step": 113
+    },
+    {
+      "epoch": 0.17040358744394618,
+      "grad_norm": 0.38967138781904614,
+      "learning_rate": 2.835820895522388e-05,
+      "loss": 0.5876,
+      "step": 114
+    },
+    {
+      "epoch": 0.17189835575485798,
+      "grad_norm": 0.34234490108312876,
+      "learning_rate": 2.8606965174129354e-05,
+      "loss": 0.5647,
+      "step": 115
+    },
+    {
+      "epoch": 0.17339312406576982,
+      "grad_norm": 0.46165844735342804,
+      "learning_rate": 2.885572139303483e-05,
+      "loss": 0.5986,
+      "step": 116
+    },
+    {
+      "epoch": 0.17488789237668162,
+      "grad_norm": 0.3464611346887158,
+      "learning_rate": 2.91044776119403e-05,
+      "loss": 0.5895,
+      "step": 117
+    },
+    {
+      "epoch": 0.17638266068759342,
+      "grad_norm": 0.39770588680748536,
+      "learning_rate": 2.935323383084577e-05,
+      "loss": 0.6072,
+      "step": 118
+    },
+    {
+      "epoch": 0.17787742899850523,
+      "grad_norm": 0.3580213541921293,
+      "learning_rate": 2.9601990049751244e-05,
+      "loss": 0.5899,
+      "step": 119
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.4054164887208958,
+      "learning_rate": 2.9850746268656714e-05,
+      "loss": 0.6124,
+      "step": 120
+    },
+    {
+      "epoch": 0.18086696562032886,
+      "grad_norm": 0.3934051348598669,
+      "learning_rate": 3.009950248756219e-05,
+      "loss": 0.5908,
+      "step": 121
+    },
+    {
+      "epoch": 0.18236173393124067,
+      "grad_norm": 0.4246010218655408,
+      "learning_rate": 3.0348258706467664e-05,
+      "loss": 0.5698,
+      "step": 122
+    },
+    {
+      "epoch": 0.18385650224215247,
+      "grad_norm": 0.38948720240626067,
+      "learning_rate": 3.059701492537314e-05,
+      "loss": 0.6066,
+      "step": 123
+    },
+    {
+      "epoch": 0.18535127055306427,
+      "grad_norm": 0.44655155335953617,
+      "learning_rate": 3.084577114427861e-05,
+      "loss": 0.5832,
+      "step": 124
+    },
+    {
+      "epoch": 0.18684603886397608,
+      "grad_norm": 0.46955303512497965,
+      "learning_rate": 3.109452736318408e-05,
+      "loss": 0.5783,
+      "step": 125
+    },
+    {
+      "epoch": 0.18834080717488788,
+      "grad_norm": 0.4210305232308902,
+      "learning_rate": 3.1343283582089554e-05,
+      "loss": 0.5924,
+      "step": 126
+    },
+    {
+      "epoch": 0.1898355754857997,
+      "grad_norm": 0.4066435948137076,
+      "learning_rate": 3.1592039800995024e-05,
+      "loss": 0.5803,
+      "step": 127
+    },
+    {
+      "epoch": 0.19133034379671152,
+      "grad_norm": 0.38625347487957223,
+      "learning_rate": 3.18407960199005e-05,
+      "loss": 0.5855,
+      "step": 128
+    },
+    {
+      "epoch": 0.19282511210762332,
+      "grad_norm": 0.48491717274057416,
+      "learning_rate": 3.208955223880597e-05,
+      "loss": 0.5912,
+      "step": 129
+    },
+    {
+      "epoch": 0.19431988041853512,
+      "grad_norm": 0.4340091616752196,
+      "learning_rate": 3.233830845771145e-05,
+      "loss": 0.5941,
+      "step": 130
+    },
+    {
+      "epoch": 0.19581464872944693,
+      "grad_norm": 0.49113623008625357,
+      "learning_rate": 3.258706467661692e-05,
+      "loss": 0.563,
+      "step": 131
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.4164259059539615,
+      "learning_rate": 3.283582089552239e-05,
+      "loss": 0.5789,
+      "step": 132
+    },
+    {
+      "epoch": 0.19880418535127056,
+      "grad_norm": 0.4407478277564631,
+      "learning_rate": 3.3084577114427865e-05,
+      "loss": 0.5648,
+      "step": 133
+    },
+    {
+      "epoch": 0.20029895366218237,
+      "grad_norm": 0.447706498393888,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5805,
+      "step": 134
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 0.4171185382966137,
+      "learning_rate": 3.358208955223881e-05,
+      "loss": 0.6144,
+      "step": 135
+    },
+    {
+      "epoch": 0.20328849028400597,
+      "grad_norm": 0.48179329985480596,
+      "learning_rate": 3.383084577114428e-05,
+      "loss": 0.6034,
+      "step": 136
+    },
+    {
+      "epoch": 0.20478325859491778,
+      "grad_norm": 0.5045084460077158,
+      "learning_rate": 3.407960199004975e-05,
+      "loss": 0.547,
+      "step": 137
+    },
+    {
+      "epoch": 0.2062780269058296,
+      "grad_norm": 0.3900068984102361,
+      "learning_rate": 3.432835820895522e-05,
+      "loss": 0.6373,
+      "step": 138
+    },
+    {
+      "epoch": 0.2077727952167414,
+      "grad_norm": 0.48719241536115465,
+      "learning_rate": 3.45771144278607e-05,
+      "loss": 0.5686,
+      "step": 139
+    },
+    {
+      "epoch": 0.20926756352765322,
+      "grad_norm": 0.4247790945366846,
+      "learning_rate": 3.4825870646766175e-05,
+      "loss": 0.5834,
+      "step": 140
+    },
+    {
+      "epoch": 0.21076233183856502,
+      "grad_norm": 0.4983520328060873,
+      "learning_rate": 3.5074626865671645e-05,
+      "loss": 0.5769,
+      "step": 141
+    },
+    {
+      "epoch": 0.21225710014947682,
+      "grad_norm": 0.4187706449281194,
+      "learning_rate": 3.5323383084577115e-05,
+      "loss": 0.5556,
+      "step": 142
+    },
+    {
+      "epoch": 0.21375186846038863,
+      "grad_norm": 0.5236457602898553,
+      "learning_rate": 3.5572139303482585e-05,
+      "loss": 0.5747,
+      "step": 143
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.404886267098541,
+      "learning_rate": 3.582089552238806e-05,
+      "loss": 0.574,
+      "step": 144
+    },
+    {
+      "epoch": 0.21674140508221226,
+      "grad_norm": 0.6484921491071229,
+      "learning_rate": 3.606965174129354e-05,
+      "loss": 0.6032,
+      "step": 145
+    },
+    {
+      "epoch": 0.21823617339312407,
+      "grad_norm": 0.43170659721446325,
+      "learning_rate": 3.631840796019901e-05,
+      "loss": 0.6071,
+      "step": 146
+    },
+    {
+      "epoch": 0.21973094170403587,
+      "grad_norm": 0.7210244524821048,
+      "learning_rate": 3.656716417910448e-05,
+      "loss": 0.6034,
+      "step": 147
+    },
+    {
+      "epoch": 0.22122571001494767,
+      "grad_norm": 0.6318210242142426,
+      "learning_rate": 3.681592039800995e-05,
+      "loss": 0.5736,
+      "step": 148
+    },
+    {
+      "epoch": 0.2227204783258595,
+      "grad_norm": 0.5113480200648777,
+      "learning_rate": 3.706467661691542e-05,
+      "loss": 0.5942,
+      "step": 149
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 0.531580680382394,
+      "learning_rate": 3.73134328358209e-05,
+      "loss": 0.5711,
+      "step": 150
+    },
+    {
+      "epoch": 0.2257100149476831,
+      "grad_norm": 0.5683618575934614,
+      "learning_rate": 3.756218905472637e-05,
+      "loss": 0.5651,
+      "step": 151
+    },
+    {
+      "epoch": 0.22720478325859492,
+      "grad_norm": 0.5483869663999059,
+      "learning_rate": 3.781094527363184e-05,
+      "loss": 0.5711,
+      "step": 152
+    },
+    {
+      "epoch": 0.22869955156950672,
+      "grad_norm": 0.4310166414233741,
+      "learning_rate": 3.805970149253731e-05,
+      "loss": 0.5735,
+      "step": 153
+    },
+    {
+      "epoch": 0.23019431988041852,
+      "grad_norm": 0.472883962541052,
+      "learning_rate": 3.830845771144278e-05,
+      "loss": 0.5533,
+      "step": 154
+    },
+    {
+      "epoch": 0.23168908819133036,
+      "grad_norm": 0.4126945115875031,
+      "learning_rate": 3.855721393034826e-05,
+      "loss": 0.573,
+      "step": 155
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.4590816624548245,
+      "learning_rate": 3.8805970149253736e-05,
+      "loss": 0.5614,
+      "step": 156
+    },
+    {
+      "epoch": 0.23467862481315396,
+      "grad_norm": 0.38836265908784395,
+      "learning_rate": 3.9054726368159206e-05,
+      "loss": 0.5487,
+      "step": 157
+    },
+    {
+      "epoch": 0.23617339312406577,
+      "grad_norm": 0.4777658026192644,
+      "learning_rate": 3.9303482587064676e-05,
+      "loss": 0.5553,
+      "step": 158
+    },
+    {
+      "epoch": 0.23766816143497757,
+      "grad_norm": 0.4962732456724574,
+      "learning_rate": 3.9552238805970146e-05,
+      "loss": 0.5787,
+      "step": 159
+    },
+    {
+      "epoch": 0.23916292974588937,
+      "grad_norm": 0.4872177275465212,
+      "learning_rate": 3.980099502487562e-05,
+      "loss": 0.5791,
+      "step": 160
+    },
+    {
+      "epoch": 0.2406576980568012,
+      "grad_norm": 0.5022543171763286,
+      "learning_rate": 4.00497512437811e-05,
+      "loss": 0.5644,
+      "step": 161
+    },
+    {
+      "epoch": 0.242152466367713,
+      "grad_norm": 0.4842497060337267,
+      "learning_rate": 4.029850746268657e-05,
+      "loss": 0.6114,
+      "step": 162
+    },
+    {
+      "epoch": 0.2436472346786248,
+      "grad_norm": 0.42993669718044775,
+      "learning_rate": 4.054726368159204e-05,
+      "loss": 0.561,
+      "step": 163
+    },
+    {
+      "epoch": 0.24514200298953662,
+      "grad_norm": 0.46049224104609243,
+      "learning_rate": 4.079601990049751e-05,
+      "loss": 0.5671,
+      "step": 164
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 0.5647129459508742,
+      "learning_rate": 4.104477611940299e-05,
+      "loss": 0.5842,
+      "step": 165
+    },
+    {
+      "epoch": 0.24813153961136025,
+      "grad_norm": 0.5252445425503527,
+      "learning_rate": 4.1293532338308464e-05,
+      "loss": 0.5651,
+      "step": 166
+    },
+    {
+      "epoch": 0.24962630792227206,
+      "grad_norm": 0.47258085263098065,
+      "learning_rate": 4.1542288557213934e-05,
+      "loss": 0.5973,
+      "step": 167
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.4609840323233955,
+      "learning_rate": 4.1791044776119404e-05,
+      "loss": 0.577,
+      "step": 168
+    },
+    {
+      "epoch": 0.2526158445440957,
+      "grad_norm": 0.5704008712392566,
+      "learning_rate": 4.2039800995024874e-05,
+      "loss": 0.5792,
+      "step": 169
+    },
+    {
+      "epoch": 0.25411061285500747,
+      "grad_norm": 0.6187829167018652,
+      "learning_rate": 4.228855721393035e-05,
+      "loss": 0.531,
+      "step": 170
+    },
+    {
+      "epoch": 0.2556053811659193,
+      "grad_norm": 0.42425186997370345,
+      "learning_rate": 4.253731343283582e-05,
+      "loss": 0.5703,
+      "step": 171
+    },
+    {
+      "epoch": 0.2571001494768311,
+      "grad_norm": 0.47780953498009376,
+      "learning_rate": 4.27860696517413e-05,
+      "loss": 0.5517,
+      "step": 172
+    },
+    {
+      "epoch": 0.2585949177877429,
+      "grad_norm": 0.587974041546309,
+      "learning_rate": 4.303482587064677e-05,
+      "loss": 0.5459,
+      "step": 173
+    },
+    {
+      "epoch": 0.2600896860986547,
+      "grad_norm": 0.42077566038565806,
+      "learning_rate": 4.328358208955224e-05,
+      "loss": 0.5656,
+      "step": 174
+    },
+    {
+      "epoch": 0.2615844544095665,
+      "grad_norm": 0.5093785863170577,
+      "learning_rate": 4.3532338308457714e-05,
+      "loss": 0.557,
+      "step": 175
+    },
+    {
+      "epoch": 0.26307922272047835,
+      "grad_norm": 0.4758091446348775,
+      "learning_rate": 4.3781094527363184e-05,
+      "loss": 0.5671,
+      "step": 176
+    },
+    {
+      "epoch": 0.2645739910313901,
+      "grad_norm": 0.44305965607616465,
+      "learning_rate": 4.402985074626866e-05,
+      "loss": 0.5457,
+      "step": 177
+    },
+    {
+      "epoch": 0.26606875934230195,
+      "grad_norm": 0.4888419046451185,
+      "learning_rate": 4.427860696517413e-05,
+      "loss": 0.5797,
+      "step": 178
+    },
+    {
+      "epoch": 0.26756352765321373,
+      "grad_norm": 0.4439919781858181,
+      "learning_rate": 4.452736318407961e-05,
+      "loss": 0.5564,
+      "step": 179
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.4121734791764489,
+      "learning_rate": 4.477611940298508e-05,
+      "loss": 0.5785,
+      "step": 180
+    },
+    {
+      "epoch": 0.2705530642750374,
+      "grad_norm": 0.41909525690185084,
+      "learning_rate": 4.502487562189055e-05,
+      "loss": 0.5356,
+      "step": 181
+    },
+    {
+      "epoch": 0.27204783258594917,
+      "grad_norm": 0.4398076693060589,
+      "learning_rate": 4.5273631840796025e-05,
+      "loss": 0.5438,
+      "step": 182
+    },
+    {
+      "epoch": 0.273542600896861,
+      "grad_norm": 0.4410909180439866,
+      "learning_rate": 4.5522388059701495e-05,
+      "loss": 0.553,
+      "step": 183
+    },
+    {
+      "epoch": 0.2750373692077728,
+      "grad_norm": 0.5297395219906865,
+      "learning_rate": 4.577114427860697e-05,
+      "loss": 0.5263,
+      "step": 184
+    },
+    {
+      "epoch": 0.2765321375186846,
+      "grad_norm": 0.49397700042713455,
+      "learning_rate": 4.601990049751244e-05,
+      "loss": 0.5764,
+      "step": 185
+    },
+    {
+      "epoch": 0.27802690582959644,
+      "grad_norm": 0.4415787220031261,
+      "learning_rate": 4.626865671641791e-05,
+      "loss": 0.5838,
+      "step": 186
+    },
+    {
+      "epoch": 0.2795216741405082,
+      "grad_norm": 0.4775216039188186,
+      "learning_rate": 4.651741293532338e-05,
+      "loss": 0.5755,
+      "step": 187
+    },
+    {
+      "epoch": 0.28101644245142005,
+      "grad_norm": 0.590741237534494,
+      "learning_rate": 4.676616915422886e-05,
+      "loss": 0.5563,
+      "step": 188
+    },
+    {
+      "epoch": 0.2825112107623318,
+      "grad_norm": 0.5349978506879617,
+      "learning_rate": 4.7014925373134335e-05,
+      "loss": 0.5825,
+      "step": 189
+    },
+    {
+      "epoch": 0.28400597907324365,
+      "grad_norm": 0.557455109148311,
+      "learning_rate": 4.7263681592039805e-05,
+      "loss": 0.5564,
+      "step": 190
+    },
+    {
+      "epoch": 0.28550074738415543,
+      "grad_norm": 0.4695346138935822,
+      "learning_rate": 4.7512437810945275e-05,
+      "loss": 0.5886,
+      "step": 191
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.44307669780621095,
+      "learning_rate": 4.7761194029850745e-05,
+      "loss": 0.5732,
+      "step": 192
+    },
+    {
+      "epoch": 0.2884902840059791,
+      "grad_norm": 0.49982271156238806,
+      "learning_rate": 4.800995024875622e-05,
+      "loss": 0.5826,
+      "step": 193
+    },
+    {
+      "epoch": 0.28998505231689087,
+      "grad_norm": 0.5046159079314815,
+      "learning_rate": 4.82587064676617e-05,
+      "loss": 0.5614,
+      "step": 194
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 0.5095445920015009,
+      "learning_rate": 4.850746268656717e-05,
+      "loss": 0.5407,
+      "step": 195
+    },
+    {
+      "epoch": 0.2929745889387145,
+      "grad_norm": 0.7568966332572932,
+      "learning_rate": 4.875621890547264e-05,
+      "loss": 0.5432,
+      "step": 196
+    },
+    {
+      "epoch": 0.2944693572496263,
+      "grad_norm": 0.4757178066225346,
+      "learning_rate": 4.900497512437811e-05,
+      "loss": 0.5561,
+      "step": 197
+    },
+    {
+      "epoch": 0.29596412556053814,
+      "grad_norm": 0.4479840950748226,
+      "learning_rate": 4.9253731343283586e-05,
+      "loss": 0.565,
+      "step": 198
+    },
+    {
+      "epoch": 0.2974588938714499,
+      "grad_norm": 0.5209014498155967,
+      "learning_rate": 4.950248756218906e-05,
+      "loss": 0.5578,
+      "step": 199
+    },
+    {
+      "epoch": 0.29895366218236175,
+      "grad_norm": 0.5745750496545116,
+      "learning_rate": 4.975124378109453e-05,
+      "loss": 0.5662,
+      "step": 200
+    },
+    {
+      "epoch": 0.3004484304932735,
+      "grad_norm": 0.6178906361203442,
+      "learning_rate": 5e-05,
+      "loss": 0.572,
+      "step": 201
+    },
+    {
+      "epoch": 0.30194319880418535,
+      "grad_norm": 0.8041360088140427,
+      "learning_rate": 4.9972314507198234e-05,
+      "loss": 0.5624,
+      "step": 202
+    },
+    {
+      "epoch": 0.3034379671150972,
+      "grad_norm": 1.030128196909196,
+      "learning_rate": 4.994462901439646e-05,
+      "loss": 0.5493,
+      "step": 203
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.6170124120595559,
+      "learning_rate": 4.991694352159469e-05,
+      "loss": 0.5463,
+      "step": 204
+    },
+    {
+      "epoch": 0.3064275037369208,
+      "grad_norm": 0.5595040313662786,
+      "learning_rate": 4.9889258028792915e-05,
+      "loss": 0.5595,
+      "step": 205
+    },
+    {
+      "epoch": 0.30792227204783257,
+      "grad_norm": 0.87316814969042,
+      "learning_rate": 4.986157253599114e-05,
+      "loss": 0.5427,
+      "step": 206
+    },
+    {
+      "epoch": 0.3094170403587444,
+      "grad_norm": 0.6063278615671696,
+      "learning_rate": 4.983388704318937e-05,
+      "loss": 0.5439,
+      "step": 207
+    },
+    {
+      "epoch": 0.3109118086696562,
+      "grad_norm": 0.674100446223187,
+      "learning_rate": 4.9806201550387596e-05,
+      "loss": 0.5603,
+      "step": 208
+    },
+    {
+      "epoch": 0.312406576980568,
+      "grad_norm": 0.5982708622456756,
+      "learning_rate": 4.977851605758583e-05,
+      "loss": 0.5579,
+      "step": 209
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 0.4856123302080704,
+      "learning_rate": 4.975083056478406e-05,
+      "loss": 0.5508,
+      "step": 210
+    },
+    {
+      "epoch": 0.3153961136023916,
+      "grad_norm": 0.5698556069069233,
+      "learning_rate": 4.9723145071982284e-05,
+      "loss": 0.5639,
+      "step": 211
+    },
+    {
+      "epoch": 0.31689088191330345,
+      "grad_norm": 0.47790495685772544,
+      "learning_rate": 4.9695459579180516e-05,
+      "loss": 0.5469,
+      "step": 212
+    },
+    {
+      "epoch": 0.3183856502242152,
+      "grad_norm": 0.43944886084790225,
+      "learning_rate": 4.966777408637874e-05,
+      "loss": 0.5714,
+      "step": 213
+    },
+    {
+      "epoch": 0.31988041853512705,
+      "grad_norm": 0.46588273386296014,
+      "learning_rate": 4.9640088593576966e-05,
+      "loss": 0.5411,
+      "step": 214
+    },
+    {
+      "epoch": 0.3213751868460389,
+      "grad_norm": 0.4233576139148261,
+      "learning_rate": 4.96124031007752e-05,
+      "loss": 0.534,
+      "step": 215
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.4233305494196149,
+      "learning_rate": 4.958471760797342e-05,
+      "loss": 0.5514,
+      "step": 216
+    },
+    {
+      "epoch": 0.3243647234678625,
+      "grad_norm": 0.47209821693907145,
+      "learning_rate": 4.9557032115171654e-05,
+      "loss": 0.5388,
+      "step": 217
+    },
+    {
+      "epoch": 0.32585949177877427,
+      "grad_norm": 0.43368001474127227,
+      "learning_rate": 4.9529346622369885e-05,
+      "loss": 0.5629,
+      "step": 218
+    },
+    {
+      "epoch": 0.3273542600896861,
+      "grad_norm": 0.5144813918737933,
+      "learning_rate": 4.950166112956811e-05,
+      "loss": 0.5675,
+      "step": 219
+    },
+    {
+      "epoch": 0.32884902840059793,
+      "grad_norm": 0.4332511648290744,
+      "learning_rate": 4.9473975636766335e-05,
+      "loss": 0.5262,
+      "step": 220
+    },
+    {
+      "epoch": 0.3303437967115097,
+      "grad_norm": 0.5007231130147376,
+      "learning_rate": 4.9446290143964566e-05,
+      "loss": 0.5619,
+      "step": 221
+    },
+    {
+      "epoch": 0.33183856502242154,
+      "grad_norm": 0.48707792323764826,
+      "learning_rate": 4.941860465116279e-05,
+      "loss": 0.5615,
+      "step": 222
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5119188217509304,
+      "learning_rate": 4.939091915836102e-05,
+      "loss": 0.5643,
+      "step": 223
+    },
+    {
+      "epoch": 0.33482810164424515,
+      "grad_norm": 0.5061161514954574,
+      "learning_rate": 4.936323366555925e-05,
+      "loss": 0.569,
+      "step": 224
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 0.6153333868135146,
+      "learning_rate": 4.933554817275748e-05,
+      "loss": 0.5399,
+      "step": 225
+    },
+    {
+      "epoch": 0.33781763826606875,
+      "grad_norm": 0.589905842198286,
+      "learning_rate": 4.930786267995571e-05,
+      "loss": 0.537,
+      "step": 226
+    },
+    {
+      "epoch": 0.3393124065769806,
+      "grad_norm": 0.46718732997133444,
+      "learning_rate": 4.9280177187153935e-05,
+      "loss": 0.5361,
+      "step": 227
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 0.6321634343130755,
+      "learning_rate": 4.925249169435216e-05,
+      "loss": 0.5437,
+      "step": 228
+    },
+    {
+      "epoch": 0.3423019431988042,
+      "grad_norm": 0.6180133391964525,
+      "learning_rate": 4.922480620155039e-05,
+      "loss": 0.5487,
+      "step": 229
+    },
+    {
+      "epoch": 0.34379671150971597,
+      "grad_norm": 0.7157297209419416,
+      "learning_rate": 4.919712070874862e-05,
+      "loss": 0.5625,
+      "step": 230
+    },
+    {
+      "epoch": 0.3452914798206278,
+      "grad_norm": 0.600567747233388,
+      "learning_rate": 4.916943521594685e-05,
+      "loss": 0.5608,
+      "step": 231
+    },
+    {
+      "epoch": 0.34678624813153963,
+      "grad_norm": 0.45848254397744986,
+      "learning_rate": 4.914174972314507e-05,
+      "loss": 0.5555,
+      "step": 232
+    },
+    {
+      "epoch": 0.3482810164424514,
+      "grad_norm": 0.6631739998325054,
+      "learning_rate": 4.9114064230343305e-05,
+      "loss": 0.56,
+      "step": 233
+    },
+    {
+      "epoch": 0.34977578475336324,
+      "grad_norm": 0.79475800967058,
+      "learning_rate": 4.9086378737541536e-05,
+      "loss": 0.5498,
+      "step": 234
+    },
+    {
+      "epoch": 0.351270553064275,
+      "grad_norm": 0.5587357844698289,
+      "learning_rate": 4.9058693244739754e-05,
+      "loss": 0.5712,
+      "step": 235
+    },
+    {
+      "epoch": 0.35276532137518685,
+      "grad_norm": 0.7836614658710158,
+      "learning_rate": 4.9031007751937986e-05,
+      "loss": 0.5754,
+      "step": 236
+    },
+    {
+      "epoch": 0.3542600896860987,
+      "grad_norm": 0.9208318911774509,
+      "learning_rate": 4.900332225913621e-05,
+      "loss": 0.5554,
+      "step": 237
+    },
+    {
+      "epoch": 0.35575485799701045,
+      "grad_norm": 0.7384374824026935,
+      "learning_rate": 4.897563676633444e-05,
+      "loss": 0.5514,
+      "step": 238
+    },
+    {
+      "epoch": 0.3572496263079223,
+      "grad_norm": 0.7481719623946408,
+      "learning_rate": 4.8947951273532674e-05,
+      "loss": 0.5673,
+      "step": 239
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 1.0312769053344135,
+      "learning_rate": 4.89202657807309e-05,
+      "loss": 0.5414,
+      "step": 240
+    },
+    {
+      "epoch": 0.3602391629297459,
+      "grad_norm": 0.5686276509720866,
+      "learning_rate": 4.889258028792913e-05,
+      "loss": 0.5581,
+      "step": 241
+    },
+    {
+      "epoch": 0.3617339312406577,
+      "grad_norm": 0.801031574735393,
+      "learning_rate": 4.8864894795127355e-05,
+      "loss": 0.5332,
+      "step": 242
+    },
+    {
+      "epoch": 0.3632286995515695,
+      "grad_norm": 0.8969995536473471,
+      "learning_rate": 4.883720930232558e-05,
+      "loss": 0.5495,
+      "step": 243
+    },
+    {
+      "epoch": 0.36472346786248133,
+      "grad_norm": 0.48625416724705833,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 0.5548,
+      "step": 244
+    },
+    {
+      "epoch": 0.3662182361733931,
+      "grad_norm": 0.8851629297238233,
+      "learning_rate": 4.8781838316722036e-05,
+      "loss": 0.5293,
+      "step": 245
+    },
+    {
+      "epoch": 0.36771300448430494,
+      "grad_norm": 0.5511712930516371,
+      "learning_rate": 4.875415282392027e-05,
+      "loss": 0.5318,
+      "step": 246
+    },
+    {
+      "epoch": 0.3692077727952167,
+      "grad_norm": 0.7142071729182335,
+      "learning_rate": 4.87264673311185e-05,
+      "loss": 0.537,
+      "step": 247
+    },
+    {
+      "epoch": 0.37070254110612855,
+      "grad_norm": 0.6930644980395129,
+      "learning_rate": 4.8698781838316724e-05,
+      "loss": 0.5515,
+      "step": 248
+    },
+    {
+      "epoch": 0.3721973094170404,
+      "grad_norm": 0.44585410747066656,
+      "learning_rate": 4.8671096345514956e-05,
+      "loss": 0.5586,
+      "step": 249
+    },
+    {
+      "epoch": 0.37369207772795215,
+      "grad_norm": 0.7683261791308162,
+      "learning_rate": 4.864341085271318e-05,
+      "loss": 0.5247,
+      "step": 250
+    },
+    {
+      "epoch": 0.375186846038864,
+      "grad_norm": 0.5299048185552507,
+      "learning_rate": 4.8615725359911405e-05,
+      "loss": 0.5496,
+      "step": 251
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.6148990293448212,
+      "learning_rate": 4.858803986710964e-05,
+      "loss": 0.5478,
+      "step": 252
+    },
+    {
+      "epoch": 0.3781763826606876,
+      "grad_norm": 0.5964659616404414,
+      "learning_rate": 4.856035437430786e-05,
+      "loss": 0.509,
+      "step": 253
+    },
+    {
+      "epoch": 0.3796711509715994,
+      "grad_norm": 0.5179236653472427,
+      "learning_rate": 4.853266888150609e-05,
+      "loss": 0.5453,
+      "step": 254
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 0.6685060507622277,
+      "learning_rate": 4.8504983388704325e-05,
+      "loss": 0.5437,
+      "step": 255
+    },
+    {
+      "epoch": 0.38266068759342303,
+      "grad_norm": 0.4260349525197797,
+      "learning_rate": 4.847729789590255e-05,
+      "loss": 0.5464,
+      "step": 256
+    },
+    {
+      "epoch": 0.3841554559043348,
+      "grad_norm": 0.6970479375261545,
+      "learning_rate": 4.8449612403100775e-05,
+      "loss": 0.5488,
+      "step": 257
+    },
+    {
+      "epoch": 0.38565022421524664,
+      "grad_norm": 0.5694402215752169,
+      "learning_rate": 4.8421926910299006e-05,
+      "loss": 0.5397,
+      "step": 258
+    },
+    {
+      "epoch": 0.38714499252615847,
+      "grad_norm": 0.5207744363345348,
+      "learning_rate": 4.839424141749723e-05,
+      "loss": 0.5638,
+      "step": 259
+    },
+    {
+      "epoch": 0.38863976083707025,
+      "grad_norm": 0.5344354520971135,
+      "learning_rate": 4.836655592469546e-05,
+      "loss": 0.5283,
+      "step": 260
+    },
+    {
+      "epoch": 0.3901345291479821,
+      "grad_norm": 0.4928977414914228,
+      "learning_rate": 4.833887043189369e-05,
+      "loss": 0.5332,
+      "step": 261
+    },
+    {
+      "epoch": 0.39162929745889385,
+      "grad_norm": 0.3835710476768639,
+      "learning_rate": 4.831118493909192e-05,
+      "loss": 0.5473,
+      "step": 262
+    },
+    {
+      "epoch": 0.3931240657698057,
+      "grad_norm": 0.411629260826527,
+      "learning_rate": 4.828349944629015e-05,
+      "loss": 0.5504,
+      "step": 263
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.39443040719574457,
+      "learning_rate": 4.8255813953488375e-05,
+      "loss": 0.5439,
+      "step": 264
+    },
+    {
+      "epoch": 0.3961136023916293,
+      "grad_norm": 0.42607428799155017,
+      "learning_rate": 4.82281284606866e-05,
+      "loss": 0.5551,
+      "step": 265
+    },
+    {
+      "epoch": 0.3976083707025411,
+      "grad_norm": 0.4087803601843277,
+      "learning_rate": 4.820044296788483e-05,
+      "loss": 0.5683,
+      "step": 266
+    },
+    {
+      "epoch": 0.3991031390134529,
+      "grad_norm": 0.39563023093060706,
+      "learning_rate": 4.8172757475083056e-05,
+      "loss": 0.5297,
+      "step": 267
+    },
+    {
+      "epoch": 0.40059790732436473,
+      "grad_norm": 0.36283793011271837,
+      "learning_rate": 4.814507198228129e-05,
+      "loss": 0.5425,
+      "step": 268
+    },
+    {
+      "epoch": 0.4020926756352765,
+      "grad_norm": 0.5376391562515126,
+      "learning_rate": 4.811738648947951e-05,
+      "loss": 0.5653,
+      "step": 269
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 0.4542214775277742,
+      "learning_rate": 4.8089700996677744e-05,
+      "loss": 0.5383,
+      "step": 270
+    },
+    {
+      "epoch": 0.40508221225710017,
+      "grad_norm": 0.43031900676366025,
+      "learning_rate": 4.8062015503875976e-05,
+      "loss": 0.5349,
+      "step": 271
+    },
+    {
+      "epoch": 0.40657698056801195,
+      "grad_norm": 0.5851691250109587,
+      "learning_rate": 4.8034330011074194e-05,
+      "loss": 0.5395,
+      "step": 272
+    },
+    {
+      "epoch": 0.4080717488789238,
+      "grad_norm": 0.4861909115046238,
+      "learning_rate": 4.8006644518272426e-05,
+      "loss": 0.5258,
+      "step": 273
+    },
+    {
+      "epoch": 0.40956651718983555,
+      "grad_norm": 0.43409993837477895,
+      "learning_rate": 4.797895902547066e-05,
+      "loss": 0.5154,
+      "step": 274
+    },
+    {
+      "epoch": 0.4110612855007474,
+      "grad_norm": 0.6036256877756455,
+      "learning_rate": 4.795127353266888e-05,
+      "loss": 0.5442,
+      "step": 275
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.46925003594307324,
+      "learning_rate": 4.7923588039867114e-05,
+      "loss": 0.5409,
+      "step": 276
+    },
+    {
+      "epoch": 0.414050822122571,
+      "grad_norm": 0.5625141407586433,
+      "learning_rate": 4.789590254706534e-05,
+      "loss": 0.5315,
+      "step": 277
+    },
+    {
+      "epoch": 0.4155455904334828,
+      "grad_norm": 0.5528630032978703,
+      "learning_rate": 4.786821705426357e-05,
+      "loss": 0.5204,
+      "step": 278
+    },
+    {
+      "epoch": 0.4170403587443946,
+      "grad_norm": 0.4376863428525397,
+      "learning_rate": 4.78405315614618e-05,
+      "loss": 0.5541,
+      "step": 279
+    },
+    {
+      "epoch": 0.41853512705530643,
+      "grad_norm": 0.6411895516894164,
+      "learning_rate": 4.781284606866002e-05,
+      "loss": 0.5258,
+      "step": 280
+    },
+    {
+      "epoch": 0.4200298953662182,
+      "grad_norm": 0.49808391371952165,
+      "learning_rate": 4.778516057585825e-05,
+      "loss": 0.5503,
+      "step": 281
+    },
+    {
+      "epoch": 0.42152466367713004,
+      "grad_norm": 0.5813255086124487,
+      "learning_rate": 4.775747508305648e-05,
+      "loss": 0.5163,
+      "step": 282
+    },
+    {
+      "epoch": 0.42301943198804187,
+      "grad_norm": 0.5595407793713789,
+      "learning_rate": 4.772978959025471e-05,
+      "loss": 0.5281,
+      "step": 283
+    },
+    {
+      "epoch": 0.42451420029895365,
+      "grad_norm": 0.4306515610681992,
+      "learning_rate": 4.770210409745294e-05,
+      "loss": 0.5406,
+      "step": 284
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 0.5620983559960172,
+      "learning_rate": 4.7674418604651164e-05,
+      "loss": 0.5333,
+      "step": 285
+    },
+    {
+      "epoch": 0.42750373692077726,
+      "grad_norm": 0.44245341793438336,
+      "learning_rate": 4.7646733111849396e-05,
+      "loss": 0.5047,
+      "step": 286
+    },
+    {
+      "epoch": 0.4289985052316891,
+      "grad_norm": 0.4295485855380789,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.5538,
+      "step": 287
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.5341712049744588,
+      "learning_rate": 4.7591362126245845e-05,
+      "loss": 0.5307,
+      "step": 288
+    },
+    {
+      "epoch": 0.4319880418535127,
+      "grad_norm": 0.4215209207606345,
+      "learning_rate": 4.756367663344408e-05,
+      "loss": 0.5357,
+      "step": 289
+    },
+    {
+      "epoch": 0.4334828101644245,
+      "grad_norm": 0.4235046823429801,
+      "learning_rate": 4.753599114064231e-05,
+      "loss": 0.5438,
+      "step": 290
+    },
+    {
+      "epoch": 0.4349775784753363,
+      "grad_norm": 0.5193213910721526,
+      "learning_rate": 4.750830564784053e-05,
+      "loss": 0.5373,
+      "step": 291
+    },
+    {
+      "epoch": 0.43647234678624813,
+      "grad_norm": 0.47182451810168974,
+      "learning_rate": 4.7480620155038765e-05,
+      "loss": 0.5463,
+      "step": 292
+    },
+    {
+      "epoch": 0.43796711509715996,
+      "grad_norm": 0.4083506476367757,
+      "learning_rate": 4.745293466223699e-05,
+      "loss": 0.5319,
+      "step": 293
+    },
+    {
+      "epoch": 0.43946188340807174,
+      "grad_norm": 0.5880663401469437,
+      "learning_rate": 4.742524916943522e-05,
+      "loss": 0.5431,
+      "step": 294
+    },
+    {
+      "epoch": 0.44095665171898357,
+      "grad_norm": 0.4166651480779516,
+      "learning_rate": 4.7397563676633446e-05,
+      "loss": 0.5094,
+      "step": 295
+    },
+    {
+      "epoch": 0.44245142002989535,
+      "grad_norm": 0.4606995396018442,
+      "learning_rate": 4.736987818383167e-05,
+      "loss": 0.5213,
+      "step": 296
+    },
+    {
+      "epoch": 0.4439461883408072,
+      "grad_norm": 0.37534952631003465,
+      "learning_rate": 4.73421926910299e-05,
+      "loss": 0.5251,
+      "step": 297
+    },
+    {
+      "epoch": 0.445440956651719,
+      "grad_norm": 0.5041502258166594,
+      "learning_rate": 4.7314507198228134e-05,
+      "loss": 0.5393,
+      "step": 298
+    },
+    {
+      "epoch": 0.4469357249626308,
+      "grad_norm": 0.4795301403485978,
+      "learning_rate": 4.728682170542636e-05,
+      "loss": 0.5406,
+      "step": 299
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.45796944644288695,
+      "learning_rate": 4.725913621262459e-05,
+      "loss": 0.5438,
+      "step": 300
+    },
+    {
+      "epoch": 0.4499252615844544,
+      "grad_norm": 0.47114546088781706,
+      "learning_rate": 4.7231450719822815e-05,
+      "loss": 0.5438,
+      "step": 301
+    },
+    {
+      "epoch": 0.4514200298953662,
+      "grad_norm": 0.4315117274760818,
+      "learning_rate": 4.720376522702104e-05,
+      "loss": 0.5193,
+      "step": 302
+    },
+    {
+      "epoch": 0.452914798206278,
+      "grad_norm": 0.528868752788875,
+      "learning_rate": 4.717607973421927e-05,
+      "loss": 0.5416,
+      "step": 303
+    },
+    {
+      "epoch": 0.45440956651718983,
+      "grad_norm": 0.4374692023105216,
+      "learning_rate": 4.7148394241417496e-05,
+      "loss": 0.5487,
+      "step": 304
+    },
+    {
+      "epoch": 0.45590433482810166,
+      "grad_norm": 0.47245411223581774,
+      "learning_rate": 4.712070874861573e-05,
+      "loss": 0.5371,
+      "step": 305
+    },
+    {
+      "epoch": 0.45739910313901344,
+      "grad_norm": 0.5331388939250283,
+      "learning_rate": 4.709302325581396e-05,
+      "loss": 0.5415,
+      "step": 306
+    },
+    {
+      "epoch": 0.4588938714499253,
+      "grad_norm": 0.4236010032638064,
+      "learning_rate": 4.7065337763012184e-05,
+      "loss": 0.5564,
+      "step": 307
+    },
+    {
+      "epoch": 0.46038863976083705,
+      "grad_norm": 0.602923882103723,
+      "learning_rate": 4.7037652270210416e-05,
+      "loss": 0.5557,
+      "step": 308
+    },
+    {
+      "epoch": 0.4618834080717489,
+      "grad_norm": 0.48005292520935006,
+      "learning_rate": 4.700996677740864e-05,
+      "loss": 0.5451,
+      "step": 309
+    },
+    {
+      "epoch": 0.4633781763826607,
+      "grad_norm": 0.6417184066953241,
+      "learning_rate": 4.6982281284606865e-05,
+      "loss": 0.5324,
+      "step": 310
+    },
+    {
+      "epoch": 0.4648729446935725,
+      "grad_norm": 0.40017220842408485,
+      "learning_rate": 4.69545957918051e-05,
+      "loss": 0.5251,
+      "step": 311
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.6365287013766823,
+      "learning_rate": 4.692691029900332e-05,
+      "loss": 0.5188,
+      "step": 312
+    },
+    {
+      "epoch": 0.4678624813153961,
+      "grad_norm": 0.4244992037478277,
+      "learning_rate": 4.6899224806201553e-05,
+      "loss": 0.5214,
+      "step": 313
+    },
+    {
+      "epoch": 0.4693572496263079,
+      "grad_norm": 0.5925061457717842,
+      "learning_rate": 4.687153931339978e-05,
+      "loss": 0.527,
+      "step": 314
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 0.4085508899046393,
+      "learning_rate": 4.684385382059801e-05,
+      "loss": 0.5272,
+      "step": 315
+    },
+    {
+      "epoch": 0.47234678624813153,
+      "grad_norm": 0.4589621198899089,
+      "learning_rate": 4.681616832779624e-05,
+      "loss": 0.5065,
+      "step": 316
+    },
+    {
+      "epoch": 0.47384155455904337,
+      "grad_norm": 0.4525220417623343,
+      "learning_rate": 4.678848283499446e-05,
+      "loss": 0.5369,
+      "step": 317
+    },
+    {
+      "epoch": 0.47533632286995514,
+      "grad_norm": 0.4160216189754276,
+      "learning_rate": 4.676079734219269e-05,
+      "loss": 0.5221,
+      "step": 318
+    },
+    {
+      "epoch": 0.476831091180867,
+      "grad_norm": 0.4852950698450724,
+      "learning_rate": 4.673311184939092e-05,
+      "loss": 0.521,
+      "step": 319
+    },
+    {
+      "epoch": 0.47832585949177875,
+      "grad_norm": 0.4343707533907455,
+      "learning_rate": 4.670542635658915e-05,
+      "loss": 0.5325,
+      "step": 320
+    },
+    {
+      "epoch": 0.4798206278026906,
+      "grad_norm": 0.4412930423381022,
+      "learning_rate": 4.667774086378738e-05,
+      "loss": 0.5465,
+      "step": 321
+    },
+    {
+      "epoch": 0.4813153961136024,
+      "grad_norm": 0.4092766191076502,
+      "learning_rate": 4.6650055370985604e-05,
+      "loss": 0.5343,
+      "step": 322
+    },
+    {
+      "epoch": 0.4828101644245142,
+      "grad_norm": 0.4377560814216626,
+      "learning_rate": 4.6622369878183835e-05,
+      "loss": 0.527,
+      "step": 323
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.45490712745880213,
+      "learning_rate": 4.659468438538207e-05,
+      "loss": 0.5459,
+      "step": 324
+    },
+    {
+      "epoch": 0.4857997010463378,
+      "grad_norm": 0.509479909866683,
+      "learning_rate": 4.6566998892580285e-05,
+      "loss": 0.5225,
+      "step": 325
+    },
+    {
+      "epoch": 0.4872944693572496,
+      "grad_norm": 0.49764332953228047,
+      "learning_rate": 4.6539313399778517e-05,
+      "loss": 0.5196,
+      "step": 326
+    },
+    {
+      "epoch": 0.48878923766816146,
+      "grad_norm": 0.4253965416084026,
+      "learning_rate": 4.651162790697675e-05,
+      "loss": 0.5206,
+      "step": 327
+    },
+    {
+      "epoch": 0.49028400597907323,
+      "grad_norm": 0.4235763145759592,
+      "learning_rate": 4.648394241417497e-05,
+      "loss": 0.5284,
+      "step": 328
+    },
+    {
+      "epoch": 0.49177877428998507,
+      "grad_norm": 0.46225112800572166,
+      "learning_rate": 4.6456256921373205e-05,
+      "loss": 0.5482,
+      "step": 329
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 0.4040897005381275,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.5327,
+      "step": 330
+    },
+    {
+      "epoch": 0.4947683109118087,
+      "grad_norm": 0.5571437783882738,
+      "learning_rate": 4.640088593576966e-05,
+      "loss": 0.5385,
+      "step": 331
+    },
+    {
+      "epoch": 0.4962630792227205,
+      "grad_norm": 0.4132036316420479,
+      "learning_rate": 4.6373200442967886e-05,
+      "loss": 0.5013,
+      "step": 332
+    },
+    {
+      "epoch": 0.4977578475336323,
+      "grad_norm": 0.5702210997188409,
+      "learning_rate": 4.634551495016611e-05,
+      "loss": 0.5527,
+      "step": 333
+    },
+    {
+      "epoch": 0.4992526158445441,
+      "grad_norm": 0.5107876384268051,
+      "learning_rate": 4.631782945736434e-05,
+      "loss": 0.5461,
+      "step": 334
+    },
+    {
+      "epoch": 0.5007473841554559,
+      "grad_norm": 0.45722462543342446,
+      "learning_rate": 4.6290143964562574e-05,
+      "loss": 0.5356,
+      "step": 335
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.5300073257568595,
+      "learning_rate": 4.62624584717608e-05,
+      "loss": 0.5273,
+      "step": 336
+    },
+    {
+      "epoch": 0.5037369207772795,
+      "grad_norm": 0.44932362927426167,
+      "learning_rate": 4.623477297895903e-05,
+      "loss": 0.5494,
+      "step": 337
+    },
+    {
+      "epoch": 0.5052316890881914,
+      "grad_norm": 0.5136286886325566,
+      "learning_rate": 4.6207087486157255e-05,
+      "loss": 0.541,
+      "step": 338
+    },
+    {
+      "epoch": 0.5067264573991032,
+      "grad_norm": 0.453962903159194,
+      "learning_rate": 4.6179401993355486e-05,
+      "loss": 0.5226,
+      "step": 339
+    },
+    {
+      "epoch": 0.5082212257100149,
+      "grad_norm": 0.4798844782967225,
+      "learning_rate": 4.615171650055371e-05,
+      "loss": 0.5282,
+      "step": 340
+    },
+    {
+      "epoch": 0.5097159940209267,
+      "grad_norm": 0.44164443870518455,
+      "learning_rate": 4.6124031007751936e-05,
+      "loss": 0.5465,
+      "step": 341
+    },
+    {
+      "epoch": 0.5112107623318386,
+      "grad_norm": 0.5196104090480164,
+      "learning_rate": 4.609634551495017e-05,
+      "loss": 0.5352,
+      "step": 342
+    },
+    {
+      "epoch": 0.5127055306427504,
+      "grad_norm": 0.4963901102286841,
+      "learning_rate": 4.60686600221484e-05,
+      "loss": 0.5197,
+      "step": 343
+    },
+    {
+      "epoch": 0.5142002989536621,
+      "grad_norm": 0.43984471935701536,
+      "learning_rate": 4.6040974529346624e-05,
+      "loss": 0.5171,
+      "step": 344
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 0.45931833278231715,
+      "learning_rate": 4.6013289036544856e-05,
+      "loss": 0.5302,
+      "step": 345
+    },
+    {
+      "epoch": 0.5171898355754858,
+      "grad_norm": 0.4269919381456779,
+      "learning_rate": 4.598560354374308e-05,
+      "loss": 0.521,
+      "step": 346
+    },
+    {
+      "epoch": 0.5186846038863976,
+      "grad_norm": 0.41744098090370046,
+      "learning_rate": 4.5957918050941305e-05,
+      "loss": 0.5326,
+      "step": 347
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.43413933267081767,
+      "learning_rate": 4.593023255813954e-05,
+      "loss": 0.5259,
+      "step": 348
+    },
+    {
+      "epoch": 0.5216741405082213,
+      "grad_norm": 0.4273757371156908,
+      "learning_rate": 4.590254706533776e-05,
+      "loss": 0.5089,
+      "step": 349
+    },
+    {
+      "epoch": 0.523168908819133,
+      "grad_norm": 0.37474944751755457,
+      "learning_rate": 4.587486157253599e-05,
+      "loss": 0.5216,
+      "step": 350
+    },
+    {
+      "epoch": 0.5246636771300448,
+      "grad_norm": 0.4394924940577488,
+      "learning_rate": 4.5847176079734225e-05,
+      "loss": 0.5259,
+      "step": 351
+    },
+    {
+      "epoch": 0.5261584454409567,
+      "grad_norm": 0.43878136909501403,
+      "learning_rate": 4.581949058693245e-05,
+      "loss": 0.5245,
+      "step": 352
+    },
+    {
+      "epoch": 0.5276532137518685,
+      "grad_norm": 0.3630422364903993,
+      "learning_rate": 4.579180509413068e-05,
+      "loss": 0.5415,
+      "step": 353
+    },
+    {
+      "epoch": 0.5291479820627802,
+      "grad_norm": 0.4479279785841185,
+      "learning_rate": 4.5764119601328906e-05,
+      "loss": 0.521,
+      "step": 354
+    },
+    {
+      "epoch": 0.5306427503736921,
+      "grad_norm": 0.3773854691864007,
+      "learning_rate": 4.573643410852713e-05,
+      "loss": 0.5264,
+      "step": 355
+    },
+    {
+      "epoch": 0.5321375186846039,
+      "grad_norm": 0.5297577211396239,
+      "learning_rate": 4.570874861572536e-05,
+      "loss": 0.5512,
+      "step": 356
+    },
+    {
+      "epoch": 0.5336322869955157,
+      "grad_norm": 0.532569688956907,
+      "learning_rate": 4.568106312292359e-05,
+      "loss": 0.5409,
+      "step": 357
+    },
+    {
+      "epoch": 0.5351270553064275,
+      "grad_norm": 0.3938424634699327,
+      "learning_rate": 4.565337763012182e-05,
+      "loss": 0.5342,
+      "step": 358
+    },
+    {
+      "epoch": 0.5366218236173393,
+      "grad_norm": 0.5490407784437218,
+      "learning_rate": 4.562569213732005e-05,
+      "loss": 0.548,
+      "step": 359
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.3935732231886853,
+      "learning_rate": 4.5598006644518275e-05,
+      "loss": 0.5155,
+      "step": 360
+    },
+    {
+      "epoch": 0.5396113602391629,
+      "grad_norm": 0.5205658720090466,
+      "learning_rate": 4.557032115171651e-05,
+      "loss": 0.541,
+      "step": 361
+    },
+    {
+      "epoch": 0.5411061285500748,
+      "grad_norm": 0.4333101504455433,
+      "learning_rate": 4.554263565891473e-05,
+      "loss": 0.5255,
+      "step": 362
+    },
+    {
+      "epoch": 0.5426008968609866,
+      "grad_norm": 0.4339381002427947,
+      "learning_rate": 4.5514950166112956e-05,
+      "loss": 0.5443,
+      "step": 363
+    },
+    {
+      "epoch": 0.5440956651718983,
+      "grad_norm": 0.5618240703354906,
+      "learning_rate": 4.548726467331119e-05,
+      "loss": 0.5417,
+      "step": 364
+    },
+    {
+      "epoch": 0.5455904334828101,
+      "grad_norm": 0.4628042814793409,
+      "learning_rate": 4.545957918050941e-05,
+      "loss": 0.5373,
+      "step": 365
+    },
+    {
+      "epoch": 0.547085201793722,
+      "grad_norm": 0.5103447365650151,
+      "learning_rate": 4.5431893687707644e-05,
+      "loss": 0.5328,
+      "step": 366
+    },
+    {
+      "epoch": 0.5485799701046338,
+      "grad_norm": 0.4840605212286738,
+      "learning_rate": 4.5404208194905876e-05,
+      "loss": 0.5047,
+      "step": 367
+    },
+    {
+      "epoch": 0.5500747384155455,
+      "grad_norm": 0.5031398184620589,
+      "learning_rate": 4.53765227021041e-05,
+      "loss": 0.5399,
+      "step": 368
+    },
+    {
+      "epoch": 0.5515695067264574,
+      "grad_norm": 0.510978023524668,
+      "learning_rate": 4.5348837209302326e-05,
+      "loss": 0.5419,
+      "step": 369
+    },
+    {
+      "epoch": 0.5530642750373692,
+      "grad_norm": 0.44906318612611507,
+      "learning_rate": 4.532115171650056e-05,
+      "loss": 0.5197,
+      "step": 370
+    },
+    {
+      "epoch": 0.554559043348281,
+      "grad_norm": 0.5002781496370401,
+      "learning_rate": 4.529346622369878e-05,
+      "loss": 0.5349,
+      "step": 371
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.4632478554998675,
+      "learning_rate": 4.5265780730897014e-05,
+      "loss": 0.5399,
+      "step": 372
+    },
+    {
+      "epoch": 0.5575485799701047,
+      "grad_norm": 0.4342351823887393,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.5312,
+      "step": 373
+    },
+    {
+      "epoch": 0.5590433482810164,
+      "grad_norm": 0.5167668392870697,
+      "learning_rate": 4.521040974529347e-05,
+      "loss": 0.5215,
+      "step": 374
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 0.4297296545124882,
+      "learning_rate": 4.51827242524917e-05,
+      "loss": 0.5123,
+      "step": 375
+    },
+    {
+      "epoch": 0.5620328849028401,
+      "grad_norm": 0.5454516303359211,
+      "learning_rate": 4.5155038759689926e-05,
+      "loss": 0.5116,
+      "step": 376
+    },
+    {
+      "epoch": 0.5635276532137519,
+      "grad_norm": 0.4633260298797532,
+      "learning_rate": 4.512735326688815e-05,
+      "loss": 0.5069,
+      "step": 377
+    },
+    {
+      "epoch": 0.5650224215246636,
+      "grad_norm": 0.4690736114549489,
+      "learning_rate": 4.509966777408638e-05,
+      "loss": 0.5357,
+      "step": 378
+    },
+    {
+      "epoch": 0.5665171898355755,
+      "grad_norm": 0.5855501309074868,
+      "learning_rate": 4.507198228128461e-05,
+      "loss": 0.5308,
+      "step": 379
+    },
+    {
+      "epoch": 0.5680119581464873,
+      "grad_norm": 0.4422252132161514,
+      "learning_rate": 4.504429678848284e-05,
+      "loss": 0.5139,
+      "step": 380
+    },
+    {
+      "epoch": 0.5695067264573991,
+      "grad_norm": 0.582361932955961,
+      "learning_rate": 4.5016611295681064e-05,
+      "loss": 0.5235,
+      "step": 381
+    },
+    {
+      "epoch": 0.5710014947683109,
+      "grad_norm": 0.5658206270555286,
+      "learning_rate": 4.4988925802879295e-05,
+      "loss": 0.5235,
+      "step": 382
+    },
+    {
+      "epoch": 0.5724962630792227,
+      "grad_norm": 0.4434155599452413,
+      "learning_rate": 4.496124031007753e-05,
+      "loss": 0.537,
+      "step": 383
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.5938141471642983,
+      "learning_rate": 4.4933554817275745e-05,
+      "loss": 0.5002,
+      "step": 384
+    },
+    {
+      "epoch": 0.5754857997010463,
+      "grad_norm": 0.4391147698674983,
+      "learning_rate": 4.490586932447398e-05,
+      "loss": 0.5255,
+      "step": 385
+    },
+    {
+      "epoch": 0.5769805680119582,
+      "grad_norm": 0.525194291444884,
+      "learning_rate": 4.48781838316722e-05,
+      "loss": 0.5263,
+      "step": 386
+    },
+    {
+      "epoch": 0.57847533632287,
+      "grad_norm": 0.5084770312479528,
+      "learning_rate": 4.485049833887043e-05,
+      "loss": 0.5233,
+      "step": 387
+    },
+    {
+      "epoch": 0.5799701046337817,
+      "grad_norm": 0.4031705389660936,
+      "learning_rate": 4.4822812846068665e-05,
+      "loss": 0.5011,
+      "step": 388
+    },
+    {
+      "epoch": 0.5814648729446936,
+      "grad_norm": 0.5138169240657705,
+      "learning_rate": 4.479512735326689e-05,
+      "loss": 0.5289,
+      "step": 389
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 0.4464907758975677,
+      "learning_rate": 4.476744186046512e-05,
+      "loss": 0.5432,
+      "step": 390
+    },
+    {
+      "epoch": 0.5844544095665172,
+      "grad_norm": 0.4655116745820642,
+      "learning_rate": 4.4739756367663346e-05,
+      "loss": 0.5094,
+      "step": 391
+    },
+    {
+      "epoch": 0.585949177877429,
+      "grad_norm": 0.5563043114214494,
+      "learning_rate": 4.471207087486157e-05,
+      "loss": 0.5398,
+      "step": 392
+    },
+    {
+      "epoch": 0.5874439461883408,
+      "grad_norm": 0.3683276987689894,
+      "learning_rate": 4.46843853820598e-05,
+      "loss": 0.5156,
+      "step": 393
+    },
+    {
+      "epoch": 0.5889387144992526,
+      "grad_norm": 0.5102906534561898,
+      "learning_rate": 4.465669988925803e-05,
+      "loss": 0.5303,
+      "step": 394
+    },
+    {
+      "epoch": 0.5904334828101644,
+      "grad_norm": 0.42232952493189435,
+      "learning_rate": 4.462901439645626e-05,
+      "loss": 0.518,
+      "step": 395
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.43047822683455383,
+      "learning_rate": 4.460132890365449e-05,
+      "loss": 0.5247,
+      "step": 396
+    },
+    {
+      "epoch": 0.593423019431988,
+      "grad_norm": 0.4549819318662644,
+      "learning_rate": 4.4573643410852715e-05,
+      "loss": 0.5134,
+      "step": 397
+    },
+    {
+      "epoch": 0.5949177877428998,
+      "grad_norm": 0.3726469148315398,
+      "learning_rate": 4.4545957918050947e-05,
+      "loss": 0.5306,
+      "step": 398
+    },
+    {
+      "epoch": 0.5964125560538116,
+      "grad_norm": 0.4956866177359307,
+      "learning_rate": 4.451827242524917e-05,
+      "loss": 0.5289,
+      "step": 399
+    },
+    {
+      "epoch": 0.5979073243647235,
+      "grad_norm": 0.38387301452888895,
+      "learning_rate": 4.4490586932447396e-05,
+      "loss": 0.5218,
+      "step": 400
+    },
+    {
+      "epoch": 0.5994020926756353,
+      "grad_norm": 0.4050379888303409,
+      "learning_rate": 4.446290143964563e-05,
+      "loss": 0.5116,
+      "step": 401
+    },
+    {
+      "epoch": 0.600896860986547,
+      "grad_norm": 0.4022535768162567,
+      "learning_rate": 4.443521594684385e-05,
+      "loss": 0.5179,
+      "step": 402
+    },
+    {
+      "epoch": 0.6023916292974589,
+      "grad_norm": 0.3961746271831333,
+      "learning_rate": 4.4407530454042084e-05,
+      "loss": 0.4996,
+      "step": 403
+    },
+    {
+      "epoch": 0.6038863976083707,
+      "grad_norm": 0.45297085942986404,
+      "learning_rate": 4.4379844961240316e-05,
+      "loss": 0.5333,
+      "step": 404
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 0.3870838985957634,
+      "learning_rate": 4.435215946843854e-05,
+      "loss": 0.525,
+      "step": 405
+    },
+    {
+      "epoch": 0.6068759342301944,
+      "grad_norm": 0.43820933217716346,
+      "learning_rate": 4.432447397563677e-05,
+      "loss": 0.5119,
+      "step": 406
+    },
+    {
+      "epoch": 0.6083707025411061,
+      "grad_norm": 0.49310218192616156,
+      "learning_rate": 4.4296788482835e-05,
+      "loss": 0.4997,
+      "step": 407
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.3382108101397666,
+      "learning_rate": 4.426910299003322e-05,
+      "loss": 0.5384,
+      "step": 408
+    },
+    {
+      "epoch": 0.6113602391629297,
+      "grad_norm": 0.4685376128615174,
+      "learning_rate": 4.424141749723145e-05,
+      "loss": 0.5218,
+      "step": 409
+    },
+    {
+      "epoch": 0.6128550074738416,
+      "grad_norm": 0.43209802371532424,
+      "learning_rate": 4.421373200442968e-05,
+      "loss": 0.5231,
+      "step": 410
+    },
+    {
+      "epoch": 0.6143497757847534,
+      "grad_norm": 0.3780611893510471,
+      "learning_rate": 4.418604651162791e-05,
+      "loss": 0.5061,
+      "step": 411
+    },
+    {
+      "epoch": 0.6158445440956651,
+      "grad_norm": 0.4991043834476504,
+      "learning_rate": 4.415836101882614e-05,
+      "loss": 0.5262,
+      "step": 412
+    },
+    {
+      "epoch": 0.617339312406577,
+      "grad_norm": 0.3650050971446354,
+      "learning_rate": 4.4130675526024366e-05,
+      "loss": 0.5217,
+      "step": 413
+    },
+    {
+      "epoch": 0.6188340807174888,
+      "grad_norm": 0.45727803369578446,
+      "learning_rate": 4.410299003322259e-05,
+      "loss": 0.5165,
+      "step": 414
+    },
+    {
+      "epoch": 0.6203288490284006,
+      "grad_norm": 0.4258496591067367,
+      "learning_rate": 4.407530454042082e-05,
+      "loss": 0.4942,
+      "step": 415
+    },
+    {
+      "epoch": 0.6218236173393124,
+      "grad_norm": 0.38820837435075223,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 0.5166,
+      "step": 416
+    },
+    {
+      "epoch": 0.6233183856502242,
+      "grad_norm": 0.4016987068908348,
+      "learning_rate": 4.401993355481728e-05,
+      "loss": 0.5347,
+      "step": 417
+    },
+    {
+      "epoch": 0.624813153961136,
+      "grad_norm": 0.3821397323315488,
+      "learning_rate": 4.3992248062015504e-05,
+      "loss": 0.5201,
+      "step": 418
+    },
+    {
+      "epoch": 0.6263079222720478,
+      "grad_norm": 0.3979299692330797,
+      "learning_rate": 4.3964562569213735e-05,
+      "loss": 0.5221,
+      "step": 419
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.38198797556793757,
+      "learning_rate": 4.393687707641197e-05,
+      "loss": 0.5136,
+      "step": 420
+    },
+    {
+      "epoch": 0.6292974588938715,
+      "grad_norm": 0.3927533574799816,
+      "learning_rate": 4.390919158361019e-05,
+      "loss": 0.509,
+      "step": 421
+    },
+    {
+      "epoch": 0.6307922272047832,
+      "grad_norm": 0.4003196859744738,
+      "learning_rate": 4.3881506090808416e-05,
+      "loss": 0.5151,
+      "step": 422
+    },
+    {
+      "epoch": 0.6322869955156951,
+      "grad_norm": 0.40165694404393115,
+      "learning_rate": 4.385382059800665e-05,
+      "loss": 0.5108,
+      "step": 423
+    },
+    {
+      "epoch": 0.6337817638266069,
+      "grad_norm": 0.379114033319736,
+      "learning_rate": 4.382613510520487e-05,
+      "loss": 0.5419,
+      "step": 424
+    },
+    {
+      "epoch": 0.6352765321375187,
+      "grad_norm": 0.4181769138283292,
+      "learning_rate": 4.3798449612403104e-05,
+      "loss": 0.5548,
+      "step": 425
+    },
+    {
+      "epoch": 0.6367713004484304,
+      "grad_norm": 0.3947896581680498,
+      "learning_rate": 4.377076411960133e-05,
+      "loss": 0.5283,
+      "step": 426
+    },
+    {
+      "epoch": 0.6382660687593423,
+      "grad_norm": 0.41767188245258996,
+      "learning_rate": 4.374307862679956e-05,
+      "loss": 0.5106,
+      "step": 427
+    },
+    {
+      "epoch": 0.6397608370702541,
+      "grad_norm": 0.42698321739927153,
+      "learning_rate": 4.371539313399779e-05,
+      "loss": 0.5146,
+      "step": 428
+    },
+    {
+      "epoch": 0.6412556053811659,
+      "grad_norm": 0.4874691197625313,
+      "learning_rate": 4.368770764119601e-05,
+      "loss": 0.5147,
+      "step": 429
+    },
+    {
+      "epoch": 0.6427503736920778,
+      "grad_norm": 0.3682854131996429,
+      "learning_rate": 4.366002214839424e-05,
+      "loss": 0.5076,
+      "step": 430
+    },
+    {
+      "epoch": 0.6442451420029895,
+      "grad_norm": 0.4534473098902625,
+      "learning_rate": 4.3632336655592474e-05,
+      "loss": 0.516,
+      "step": 431
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.39716862011359866,
+      "learning_rate": 4.36046511627907e-05,
+      "loss": 0.4831,
+      "step": 432
+    },
+    {
+      "epoch": 0.6472346786248132,
+      "grad_norm": 0.44090184597824084,
+      "learning_rate": 4.357696566998893e-05,
+      "loss": 0.5121,
+      "step": 433
+    },
+    {
+      "epoch": 0.648729446935725,
+      "grad_norm": 0.45602328532228237,
+      "learning_rate": 4.3549280177187155e-05,
+      "loss": 0.5121,
+      "step": 434
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 0.41498994057663086,
+      "learning_rate": 4.3521594684385386e-05,
+      "loss": 0.5298,
+      "step": 435
+    },
+    {
+      "epoch": 0.6517189835575485,
+      "grad_norm": 0.4550452372119635,
+      "learning_rate": 4.349390919158362e-05,
+      "loss": 0.5184,
+      "step": 436
+    },
+    {
+      "epoch": 0.6532137518684604,
+      "grad_norm": 0.3774547948584679,
+      "learning_rate": 4.3466223698781836e-05,
+      "loss": 0.5346,
+      "step": 437
+    },
+    {
+      "epoch": 0.6547085201793722,
+      "grad_norm": 0.5446510146164512,
+      "learning_rate": 4.343853820598007e-05,
+      "loss": 0.5472,
+      "step": 438
+    },
+    {
+      "epoch": 0.656203288490284,
+      "grad_norm": 0.3485798358068613,
+      "learning_rate": 4.34108527131783e-05,
+      "loss": 0.5103,
+      "step": 439
+    },
+    {
+      "epoch": 0.6576980568011959,
+      "grad_norm": 0.5732013014525139,
+      "learning_rate": 4.3383167220376524e-05,
+      "loss": 0.5066,
+      "step": 440
+    },
+    {
+      "epoch": 0.6591928251121076,
+      "grad_norm": 0.3780653863007362,
+      "learning_rate": 4.3355481727574756e-05,
+      "loss": 0.5193,
+      "step": 441
+    },
+    {
+      "epoch": 0.6606875934230194,
+      "grad_norm": 0.4339106511601642,
+      "learning_rate": 4.332779623477298e-05,
+      "loss": 0.5297,
+      "step": 442
+    },
+    {
+      "epoch": 0.6621823617339312,
+      "grad_norm": 0.40882340558583663,
+      "learning_rate": 4.330011074197121e-05,
+      "loss": 0.5303,
+      "step": 443
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.4651900906094084,
+      "learning_rate": 4.327242524916944e-05,
+      "loss": 0.5119,
+      "step": 444
+    },
+    {
+      "epoch": 0.6651718983557549,
+      "grad_norm": 0.3935719496853382,
+      "learning_rate": 4.324473975636766e-05,
+      "loss": 0.5192,
+      "step": 445
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.4510750353285506,
+      "learning_rate": 4.321705426356589e-05,
+      "loss": 0.5219,
+      "step": 446
+    },
+    {
+      "epoch": 0.6681614349775785,
+      "grad_norm": 0.3863956843242051,
+      "learning_rate": 4.3189368770764125e-05,
+      "loss": 0.5018,
+      "step": 447
+    },
+    {
+      "epoch": 0.6696562032884903,
+      "grad_norm": 0.5676982840980781,
+      "learning_rate": 4.316168327796235e-05,
+      "loss": 0.5279,
+      "step": 448
+    },
+    {
+      "epoch": 0.6711509715994021,
+      "grad_norm": 0.41165878106375514,
+      "learning_rate": 4.313399778516058e-05,
+      "loss": 0.525,
+      "step": 449
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 0.464731812191866,
+      "learning_rate": 4.3106312292358806e-05,
+      "loss": 0.5027,
+      "step": 450
+    },
+    {
+      "epoch": 0.6741405082212257,
+      "grad_norm": 0.3818308445522795,
+      "learning_rate": 4.307862679955704e-05,
+      "loss": 0.494,
+      "step": 451
+    },
+    {
+      "epoch": 0.6756352765321375,
+      "grad_norm": 0.41393360577086935,
+      "learning_rate": 4.305094130675526e-05,
+      "loss": 0.5029,
+      "step": 452
+    },
+    {
+      "epoch": 0.6771300448430493,
+      "grad_norm": 0.37685022535737434,
+      "learning_rate": 4.302325581395349e-05,
+      "loss": 0.527,
+      "step": 453
+    },
+    {
+      "epoch": 0.6786248131539612,
+      "grad_norm": 0.45976902444548984,
+      "learning_rate": 4.299557032115172e-05,
+      "loss": 0.4942,
+      "step": 454
+    },
+    {
+      "epoch": 0.680119581464873,
+      "grad_norm": 0.3575303079797226,
+      "learning_rate": 4.2967884828349944e-05,
+      "loss": 0.5076,
+      "step": 455
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.392031415299761,
+      "learning_rate": 4.2940199335548175e-05,
+      "loss": 0.5395,
+      "step": 456
+    },
+    {
+      "epoch": 0.6831091180866966,
+      "grad_norm": 0.43451658481340427,
+      "learning_rate": 4.291251384274641e-05,
+      "loss": 0.5002,
+      "step": 457
+    },
+    {
+      "epoch": 0.6846038863976084,
+      "grad_norm": 0.39576309745251165,
+      "learning_rate": 4.288482834994463e-05,
+      "loss": 0.532,
+      "step": 458
+    },
+    {
+      "epoch": 0.6860986547085202,
+      "grad_norm": 0.3674743929295811,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.5339,
+      "step": 459
+    },
+    {
+      "epoch": 0.6875934230194319,
+      "grad_norm": 0.4333460322016697,
+      "learning_rate": 4.282945736434109e-05,
+      "loss": 0.4989,
+      "step": 460
+    },
+    {
+      "epoch": 0.6890881913303438,
+      "grad_norm": 0.4354804323720021,
+      "learning_rate": 4.280177187153931e-05,
+      "loss": 0.5152,
+      "step": 461
+    },
+    {
+      "epoch": 0.6905829596412556,
+      "grad_norm": 0.5231022542065054,
+      "learning_rate": 4.2774086378737544e-05,
+      "loss": 0.5705,
+      "step": 462
+    },
+    {
+      "epoch": 0.6920777279521674,
+      "grad_norm": 0.4217540228509454,
+      "learning_rate": 4.274640088593577e-05,
+      "loss": 0.5221,
+      "step": 463
+    },
+    {
+      "epoch": 0.6935724962630793,
+      "grad_norm": 0.4196727243386211,
+      "learning_rate": 4.2718715393134e-05,
+      "loss": 0.5036,
+      "step": 464
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 0.38291429273426786,
+      "learning_rate": 4.269102990033223e-05,
+      "loss": 0.5174,
+      "step": 465
+    },
+    {
+      "epoch": 0.6965620328849028,
+      "grad_norm": 0.44218893252998387,
+      "learning_rate": 4.266334440753045e-05,
+      "loss": 0.5071,
+      "step": 466
+    },
+    {
+      "epoch": 0.6980568011958147,
+      "grad_norm": 0.4546282481391166,
+      "learning_rate": 4.263565891472868e-05,
+      "loss": 0.5068,
+      "step": 467
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 0.38647239315239096,
+      "learning_rate": 4.2607973421926913e-05,
+      "loss": 0.5227,
+      "step": 468
+    },
+    {
+      "epoch": 0.7010463378176383,
+      "grad_norm": 0.48928288192215114,
+      "learning_rate": 4.258028792912514e-05,
+      "loss": 0.5193,
+      "step": 469
+    },
+    {
+      "epoch": 0.70254110612855,
+      "grad_norm": 0.38959864749174067,
+      "learning_rate": 4.255260243632337e-05,
+      "loss": 0.4964,
+      "step": 470
+    },
+    {
+      "epoch": 0.7040358744394619,
+      "grad_norm": 0.4215212445169176,
+      "learning_rate": 4.2524916943521595e-05,
+      "loss": 0.5305,
+      "step": 471
+    },
+    {
+      "epoch": 0.7055306427503737,
+      "grad_norm": 0.3679819066246997,
+      "learning_rate": 4.2497231450719826e-05,
+      "loss": 0.5308,
+      "step": 472
+    },
+    {
+      "epoch": 0.7070254110612855,
+      "grad_norm": 0.4349746070233607,
+      "learning_rate": 4.246954595791806e-05,
+      "loss": 0.5112,
+      "step": 473
+    },
+    {
+      "epoch": 0.7085201793721974,
+      "grad_norm": 0.4936666709213464,
+      "learning_rate": 4.2441860465116276e-05,
+      "loss": 0.5197,
+      "step": 474
+    },
+    {
+      "epoch": 0.7100149476831091,
+      "grad_norm": 0.3685636094832015,
+      "learning_rate": 4.241417497231451e-05,
+      "loss": 0.487,
+      "step": 475
+    },
+    {
+      "epoch": 0.7115097159940209,
+      "grad_norm": 0.5281749571738503,
+      "learning_rate": 4.238648947951274e-05,
+      "loss": 0.515,
+      "step": 476
+    },
+    {
+      "epoch": 0.7130044843049327,
+      "grad_norm": 0.4437088194279032,
+      "learning_rate": 4.2358803986710964e-05,
+      "loss": 0.5122,
+      "step": 477
+    },
+    {
+      "epoch": 0.7144992526158446,
+      "grad_norm": 0.46040052294123207,
+      "learning_rate": 4.2331118493909195e-05,
+      "loss": 0.5335,
+      "step": 478
+    },
+    {
+      "epoch": 0.7159940209267563,
+      "grad_norm": 0.41162026392895307,
+      "learning_rate": 4.230343300110742e-05,
+      "loss": 0.5079,
+      "step": 479
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.39900839092279733,
+      "learning_rate": 4.227574750830565e-05,
+      "loss": 0.5119,
+      "step": 480
+    },
+    {
+      "epoch": 0.71898355754858,
+      "grad_norm": 0.5138580904375739,
+      "learning_rate": 4.2248062015503877e-05,
+      "loss": 0.5213,
+      "step": 481
+    },
+    {
+      "epoch": 0.7204783258594918,
+      "grad_norm": 0.40334631121948045,
+      "learning_rate": 4.22203765227021e-05,
+      "loss": 0.5033,
+      "step": 482
+    },
+    {
+      "epoch": 0.7219730941704036,
+      "grad_norm": 0.5377765217794374,
+      "learning_rate": 4.219269102990033e-05,
+      "loss": 0.5437,
+      "step": 483
+    },
+    {
+      "epoch": 0.7234678624813154,
+      "grad_norm": 0.4668533414210699,
+      "learning_rate": 4.2165005537098565e-05,
+      "loss": 0.5075,
+      "step": 484
+    },
+    {
+      "epoch": 0.7249626307922272,
+      "grad_norm": 0.4971375063041717,
+      "learning_rate": 4.213732004429679e-05,
+      "loss": 0.5106,
+      "step": 485
+    },
+    {
+      "epoch": 0.726457399103139,
+      "grad_norm": 0.5172111587977276,
+      "learning_rate": 4.210963455149502e-05,
+      "loss": 0.51,
+      "step": 486
+    },
+    {
+      "epoch": 0.7279521674140508,
+      "grad_norm": 0.4301345180527731,
+      "learning_rate": 4.2081949058693246e-05,
+      "loss": 0.5311,
+      "step": 487
+    },
+    {
+      "epoch": 0.7294469357249627,
+      "grad_norm": 0.472145404867057,
+      "learning_rate": 4.205426356589148e-05,
+      "loss": 0.4997,
+      "step": 488
+    },
+    {
+      "epoch": 0.7309417040358744,
+      "grad_norm": 0.42818612240850873,
+      "learning_rate": 4.20265780730897e-05,
+      "loss": 0.5173,
+      "step": 489
+    },
+    {
+      "epoch": 0.7324364723467862,
+      "grad_norm": 0.424078264101112,
+      "learning_rate": 4.199889258028793e-05,
+      "loss": 0.5023,
+      "step": 490
+    },
+    {
+      "epoch": 0.7339312406576981,
+      "grad_norm": 0.38556606584532255,
+      "learning_rate": 4.197120708748616e-05,
+      "loss": 0.5093,
+      "step": 491
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 0.39483386944982607,
+      "learning_rate": 4.194352159468439e-05,
+      "loss": 0.4909,
+      "step": 492
+    },
+    {
+      "epoch": 0.7369207772795217,
+      "grad_norm": 0.4721211602570245,
+      "learning_rate": 4.1915836101882615e-05,
+      "loss": 0.5214,
+      "step": 493
+    },
+    {
+      "epoch": 0.7384155455904334,
+      "grad_norm": 0.4262811747610391,
+      "learning_rate": 4.1888150609080846e-05,
+      "loss": 0.5165,
+      "step": 494
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 0.4347121159238024,
+      "learning_rate": 4.186046511627907e-05,
+      "loss": 0.5245,
+      "step": 495
+    },
+    {
+      "epoch": 0.7414050822122571,
+      "grad_norm": 0.3852401093978401,
+      "learning_rate": 4.1832779623477296e-05,
+      "loss": 0.5103,
+      "step": 496
+    },
+    {
+      "epoch": 0.7428998505231689,
+      "grad_norm": 0.41136654420830326,
+      "learning_rate": 4.180509413067553e-05,
+      "loss": 0.4981,
+      "step": 497
+    },
+    {
+      "epoch": 0.7443946188340808,
+      "grad_norm": 0.46731197568732047,
+      "learning_rate": 4.177740863787375e-05,
+      "loss": 0.5072,
+      "step": 498
+    },
+    {
+      "epoch": 0.7458893871449925,
+      "grad_norm": 0.3747683877132901,
+      "learning_rate": 4.1749723145071984e-05,
+      "loss": 0.4979,
+      "step": 499
+    },
+    {
+      "epoch": 0.7473841554559043,
+      "grad_norm": 0.4047345620387918,
+      "learning_rate": 4.1722037652270216e-05,
+      "loss": 0.5125,
+      "step": 500
+    },
+    {
+      "epoch": 0.7488789237668162,
+      "grad_norm": 0.37220406668785466,
+      "learning_rate": 4.169435215946844e-05,
+      "loss": 0.5233,
+      "step": 501
+    },
+    {
+      "epoch": 0.750373692077728,
+      "grad_norm": 0.4599502495062675,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4952,
+      "step": 502
+    },
+    {
+      "epoch": 0.7518684603886397,
+      "grad_norm": 0.3705135062370848,
+      "learning_rate": 4.16389811738649e-05,
+      "loss": 0.5041,
+      "step": 503
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.4710494119084625,
+      "learning_rate": 4.161129568106312e-05,
+      "loss": 0.5021,
+      "step": 504
+    },
+    {
+      "epoch": 0.7548579970104634,
+      "grad_norm": 0.42431539871718355,
+      "learning_rate": 4.158361018826135e-05,
+      "loss": 0.5114,
+      "step": 505
+    },
+    {
+      "epoch": 0.7563527653213752,
+      "grad_norm": 0.5643325972045106,
+      "learning_rate": 4.155592469545958e-05,
+      "loss": 0.515,
+      "step": 506
+    },
+    {
+      "epoch": 0.757847533632287,
+      "grad_norm": 0.5552640356039347,
+      "learning_rate": 4.152823920265781e-05,
+      "loss": 0.5256,
+      "step": 507
+    },
+    {
+      "epoch": 0.7593423019431988,
+      "grad_norm": 0.4875339692322978,
+      "learning_rate": 4.150055370985604e-05,
+      "loss": 0.5239,
+      "step": 508
+    },
+    {
+      "epoch": 0.7608370702541106,
+      "grad_norm": 0.5046252537109269,
+      "learning_rate": 4.1472868217054266e-05,
+      "loss": 0.5344,
+      "step": 509
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 0.45818957099859786,
+      "learning_rate": 4.14451827242525e-05,
+      "loss": 0.4939,
+      "step": 510
+    },
+    {
+      "epoch": 0.7638266068759342,
+      "grad_norm": 0.44101183218092854,
+      "learning_rate": 4.141749723145072e-05,
+      "loss": 0.5221,
+      "step": 511
+    },
+    {
+      "epoch": 0.7653213751868461,
+      "grad_norm": 0.6197376254875241,
+      "learning_rate": 4.138981173864895e-05,
+      "loss": 0.4956,
+      "step": 512
+    },
+    {
+      "epoch": 0.7668161434977578,
+      "grad_norm": 0.4403867826705504,
+      "learning_rate": 4.136212624584718e-05,
+      "loss": 0.5334,
+      "step": 513
+    },
+    {
+      "epoch": 0.7683109118086696,
+      "grad_norm": 0.6110980701753278,
+      "learning_rate": 4.1334440753045404e-05,
+      "loss": 0.5255,
+      "step": 514
+    },
+    {
+      "epoch": 0.7698056801195815,
+      "grad_norm": 0.4346299960182701,
+      "learning_rate": 4.1306755260243635e-05,
+      "loss": 0.5214,
+      "step": 515
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.6514371867292363,
+      "learning_rate": 4.127906976744187e-05,
+      "loss": 0.5027,
+      "step": 516
+    },
+    {
+      "epoch": 0.772795216741405,
+      "grad_norm": 0.37059553064618117,
+      "learning_rate": 4.125138427464009e-05,
+      "loss": 0.4962,
+      "step": 517
+    },
+    {
+      "epoch": 0.7742899850523169,
+      "grad_norm": 0.5150051920625823,
+      "learning_rate": 4.122369878183832e-05,
+      "loss": 0.5189,
+      "step": 518
+    },
+    {
+      "epoch": 0.7757847533632287,
+      "grad_norm": 0.39786915768675934,
+      "learning_rate": 4.119601328903655e-05,
+      "loss": 0.518,
+      "step": 519
+    },
+    {
+      "epoch": 0.7772795216741405,
+      "grad_norm": 0.49739233527968585,
+      "learning_rate": 4.116832779623477e-05,
+      "loss": 0.5088,
+      "step": 520
+    },
+    {
+      "epoch": 0.7787742899850523,
+      "grad_norm": 0.4628976246383883,
+      "learning_rate": 4.1140642303433004e-05,
+      "loss": 0.4981,
+      "step": 521
+    },
+    {
+      "epoch": 0.7802690582959642,
+      "grad_norm": 0.38760600389093824,
+      "learning_rate": 4.111295681063123e-05,
+      "loss": 0.5372,
+      "step": 522
+    },
+    {
+      "epoch": 0.7817638266068759,
+      "grad_norm": 0.5192593805259457,
+      "learning_rate": 4.108527131782946e-05,
+      "loss": 0.507,
+      "step": 523
+    },
+    {
+      "epoch": 0.7832585949177877,
+      "grad_norm": 0.3600337098638405,
+      "learning_rate": 4.105758582502769e-05,
+      "loss": 0.5083,
+      "step": 524
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 0.42075690600504567,
+      "learning_rate": 4.102990033222592e-05,
+      "loss": 0.5133,
+      "step": 525
+    },
+    {
+      "epoch": 0.7862481315396114,
+      "grad_norm": 0.37575070388130516,
+      "learning_rate": 4.100221483942414e-05,
+      "loss": 0.5201,
+      "step": 526
+    },
+    {
+      "epoch": 0.7877428998505231,
+      "grad_norm": 0.40265034224543167,
+      "learning_rate": 4.097452934662237e-05,
+      "loss": 0.5068,
+      "step": 527
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.349257970629535,
+      "learning_rate": 4.09468438538206e-05,
+      "loss": 0.4849,
+      "step": 528
+    },
+    {
+      "epoch": 0.7907324364723468,
+      "grad_norm": 0.3659865290440735,
+      "learning_rate": 4.091915836101883e-05,
+      "loss": 0.4908,
+      "step": 529
+    },
+    {
+      "epoch": 0.7922272047832586,
+      "grad_norm": 0.3894642242427512,
+      "learning_rate": 4.0891472868217055e-05,
+      "loss": 0.472,
+      "step": 530
+    },
+    {
+      "epoch": 0.7937219730941704,
+      "grad_norm": 0.43592056597354883,
+      "learning_rate": 4.0863787375415286e-05,
+      "loss": 0.5034,
+      "step": 531
+    },
+    {
+      "epoch": 0.7952167414050823,
+      "grad_norm": 0.3567532613138878,
+      "learning_rate": 4.083610188261351e-05,
+      "loss": 0.5299,
+      "step": 532
+    },
+    {
+      "epoch": 0.796711509715994,
+      "grad_norm": 0.4632091147408238,
+      "learning_rate": 4.080841638981174e-05,
+      "loss": 0.5292,
+      "step": 533
+    },
+    {
+      "epoch": 0.7982062780269058,
+      "grad_norm": 0.37011598059169587,
+      "learning_rate": 4.078073089700997e-05,
+      "loss": 0.4967,
+      "step": 534
+    },
+    {
+      "epoch": 0.7997010463378177,
+      "grad_norm": 0.44396281343302413,
+      "learning_rate": 4.075304540420819e-05,
+      "loss": 0.5301,
+      "step": 535
+    },
+    {
+      "epoch": 0.8011958146487295,
+      "grad_norm": 0.37323039989156426,
+      "learning_rate": 4.0725359911406424e-05,
+      "loss": 0.5066,
+      "step": 536
+    },
+    {
+      "epoch": 0.8026905829596412,
+      "grad_norm": 0.3558999969643179,
+      "learning_rate": 4.0697674418604655e-05,
+      "loss": 0.4961,
+      "step": 537
+    },
+    {
+      "epoch": 0.804185351270553,
+      "grad_norm": 0.4033449036167874,
+      "learning_rate": 4.066998892580288e-05,
+      "loss": 0.5346,
+      "step": 538
+    },
+    {
+      "epoch": 0.8056801195814649,
+      "grad_norm": 0.394391205103531,
+      "learning_rate": 4.064230343300111e-05,
+      "loss": 0.4915,
+      "step": 539
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 0.38421796282754245,
+      "learning_rate": 4.061461794019934e-05,
+      "loss": 0.5245,
+      "step": 540
+    },
+    {
+      "epoch": 0.8086696562032885,
+      "grad_norm": 0.4028069484854027,
+      "learning_rate": 4.058693244739756e-05,
+      "loss": 0.5094,
+      "step": 541
+    },
+    {
+      "epoch": 0.8101644245142003,
+      "grad_norm": 0.4137496132260613,
+      "learning_rate": 4.055924695459579e-05,
+      "loss": 0.5101,
+      "step": 542
+    },
+    {
+      "epoch": 0.8116591928251121,
+      "grad_norm": 0.38912677579720745,
+      "learning_rate": 4.053156146179402e-05,
+      "loss": 0.5087,
+      "step": 543
+    },
+    {
+      "epoch": 0.8131539611360239,
+      "grad_norm": 0.4244176906570153,
+      "learning_rate": 4.050387596899225e-05,
+      "loss": 0.5304,
+      "step": 544
+    },
+    {
+      "epoch": 0.8146487294469357,
+      "grad_norm": 0.43014248825003276,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.5024,
+      "step": 545
+    },
+    {
+      "epoch": 0.8161434977578476,
+      "grad_norm": 0.5438883318571744,
+      "learning_rate": 4.0448504983388706e-05,
+      "loss": 0.4891,
+      "step": 546
+    },
+    {
+      "epoch": 0.8176382660687593,
+      "grad_norm": 0.40605876186404116,
+      "learning_rate": 4.042081949058694e-05,
+      "loss": 0.493,
+      "step": 547
+    },
+    {
+      "epoch": 0.8191330343796711,
+      "grad_norm": 0.5124328351064973,
+      "learning_rate": 4.039313399778516e-05,
+      "loss": 0.5087,
+      "step": 548
+    },
+    {
+      "epoch": 0.820627802690583,
+      "grad_norm": 0.48118619711943483,
+      "learning_rate": 4.036544850498339e-05,
+      "loss": 0.5124,
+      "step": 549
+    },
+    {
+      "epoch": 0.8221225710014948,
+      "grad_norm": 0.48737526079008353,
+      "learning_rate": 4.033776301218162e-05,
+      "loss": 0.501,
+      "step": 550
+    },
+    {
+      "epoch": 0.8236173393124065,
+      "grad_norm": 0.42086316964418174,
+      "learning_rate": 4.0310077519379843e-05,
+      "loss": 0.5121,
+      "step": 551
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.5263195781953447,
+      "learning_rate": 4.0282392026578075e-05,
+      "loss": 0.5169,
+      "step": 552
+    },
+    {
+      "epoch": 0.8266068759342302,
+      "grad_norm": 0.41084371596005664,
+      "learning_rate": 4.0254706533776307e-05,
+      "loss": 0.5018,
+      "step": 553
+    },
+    {
+      "epoch": 0.828101644245142,
+      "grad_norm": 0.5407229346604753,
+      "learning_rate": 4.022702104097453e-05,
+      "loss": 0.5231,
+      "step": 554
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 0.4078037137243094,
+      "learning_rate": 4.019933554817276e-05,
+      "loss": 0.4957,
+      "step": 555
+    },
+    {
+      "epoch": 0.8310911808669657,
+      "grad_norm": 0.3912588964967917,
+      "learning_rate": 4.017165005537099e-05,
+      "loss": 0.493,
+      "step": 556
+    },
+    {
+      "epoch": 0.8325859491778774,
+      "grad_norm": 0.3747855472189801,
+      "learning_rate": 4.014396456256921e-05,
+      "loss": 0.5031,
+      "step": 557
+    },
+    {
+      "epoch": 0.8340807174887892,
+      "grad_norm": 0.38969686951133,
+      "learning_rate": 4.0116279069767444e-05,
+      "loss": 0.4861,
+      "step": 558
+    },
+    {
+      "epoch": 0.8355754857997011,
+      "grad_norm": 0.35049368027927535,
+      "learning_rate": 4.008859357696567e-05,
+      "loss": 0.4915,
+      "step": 559
+    },
+    {
+      "epoch": 0.8370702541106129,
+      "grad_norm": 0.3852207794414822,
+      "learning_rate": 4.00609080841639e-05,
+      "loss": 0.5224,
+      "step": 560
+    },
+    {
+      "epoch": 0.8385650224215246,
+      "grad_norm": 0.40737677141611556,
+      "learning_rate": 4.003322259136213e-05,
+      "loss": 0.5095,
+      "step": 561
+    },
+    {
+      "epoch": 0.8400597907324364,
+      "grad_norm": 0.3889726004496969,
+      "learning_rate": 4.000553709856036e-05,
+      "loss": 0.495,
+      "step": 562
+    },
+    {
+      "epoch": 0.8415545590433483,
+      "grad_norm": 0.4541976715831387,
+      "learning_rate": 3.997785160575858e-05,
+      "loss": 0.5017,
+      "step": 563
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.3285862002262236,
+      "learning_rate": 3.995016611295681e-05,
+      "loss": 0.4887,
+      "step": 564
+    },
+    {
+      "epoch": 0.8445440956651719,
+      "grad_norm": 0.45932687734741345,
+      "learning_rate": 3.992248062015504e-05,
+      "loss": 0.485,
+      "step": 565
+    },
+    {
+      "epoch": 0.8460388639760837,
+      "grad_norm": 0.4036930511901695,
+      "learning_rate": 3.989479512735327e-05,
+      "loss": 0.5026,
+      "step": 566
+    },
+    {
+      "epoch": 0.8475336322869955,
+      "grad_norm": 0.4012358035600719,
+      "learning_rate": 3.9867109634551495e-05,
+      "loss": 0.4978,
+      "step": 567
+    },
+    {
+      "epoch": 0.8490284005979073,
+      "grad_norm": 0.47126058883924277,
+      "learning_rate": 3.9839424141749726e-05,
+      "loss": 0.5187,
+      "step": 568
+    },
+    {
+      "epoch": 0.8505231689088192,
+      "grad_norm": 0.44689232473392176,
+      "learning_rate": 3.981173864894796e-05,
+      "loss": 0.4927,
+      "step": 569
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 0.416922420570468,
+      "learning_rate": 3.978405315614618e-05,
+      "loss": 0.5071,
+      "step": 570
+    },
+    {
+      "epoch": 0.8535127055306427,
+      "grad_norm": 0.4004422194714132,
+      "learning_rate": 3.975636766334441e-05,
+      "loss": 0.4932,
+      "step": 571
+    },
+    {
+      "epoch": 0.8550074738415545,
+      "grad_norm": 0.49605994896377936,
+      "learning_rate": 3.972868217054264e-05,
+      "loss": 0.4874,
+      "step": 572
+    },
+    {
+      "epoch": 0.8565022421524664,
+      "grad_norm": 0.3745032463212653,
+      "learning_rate": 3.9700996677740864e-05,
+      "loss": 0.4917,
+      "step": 573
+    },
+    {
+      "epoch": 0.8579970104633782,
+      "grad_norm": 0.39338984483800893,
+      "learning_rate": 3.9673311184939095e-05,
+      "loss": 0.4786,
+      "step": 574
+    },
+    {
+      "epoch": 0.85949177877429,
+      "grad_norm": 0.48932684330865234,
+      "learning_rate": 3.964562569213732e-05,
+      "loss": 0.5175,
+      "step": 575
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.349937751343075,
+      "learning_rate": 3.961794019933555e-05,
+      "loss": 0.5309,
+      "step": 576
+    },
+    {
+      "epoch": 0.8624813153961136,
+      "grad_norm": 0.436408378453945,
+      "learning_rate": 3.959025470653378e-05,
+      "loss": 0.4944,
+      "step": 577
+    },
+    {
+      "epoch": 0.8639760837070254,
+      "grad_norm": 0.4384384738455648,
+      "learning_rate": 3.9562569213732e-05,
+      "loss": 0.532,
+      "step": 578
+    },
+    {
+      "epoch": 0.8654708520179372,
+      "grad_norm": 0.4338390618373011,
+      "learning_rate": 3.953488372093023e-05,
+      "loss": 0.5038,
+      "step": 579
+    },
+    {
+      "epoch": 0.866965620328849,
+      "grad_norm": 0.4042202712923035,
+      "learning_rate": 3.9507198228128464e-05,
+      "loss": 0.509,
+      "step": 580
+    },
+    {
+      "epoch": 0.8684603886397608,
+      "grad_norm": 0.3580680432665876,
+      "learning_rate": 3.947951273532669e-05,
+      "loss": 0.5402,
+      "step": 581
+    },
+    {
+      "epoch": 0.8699551569506726,
+      "grad_norm": 0.4203635547451078,
+      "learning_rate": 3.945182724252492e-05,
+      "loss": 0.5124,
+      "step": 582
+    },
+    {
+      "epoch": 0.8714499252615845,
+      "grad_norm": 0.38721257104748563,
+      "learning_rate": 3.9424141749723146e-05,
+      "loss": 0.4908,
+      "step": 583
+    },
+    {
+      "epoch": 0.8729446935724963,
+      "grad_norm": 0.4512800560194969,
+      "learning_rate": 3.939645625692138e-05,
+      "loss": 0.4988,
+      "step": 584
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 0.45439800087420723,
+      "learning_rate": 3.936877076411961e-05,
+      "loss": 0.5067,
+      "step": 585
+    },
+    {
+      "epoch": 0.8759342301943199,
+      "grad_norm": 0.3910692385743146,
+      "learning_rate": 3.934108527131783e-05,
+      "loss": 0.4888,
+      "step": 586
+    },
+    {
+      "epoch": 0.8774289985052317,
+      "grad_norm": 0.5420791885519113,
+      "learning_rate": 3.931339977851606e-05,
+      "loss": 0.4977,
+      "step": 587
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.3565652262497022,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.4947,
+      "step": 588
+    },
+    {
+      "epoch": 0.8804185351270553,
+      "grad_norm": 0.4326955572263076,
+      "learning_rate": 3.9258028792912515e-05,
+      "loss": 0.4974,
+      "step": 589
+    },
+    {
+      "epoch": 0.8819133034379671,
+      "grad_norm": 0.4197068397128714,
+      "learning_rate": 3.9230343300110746e-05,
+      "loss": 0.517,
+      "step": 590
+    },
+    {
+      "epoch": 0.8834080717488789,
+      "grad_norm": 0.3360409288444025,
+      "learning_rate": 3.920265780730897e-05,
+      "loss": 0.4835,
+      "step": 591
+    },
+    {
+      "epoch": 0.8849028400597907,
+      "grad_norm": 0.4367181121439196,
+      "learning_rate": 3.91749723145072e-05,
+      "loss": 0.5155,
+      "step": 592
+    },
+    {
+      "epoch": 0.8863976083707026,
+      "grad_norm": 0.31252935913920926,
+      "learning_rate": 3.914728682170543e-05,
+      "loss": 0.5084,
+      "step": 593
+    },
+    {
+      "epoch": 0.8878923766816144,
+      "grad_norm": 0.4559464625954473,
+      "learning_rate": 3.911960132890365e-05,
+      "loss": 0.5034,
+      "step": 594
+    },
+    {
+      "epoch": 0.8893871449925261,
+      "grad_norm": 0.3354816918736476,
+      "learning_rate": 3.9091915836101884e-05,
+      "loss": 0.5055,
+      "step": 595
+    },
+    {
+      "epoch": 0.890881913303438,
+      "grad_norm": 0.3866942391297095,
+      "learning_rate": 3.9064230343300116e-05,
+      "loss": 0.4825,
+      "step": 596
+    },
+    {
+      "epoch": 0.8923766816143498,
+      "grad_norm": 0.41256067418047304,
+      "learning_rate": 3.903654485049834e-05,
+      "loss": 0.5263,
+      "step": 597
+    },
+    {
+      "epoch": 0.8938714499252616,
+      "grad_norm": 0.363915536640541,
+      "learning_rate": 3.900885935769657e-05,
+      "loss": 0.5019,
+      "step": 598
+    },
+    {
+      "epoch": 0.8953662182361734,
+      "grad_norm": 0.49171160345494414,
+      "learning_rate": 3.89811738648948e-05,
+      "loss": 0.5256,
+      "step": 599
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.33738284457193185,
+      "learning_rate": 3.895348837209303e-05,
+      "loss": 0.4972,
+      "step": 600
+    },
+    {
+      "epoch": 0.898355754857997,
+      "grad_norm": 0.3932096128066618,
+      "learning_rate": 3.892580287929125e-05,
+      "loss": 0.5134,
+      "step": 601
+    },
+    {
+      "epoch": 0.8998505231689088,
+      "grad_norm": 0.4662536088293291,
+      "learning_rate": 3.889811738648948e-05,
+      "loss": 0.5144,
+      "step": 602
+    },
+    {
+      "epoch": 0.9013452914798207,
+      "grad_norm": 0.4039530594619555,
+      "learning_rate": 3.887043189368771e-05,
+      "loss": 0.491,
+      "step": 603
+    },
+    {
+      "epoch": 0.9028400597907325,
+      "grad_norm": 0.40930883347351643,
+      "learning_rate": 3.8842746400885934e-05,
+      "loss": 0.4998,
+      "step": 604
+    },
+    {
+      "epoch": 0.9043348281016442,
+      "grad_norm": 0.5273103642477311,
+      "learning_rate": 3.8815060908084166e-05,
+      "loss": 0.5285,
+      "step": 605
+    },
+    {
+      "epoch": 0.905829596412556,
+      "grad_norm": 0.519990121160213,
+      "learning_rate": 3.87873754152824e-05,
+      "loss": 0.5177,
+      "step": 606
+    },
+    {
+      "epoch": 0.9073243647234679,
+      "grad_norm": 0.4029775362245476,
+      "learning_rate": 3.875968992248062e-05,
+      "loss": 0.4949,
+      "step": 607
+    },
+    {
+      "epoch": 0.9088191330343797,
+      "grad_norm": 0.40279622669917897,
+      "learning_rate": 3.873200442967885e-05,
+      "loss": 0.5151,
+      "step": 608
+    },
+    {
+      "epoch": 0.9103139013452914,
+      "grad_norm": 0.4012459999222105,
+      "learning_rate": 3.870431893687708e-05,
+      "loss": 0.506,
+      "step": 609
+    },
+    {
+      "epoch": 0.9118086696562033,
+      "grad_norm": 0.35503045823792073,
+      "learning_rate": 3.8676633444075304e-05,
+      "loss": 0.4991,
+      "step": 610
+    },
+    {
+      "epoch": 0.9133034379671151,
+      "grad_norm": 0.45008752650413925,
+      "learning_rate": 3.8648947951273535e-05,
+      "loss": 0.5016,
+      "step": 611
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 0.4024587006061886,
+      "learning_rate": 3.862126245847176e-05,
+      "loss": 0.518,
+      "step": 612
+    },
+    {
+      "epoch": 0.9162929745889388,
+      "grad_norm": 0.37095322245264445,
+      "learning_rate": 3.859357696566999e-05,
+      "loss": 0.4972,
+      "step": 613
+    },
+    {
+      "epoch": 0.9177877428998505,
+      "grad_norm": 0.4510233091905665,
+      "learning_rate": 3.856589147286822e-05,
+      "loss": 0.5183,
+      "step": 614
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 0.36005415661230844,
+      "learning_rate": 3.853820598006645e-05,
+      "loss": 0.5196,
+      "step": 615
+    },
+    {
+      "epoch": 0.9207772795216741,
+      "grad_norm": 0.45543353843542506,
+      "learning_rate": 3.851052048726467e-05,
+      "loss": 0.5133,
+      "step": 616
+    },
+    {
+      "epoch": 0.922272047832586,
+      "grad_norm": 0.3095961023524229,
+      "learning_rate": 3.8482834994462904e-05,
+      "loss": 0.4956,
+      "step": 617
+    },
+    {
+      "epoch": 0.9237668161434978,
+      "grad_norm": 0.45734226837468783,
+      "learning_rate": 3.845514950166113e-05,
+      "loss": 0.4899,
+      "step": 618
+    },
+    {
+      "epoch": 0.9252615844544095,
+      "grad_norm": 0.42163507090477986,
+      "learning_rate": 3.842746400885936e-05,
+      "loss": 0.4988,
+      "step": 619
+    },
+    {
+      "epoch": 0.9267563527653214,
+      "grad_norm": 0.4255605796066281,
+      "learning_rate": 3.8399778516057585e-05,
+      "loss": 0.4772,
+      "step": 620
+    },
+    {
+      "epoch": 0.9282511210762332,
+      "grad_norm": 0.3433040479193587,
+      "learning_rate": 3.837209302325582e-05,
+      "loss": 0.473,
+      "step": 621
+    },
+    {
+      "epoch": 0.929745889387145,
+      "grad_norm": 0.36010475944699977,
+      "learning_rate": 3.834440753045405e-05,
+      "loss": 0.5163,
+      "step": 622
+    },
+    {
+      "epoch": 0.9312406576980568,
+      "grad_norm": 0.4659043652574626,
+      "learning_rate": 3.831672203765227e-05,
+      "loss": 0.4964,
+      "step": 623
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.35209183719832093,
+      "learning_rate": 3.82890365448505e-05,
+      "loss": 0.4981,
+      "step": 624
+    },
+    {
+      "epoch": 0.9342301943198804,
+      "grad_norm": 0.43484047665772324,
+      "learning_rate": 3.826135105204873e-05,
+      "loss": 0.4982,
+      "step": 625
+    },
+    {
+      "epoch": 0.9357249626307922,
+      "grad_norm": 0.40577314915615453,
+      "learning_rate": 3.8233665559246955e-05,
+      "loss": 0.5107,
+      "step": 626
+    },
+    {
+      "epoch": 0.9372197309417041,
+      "grad_norm": 0.39193317947286915,
+      "learning_rate": 3.8205980066445186e-05,
+      "loss": 0.5117,
+      "step": 627
+    },
+    {
+      "epoch": 0.9387144992526159,
+      "grad_norm": 0.44574532474545003,
+      "learning_rate": 3.817829457364341e-05,
+      "loss": 0.503,
+      "step": 628
+    },
+    {
+      "epoch": 0.9402092675635276,
+      "grad_norm": 0.4125776246746015,
+      "learning_rate": 3.815060908084164e-05,
+      "loss": 0.5,
+      "step": 629
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 0.3889490710028185,
+      "learning_rate": 3.8122923588039874e-05,
+      "loss": 0.5144,
+      "step": 630
+    },
+    {
+      "epoch": 0.9431988041853513,
+      "grad_norm": 0.4967450719563722,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.4979,
+      "step": 631
+    },
+    {
+      "epoch": 0.9446935724962631,
+      "grad_norm": 0.397949157812168,
+      "learning_rate": 3.8067552602436324e-05,
+      "loss": 0.4982,
+      "step": 632
+    },
+    {
+      "epoch": 0.9461883408071748,
+      "grad_norm": 0.49691388504513045,
+      "learning_rate": 3.8039867109634555e-05,
+      "loss": 0.509,
+      "step": 633
+    },
+    {
+      "epoch": 0.9476831091180867,
+      "grad_norm": 0.4029073507227645,
+      "learning_rate": 3.801218161683278e-05,
+      "loss": 0.5053,
+      "step": 634
+    },
+    {
+      "epoch": 0.9491778774289985,
+      "grad_norm": 0.4421016286836699,
+      "learning_rate": 3.798449612403101e-05,
+      "loss": 0.505,
+      "step": 635
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 0.4140802929082004,
+      "learning_rate": 3.7956810631229237e-05,
+      "loss": 0.4866,
+      "step": 636
+    },
+    {
+      "epoch": 0.9521674140508222,
+      "grad_norm": 0.41289650511242837,
+      "learning_rate": 3.792912513842747e-05,
+      "loss": 0.5049,
+      "step": 637
+    },
+    {
+      "epoch": 0.953662182361734,
+      "grad_norm": 0.4000366534289744,
+      "learning_rate": 3.790143964562569e-05,
+      "loss": 0.497,
+      "step": 638
+    },
+    {
+      "epoch": 0.9551569506726457,
+      "grad_norm": 0.445620683106817,
+      "learning_rate": 3.787375415282392e-05,
+      "loss": 0.5006,
+      "step": 639
+    },
+    {
+      "epoch": 0.9566517189835575,
+      "grad_norm": 0.4591490437123157,
+      "learning_rate": 3.784606866002215e-05,
+      "loss": 0.4967,
+      "step": 640
+    },
+    {
+      "epoch": 0.9581464872944694,
+      "grad_norm": 0.41574459674272113,
+      "learning_rate": 3.781838316722038e-05,
+      "loss": 0.5227,
+      "step": 641
+    },
+    {
+      "epoch": 0.9596412556053812,
+      "grad_norm": 0.3708217482090666,
+      "learning_rate": 3.7790697674418606e-05,
+      "loss": 0.4915,
+      "step": 642
+    },
+    {
+      "epoch": 0.9611360239162929,
+      "grad_norm": 0.4163993003589286,
+      "learning_rate": 3.776301218161684e-05,
+      "loss": 0.4915,
+      "step": 643
+    },
+    {
+      "epoch": 0.9626307922272048,
+      "grad_norm": 0.3978946030947043,
+      "learning_rate": 3.773532668881506e-05,
+      "loss": 0.5017,
+      "step": 644
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 0.3952616272092666,
+      "learning_rate": 3.7707641196013294e-05,
+      "loss": 0.4963,
+      "step": 645
+    },
+    {
+      "epoch": 0.9656203288490284,
+      "grad_norm": 0.3748296486698132,
+      "learning_rate": 3.767995570321152e-05,
+      "loss": 0.4824,
+      "step": 646
+    },
+    {
+      "epoch": 0.9671150971599403,
+      "grad_norm": 0.31986781390721636,
+      "learning_rate": 3.765227021040974e-05,
+      "loss": 0.5014,
+      "step": 647
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.41299572998373624,
+      "learning_rate": 3.7624584717607975e-05,
+      "loss": 0.5272,
+      "step": 648
+    },
+    {
+      "epoch": 0.9701046337817638,
+      "grad_norm": 0.3665924599601556,
+      "learning_rate": 3.7596899224806207e-05,
+      "loss": 0.5161,
+      "step": 649
+    },
+    {
+      "epoch": 0.9715994020926756,
+      "grad_norm": 0.36694653532742666,
+      "learning_rate": 3.756921373200443e-05,
+      "loss": 0.487,
+      "step": 650
+    },
+    {
+      "epoch": 0.9730941704035875,
+      "grad_norm": 0.47061331842752996,
+      "learning_rate": 3.754152823920266e-05,
+      "loss": 0.5313,
+      "step": 651
+    },
+    {
+      "epoch": 0.9745889387144993,
+      "grad_norm": 0.42241526158413856,
+      "learning_rate": 3.751384274640089e-05,
+      "loss": 0.4942,
+      "step": 652
+    },
+    {
+      "epoch": 0.976083707025411,
+      "grad_norm": 0.44816063140729956,
+      "learning_rate": 3.748615725359911e-05,
+      "loss": 0.4951,
+      "step": 653
+    },
+    {
+      "epoch": 0.9775784753363229,
+      "grad_norm": 0.39293220291685277,
+      "learning_rate": 3.7458471760797344e-05,
+      "loss": 0.4915,
+      "step": 654
+    },
+    {
+      "epoch": 0.9790732436472347,
+      "grad_norm": 0.429211903187621,
+      "learning_rate": 3.743078626799557e-05,
+      "loss": 0.5025,
+      "step": 655
+    },
+    {
+      "epoch": 0.9805680119581465,
+      "grad_norm": 0.4341371734284874,
+      "learning_rate": 3.74031007751938e-05,
+      "loss": 0.5093,
+      "step": 656
+    },
+    {
+      "epoch": 0.9820627802690582,
+      "grad_norm": 0.3559853694471361,
+      "learning_rate": 3.737541528239203e-05,
+      "loss": 0.4979,
+      "step": 657
+    },
+    {
+      "epoch": 0.9835575485799701,
+      "grad_norm": 0.4105542585596947,
+      "learning_rate": 3.734772978959026e-05,
+      "loss": 0.5006,
+      "step": 658
+    },
+    {
+      "epoch": 0.9850523168908819,
+      "grad_norm": 0.4025001115572041,
+      "learning_rate": 3.732004429678849e-05,
+      "loss": 0.4949,
+      "step": 659
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 0.38527039730127766,
+      "learning_rate": 3.729235880398671e-05,
+      "loss": 0.5138,
+      "step": 660
+    },
+    {
+      "epoch": 0.9880418535127056,
+      "grad_norm": 0.3543774089432186,
+      "learning_rate": 3.726467331118494e-05,
+      "loss": 0.4875,
+      "step": 661
+    },
+    {
+      "epoch": 0.9895366218236173,
+      "grad_norm": 0.4034228429254513,
+      "learning_rate": 3.723698781838317e-05,
+      "loss": 0.4946,
+      "step": 662
+    },
+    {
+      "epoch": 0.9910313901345291,
+      "grad_norm": 0.39852112473172785,
+      "learning_rate": 3.7209302325581394e-05,
+      "loss": 0.5224,
+      "step": 663
+    },
+    {
+      "epoch": 0.992526158445441,
+      "grad_norm": 0.4186588834164515,
+      "learning_rate": 3.7181616832779626e-05,
+      "loss": 0.4978,
+      "step": 664
+    },
+    {
+      "epoch": 0.9940209267563528,
+      "grad_norm": 0.3522705163340096,
+      "learning_rate": 3.715393133997786e-05,
+      "loss": 0.495,
+      "step": 665
+    },
+    {
+      "epoch": 0.9955156950672646,
+      "grad_norm": 0.3452957107936995,
+      "learning_rate": 3.712624584717608e-05,
+      "loss": 0.5019,
+      "step": 666
+    },
+    {
+      "epoch": 0.9970104633781763,
+      "grad_norm": 0.4127944722022095,
+      "learning_rate": 3.7098560354374314e-05,
+      "loss": 0.5147,
+      "step": 667
+    },
+    {
+      "epoch": 0.9985052316890882,
+      "grad_norm": 0.4595161067492149,
+      "learning_rate": 3.707087486157253e-05,
+      "loss": 0.5152,
+      "step": 668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31413463926317925,
+      "learning_rate": 3.7043189368770764e-05,
+      "loss": 0.5076,
+      "step": 669
+    },
+    {
+      "epoch": 1.0014947683109119,
+      "grad_norm": 0.5041376236545296,
+      "learning_rate": 3.7015503875968995e-05,
+      "loss": 0.4202,
+      "step": 670
+    },
+    {
+      "epoch": 1.0029895366218236,
+      "grad_norm": 0.35737291921931275,
+      "learning_rate": 3.698781838316722e-05,
+      "loss": 0.424,
+      "step": 671
+    },
+    {
+      "epoch": 1.0044843049327354,
+      "grad_norm": 0.5464835461846216,
+      "learning_rate": 3.696013289036545e-05,
+      "loss": 0.452,
+      "step": 672
+    },
+    {
+      "epoch": 1.0059790732436473,
+      "grad_norm": 0.4608584704702098,
+      "learning_rate": 3.6932447397563676e-05,
+      "loss": 0.4377,
+      "step": 673
+    },
+    {
+      "epoch": 1.007473841554559,
+      "grad_norm": 0.48735696315597704,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 0.4564,
+      "step": 674
+    },
+    {
+      "epoch": 1.0089686098654709,
+      "grad_norm": 0.39087368912281245,
+      "learning_rate": 3.687707641196013e-05,
+      "loss": 0.43,
+      "step": 675
+    },
+    {
+      "epoch": 1.0104633781763828,
+      "grad_norm": 0.46344375888510797,
+      "learning_rate": 3.684939091915836e-05,
+      "loss": 0.4239,
+      "step": 676
+    },
+    {
+      "epoch": 1.0119581464872944,
+      "grad_norm": 0.4102725518470781,
+      "learning_rate": 3.682170542635659e-05,
+      "loss": 0.4408,
+      "step": 677
+    },
+    {
+      "epoch": 1.0134529147982063,
+      "grad_norm": 0.3771185707085356,
+      "learning_rate": 3.679401993355482e-05,
+      "loss": 0.4385,
+      "step": 678
+    },
+    {
+      "epoch": 1.014947683109118,
+      "grad_norm": 0.427844250737845,
+      "learning_rate": 3.6766334440753046e-05,
+      "loss": 0.4481,
+      "step": 679
+    },
+    {
+      "epoch": 1.0164424514200299,
+      "grad_norm": 0.373791300282762,
+      "learning_rate": 3.673864894795128e-05,
+      "loss": 0.4333,
+      "step": 680
+    },
+    {
+      "epoch": 1.0179372197309418,
+      "grad_norm": 0.36520233669292723,
+      "learning_rate": 3.67109634551495e-05,
+      "loss": 0.446,
+      "step": 681
+    },
+    {
+      "epoch": 1.0194319880418534,
+      "grad_norm": 0.3911455238301078,
+      "learning_rate": 3.6683277962347734e-05,
+      "loss": 0.4325,
+      "step": 682
+    },
+    {
+      "epoch": 1.0209267563527653,
+      "grad_norm": 0.3414252853028562,
+      "learning_rate": 3.665559246954596e-05,
+      "loss": 0.4206,
+      "step": 683
+    },
+    {
+      "epoch": 1.0224215246636772,
+      "grad_norm": 0.3895262974942302,
+      "learning_rate": 3.662790697674418e-05,
+      "loss": 0.4564,
+      "step": 684
+    },
+    {
+      "epoch": 1.0239162929745889,
+      "grad_norm": 0.3132766684986166,
+      "learning_rate": 3.6600221483942415e-05,
+      "loss": 0.4282,
+      "step": 685
+    },
+    {
+      "epoch": 1.0254110612855007,
+      "grad_norm": 0.3512155534049176,
+      "learning_rate": 3.6572535991140646e-05,
+      "loss": 0.4386,
+      "step": 686
+    },
+    {
+      "epoch": 1.0269058295964126,
+      "grad_norm": 0.3739868421335463,
+      "learning_rate": 3.654485049833887e-05,
+      "loss": 0.4424,
+      "step": 687
+    },
+    {
+      "epoch": 1.0284005979073243,
+      "grad_norm": 0.35865755069393196,
+      "learning_rate": 3.65171650055371e-05,
+      "loss": 0.4468,
+      "step": 688
+    },
+    {
+      "epoch": 1.0298953662182362,
+      "grad_norm": 0.45549152633822565,
+      "learning_rate": 3.648947951273533e-05,
+      "loss": 0.4637,
+      "step": 689
+    },
+    {
+      "epoch": 1.031390134529148,
+      "grad_norm": 0.43934044120486765,
+      "learning_rate": 3.646179401993355e-05,
+      "loss": 0.4212,
+      "step": 690
+    },
+    {
+      "epoch": 1.0328849028400597,
+      "grad_norm": 0.36790843651788896,
+      "learning_rate": 3.6434108527131784e-05,
+      "loss": 0.4285,
+      "step": 691
+    },
+    {
+      "epoch": 1.0343796711509716,
+      "grad_norm": 0.33035264956551924,
+      "learning_rate": 3.640642303433001e-05,
+      "loss": 0.4383,
+      "step": 692
+    },
+    {
+      "epoch": 1.0358744394618835,
+      "grad_norm": 0.39145475906090027,
+      "learning_rate": 3.637873754152824e-05,
+      "loss": 0.4258,
+      "step": 693
+    },
+    {
+      "epoch": 1.0373692077727952,
+      "grad_norm": 0.36112448733507163,
+      "learning_rate": 3.635105204872647e-05,
+      "loss": 0.4297,
+      "step": 694
+    },
+    {
+      "epoch": 1.038863976083707,
+      "grad_norm": 0.344108163855027,
+      "learning_rate": 3.63233665559247e-05,
+      "loss": 0.4204,
+      "step": 695
+    },
+    {
+      "epoch": 1.0403587443946187,
+      "grad_norm": 0.32001931550229606,
+      "learning_rate": 3.629568106312293e-05,
+      "loss": 0.4232,
+      "step": 696
+    },
+    {
+      "epoch": 1.0418535127055306,
+      "grad_norm": 0.361538973390565,
+      "learning_rate": 3.626799557032115e-05,
+      "loss": 0.4428,
+      "step": 697
+    },
+    {
+      "epoch": 1.0433482810164425,
+      "grad_norm": 0.31428626470694754,
+      "learning_rate": 3.624031007751938e-05,
+      "loss": 0.4305,
+      "step": 698
+    },
+    {
+      "epoch": 1.0448430493273542,
+      "grad_norm": 0.37905484151399893,
+      "learning_rate": 3.621262458471761e-05,
+      "loss": 0.4271,
+      "step": 699
+    },
+    {
+      "epoch": 1.046337817638266,
+      "grad_norm": 0.436092246625112,
+      "learning_rate": 3.6184939091915834e-05,
+      "loss": 0.4421,
+      "step": 700
+    },
+    {
+      "epoch": 1.047832585949178,
+      "grad_norm": 0.3728908527561458,
+      "learning_rate": 3.6157253599114066e-05,
+      "loss": 0.4254,
+      "step": 701
+    },
+    {
+      "epoch": 1.0493273542600896,
+      "grad_norm": 0.347863049405835,
+      "learning_rate": 3.61295681063123e-05,
+      "loss": 0.4287,
+      "step": 702
+    },
+    {
+      "epoch": 1.0508221225710015,
+      "grad_norm": 0.3718997705509668,
+      "learning_rate": 3.610188261351052e-05,
+      "loss": 0.4133,
+      "step": 703
+    },
+    {
+      "epoch": 1.0523168908819134,
+      "grad_norm": 0.4526039702156657,
+      "learning_rate": 3.6074197120708754e-05,
+      "loss": 0.4399,
+      "step": 704
+    },
+    {
+      "epoch": 1.053811659192825,
+      "grad_norm": 0.3618577127423069,
+      "learning_rate": 3.604651162790698e-05,
+      "loss": 0.4526,
+      "step": 705
+    },
+    {
+      "epoch": 1.055306427503737,
+      "grad_norm": 0.41237490208157834,
+      "learning_rate": 3.6018826135105203e-05,
+      "loss": 0.4226,
+      "step": 706
+    },
+    {
+      "epoch": 1.0568011958146488,
+      "grad_norm": 0.40598366190476054,
+      "learning_rate": 3.5991140642303435e-05,
+      "loss": 0.4169,
+      "step": 707
+    },
+    {
+      "epoch": 1.0582959641255605,
+      "grad_norm": 0.3919399186494091,
+      "learning_rate": 3.596345514950166e-05,
+      "loss": 0.4355,
+      "step": 708
+    },
+    {
+      "epoch": 1.0597907324364724,
+      "grad_norm": 0.3991193470181896,
+      "learning_rate": 3.593576965669989e-05,
+      "loss": 0.4343,
+      "step": 709
+    },
+    {
+      "epoch": 1.0612855007473843,
+      "grad_norm": 0.41583601924374863,
+      "learning_rate": 3.590808416389812e-05,
+      "loss": 0.4251,
+      "step": 710
+    },
+    {
+      "epoch": 1.062780269058296,
+      "grad_norm": 0.3536062252937738,
+      "learning_rate": 3.588039867109635e-05,
+      "loss": 0.4223,
+      "step": 711
+    },
+    {
+      "epoch": 1.0642750373692078,
+      "grad_norm": 0.4157471722931038,
+      "learning_rate": 3.585271317829458e-05,
+      "loss": 0.4227,
+      "step": 712
+    },
+    {
+      "epoch": 1.0657698056801195,
+      "grad_norm": 0.39679435149593034,
+      "learning_rate": 3.5825027685492804e-05,
+      "loss": 0.4331,
+      "step": 713
+    },
+    {
+      "epoch": 1.0672645739910314,
+      "grad_norm": 0.34129691907993776,
+      "learning_rate": 3.579734219269103e-05,
+      "loss": 0.4036,
+      "step": 714
+    },
+    {
+      "epoch": 1.0687593423019432,
+      "grad_norm": 0.3485471972609879,
+      "learning_rate": 3.576965669988926e-05,
+      "loss": 0.4427,
+      "step": 715
+    },
+    {
+      "epoch": 1.070254110612855,
+      "grad_norm": 0.4113264500473381,
+      "learning_rate": 3.5741971207087485e-05,
+      "loss": 0.4548,
+      "step": 716
+    },
+    {
+      "epoch": 1.0717488789237668,
+      "grad_norm": 0.32363942136004503,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.4461,
+      "step": 717
+    },
+    {
+      "epoch": 1.0732436472346787,
+      "grad_norm": 0.4043688763899891,
+      "learning_rate": 3.568660022148395e-05,
+      "loss": 0.4555,
+      "step": 718
+    },
+    {
+      "epoch": 1.0747384155455904,
+      "grad_norm": 0.36451787004242725,
+      "learning_rate": 3.565891472868217e-05,
+      "loss": 0.451,
+      "step": 719
+    },
+    {
+      "epoch": 1.0762331838565022,
+      "grad_norm": 0.3613764343959086,
+      "learning_rate": 3.56312292358804e-05,
+      "loss": 0.4192,
+      "step": 720
+    },
+    {
+      "epoch": 1.0777279521674141,
+      "grad_norm": 0.3884702461350999,
+      "learning_rate": 3.560354374307863e-05,
+      "loss": 0.4173,
+      "step": 721
+    },
+    {
+      "epoch": 1.0792227204783258,
+      "grad_norm": 0.36010259554238644,
+      "learning_rate": 3.5575858250276855e-05,
+      "loss": 0.4348,
+      "step": 722
+    },
+    {
+      "epoch": 1.0807174887892377,
+      "grad_norm": 0.4674837632132591,
+      "learning_rate": 3.5548172757475086e-05,
+      "loss": 0.4566,
+      "step": 723
+    },
+    {
+      "epoch": 1.0822122571001496,
+      "grad_norm": 0.3525158318247762,
+      "learning_rate": 3.552048726467331e-05,
+      "loss": 0.4468,
+      "step": 724
+    },
+    {
+      "epoch": 1.0837070254110612,
+      "grad_norm": 0.3823371146032736,
+      "learning_rate": 3.549280177187154e-05,
+      "loss": 0.4317,
+      "step": 725
+    },
+    {
+      "epoch": 1.0852017937219731,
+      "grad_norm": 0.36446153137099374,
+      "learning_rate": 3.5465116279069774e-05,
+      "loss": 0.4436,
+      "step": 726
+    },
+    {
+      "epoch": 1.086696562032885,
+      "grad_norm": 0.40277847828274455,
+      "learning_rate": 3.5437430786268e-05,
+      "loss": 0.4085,
+      "step": 727
+    },
+    {
+      "epoch": 1.0881913303437967,
+      "grad_norm": 0.4196135836852605,
+      "learning_rate": 3.5409745293466224e-05,
+      "loss": 0.4795,
+      "step": 728
+    },
+    {
+      "epoch": 1.0896860986547086,
+      "grad_norm": 0.4056441922269939,
+      "learning_rate": 3.5382059800664455e-05,
+      "loss": 0.4263,
+      "step": 729
+    },
+    {
+      "epoch": 1.0911808669656202,
+      "grad_norm": 0.3875662490273799,
+      "learning_rate": 3.535437430786268e-05,
+      "loss": 0.4321,
+      "step": 730
+    },
+    {
+      "epoch": 1.092675635276532,
+      "grad_norm": 0.4421128611189418,
+      "learning_rate": 3.532668881506091e-05,
+      "loss": 0.4389,
+      "step": 731
+    },
+    {
+      "epoch": 1.094170403587444,
+      "grad_norm": 0.35068914683579555,
+      "learning_rate": 3.5299003322259136e-05,
+      "loss": 0.4247,
+      "step": 732
+    },
+    {
+      "epoch": 1.0956651718983557,
+      "grad_norm": 0.5872002027187907,
+      "learning_rate": 3.527131782945737e-05,
+      "loss": 0.4357,
+      "step": 733
+    },
+    {
+      "epoch": 1.0971599402092675,
+      "grad_norm": 0.32467735209962073,
+      "learning_rate": 3.52436323366556e-05,
+      "loss": 0.4475,
+      "step": 734
+    },
+    {
+      "epoch": 1.0986547085201794,
+      "grad_norm": 0.5816013572801377,
+      "learning_rate": 3.521594684385382e-05,
+      "loss": 0.4434,
+      "step": 735
+    },
+    {
+      "epoch": 1.100149476831091,
+      "grad_norm": 0.36647300810756767,
+      "learning_rate": 3.518826135105205e-05,
+      "loss": 0.4468,
+      "step": 736
+    },
+    {
+      "epoch": 1.101644245142003,
+      "grad_norm": 0.5700525639921802,
+      "learning_rate": 3.516057585825028e-05,
+      "loss": 0.4393,
+      "step": 737
+    },
+    {
+      "epoch": 1.1031390134529149,
+      "grad_norm": 0.366218097434565,
+      "learning_rate": 3.5132890365448506e-05,
+      "loss": 0.4433,
+      "step": 738
+    },
+    {
+      "epoch": 1.1046337817638265,
+      "grad_norm": 0.4995801814388927,
+      "learning_rate": 3.510520487264674e-05,
+      "loss": 0.4382,
+      "step": 739
+    },
+    {
+      "epoch": 1.1061285500747384,
+      "grad_norm": 0.3686961249976862,
+      "learning_rate": 3.507751937984496e-05,
+      "loss": 0.4382,
+      "step": 740
+    },
+    {
+      "epoch": 1.1076233183856503,
+      "grad_norm": 0.5231626284520505,
+      "learning_rate": 3.5049833887043194e-05,
+      "loss": 0.4177,
+      "step": 741
+    },
+    {
+      "epoch": 1.109118086696562,
+      "grad_norm": 0.3299404296450683,
+      "learning_rate": 3.5022148394241425e-05,
+      "loss": 0.4239,
+      "step": 742
+    },
+    {
+      "epoch": 1.1106128550074739,
+      "grad_norm": 0.37966153302930483,
+      "learning_rate": 3.499446290143964e-05,
+      "loss": 0.4119,
+      "step": 743
+    },
+    {
+      "epoch": 1.1121076233183858,
+      "grad_norm": 0.3146075483481823,
+      "learning_rate": 3.4966777408637875e-05,
+      "loss": 0.4279,
+      "step": 744
+    },
+    {
+      "epoch": 1.1136023916292974,
+      "grad_norm": 0.3608697985483008,
+      "learning_rate": 3.49390919158361e-05,
+      "loss": 0.4112,
+      "step": 745
+    },
+    {
+      "epoch": 1.1150971599402093,
+      "grad_norm": 0.31376991464037846,
+      "learning_rate": 3.491140642303433e-05,
+      "loss": 0.4228,
+      "step": 746
+    },
+    {
+      "epoch": 1.116591928251121,
+      "grad_norm": 0.3384681556300444,
+      "learning_rate": 3.488372093023256e-05,
+      "loss": 0.4287,
+      "step": 747
+    },
+    {
+      "epoch": 1.1180866965620329,
+      "grad_norm": 0.3996920835134498,
+      "learning_rate": 3.485603543743079e-05,
+      "loss": 0.433,
+      "step": 748
+    },
+    {
+      "epoch": 1.1195814648729447,
+      "grad_norm": 0.4296995569618953,
+      "learning_rate": 3.482834994462902e-05,
+      "loss": 0.4419,
+      "step": 749
+    },
+    {
+      "epoch": 1.1210762331838564,
+      "grad_norm": 0.3244557769791778,
+      "learning_rate": 3.4800664451827244e-05,
+      "loss": 0.4209,
+      "step": 750
+    },
+    {
+      "epoch": 1.1225710014947683,
+      "grad_norm": 0.4146109537143521,
+      "learning_rate": 3.477297895902547e-05,
+      "loss": 0.4417,
+      "step": 751
+    },
+    {
+      "epoch": 1.1240657698056802,
+      "grad_norm": 0.3569587349641959,
+      "learning_rate": 3.47452934662237e-05,
+      "loss": 0.4267,
+      "step": 752
+    },
+    {
+      "epoch": 1.1255605381165918,
+      "grad_norm": 0.38956602993050515,
+      "learning_rate": 3.4717607973421925e-05,
+      "loss": 0.4349,
+      "step": 753
+    },
+    {
+      "epoch": 1.1270553064275037,
+      "grad_norm": 0.3863854748642883,
+      "learning_rate": 3.468992248062016e-05,
+      "loss": 0.4298,
+      "step": 754
+    },
+    {
+      "epoch": 1.1285500747384156,
+      "grad_norm": 0.3931959405842961,
+      "learning_rate": 3.466223698781839e-05,
+      "loss": 0.4348,
+      "step": 755
+    },
+    {
+      "epoch": 1.1300448430493273,
+      "grad_norm": 0.342743363292355,
+      "learning_rate": 3.463455149501661e-05,
+      "loss": 0.452,
+      "step": 756
+    },
+    {
+      "epoch": 1.1315396113602392,
+      "grad_norm": 0.38363191313307504,
+      "learning_rate": 3.460686600221484e-05,
+      "loss": 0.4152,
+      "step": 757
+    },
+    {
+      "epoch": 1.133034379671151,
+      "grad_norm": 0.3804563192219821,
+      "learning_rate": 3.457918050941307e-05,
+      "loss": 0.4176,
+      "step": 758
+    },
+    {
+      "epoch": 1.1345291479820627,
+      "grad_norm": 0.37167470235761996,
+      "learning_rate": 3.4551495016611294e-05,
+      "loss": 0.4275,
+      "step": 759
+    },
+    {
+      "epoch": 1.1360239162929746,
+      "grad_norm": 0.3499816624979574,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 0.4208,
+      "step": 760
+    },
+    {
+      "epoch": 1.1375186846038865,
+      "grad_norm": 0.4104354913051497,
+      "learning_rate": 3.449612403100775e-05,
+      "loss": 0.4525,
+      "step": 761
+    },
+    {
+      "epoch": 1.1390134529147982,
+      "grad_norm": 0.3609401560128115,
+      "learning_rate": 3.446843853820598e-05,
+      "loss": 0.4281,
+      "step": 762
+    },
+    {
+      "epoch": 1.14050822122571,
+      "grad_norm": 0.42554025007354335,
+      "learning_rate": 3.4440753045404214e-05,
+      "loss": 0.4271,
+      "step": 763
+    },
+    {
+      "epoch": 1.1420029895366217,
+      "grad_norm": 0.36517366706890364,
+      "learning_rate": 3.441306755260244e-05,
+      "loss": 0.4206,
+      "step": 764
+    },
+    {
+      "epoch": 1.1434977578475336,
+      "grad_norm": 0.3494858927390298,
+      "learning_rate": 3.4385382059800664e-05,
+      "loss": 0.4339,
+      "step": 765
+    },
+    {
+      "epoch": 1.1449925261584455,
+      "grad_norm": 0.3605751004533258,
+      "learning_rate": 3.4357696566998895e-05,
+      "loss": 0.4427,
+      "step": 766
+    },
+    {
+      "epoch": 1.1464872944693572,
+      "grad_norm": 0.3450735778779831,
+      "learning_rate": 3.433001107419712e-05,
+      "loss": 0.433,
+      "step": 767
+    },
+    {
+      "epoch": 1.147982062780269,
+      "grad_norm": 0.2825658425945988,
+      "learning_rate": 3.430232558139535e-05,
+      "loss": 0.4104,
+      "step": 768
+    },
+    {
+      "epoch": 1.149476831091181,
+      "grad_norm": 0.3514153609129661,
+      "learning_rate": 3.4274640088593576e-05,
+      "loss": 0.4466,
+      "step": 769
+    },
+    {
+      "epoch": 1.1509715994020926,
+      "grad_norm": 0.36032898153630816,
+      "learning_rate": 3.424695459579181e-05,
+      "loss": 0.4542,
+      "step": 770
+    },
+    {
+      "epoch": 1.1524663677130045,
+      "grad_norm": 0.3179593214577448,
+      "learning_rate": 3.421926910299004e-05,
+      "loss": 0.4322,
+      "step": 771
+    },
+    {
+      "epoch": 1.1539611360239164,
+      "grad_norm": 0.31509282937529887,
+      "learning_rate": 3.419158361018826e-05,
+      "loss": 0.4219,
+      "step": 772
+    },
+    {
+      "epoch": 1.155455904334828,
+      "grad_norm": 0.3798833989018134,
+      "learning_rate": 3.416389811738649e-05,
+      "loss": 0.4567,
+      "step": 773
+    },
+    {
+      "epoch": 1.15695067264574,
+      "grad_norm": 0.29757318703135116,
+      "learning_rate": 3.413621262458472e-05,
+      "loss": 0.4417,
+      "step": 774
+    },
+    {
+      "epoch": 1.1584454409566518,
+      "grad_norm": 0.3308361064395809,
+      "learning_rate": 3.4108527131782945e-05,
+      "loss": 0.4175,
+      "step": 775
+    },
+    {
+      "epoch": 1.1599402092675635,
+      "grad_norm": 0.32229579880821946,
+      "learning_rate": 3.408084163898118e-05,
+      "loss": 0.4325,
+      "step": 776
+    },
+    {
+      "epoch": 1.1614349775784754,
+      "grad_norm": 0.32446484248564145,
+      "learning_rate": 3.40531561461794e-05,
+      "loss": 0.4274,
+      "step": 777
+    },
+    {
+      "epoch": 1.1629297458893872,
+      "grad_norm": 0.31980373198363277,
+      "learning_rate": 3.4025470653377633e-05,
+      "loss": 0.4182,
+      "step": 778
+    },
+    {
+      "epoch": 1.164424514200299,
+      "grad_norm": 0.3966778569713813,
+      "learning_rate": 3.3997785160575865e-05,
+      "loss": 0.4457,
+      "step": 779
+    },
+    {
+      "epoch": 1.1659192825112108,
+      "grad_norm": 0.39973920704337973,
+      "learning_rate": 3.397009966777408e-05,
+      "loss": 0.4403,
+      "step": 780
+    },
+    {
+      "epoch": 1.1674140508221225,
+      "grad_norm": 0.34930016685588317,
+      "learning_rate": 3.3942414174972315e-05,
+      "loss": 0.4209,
+      "step": 781
+    },
+    {
+      "epoch": 1.1689088191330343,
+      "grad_norm": 0.4095286219205645,
+      "learning_rate": 3.3914728682170546e-05,
+      "loss": 0.4242,
+      "step": 782
+    },
+    {
+      "epoch": 1.1704035874439462,
+      "grad_norm": 0.36617639909235233,
+      "learning_rate": 3.388704318936877e-05,
+      "loss": 0.4234,
+      "step": 783
+    },
+    {
+      "epoch": 1.171898355754858,
+      "grad_norm": 0.3501733180720502,
+      "learning_rate": 3.3859357696567e-05,
+      "loss": 0.4263,
+      "step": 784
+    },
+    {
+      "epoch": 1.1733931240657698,
+      "grad_norm": 0.35801306116849746,
+      "learning_rate": 3.383167220376523e-05,
+      "loss": 0.435,
+      "step": 785
+    },
+    {
+      "epoch": 1.1748878923766817,
+      "grad_norm": 0.38492473756650314,
+      "learning_rate": 3.380398671096346e-05,
+      "loss": 0.4265,
+      "step": 786
+    },
+    {
+      "epoch": 1.1763826606875933,
+      "grad_norm": 0.41653851931328867,
+      "learning_rate": 3.3776301218161684e-05,
+      "loss": 0.4467,
+      "step": 787
+    },
+    {
+      "epoch": 1.1778774289985052,
+      "grad_norm": 0.39315401568214475,
+      "learning_rate": 3.374861572535991e-05,
+      "loss": 0.4425,
+      "step": 788
+    },
+    {
+      "epoch": 1.1793721973094171,
+      "grad_norm": 0.38407193705883685,
+      "learning_rate": 3.372093023255814e-05,
+      "loss": 0.4191,
+      "step": 789
+    },
+    {
+      "epoch": 1.1808669656203288,
+      "grad_norm": 0.39492032657507836,
+      "learning_rate": 3.369324473975637e-05,
+      "loss": 0.4532,
+      "step": 790
+    },
+    {
+      "epoch": 1.1823617339312407,
+      "grad_norm": 0.36885178855281875,
+      "learning_rate": 3.3665559246954597e-05,
+      "loss": 0.4392,
+      "step": 791
+    },
+    {
+      "epoch": 1.1838565022421526,
+      "grad_norm": 0.376492179930556,
+      "learning_rate": 3.363787375415283e-05,
+      "loss": 0.4386,
+      "step": 792
+    },
+    {
+      "epoch": 1.1853512705530642,
+      "grad_norm": 0.3140913893989804,
+      "learning_rate": 3.361018826135105e-05,
+      "loss": 0.4408,
+      "step": 793
+    },
+    {
+      "epoch": 1.186846038863976,
+      "grad_norm": 0.44675465075234416,
+      "learning_rate": 3.3582502768549285e-05,
+      "loss": 0.4558,
+      "step": 794
+    },
+    {
+      "epoch": 1.188340807174888,
+      "grad_norm": 0.3404337311651932,
+      "learning_rate": 3.355481727574751e-05,
+      "loss": 0.4209,
+      "step": 795
+    },
+    {
+      "epoch": 1.1898355754857997,
+      "grad_norm": 0.3612850045742891,
+      "learning_rate": 3.3527131782945734e-05,
+      "loss": 0.4132,
+      "step": 796
+    },
+    {
+      "epoch": 1.1913303437967115,
+      "grad_norm": 0.45644554550703725,
+      "learning_rate": 3.3499446290143966e-05,
+      "loss": 0.4495,
+      "step": 797
+    },
+    {
+      "epoch": 1.1928251121076232,
+      "grad_norm": 0.3302244465755857,
+      "learning_rate": 3.34717607973422e-05,
+      "loss": 0.4159,
+      "step": 798
+    },
+    {
+      "epoch": 1.194319880418535,
+      "grad_norm": 0.4025924800083753,
+      "learning_rate": 3.344407530454042e-05,
+      "loss": 0.4241,
+      "step": 799
+    },
+    {
+      "epoch": 1.195814648729447,
+      "grad_norm": 0.3633384338378567,
+      "learning_rate": 3.3416389811738654e-05,
+      "loss": 0.4333,
+      "step": 800
+    },
+    {
+      "epoch": 1.1973094170403586,
+      "grad_norm": 0.3033490031296353,
+      "learning_rate": 3.338870431893688e-05,
+      "loss": 0.4095,
+      "step": 801
+    },
+    {
+      "epoch": 1.1988041853512705,
+      "grad_norm": 0.32533975519955866,
+      "learning_rate": 3.33610188261351e-05,
+      "loss": 0.444,
+      "step": 802
+    },
+    {
+      "epoch": 1.2002989536621824,
+      "grad_norm": 0.34345656102140476,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.433,
+      "step": 803
+    },
+    {
+      "epoch": 1.201793721973094,
+      "grad_norm": 0.32151334654727504,
+      "learning_rate": 3.330564784053156e-05,
+      "loss": 0.4066,
+      "step": 804
+    },
+    {
+      "epoch": 1.203288490284006,
+      "grad_norm": 0.3400199384542232,
+      "learning_rate": 3.327796234772979e-05,
+      "loss": 0.4292,
+      "step": 805
+    },
+    {
+      "epoch": 1.2047832585949179,
+      "grad_norm": 0.3383756237090503,
+      "learning_rate": 3.325027685492802e-05,
+      "loss": 0.4473,
+      "step": 806
+    },
+    {
+      "epoch": 1.2062780269058295,
+      "grad_norm": 0.3433515937451186,
+      "learning_rate": 3.322259136212625e-05,
+      "loss": 0.4284,
+      "step": 807
+    },
+    {
+      "epoch": 1.2077727952167414,
+      "grad_norm": 0.4495509217423976,
+      "learning_rate": 3.319490586932448e-05,
+      "loss": 0.434,
+      "step": 808
+    },
+    {
+      "epoch": 1.2092675635276533,
+      "grad_norm": 0.3255927060311819,
+      "learning_rate": 3.3167220376522704e-05,
+      "loss": 0.4387,
+      "step": 809
+    },
+    {
+      "epoch": 1.210762331838565,
+      "grad_norm": 0.42008320099741575,
+      "learning_rate": 3.313953488372093e-05,
+      "loss": 0.4349,
+      "step": 810
+    },
+    {
+      "epoch": 1.2122571001494769,
+      "grad_norm": 0.361196451546201,
+      "learning_rate": 3.311184939091916e-05,
+      "loss": 0.4209,
+      "step": 811
+    },
+    {
+      "epoch": 1.2137518684603887,
+      "grad_norm": 0.31318554457860354,
+      "learning_rate": 3.3084163898117385e-05,
+      "loss": 0.4362,
+      "step": 812
+    },
+    {
+      "epoch": 1.2152466367713004,
+      "grad_norm": 0.4513666410481421,
+      "learning_rate": 3.305647840531562e-05,
+      "loss": 0.4383,
+      "step": 813
+    },
+    {
+      "epoch": 1.2167414050822123,
+      "grad_norm": 0.3196473786269645,
+      "learning_rate": 3.302879291251385e-05,
+      "loss": 0.4298,
+      "step": 814
+    },
+    {
+      "epoch": 1.218236173393124,
+      "grad_norm": 0.42278616099402433,
+      "learning_rate": 3.300110741971207e-05,
+      "loss": 0.4455,
+      "step": 815
+    },
+    {
+      "epoch": 1.2197309417040358,
+      "grad_norm": 0.3355325933505012,
+      "learning_rate": 3.2973421926910305e-05,
+      "loss": 0.4381,
+      "step": 816
+    },
+    {
+      "epoch": 1.2212257100149477,
+      "grad_norm": 0.44026141123895746,
+      "learning_rate": 3.294573643410852e-05,
+      "loss": 0.4436,
+      "step": 817
+    },
+    {
+      "epoch": 1.2227204783258596,
+      "grad_norm": 0.3415228385243287,
+      "learning_rate": 3.2918050941306754e-05,
+      "loss": 0.4433,
+      "step": 818
+    },
+    {
+      "epoch": 1.2242152466367713,
+      "grad_norm": 0.3850332941280085,
+      "learning_rate": 3.2890365448504986e-05,
+      "loss": 0.4556,
+      "step": 819
+    },
+    {
+      "epoch": 1.2257100149476832,
+      "grad_norm": 0.3466900190119601,
+      "learning_rate": 3.286267995570321e-05,
+      "loss": 0.4284,
+      "step": 820
+    },
+    {
+      "epoch": 1.2272047832585948,
+      "grad_norm": 0.37813620935169434,
+      "learning_rate": 3.283499446290144e-05,
+      "loss": 0.4461,
+      "step": 821
+    },
+    {
+      "epoch": 1.2286995515695067,
+      "grad_norm": 0.372431623788217,
+      "learning_rate": 3.280730897009967e-05,
+      "loss": 0.4297,
+      "step": 822
+    },
+    {
+      "epoch": 1.2301943198804186,
+      "grad_norm": 0.33805022419015546,
+      "learning_rate": 3.27796234772979e-05,
+      "loss": 0.4297,
+      "step": 823
+    },
+    {
+      "epoch": 1.2316890881913303,
+      "grad_norm": 0.3076238228935569,
+      "learning_rate": 3.275193798449613e-05,
+      "loss": 0.4396,
+      "step": 824
+    },
+    {
+      "epoch": 1.2331838565022422,
+      "grad_norm": 0.406268481699972,
+      "learning_rate": 3.272425249169435e-05,
+      "loss": 0.453,
+      "step": 825
+    },
+    {
+      "epoch": 1.234678624813154,
+      "grad_norm": 0.3051411965650107,
+      "learning_rate": 3.269656699889258e-05,
+      "loss": 0.4206,
+      "step": 826
+    },
+    {
+      "epoch": 1.2361733931240657,
+      "grad_norm": 0.4291005125645932,
+      "learning_rate": 3.266888150609081e-05,
+      "loss": 0.455,
+      "step": 827
+    },
+    {
+      "epoch": 1.2376681614349776,
+      "grad_norm": 0.35234770913229896,
+      "learning_rate": 3.2641196013289036e-05,
+      "loss": 0.4614,
+      "step": 828
+    },
+    {
+      "epoch": 1.2391629297458895,
+      "grad_norm": 0.3771390713582011,
+      "learning_rate": 3.261351052048727e-05,
+      "loss": 0.4095,
+      "step": 829
+    },
+    {
+      "epoch": 1.2406576980568012,
+      "grad_norm": 0.32026677478076715,
+      "learning_rate": 3.258582502768549e-05,
+      "loss": 0.421,
+      "step": 830
+    },
+    {
+      "epoch": 1.242152466367713,
+      "grad_norm": 0.4208181572133447,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 0.439,
+      "step": 831
+    },
+    {
+      "epoch": 1.2436472346786247,
+      "grad_norm": 0.2725578483097924,
+      "learning_rate": 3.253045404208195e-05,
+      "loss": 0.439,
+      "step": 832
+    },
+    {
+      "epoch": 1.2451420029895366,
+      "grad_norm": 0.3592427542647874,
+      "learning_rate": 3.2502768549280174e-05,
+      "loss": 0.4333,
+      "step": 833
+    },
+    {
+      "epoch": 1.2466367713004485,
+      "grad_norm": 0.35344703985056314,
+      "learning_rate": 3.2475083056478406e-05,
+      "loss": 0.4244,
+      "step": 834
+    },
+    {
+      "epoch": 1.2481315396113604,
+      "grad_norm": 0.3725832140637695,
+      "learning_rate": 3.244739756367664e-05,
+      "loss": 0.4381,
+      "step": 835
+    },
+    {
+      "epoch": 1.249626307922272,
+      "grad_norm": 0.30591417134249227,
+      "learning_rate": 3.241971207087486e-05,
+      "loss": 0.4427,
+      "step": 836
+    },
+    {
+      "epoch": 1.251121076233184,
+      "grad_norm": 0.3823918351091783,
+      "learning_rate": 3.2392026578073094e-05,
+      "loss": 0.4498,
+      "step": 837
+    },
+    {
+      "epoch": 1.2526158445440956,
+      "grad_norm": 0.3245028908133399,
+      "learning_rate": 3.236434108527132e-05,
+      "loss": 0.4514,
+      "step": 838
+    },
+    {
+      "epoch": 1.2541106128550075,
+      "grad_norm": 0.32534808977911633,
+      "learning_rate": 3.233665559246955e-05,
+      "loss": 0.429,
+      "step": 839
+    },
+    {
+      "epoch": 1.2556053811659194,
+      "grad_norm": 0.30685652831288096,
+      "learning_rate": 3.2308970099667775e-05,
+      "loss": 0.451,
+      "step": 840
+    },
+    {
+      "epoch": 1.257100149476831,
+      "grad_norm": 0.34360497159870224,
+      "learning_rate": 3.2281284606866e-05,
+      "loss": 0.4191,
+      "step": 841
+    },
+    {
+      "epoch": 1.258594917787743,
+      "grad_norm": 0.2527148678416448,
+      "learning_rate": 3.225359911406423e-05,
+      "loss": 0.4029,
+      "step": 842
+    },
+    {
+      "epoch": 1.2600896860986546,
+      "grad_norm": 0.3214997196399334,
+      "learning_rate": 3.222591362126246e-05,
+      "loss": 0.4192,
+      "step": 843
+    },
+    {
+      "epoch": 1.2615844544095665,
+      "grad_norm": 0.3325479697452683,
+      "learning_rate": 3.219822812846069e-05,
+      "loss": 0.4059,
+      "step": 844
+    },
+    {
+      "epoch": 1.2630792227204783,
+      "grad_norm": 0.2770992062809255,
+      "learning_rate": 3.217054263565892e-05,
+      "loss": 0.4625,
+      "step": 845
+    },
+    {
+      "epoch": 1.2645739910313902,
+      "grad_norm": 0.32033695193962197,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.429,
+      "step": 846
+    },
+    {
+      "epoch": 1.266068759342302,
+      "grad_norm": 0.33518422904465395,
+      "learning_rate": 3.211517165005537e-05,
+      "loss": 0.4251,
+      "step": 847
+    },
+    {
+      "epoch": 1.2675635276532138,
+      "grad_norm": 0.2980618691797428,
+      "learning_rate": 3.20874861572536e-05,
+      "loss": 0.4393,
+      "step": 848
+    },
+    {
+      "epoch": 1.2690582959641254,
+      "grad_norm": 0.311519351777786,
+      "learning_rate": 3.2059800664451825e-05,
+      "loss": 0.4414,
+      "step": 849
+    },
+    {
+      "epoch": 1.2705530642750373,
+      "grad_norm": 0.34871920336553225,
+      "learning_rate": 3.203211517165006e-05,
+      "loss": 0.443,
+      "step": 850
+    },
+    {
+      "epoch": 1.2720478325859492,
+      "grad_norm": 0.34521043974133947,
+      "learning_rate": 3.200442967884829e-05,
+      "loss": 0.4249,
+      "step": 851
+    },
+    {
+      "epoch": 1.273542600896861,
+      "grad_norm": 0.33871883586770735,
+      "learning_rate": 3.197674418604651e-05,
+      "loss": 0.4496,
+      "step": 852
+    },
+    {
+      "epoch": 1.2750373692077728,
+      "grad_norm": 0.31353994414702063,
+      "learning_rate": 3.1949058693244745e-05,
+      "loss": 0.4328,
+      "step": 853
+    },
+    {
+      "epoch": 1.2765321375186847,
+      "grad_norm": 0.3267315963645876,
+      "learning_rate": 3.192137320044297e-05,
+      "loss": 0.4341,
+      "step": 854
+    },
+    {
+      "epoch": 1.2780269058295963,
+      "grad_norm": 0.3897546150266528,
+      "learning_rate": 3.1893687707641194e-05,
+      "loss": 0.436,
+      "step": 855
+    },
+    {
+      "epoch": 1.2795216741405082,
+      "grad_norm": 0.3237418939841499,
+      "learning_rate": 3.1866002214839426e-05,
+      "loss": 0.433,
+      "step": 856
+    },
+    {
+      "epoch": 1.28101644245142,
+      "grad_norm": 0.3831413235728547,
+      "learning_rate": 3.183831672203765e-05,
+      "loss": 0.4359,
+      "step": 857
+    },
+    {
+      "epoch": 1.2825112107623318,
+      "grad_norm": 0.3007769849944101,
+      "learning_rate": 3.181063122923588e-05,
+      "loss": 0.4078,
+      "step": 858
+    },
+    {
+      "epoch": 1.2840059790732437,
+      "grad_norm": 0.39576951806225613,
+      "learning_rate": 3.1782945736434114e-05,
+      "loss": 0.4493,
+      "step": 859
+    },
+    {
+      "epoch": 1.2855007473841553,
+      "grad_norm": 0.4052813588663847,
+      "learning_rate": 3.175526024363234e-05,
+      "loss": 0.4138,
+      "step": 860
+    },
+    {
+      "epoch": 1.2869955156950672,
+      "grad_norm": 0.3200541816381964,
+      "learning_rate": 3.172757475083057e-05,
+      "loss": 0.4309,
+      "step": 861
+    },
+    {
+      "epoch": 1.288490284005979,
+      "grad_norm": 0.4381128625399564,
+      "learning_rate": 3.1699889258028795e-05,
+      "loss": 0.443,
+      "step": 862
+    },
+    {
+      "epoch": 1.289985052316891,
+      "grad_norm": 0.32181436604722197,
+      "learning_rate": 3.167220376522702e-05,
+      "loss": 0.4416,
+      "step": 863
+    },
+    {
+      "epoch": 1.2914798206278026,
+      "grad_norm": 0.4358554863008817,
+      "learning_rate": 3.164451827242525e-05,
+      "loss": 0.4296,
+      "step": 864
+    },
+    {
+      "epoch": 1.2929745889387145,
+      "grad_norm": 0.3481342173175755,
+      "learning_rate": 3.1616832779623476e-05,
+      "loss": 0.4332,
+      "step": 865
+    },
+    {
+      "epoch": 1.2944693572496262,
+      "grad_norm": 0.38051944604143695,
+      "learning_rate": 3.158914728682171e-05,
+      "loss": 0.4298,
+      "step": 866
+    },
+    {
+      "epoch": 1.295964125560538,
+      "grad_norm": 0.35902883396150426,
+      "learning_rate": 3.156146179401994e-05,
+      "loss": 0.4351,
+      "step": 867
+    },
+    {
+      "epoch": 1.29745889387145,
+      "grad_norm": 0.35651580038149644,
+      "learning_rate": 3.1533776301218164e-05,
+      "loss": 0.4465,
+      "step": 868
+    },
+    {
+      "epoch": 1.2989536621823619,
+      "grad_norm": 0.3391628197969806,
+      "learning_rate": 3.150609080841639e-05,
+      "loss": 0.4147,
+      "step": 869
+    },
+    {
+      "epoch": 1.3004484304932735,
+      "grad_norm": 0.3374470596887314,
+      "learning_rate": 3.147840531561462e-05,
+      "loss": 0.4407,
+      "step": 870
+    },
+    {
+      "epoch": 1.3019431988041854,
+      "grad_norm": 0.31195983171175484,
+      "learning_rate": 3.1450719822812845e-05,
+      "loss": 0.4422,
+      "step": 871
+    },
+    {
+      "epoch": 1.303437967115097,
+      "grad_norm": 0.3512225668938008,
+      "learning_rate": 3.142303433001108e-05,
+      "loss": 0.4133,
+      "step": 872
+    },
+    {
+      "epoch": 1.304932735426009,
+      "grad_norm": 0.38262721719418197,
+      "learning_rate": 3.13953488372093e-05,
+      "loss": 0.4485,
+      "step": 873
+    },
+    {
+      "epoch": 1.3064275037369208,
+      "grad_norm": 0.3016060680026093,
+      "learning_rate": 3.136766334440753e-05,
+      "loss": 0.4303,
+      "step": 874
+    },
+    {
+      "epoch": 1.3079222720478325,
+      "grad_norm": 0.44050450012402187,
+      "learning_rate": 3.1339977851605765e-05,
+      "loss": 0.4325,
+      "step": 875
+    },
+    {
+      "epoch": 1.3094170403587444,
+      "grad_norm": 0.2896677698444671,
+      "learning_rate": 3.131229235880399e-05,
+      "loss": 0.4129,
+      "step": 876
+    },
+    {
+      "epoch": 1.310911808669656,
+      "grad_norm": 0.3785404316980104,
+      "learning_rate": 3.1284606866002215e-05,
+      "loss": 0.4387,
+      "step": 877
+    },
+    {
+      "epoch": 1.312406576980568,
+      "grad_norm": 0.3396950624835025,
+      "learning_rate": 3.1256921373200446e-05,
+      "loss": 0.4304,
+      "step": 878
+    },
+    {
+      "epoch": 1.3139013452914798,
+      "grad_norm": 0.43550326408935613,
+      "learning_rate": 3.122923588039867e-05,
+      "loss": 0.4263,
+      "step": 879
+    },
+    {
+      "epoch": 1.3153961136023917,
+      "grad_norm": 0.3160684576460051,
+      "learning_rate": 3.12015503875969e-05,
+      "loss": 0.4314,
+      "step": 880
+    },
+    {
+      "epoch": 1.3168908819133034,
+      "grad_norm": 0.4022392292846904,
+      "learning_rate": 3.117386489479513e-05,
+      "loss": 0.4259,
+      "step": 881
+    },
+    {
+      "epoch": 1.3183856502242153,
+      "grad_norm": 0.36742828972266767,
+      "learning_rate": 3.114617940199336e-05,
+      "loss": 0.4485,
+      "step": 882
+    },
+    {
+      "epoch": 1.319880418535127,
+      "grad_norm": 0.40992058882122884,
+      "learning_rate": 3.111849390919159e-05,
+      "loss": 0.4136,
+      "step": 883
+    },
+    {
+      "epoch": 1.3213751868460388,
+      "grad_norm": 0.35492905603516944,
+      "learning_rate": 3.109080841638981e-05,
+      "loss": 0.4349,
+      "step": 884
+    },
+    {
+      "epoch": 1.3228699551569507,
+      "grad_norm": 0.31647008674289084,
+      "learning_rate": 3.106312292358804e-05,
+      "loss": 0.4408,
+      "step": 885
+    },
+    {
+      "epoch": 1.3243647234678626,
+      "grad_norm": 0.3525665261854639,
+      "learning_rate": 3.1035437430786265e-05,
+      "loss": 0.4385,
+      "step": 886
+    },
+    {
+      "epoch": 1.3258594917787743,
+      "grad_norm": 0.3631867062472928,
+      "learning_rate": 3.1007751937984497e-05,
+      "loss": 0.444,
+      "step": 887
+    },
+    {
+      "epoch": 1.3273542600896862,
+      "grad_norm": 0.39644837151974305,
+      "learning_rate": 3.098006644518273e-05,
+      "loss": 0.432,
+      "step": 888
+    },
+    {
+      "epoch": 1.3288490284005978,
+      "grad_norm": 0.39133061527362484,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.4415,
+      "step": 889
+    },
+    {
+      "epoch": 1.3303437967115097,
+      "grad_norm": 0.3498780957657142,
+      "learning_rate": 3.0924695459579184e-05,
+      "loss": 0.451,
+      "step": 890
+    },
+    {
+      "epoch": 1.3318385650224216,
+      "grad_norm": 0.44982299564636774,
+      "learning_rate": 3.089700996677741e-05,
+      "loss": 0.4535,
+      "step": 891
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.38907326220241945,
+      "learning_rate": 3.0869324473975634e-05,
+      "loss": 0.4204,
+      "step": 892
+    },
+    {
+      "epoch": 1.3348281016442451,
+      "grad_norm": 0.3373829242506409,
+      "learning_rate": 3.0841638981173866e-05,
+      "loss": 0.4277,
+      "step": 893
+    },
+    {
+      "epoch": 1.336322869955157,
+      "grad_norm": 0.39269437287812653,
+      "learning_rate": 3.081395348837209e-05,
+      "loss": 0.4374,
+      "step": 894
+    },
+    {
+      "epoch": 1.3378176382660687,
+      "grad_norm": 0.3884723808744727,
+      "learning_rate": 3.078626799557032e-05,
+      "loss": 0.4224,
+      "step": 895
+    },
+    {
+      "epoch": 1.3393124065769806,
+      "grad_norm": 0.37691847593968747,
+      "learning_rate": 3.0758582502768554e-05,
+      "loss": 0.4343,
+      "step": 896
+    },
+    {
+      "epoch": 1.3408071748878925,
+      "grad_norm": 0.3669099338905344,
+      "learning_rate": 3.073089700996678e-05,
+      "loss": 0.4522,
+      "step": 897
+    },
+    {
+      "epoch": 1.3423019431988041,
+      "grad_norm": 0.34662143147221725,
+      "learning_rate": 3.070321151716501e-05,
+      "loss": 0.4355,
+      "step": 898
+    },
+    {
+      "epoch": 1.343796711509716,
+      "grad_norm": 0.33744939280321956,
+      "learning_rate": 3.0675526024363235e-05,
+      "loss": 0.4388,
+      "step": 899
+    },
+    {
+      "epoch": 1.3452914798206277,
+      "grad_norm": 0.36975933609677536,
+      "learning_rate": 3.064784053156146e-05,
+      "loss": 0.4091,
+      "step": 900
+    },
+    {
+      "epoch": 1.3467862481315396,
+      "grad_norm": 0.3183886735140358,
+      "learning_rate": 3.062015503875969e-05,
+      "loss": 0.4323,
+      "step": 901
+    },
+    {
+      "epoch": 1.3482810164424515,
+      "grad_norm": 0.34751440449252435,
+      "learning_rate": 3.0592469545957916e-05,
+      "loss": 0.4109,
+      "step": 902
+    },
+    {
+      "epoch": 1.3497757847533634,
+      "grad_norm": 0.33295794876053425,
+      "learning_rate": 3.056478405315615e-05,
+      "loss": 0.439,
+      "step": 903
+    },
+    {
+      "epoch": 1.351270553064275,
+      "grad_norm": 0.32032375546133307,
+      "learning_rate": 3.053709856035438e-05,
+      "loss": 0.439,
+      "step": 904
+    },
+    {
+      "epoch": 1.352765321375187,
+      "grad_norm": 0.3654415837288015,
+      "learning_rate": 3.0509413067552607e-05,
+      "loss": 0.4618,
+      "step": 905
+    },
+    {
+      "epoch": 1.3542600896860986,
+      "grad_norm": 0.3607390725330527,
+      "learning_rate": 3.0481727574750836e-05,
+      "loss": 0.4423,
+      "step": 906
+    },
+    {
+      "epoch": 1.3557548579970105,
+      "grad_norm": 0.3413518376158147,
+      "learning_rate": 3.0454042081949057e-05,
+      "loss": 0.4403,
+      "step": 907
+    },
+    {
+      "epoch": 1.3572496263079223,
+      "grad_norm": 0.37079086640733966,
+      "learning_rate": 3.0426356589147285e-05,
+      "loss": 0.4138,
+      "step": 908
+    },
+    {
+      "epoch": 1.358744394618834,
+      "grad_norm": 0.377830899131017,
+      "learning_rate": 3.0398671096345517e-05,
+      "loss": 0.4486,
+      "step": 909
+    },
+    {
+      "epoch": 1.360239162929746,
+      "grad_norm": 0.32953941800228,
+      "learning_rate": 3.0370985603543745e-05,
+      "loss": 0.4413,
+      "step": 910
+    },
+    {
+      "epoch": 1.3617339312406578,
+      "grad_norm": 0.430145165338916,
+      "learning_rate": 3.0343300110741973e-05,
+      "loss": 0.4236,
+      "step": 911
+    },
+    {
+      "epoch": 1.3632286995515694,
+      "grad_norm": 0.332660706547324,
+      "learning_rate": 3.03156146179402e-05,
+      "loss": 0.4339,
+      "step": 912
+    },
+    {
+      "epoch": 1.3647234678624813,
+      "grad_norm": 0.3969023641318391,
+      "learning_rate": 3.028792912513843e-05,
+      "loss": 0.4316,
+      "step": 913
+    },
+    {
+      "epoch": 1.3662182361733932,
+      "grad_norm": 0.37742096655595203,
+      "learning_rate": 3.0260243632336654e-05,
+      "loss": 0.4453,
+      "step": 914
+    },
+    {
+      "epoch": 1.3677130044843049,
+      "grad_norm": 0.3774316421051345,
+      "learning_rate": 3.0232558139534883e-05,
+      "loss": 0.4523,
+      "step": 915
+    },
+    {
+      "epoch": 1.3692077727952168,
+      "grad_norm": 0.32733373308528335,
+      "learning_rate": 3.020487264673311e-05,
+      "loss": 0.4267,
+      "step": 916
+    },
+    {
+      "epoch": 1.3707025411061284,
+      "grad_norm": 0.39956881695082364,
+      "learning_rate": 3.0177187153931342e-05,
+      "loss": 0.4565,
+      "step": 917
+    },
+    {
+      "epoch": 1.3721973094170403,
+      "grad_norm": 0.3275911351843889,
+      "learning_rate": 3.014950166112957e-05,
+      "loss": 0.444,
+      "step": 918
+    },
+    {
+      "epoch": 1.3736920777279522,
+      "grad_norm": 0.37217671525747525,
+      "learning_rate": 3.01218161683278e-05,
+      "loss": 0.4359,
+      "step": 919
+    },
+    {
+      "epoch": 1.375186846038864,
+      "grad_norm": 0.35214561557096363,
+      "learning_rate": 3.0094130675526027e-05,
+      "loss": 0.4371,
+      "step": 920
+    },
+    {
+      "epoch": 1.3766816143497758,
+      "grad_norm": 0.31396663001790404,
+      "learning_rate": 3.0066445182724255e-05,
+      "loss": 0.4288,
+      "step": 921
+    },
+    {
+      "epoch": 1.3781763826606876,
+      "grad_norm": 0.38576773544691656,
+      "learning_rate": 3.003875968992248e-05,
+      "loss": 0.4219,
+      "step": 922
+    },
+    {
+      "epoch": 1.3796711509715993,
+      "grad_norm": 0.30848937806886634,
+      "learning_rate": 3.0011074197120708e-05,
+      "loss": 0.431,
+      "step": 923
+    },
+    {
+      "epoch": 1.3811659192825112,
+      "grad_norm": 0.3910855381262825,
+      "learning_rate": 2.9983388704318936e-05,
+      "loss": 0.4314,
+      "step": 924
+    },
+    {
+      "epoch": 1.382660687593423,
+      "grad_norm": 0.29548047962290586,
+      "learning_rate": 2.9955703211517168e-05,
+      "loss": 0.4258,
+      "step": 925
+    },
+    {
+      "epoch": 1.3841554559043348,
+      "grad_norm": 0.38125061158640594,
+      "learning_rate": 2.9928017718715396e-05,
+      "loss": 0.4502,
+      "step": 926
+    },
+    {
+      "epoch": 1.3856502242152466,
+      "grad_norm": 0.35553179020548964,
+      "learning_rate": 2.9900332225913624e-05,
+      "loss": 0.4371,
+      "step": 927
+    },
+    {
+      "epoch": 1.3871449925261585,
+      "grad_norm": 0.27829990552490996,
+      "learning_rate": 2.9872646733111852e-05,
+      "loss": 0.4155,
+      "step": 928
+    },
+    {
+      "epoch": 1.3886397608370702,
+      "grad_norm": 0.32985633160500827,
+      "learning_rate": 2.9844961240310077e-05,
+      "loss": 0.4226,
+      "step": 929
+    },
+    {
+      "epoch": 1.390134529147982,
+      "grad_norm": 0.33158184486364717,
+      "learning_rate": 2.9817275747508305e-05,
+      "loss": 0.4329,
+      "step": 930
+    },
+    {
+      "epoch": 1.391629297458894,
+      "grad_norm": 0.3451230592380717,
+      "learning_rate": 2.9789590254706534e-05,
+      "loss": 0.4244,
+      "step": 931
+    },
+    {
+      "epoch": 1.3931240657698056,
+      "grad_norm": 0.33424382365174365,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 0.412,
+      "step": 932
+    },
+    {
+      "epoch": 1.3946188340807175,
+      "grad_norm": 0.35612929687342765,
+      "learning_rate": 2.9734219269102993e-05,
+      "loss": 0.4491,
+      "step": 933
+    },
+    {
+      "epoch": 1.3961136023916292,
+      "grad_norm": 0.31498332903379583,
+      "learning_rate": 2.970653377630122e-05,
+      "loss": 0.4358,
+      "step": 934
+    },
+    {
+      "epoch": 1.397608370702541,
+      "grad_norm": 0.3277493677995811,
+      "learning_rate": 2.967884828349945e-05,
+      "loss": 0.4384,
+      "step": 935
+    },
+    {
+      "epoch": 1.399103139013453,
+      "grad_norm": 0.36087902622703044,
+      "learning_rate": 2.9651162790697678e-05,
+      "loss": 0.4354,
+      "step": 936
+    },
+    {
+      "epoch": 1.4005979073243648,
+      "grad_norm": 0.2988116953122595,
+      "learning_rate": 2.9623477297895903e-05,
+      "loss": 0.4401,
+      "step": 937
+    },
+    {
+      "epoch": 1.4020926756352765,
+      "grad_norm": 0.3281477202251096,
+      "learning_rate": 2.959579180509413e-05,
+      "loss": 0.4304,
+      "step": 938
+    },
+    {
+      "epoch": 1.4035874439461884,
+      "grad_norm": 0.3902878867319533,
+      "learning_rate": 2.956810631229236e-05,
+      "loss": 0.4188,
+      "step": 939
+    },
+    {
+      "epoch": 1.4050822122571,
+      "grad_norm": 0.2959764395369121,
+      "learning_rate": 2.9540420819490587e-05,
+      "loss": 0.4216,
+      "step": 940
+    },
+    {
+      "epoch": 1.406576980568012,
+      "grad_norm": 0.3155295619888196,
+      "learning_rate": 2.951273532668882e-05,
+      "loss": 0.4231,
+      "step": 941
+    },
+    {
+      "epoch": 1.4080717488789238,
+      "grad_norm": 0.41051539976028856,
+      "learning_rate": 2.9485049833887047e-05,
+      "loss": 0.4483,
+      "step": 942
+    },
+    {
+      "epoch": 1.4095665171898355,
+      "grad_norm": 0.3237700860757843,
+      "learning_rate": 2.9457364341085275e-05,
+      "loss": 0.4289,
+      "step": 943
+    },
+    {
+      "epoch": 1.4110612855007474,
+      "grad_norm": 0.33103584984269224,
+      "learning_rate": 2.9429678848283497e-05,
+      "loss": 0.433,
+      "step": 944
+    },
+    {
+      "epoch": 1.4125560538116593,
+      "grad_norm": 0.31912021925111633,
+      "learning_rate": 2.940199335548173e-05,
+      "loss": 0.4282,
+      "step": 945
+    },
+    {
+      "epoch": 1.414050822122571,
+      "grad_norm": 0.3218023016066329,
+      "learning_rate": 2.9374307862679957e-05,
+      "loss": 0.4102,
+      "step": 946
+    },
+    {
+      "epoch": 1.4155455904334828,
+      "grad_norm": 0.3000056628640221,
+      "learning_rate": 2.9346622369878185e-05,
+      "loss": 0.4233,
+      "step": 947
+    },
+    {
+      "epoch": 1.4170403587443947,
+      "grad_norm": 0.3196188256353491,
+      "learning_rate": 2.9318936877076413e-05,
+      "loss": 0.4599,
+      "step": 948
+    },
+    {
+      "epoch": 1.4185351270553064,
+      "grad_norm": 0.3523839850984468,
+      "learning_rate": 2.929125138427464e-05,
+      "loss": 0.4289,
+      "step": 949
+    },
+    {
+      "epoch": 1.4200298953662183,
+      "grad_norm": 0.2964823137600726,
+      "learning_rate": 2.9263565891472873e-05,
+      "loss": 0.4344,
+      "step": 950
+    },
+    {
+      "epoch": 1.42152466367713,
+      "grad_norm": 0.2761380031688127,
+      "learning_rate": 2.92358803986711e-05,
+      "loss": 0.4201,
+      "step": 951
+    },
+    {
+      "epoch": 1.4230194319880418,
+      "grad_norm": 0.28262117213229804,
+      "learning_rate": 2.9208194905869322e-05,
+      "loss": 0.4286,
+      "step": 952
+    },
+    {
+      "epoch": 1.4245142002989537,
+      "grad_norm": 0.3229658830073446,
+      "learning_rate": 2.9180509413067554e-05,
+      "loss": 0.4451,
+      "step": 953
+    },
+    {
+      "epoch": 1.4260089686098656,
+      "grad_norm": 0.3571619595066272,
+      "learning_rate": 2.9152823920265782e-05,
+      "loss": 0.4409,
+      "step": 954
+    },
+    {
+      "epoch": 1.4275037369207773,
+      "grad_norm": 0.3063351079719421,
+      "learning_rate": 2.912513842746401e-05,
+      "loss": 0.4382,
+      "step": 955
+    },
+    {
+      "epoch": 1.4289985052316891,
+      "grad_norm": 0.3445158230797913,
+      "learning_rate": 2.909745293466224e-05,
+      "loss": 0.4355,
+      "step": 956
+    },
+    {
+      "epoch": 1.4304932735426008,
+      "grad_norm": 0.3006128988831028,
+      "learning_rate": 2.9069767441860467e-05,
+      "loss": 0.4324,
+      "step": 957
+    },
+    {
+      "epoch": 1.4319880418535127,
+      "grad_norm": 0.28106095104384243,
+      "learning_rate": 2.90420819490587e-05,
+      "loss": 0.4198,
+      "step": 958
+    },
+    {
+      "epoch": 1.4334828101644246,
+      "grad_norm": 0.2904091324958901,
+      "learning_rate": 2.901439645625692e-05,
+      "loss": 0.4383,
+      "step": 959
+    },
+    {
+      "epoch": 1.4349775784753362,
+      "grad_norm": 0.32533476959534374,
+      "learning_rate": 2.8986710963455148e-05,
+      "loss": 0.4296,
+      "step": 960
+    },
+    {
+      "epoch": 1.4364723467862481,
+      "grad_norm": 0.31042276760726384,
+      "learning_rate": 2.895902547065338e-05,
+      "loss": 0.4381,
+      "step": 961
+    },
+    {
+      "epoch": 1.43796711509716,
+      "grad_norm": 0.31936119308789207,
+      "learning_rate": 2.8931339977851608e-05,
+      "loss": 0.4163,
+      "step": 962
+    },
+    {
+      "epoch": 1.4394618834080717,
+      "grad_norm": 0.31883784298449536,
+      "learning_rate": 2.8903654485049836e-05,
+      "loss": 0.4302,
+      "step": 963
+    },
+    {
+      "epoch": 1.4409566517189836,
+      "grad_norm": 0.3161851252537244,
+      "learning_rate": 2.8875968992248064e-05,
+      "loss": 0.4441,
+      "step": 964
+    },
+    {
+      "epoch": 1.4424514200298955,
+      "grad_norm": 0.3738441285591444,
+      "learning_rate": 2.8848283499446292e-05,
+      "loss": 0.4219,
+      "step": 965
+    },
+    {
+      "epoch": 1.4439461883408071,
+      "grad_norm": 0.32338705069461743,
+      "learning_rate": 2.8820598006644517e-05,
+      "loss": 0.4179,
+      "step": 966
+    },
+    {
+      "epoch": 1.445440956651719,
+      "grad_norm": 0.36594749678131894,
+      "learning_rate": 2.8792912513842745e-05,
+      "loss": 0.4273,
+      "step": 967
+    },
+    {
+      "epoch": 1.4469357249626307,
+      "grad_norm": 0.34522709054479367,
+      "learning_rate": 2.8765227021040973e-05,
+      "loss": 0.4335,
+      "step": 968
+    },
+    {
+      "epoch": 1.4484304932735426,
+      "grad_norm": 0.3379950562455003,
+      "learning_rate": 2.8737541528239205e-05,
+      "loss": 0.4401,
+      "step": 969
+    },
+    {
+      "epoch": 1.4499252615844545,
+      "grad_norm": 0.3000713514812489,
+      "learning_rate": 2.8709856035437433e-05,
+      "loss": 0.4179,
+      "step": 970
+    },
+    {
+      "epoch": 1.4514200298953663,
+      "grad_norm": 0.3540680342254979,
+      "learning_rate": 2.868217054263566e-05,
+      "loss": 0.4297,
+      "step": 971
+    },
+    {
+      "epoch": 1.452914798206278,
+      "grad_norm": 0.32689691328989745,
+      "learning_rate": 2.865448504983389e-05,
+      "loss": 0.4369,
+      "step": 972
+    },
+    {
+      "epoch": 1.45440956651719,
+      "grad_norm": 0.3600616117476702,
+      "learning_rate": 2.8626799557032118e-05,
+      "loss": 0.4354,
+      "step": 973
+    },
+    {
+      "epoch": 1.4559043348281016,
+      "grad_norm": 0.3710287825543155,
+      "learning_rate": 2.8599114064230343e-05,
+      "loss": 0.4055,
+      "step": 974
+    },
+    {
+      "epoch": 1.4573991031390134,
+      "grad_norm": 0.38447163482314545,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.4325,
+      "step": 975
+    },
+    {
+      "epoch": 1.4588938714499253,
+      "grad_norm": 0.3654228899882485,
+      "learning_rate": 2.85437430786268e-05,
+      "loss": 0.4433,
+      "step": 976
+    },
+    {
+      "epoch": 1.460388639760837,
+      "grad_norm": 0.4311470104435173,
+      "learning_rate": 2.851605758582503e-05,
+      "loss": 0.4305,
+      "step": 977
+    },
+    {
+      "epoch": 1.4618834080717489,
+      "grad_norm": 0.32905796476032784,
+      "learning_rate": 2.848837209302326e-05,
+      "loss": 0.407,
+      "step": 978
+    },
+    {
+      "epoch": 1.4633781763826608,
+      "grad_norm": 0.4466022384759354,
+      "learning_rate": 2.8460686600221487e-05,
+      "loss": 0.4306,
+      "step": 979
+    },
+    {
+      "epoch": 1.4648729446935724,
+      "grad_norm": 0.38933665782075877,
+      "learning_rate": 2.8433001107419715e-05,
+      "loss": 0.4268,
+      "step": 980
+    },
+    {
+      "epoch": 1.4663677130044843,
+      "grad_norm": 0.32831850566437004,
+      "learning_rate": 2.840531561461794e-05,
+      "loss": 0.4052,
+      "step": 981
+    },
+    {
+      "epoch": 1.4678624813153962,
+      "grad_norm": 0.3828631853242714,
+      "learning_rate": 2.8377630121816168e-05,
+      "loss": 0.4266,
+      "step": 982
+    },
+    {
+      "epoch": 1.4693572496263079,
+      "grad_norm": 0.3174739301414559,
+      "learning_rate": 2.8349944629014396e-05,
+      "loss": 0.4295,
+      "step": 983
+    },
+    {
+      "epoch": 1.4708520179372198,
+      "grad_norm": 0.34719746467317214,
+      "learning_rate": 2.8322259136212625e-05,
+      "loss": 0.4352,
+      "step": 984
+    },
+    {
+      "epoch": 1.4723467862481314,
+      "grad_norm": 0.3022106112798685,
+      "learning_rate": 2.8294573643410853e-05,
+      "loss": 0.4176,
+      "step": 985
+    },
+    {
+      "epoch": 1.4738415545590433,
+      "grad_norm": 0.3863641083050957,
+      "learning_rate": 2.8266888150609084e-05,
+      "loss": 0.4264,
+      "step": 986
+    },
+    {
+      "epoch": 1.4753363228699552,
+      "grad_norm": 0.3338388578843077,
+      "learning_rate": 2.8239202657807313e-05,
+      "loss": 0.4112,
+      "step": 987
+    },
+    {
+      "epoch": 1.476831091180867,
+      "grad_norm": 0.4000708547241317,
+      "learning_rate": 2.821151716500554e-05,
+      "loss": 0.4428,
+      "step": 988
+    },
+    {
+      "epoch": 1.4783258594917787,
+      "grad_norm": 0.32823215831453095,
+      "learning_rate": 2.8183831672203766e-05,
+      "loss": 0.4413,
+      "step": 989
+    },
+    {
+      "epoch": 1.4798206278026906,
+      "grad_norm": 0.477317931347216,
+      "learning_rate": 2.8156146179401994e-05,
+      "loss": 0.4372,
+      "step": 990
+    },
+    {
+      "epoch": 1.4813153961136023,
+      "grad_norm": 0.30248866571055927,
+      "learning_rate": 2.8128460686600222e-05,
+      "loss": 0.4147,
+      "step": 991
+    },
+    {
+      "epoch": 1.4828101644245142,
+      "grad_norm": 0.358937840619591,
+      "learning_rate": 2.810077519379845e-05,
+      "loss": 0.4411,
+      "step": 992
+    },
+    {
+      "epoch": 1.484304932735426,
+      "grad_norm": 0.34778872510141035,
+      "learning_rate": 2.807308970099668e-05,
+      "loss": 0.4187,
+      "step": 993
+    },
+    {
+      "epoch": 1.4857997010463377,
+      "grad_norm": 0.30838627455316275,
+      "learning_rate": 2.804540420819491e-05,
+      "loss": 0.435,
+      "step": 994
+    },
+    {
+      "epoch": 1.4872944693572496,
+      "grad_norm": 0.3571237692287377,
+      "learning_rate": 2.8017718715393138e-05,
+      "loss": 0.4333,
+      "step": 995
+    },
+    {
+      "epoch": 1.4887892376681615,
+      "grad_norm": 0.3407080549211769,
+      "learning_rate": 2.799003322259136e-05,
+      "loss": 0.4241,
+      "step": 996
+    },
+    {
+      "epoch": 1.4902840059790732,
+      "grad_norm": 0.3376213729500949,
+      "learning_rate": 2.796234772978959e-05,
+      "loss": 0.4207,
+      "step": 997
+    },
+    {
+      "epoch": 1.491778774289985,
+      "grad_norm": 0.3078034082809607,
+      "learning_rate": 2.793466223698782e-05,
+      "loss": 0.4186,
+      "step": 998
+    },
+    {
+      "epoch": 1.493273542600897,
+      "grad_norm": 0.37094079152135573,
+      "learning_rate": 2.7906976744186048e-05,
+      "loss": 0.4339,
+      "step": 999
+    },
+    {
+      "epoch": 1.4947683109118086,
+      "grad_norm": 0.2823545901659568,
+      "learning_rate": 2.7879291251384276e-05,
+      "loss": 0.4313,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4962630792227205,
+      "grad_norm": 0.3348261709672509,
+      "learning_rate": 2.7851605758582504e-05,
+      "loss": 0.4201,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4977578475336322,
+      "grad_norm": 0.3116502468676231,
+      "learning_rate": 2.7823920265780736e-05,
+      "loss": 0.4109,
+      "step": 1002
+    },
+    {
+      "epoch": 1.499252615844544,
+      "grad_norm": 0.2914035964669752,
+      "learning_rate": 2.7796234772978964e-05,
+      "loss": 0.4276,
+      "step": 1003
+    },
+    {
+      "epoch": 1.500747384155456,
+      "grad_norm": 0.320348847590316,
+      "learning_rate": 2.7768549280177185e-05,
+      "loss": 0.4236,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5022421524663678,
+      "grad_norm": 0.32577754216814914,
+      "learning_rate": 2.7740863787375417e-05,
+      "loss": 0.4239,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5037369207772795,
+      "grad_norm": 0.30729403566716756,
+      "learning_rate": 2.7713178294573645e-05,
+      "loss": 0.4456,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5052316890881914,
+      "grad_norm": 0.35787332917840387,
+      "learning_rate": 2.7685492801771873e-05,
+      "loss": 0.4264,
+      "step": 1007
+    },
+    {
+      "epoch": 1.506726457399103,
+      "grad_norm": 0.2991010071319909,
+      "learning_rate": 2.76578073089701e-05,
+      "loss": 0.4434,
+      "step": 1008
+    },
+    {
+      "epoch": 1.508221225710015,
+      "grad_norm": 0.39901236157704184,
+      "learning_rate": 2.763012181616833e-05,
+      "loss": 0.4297,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5097159940209268,
+      "grad_norm": 0.38043507498640694,
+      "learning_rate": 2.760243632336656e-05,
+      "loss": 0.4009,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5112107623318387,
+      "grad_norm": 0.318508719327465,
+      "learning_rate": 2.7574750830564782e-05,
+      "loss": 0.4363,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5127055306427504,
+      "grad_norm": 0.3813702342496604,
+      "learning_rate": 2.754706533776301e-05,
+      "loss": 0.4384,
+      "step": 1012
+    },
+    {
+      "epoch": 1.514200298953662,
+      "grad_norm": 0.35735782688976764,
+      "learning_rate": 2.751937984496124e-05,
+      "loss": 0.4298,
+      "step": 1013
+    },
+    {
+      "epoch": 1.515695067264574,
+      "grad_norm": 0.28663813052916287,
+      "learning_rate": 2.749169435215947e-05,
+      "loss": 0.4073,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5171898355754858,
+      "grad_norm": 0.3545257883882148,
+      "learning_rate": 2.74640088593577e-05,
+      "loss": 0.4337,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5186846038863977,
+      "grad_norm": 0.3130304363908123,
+      "learning_rate": 2.7436323366555927e-05,
+      "loss": 0.4377,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5201793721973094,
+      "grad_norm": 0.3522611968851952,
+      "learning_rate": 2.7408637873754155e-05,
+      "loss": 0.4481,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5216741405082213,
+      "grad_norm": 0.31576978374238635,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 0.4329,
+      "step": 1018
+    },
+    {
+      "epoch": 1.523168908819133,
+      "grad_norm": 0.3277890417342227,
+      "learning_rate": 2.7353266888150608e-05,
+      "loss": 0.4271,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5246636771300448,
+      "grad_norm": 0.35989365539892754,
+      "learning_rate": 2.7325581395348836e-05,
+      "loss": 0.4238,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5261584454409567,
+      "grad_norm": 0.3584456607584094,
+      "learning_rate": 2.7297895902547064e-05,
+      "loss": 0.4226,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5276532137518686,
+      "grad_norm": 0.32947962829992744,
+      "learning_rate": 2.7270210409745296e-05,
+      "loss": 0.438,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5291479820627802,
+      "grad_norm": 0.3571495858042232,
+      "learning_rate": 2.7242524916943524e-05,
+      "loss": 0.4089,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5306427503736921,
+      "grad_norm": 0.3844571592044258,
+      "learning_rate": 2.7214839424141752e-05,
+      "loss": 0.44,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5321375186846038,
+      "grad_norm": 0.3140287632134162,
+      "learning_rate": 2.718715393133998e-05,
+      "loss": 0.4371,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5336322869955157,
+      "grad_norm": 0.39415238812880843,
+      "learning_rate": 2.7159468438538205e-05,
+      "loss": 0.4457,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5351270553064276,
+      "grad_norm": 0.3332181795527677,
+      "learning_rate": 2.7131782945736434e-05,
+      "loss": 0.4243,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5366218236173395,
+      "grad_norm": 0.2897413682099655,
+      "learning_rate": 2.7104097452934662e-05,
+      "loss": 0.4258,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5381165919282511,
+      "grad_norm": 0.3512391894011344,
+      "learning_rate": 2.707641196013289e-05,
+      "loss": 0.4374,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5396113602391628,
+      "grad_norm": 0.344665413037165,
+      "learning_rate": 2.704872646733112e-05,
+      "loss": 0.4332,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5411061285500747,
+      "grad_norm": 0.29124575584348916,
+      "learning_rate": 2.702104097452935e-05,
+      "loss": 0.4303,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5426008968609866,
+      "grad_norm": 0.3852323369831022,
+      "learning_rate": 2.6993355481727578e-05,
+      "loss": 0.4439,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5440956651718984,
+      "grad_norm": 0.2947794379544214,
+      "learning_rate": 2.6965669988925806e-05,
+      "loss": 0.4185,
+      "step": 1033
+    },
+    {
+      "epoch": 1.54559043348281,
+      "grad_norm": 0.34017260743285,
+      "learning_rate": 2.693798449612403e-05,
+      "loss": 0.4199,
+      "step": 1034
+    },
+    {
+      "epoch": 1.547085201793722,
+      "grad_norm": 0.32609932783064904,
+      "learning_rate": 2.691029900332226e-05,
+      "loss": 0.4322,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5485799701046337,
+      "grad_norm": 0.34064846070160065,
+      "learning_rate": 2.6882613510520487e-05,
+      "loss": 0.419,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5500747384155455,
+      "grad_norm": 0.30836427405727285,
+      "learning_rate": 2.6854928017718716e-05,
+      "loss": 0.4092,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5515695067264574,
+      "grad_norm": 0.3389936335441874,
+      "learning_rate": 2.6827242524916947e-05,
+      "loss": 0.4178,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5530642750373693,
+      "grad_norm": 0.3196825816813859,
+      "learning_rate": 2.6799557032115175e-05,
+      "loss": 0.4336,
+      "step": 1039
+    },
+    {
+      "epoch": 1.554559043348281,
+      "grad_norm": 0.425404374655059,
+      "learning_rate": 2.6771871539313404e-05,
+      "loss": 0.4248,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5560538116591929,
+      "grad_norm": 0.3635997013008305,
+      "learning_rate": 2.674418604651163e-05,
+      "loss": 0.4308,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5575485799701045,
+      "grad_norm": 0.3101335342767589,
+      "learning_rate": 2.6716500553709857e-05,
+      "loss": 0.4347,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5590433482810164,
+      "grad_norm": 0.3182909311491296,
+      "learning_rate": 2.6688815060908085e-05,
+      "loss": 0.4226,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5605381165919283,
+      "grad_norm": 0.29522489266008384,
+      "learning_rate": 2.6661129568106313e-05,
+      "loss": 0.4286,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5620328849028402,
+      "grad_norm": 0.30293478439662797,
+      "learning_rate": 2.663344407530454e-05,
+      "loss": 0.4347,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5635276532137519,
+      "grad_norm": 0.31616961676580746,
+      "learning_rate": 2.6605758582502773e-05,
+      "loss": 0.4592,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5650224215246635,
+      "grad_norm": 0.3132858474298203,
+      "learning_rate": 2.6578073089701e-05,
+      "loss": 0.4411,
+      "step": 1047
+    },
+    {
+      "epoch": 1.5665171898355754,
+      "grad_norm": 0.3388066237189719,
+      "learning_rate": 2.655038759689923e-05,
+      "loss": 0.4386,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5680119581464873,
+      "grad_norm": 0.39739835484657743,
+      "learning_rate": 2.652270210409745e-05,
+      "loss": 0.4668,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5695067264573992,
+      "grad_norm": 0.35672397343743495,
+      "learning_rate": 2.6495016611295682e-05,
+      "loss": 0.4342,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5710014947683109,
+      "grad_norm": 0.39012717065803054,
+      "learning_rate": 2.646733111849391e-05,
+      "loss": 0.4315,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5724962630792227,
+      "grad_norm": 0.333972341282277,
+      "learning_rate": 2.643964562569214e-05,
+      "loss": 0.4345,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5739910313901344,
+      "grad_norm": 0.43580764397456967,
+      "learning_rate": 2.6411960132890367e-05,
+      "loss": 0.4151,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5754857997010463,
+      "grad_norm": 0.31028145516595407,
+      "learning_rate": 2.6384274640088595e-05,
+      "loss": 0.4394,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5769805680119582,
+      "grad_norm": 0.4149457757197391,
+      "learning_rate": 2.6356589147286826e-05,
+      "loss": 0.4295,
+      "step": 1055
+    },
+    {
+      "epoch": 1.57847533632287,
+      "grad_norm": 0.31040302922089924,
+      "learning_rate": 2.6328903654485048e-05,
+      "loss": 0.4353,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5799701046337817,
+      "grad_norm": 0.36957905873573627,
+      "learning_rate": 2.6301218161683276e-05,
+      "loss": 0.4219,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5814648729446936,
+      "grad_norm": 0.3466980783920679,
+      "learning_rate": 2.6273532668881508e-05,
+      "loss": 0.4262,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5829596412556053,
+      "grad_norm": 0.3858762578186371,
+      "learning_rate": 2.6245847176079736e-05,
+      "loss": 0.4385,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5844544095665172,
+      "grad_norm": 0.30579576235538747,
+      "learning_rate": 2.6218161683277964e-05,
+      "loss": 0.4416,
+      "step": 1060
+    },
+    {
+      "epoch": 1.585949177877429,
+      "grad_norm": 0.4295296683355868,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.4504,
+      "step": 1061
+    },
+    {
+      "epoch": 1.587443946188341,
+      "grad_norm": 0.295834390778638,
+      "learning_rate": 2.616279069767442e-05,
+      "loss": 0.4373,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5889387144992526,
+      "grad_norm": 0.35843714224057005,
+      "learning_rate": 2.6135105204872645e-05,
+      "loss": 0.4318,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5904334828101643,
+      "grad_norm": 0.41490367538423706,
+      "learning_rate": 2.6107419712070873e-05,
+      "loss": 0.426,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5919282511210762,
+      "grad_norm": 0.4000423815230281,
+      "learning_rate": 2.60797342192691e-05,
+      "loss": 0.4225,
+      "step": 1065
+    },
+    {
+      "epoch": 1.593423019431988,
+      "grad_norm": 0.41982000336811176,
+      "learning_rate": 2.6052048726467333e-05,
+      "loss": 0.4421,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5949177877429,
+      "grad_norm": 0.40364204635630446,
+      "learning_rate": 2.602436323366556e-05,
+      "loss": 0.4536,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5964125560538116,
+      "grad_norm": 0.3675065836218711,
+      "learning_rate": 2.599667774086379e-05,
+      "loss": 0.4274,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5979073243647235,
+      "grad_norm": 0.38226766833932446,
+      "learning_rate": 2.5968992248062018e-05,
+      "loss": 0.4402,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5994020926756352,
+      "grad_norm": 0.42284152943957354,
+      "learning_rate": 2.5941306755260246e-05,
+      "loss": 0.4299,
+      "step": 1070
+    },
+    {
+      "epoch": 1.600896860986547,
+      "grad_norm": 0.3066139656141244,
+      "learning_rate": 2.591362126245847e-05,
+      "loss": 0.4462,
+      "step": 1071
+    },
+    {
+      "epoch": 1.602391629297459,
+      "grad_norm": 0.3904584875245674,
+      "learning_rate": 2.58859357696567e-05,
+      "loss": 0.4218,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6038863976083708,
+      "grad_norm": 0.41467046788938233,
+      "learning_rate": 2.5858250276854927e-05,
+      "loss": 0.4268,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6053811659192825,
+      "grad_norm": 0.362244162553987,
+      "learning_rate": 2.583056478405316e-05,
+      "loss": 0.4403,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6068759342301944,
+      "grad_norm": 0.333947081436804,
+      "learning_rate": 2.5802879291251387e-05,
+      "loss": 0.4365,
+      "step": 1075
+    },
+    {
+      "epoch": 1.608370702541106,
+      "grad_norm": 0.3686172809732321,
+      "learning_rate": 2.5775193798449615e-05,
+      "loss": 0.4475,
+      "step": 1076
+    },
+    {
+      "epoch": 1.609865470852018,
+      "grad_norm": 0.29278641379207865,
+      "learning_rate": 2.5747508305647843e-05,
+      "loss": 0.4029,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6113602391629298,
+      "grad_norm": 0.39004922316952334,
+      "learning_rate": 2.5719822812846068e-05,
+      "loss": 0.4366,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6128550074738417,
+      "grad_norm": 0.3264868508871384,
+      "learning_rate": 2.5692137320044296e-05,
+      "loss": 0.4375,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6143497757847534,
+      "grad_norm": 0.384654307737351,
+      "learning_rate": 2.5664451827242525e-05,
+      "loss": 0.4231,
+      "step": 1080
+    },
+    {
+      "epoch": 1.615844544095665,
+      "grad_norm": 0.32492943659512696,
+      "learning_rate": 2.5636766334440753e-05,
+      "loss": 0.4163,
+      "step": 1081
+    },
+    {
+      "epoch": 1.617339312406577,
+      "grad_norm": 0.320061599605898,
+      "learning_rate": 2.5609080841638984e-05,
+      "loss": 0.43,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6188340807174888,
+      "grad_norm": 0.3315180896426507,
+      "learning_rate": 2.5581395348837212e-05,
+      "loss": 0.4493,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6203288490284007,
+      "grad_norm": 0.3182455649721135,
+      "learning_rate": 2.555370985603544e-05,
+      "loss": 0.4245,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6218236173393124,
+      "grad_norm": 0.3084678650498547,
+      "learning_rate": 2.552602436323367e-05,
+      "loss": 0.4089,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6233183856502242,
+      "grad_norm": 0.3100148885463242,
+      "learning_rate": 2.5498338870431894e-05,
+      "loss": 0.4142,
+      "step": 1086
+    },
+    {
+      "epoch": 1.624813153961136,
+      "grad_norm": 0.36177789759921136,
+      "learning_rate": 2.5470653377630122e-05,
+      "loss": 0.4324,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6263079222720478,
+      "grad_norm": 0.3206245878874011,
+      "learning_rate": 2.544296788482835e-05,
+      "loss": 0.4343,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6278026905829597,
+      "grad_norm": 0.3054467589979163,
+      "learning_rate": 2.5415282392026578e-05,
+      "loss": 0.4217,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6292974588938716,
+      "grad_norm": 0.34620070290914584,
+      "learning_rate": 2.5387596899224806e-05,
+      "loss": 0.414,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6307922272047832,
+      "grad_norm": 0.33281280072061126,
+      "learning_rate": 2.5359911406423038e-05,
+      "loss": 0.4284,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6322869955156951,
+      "grad_norm": 0.3377121007619378,
+      "learning_rate": 2.5332225913621266e-05,
+      "loss": 0.4323,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6337817638266068,
+      "grad_norm": 0.3669591710046223,
+      "learning_rate": 2.5304540420819488e-05,
+      "loss": 0.434,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6352765321375187,
+      "grad_norm": 0.34167470954608214,
+      "learning_rate": 2.527685492801772e-05,
+      "loss": 0.4271,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6367713004484306,
+      "grad_norm": 0.31846837394535177,
+      "learning_rate": 2.5249169435215947e-05,
+      "loss": 0.4489,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6382660687593424,
+      "grad_norm": 0.2889694506496965,
+      "learning_rate": 2.5221483942414176e-05,
+      "loss": 0.4158,
+      "step": 1096
+    },
+    {
+      "epoch": 1.639760837070254,
+      "grad_norm": 0.32763759076436444,
+      "learning_rate": 2.5193798449612404e-05,
+      "loss": 0.4269,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6412556053811658,
+      "grad_norm": 0.29183904917918174,
+      "learning_rate": 2.5166112956810632e-05,
+      "loss": 0.431,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6427503736920777,
+      "grad_norm": 0.3880614300915995,
+      "learning_rate": 2.5138427464008864e-05,
+      "loss": 0.4426,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6442451420029895,
+      "grad_norm": 0.2734477995020025,
+      "learning_rate": 2.5110741971207092e-05,
+      "loss": 0.4126,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6457399103139014,
+      "grad_norm": 0.44142121685470137,
+      "learning_rate": 2.5083056478405313e-05,
+      "loss": 0.4424,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6472346786248133,
+      "grad_norm": 0.29074173394034253,
+      "learning_rate": 2.5055370985603545e-05,
+      "loss": 0.4258,
+      "step": 1102
+    },
+    {
+      "epoch": 1.648729446935725,
+      "grad_norm": 0.31472047775410694,
+      "learning_rate": 2.5027685492801773e-05,
+      "loss": 0.4136,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6502242152466366,
+      "grad_norm": 0.370197367069022,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4277,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6517189835575485,
+      "grad_norm": 0.31493941578759044,
+      "learning_rate": 2.497231450719823e-05,
+      "loss": 0.4421,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6532137518684604,
+      "grad_norm": 0.3271558787482336,
+      "learning_rate": 2.4944629014396458e-05,
+      "loss": 0.4398,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6547085201793723,
+      "grad_norm": 0.3568172461973672,
+      "learning_rate": 2.4916943521594686e-05,
+      "loss": 0.4451,
+      "step": 1107
+    },
+    {
+      "epoch": 1.656203288490284,
+      "grad_norm": 0.36105634577074575,
+      "learning_rate": 2.4889258028792914e-05,
+      "loss": 0.434,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6576980568011959,
+      "grad_norm": 0.35456333649955557,
+      "learning_rate": 2.4861572535991142e-05,
+      "loss": 0.4412,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6591928251121075,
+      "grad_norm": 0.3268350191108491,
+      "learning_rate": 2.483388704318937e-05,
+      "loss": 0.4146,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6606875934230194,
+      "grad_norm": 0.3073565370687034,
+      "learning_rate": 2.48062015503876e-05,
+      "loss": 0.4271,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6621823617339313,
+      "grad_norm": 0.2903849572916388,
+      "learning_rate": 2.4778516057585827e-05,
+      "loss": 0.4261,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6636771300448432,
+      "grad_norm": 0.36107071992722434,
+      "learning_rate": 2.4750830564784055e-05,
+      "loss": 0.4327,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6651718983557549,
+      "grad_norm": 0.31287819684744766,
+      "learning_rate": 2.4723145071982283e-05,
+      "loss": 0.4179,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.31682577024392056,
+      "learning_rate": 2.469545957918051e-05,
+      "loss": 0.4337,
+      "step": 1115
+    },
+    {
+      "epoch": 1.6681614349775784,
+      "grad_norm": 0.36804662684346956,
+      "learning_rate": 2.466777408637874e-05,
+      "loss": 0.4298,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6696562032884903,
+      "grad_norm": 0.3074521017213525,
+      "learning_rate": 2.4640088593576968e-05,
+      "loss": 0.4267,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6711509715994022,
+      "grad_norm": 0.3009781430578043,
+      "learning_rate": 2.4612403100775196e-05,
+      "loss": 0.4476,
+      "step": 1118
+    },
+    {
+      "epoch": 1.672645739910314,
+      "grad_norm": 0.3487709967906315,
+      "learning_rate": 2.4584717607973424e-05,
+      "loss": 0.4425,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6741405082212257,
+      "grad_norm": 0.30511404189668917,
+      "learning_rate": 2.4557032115171652e-05,
+      "loss": 0.4281,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6756352765321374,
+      "grad_norm": 0.3107017340242318,
+      "learning_rate": 2.4529346622369877e-05,
+      "loss": 0.4368,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6771300448430493,
+      "grad_norm": 0.29543565567529306,
+      "learning_rate": 2.4501661129568105e-05,
+      "loss": 0.4311,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6786248131539612,
+      "grad_norm": 0.311379924092796,
+      "learning_rate": 2.4473975636766337e-05,
+      "loss": 0.4219,
+      "step": 1123
+    },
+    {
+      "epoch": 1.680119581464873,
+      "grad_norm": 0.30209440517653297,
+      "learning_rate": 2.4446290143964565e-05,
+      "loss": 0.4294,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6816143497757847,
+      "grad_norm": 0.2977153923330207,
+      "learning_rate": 2.441860465116279e-05,
+      "loss": 0.4293,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6831091180866966,
+      "grad_norm": 0.3207906320569671,
+      "learning_rate": 2.4390919158361018e-05,
+      "loss": 0.4352,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6846038863976083,
+      "grad_norm": 0.27400115042086637,
+      "learning_rate": 2.436323366555925e-05,
+      "loss": 0.4253,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6860986547085202,
+      "grad_norm": 0.2874164025920164,
+      "learning_rate": 2.4335548172757478e-05,
+      "loss": 0.4141,
+      "step": 1128
+    },
+    {
+      "epoch": 1.687593423019432,
+      "grad_norm": 0.3160405295108985,
+      "learning_rate": 2.4307862679955703e-05,
+      "loss": 0.4436,
+      "step": 1129
+    },
+    {
+      "epoch": 1.689088191330344,
+      "grad_norm": 0.2965410563107449,
+      "learning_rate": 2.428017718715393e-05,
+      "loss": 0.4329,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6905829596412556,
+      "grad_norm": 0.3846924693043762,
+      "learning_rate": 2.4252491694352162e-05,
+      "loss": 0.4272,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6920777279521673,
+      "grad_norm": 0.29394800299905866,
+      "learning_rate": 2.4224806201550387e-05,
+      "loss": 0.4316,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6935724962630792,
+      "grad_norm": 0.2759338023421843,
+      "learning_rate": 2.4197120708748615e-05,
+      "loss": 0.4265,
+      "step": 1133
+    },
+    {
+      "epoch": 1.695067264573991,
+      "grad_norm": 0.329901465577245,
+      "learning_rate": 2.4169435215946844e-05,
+      "loss": 0.4548,
+      "step": 1134
+    },
+    {
+      "epoch": 1.696562032884903,
+      "grad_norm": 0.29692244774402154,
+      "learning_rate": 2.4141749723145075e-05,
+      "loss": 0.4231,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6980568011958148,
+      "grad_norm": 0.3031089836739788,
+      "learning_rate": 2.41140642303433e-05,
+      "loss": 0.4287,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6995515695067265,
+      "grad_norm": 0.35952166843348937,
+      "learning_rate": 2.4086378737541528e-05,
+      "loss": 0.4366,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7010463378176381,
+      "grad_norm": 0.2741386243293415,
+      "learning_rate": 2.4058693244739756e-05,
+      "loss": 0.4216,
+      "step": 1138
+    },
+    {
+      "epoch": 1.70254110612855,
+      "grad_norm": 0.3119336977983111,
+      "learning_rate": 2.4031007751937988e-05,
+      "loss": 0.4277,
+      "step": 1139
+    },
+    {
+      "epoch": 1.704035874439462,
+      "grad_norm": 0.29503845385462135,
+      "learning_rate": 2.4003322259136213e-05,
+      "loss": 0.4084,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7055306427503738,
+      "grad_norm": 0.3089589730589533,
+      "learning_rate": 2.397563676633444e-05,
+      "loss": 0.419,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7070254110612855,
+      "grad_norm": 0.28066928580621053,
+      "learning_rate": 2.394795127353267e-05,
+      "loss": 0.4136,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7085201793721974,
+      "grad_norm": 0.3198109434911019,
+      "learning_rate": 2.39202657807309e-05,
+      "loss": 0.4403,
+      "step": 1143
+    },
+    {
+      "epoch": 1.710014947683109,
+      "grad_norm": 0.2818880877890525,
+      "learning_rate": 2.3892580287929126e-05,
+      "loss": 0.4366,
+      "step": 1144
+    },
+    {
+      "epoch": 1.711509715994021,
+      "grad_norm": 0.2942432680297679,
+      "learning_rate": 2.3864894795127354e-05,
+      "loss": 0.4358,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7130044843049328,
+      "grad_norm": 0.29063391359496177,
+      "learning_rate": 2.3837209302325582e-05,
+      "loss": 0.4205,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7144992526158447,
+      "grad_norm": 0.28730375368027755,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.4192,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7159940209267563,
+      "grad_norm": 0.31662292435060274,
+      "learning_rate": 2.378183831672204e-05,
+      "loss": 0.4499,
+      "step": 1148
+    },
+    {
+      "epoch": 1.717488789237668,
+      "grad_norm": 0.28538063434972494,
+      "learning_rate": 2.3754152823920267e-05,
+      "loss": 0.4288,
+      "step": 1149
+    },
+    {
+      "epoch": 1.71898355754858,
+      "grad_norm": 0.3286365814345653,
+      "learning_rate": 2.3726467331118495e-05,
+      "loss": 0.4473,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7204783258594918,
+      "grad_norm": 0.3325958192380245,
+      "learning_rate": 2.3698781838316723e-05,
+      "loss": 0.4484,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7219730941704037,
+      "grad_norm": 0.29891339942431494,
+      "learning_rate": 2.367109634551495e-05,
+      "loss": 0.4022,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7234678624813156,
+      "grad_norm": 0.33478025372896975,
+      "learning_rate": 2.364341085271318e-05,
+      "loss": 0.4449,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7249626307922272,
+      "grad_norm": 0.3285878002143216,
+      "learning_rate": 2.3615725359911408e-05,
+      "loss": 0.4321,
+      "step": 1154
+    },
+    {
+      "epoch": 1.726457399103139,
+      "grad_norm": 0.29454226332720507,
+      "learning_rate": 2.3588039867109636e-05,
+      "loss": 0.4349,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7279521674140508,
+      "grad_norm": 0.39910421520405714,
+      "learning_rate": 2.3560354374307864e-05,
+      "loss": 0.4379,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7294469357249627,
+      "grad_norm": 0.35337931913757187,
+      "learning_rate": 2.3532668881506092e-05,
+      "loss": 0.4684,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7309417040358746,
+      "grad_norm": 0.32540147884105636,
+      "learning_rate": 2.350498338870432e-05,
+      "loss": 0.4419,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7324364723467862,
+      "grad_norm": 0.40213980984289593,
+      "learning_rate": 2.347729789590255e-05,
+      "loss": 0.4321,
+      "step": 1159
+    },
+    {
+      "epoch": 1.733931240657698,
+      "grad_norm": 0.29990401776551434,
+      "learning_rate": 2.3449612403100777e-05,
+      "loss": 0.411,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7354260089686098,
+      "grad_norm": 0.298880149696832,
+      "learning_rate": 2.3421926910299005e-05,
+      "loss": 0.4267,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7369207772795217,
+      "grad_norm": 0.3552965126585167,
+      "learning_rate": 2.339424141749723e-05,
+      "loss": 0.4274,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7384155455904335,
+      "grad_norm": 0.3338784523856532,
+      "learning_rate": 2.336655592469546e-05,
+      "loss": 0.4259,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7399103139013454,
+      "grad_norm": 0.3199521025736539,
+      "learning_rate": 2.333887043189369e-05,
+      "loss": 0.4286,
+      "step": 1164
+    },
+    {
+      "epoch": 1.741405082212257,
+      "grad_norm": 0.3786155269330912,
+      "learning_rate": 2.3311184939091918e-05,
+      "loss": 0.4116,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7428998505231688,
+      "grad_norm": 0.29959930559904296,
+      "learning_rate": 2.3283499446290142e-05,
+      "loss": 0.437,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7443946188340806,
+      "grad_norm": 0.3881026669704243,
+      "learning_rate": 2.3255813953488374e-05,
+      "loss": 0.4365,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7458893871449925,
+      "grad_norm": 0.3456531235492978,
+      "learning_rate": 2.3228128460686602e-05,
+      "loss": 0.4303,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7473841554559044,
+      "grad_norm": 0.4281429356925566,
+      "learning_rate": 2.320044296788483e-05,
+      "loss": 0.4289,
+      "step": 1169
+    },
+    {
+      "epoch": 1.7488789237668163,
+      "grad_norm": 0.3221768651418989,
+      "learning_rate": 2.3172757475083055e-05,
+      "loss": 0.4389,
+      "step": 1170
+    },
+    {
+      "epoch": 1.750373692077728,
+      "grad_norm": 0.4011254368078717,
+      "learning_rate": 2.3145071982281287e-05,
+      "loss": 0.4397,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7518684603886396,
+      "grad_norm": 0.3340054847772338,
+      "learning_rate": 2.3117386489479515e-05,
+      "loss": 0.441,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7533632286995515,
+      "grad_norm": 0.28393444400842294,
+      "learning_rate": 2.3089700996677743e-05,
+      "loss": 0.4318,
+      "step": 1173
+    },
+    {
+      "epoch": 1.7548579970104634,
+      "grad_norm": 0.33722516154130483,
+      "learning_rate": 2.3062015503875968e-05,
+      "loss": 0.4251,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7563527653213753,
+      "grad_norm": 0.30270581114898437,
+      "learning_rate": 2.30343300110742e-05,
+      "loss": 0.4168,
+      "step": 1175
+    },
+    {
+      "epoch": 1.757847533632287,
+      "grad_norm": 0.3386535681015517,
+      "learning_rate": 2.3006644518272428e-05,
+      "loss": 0.4325,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7593423019431988,
+      "grad_norm": 0.2851642141210693,
+      "learning_rate": 2.2978959025470653e-05,
+      "loss": 0.4271,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7608370702541105,
+      "grad_norm": 0.3069853006387183,
+      "learning_rate": 2.295127353266888e-05,
+      "loss": 0.4402,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7623318385650224,
+      "grad_norm": 0.3089549340717532,
+      "learning_rate": 2.2923588039867112e-05,
+      "loss": 0.4241,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7638266068759343,
+      "grad_norm": 0.3118210371158135,
+      "learning_rate": 2.289590254706534e-05,
+      "loss": 0.4328,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7653213751868462,
+      "grad_norm": 0.2854057736018241,
+      "learning_rate": 2.2868217054263565e-05,
+      "loss": 0.4226,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7668161434977578,
+      "grad_norm": 0.2893688015499784,
+      "learning_rate": 2.2840531561461794e-05,
+      "loss": 0.4126,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7683109118086695,
+      "grad_norm": 0.2794993770036586,
+      "learning_rate": 2.2812846068660025e-05,
+      "loss": 0.4362,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7698056801195814,
+      "grad_norm": 0.32037676504407353,
+      "learning_rate": 2.2785160575858253e-05,
+      "loss": 0.432,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7713004484304933,
+      "grad_norm": 0.29938457631426246,
+      "learning_rate": 2.2757475083056478e-05,
+      "loss": 0.4348,
+      "step": 1185
+    },
+    {
+      "epoch": 1.7727952167414052,
+      "grad_norm": 0.27291840530180184,
+      "learning_rate": 2.2729789590254706e-05,
+      "loss": 0.435,
+      "step": 1186
+    },
+    {
+      "epoch": 1.774289985052317,
+      "grad_norm": 0.26280025151931347,
+      "learning_rate": 2.2702104097452938e-05,
+      "loss": 0.4139,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7757847533632287,
+      "grad_norm": 0.28678648238991705,
+      "learning_rate": 2.2674418604651163e-05,
+      "loss": 0.4366,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7772795216741404,
+      "grad_norm": 0.2745450256376607,
+      "learning_rate": 2.264673311184939e-05,
+      "loss": 0.4026,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7787742899850523,
+      "grad_norm": 0.2795147321724181,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 0.4364,
+      "step": 1190
+    },
+    {
+      "epoch": 1.7802690582959642,
+      "grad_norm": 0.2744558946093191,
+      "learning_rate": 2.259136212624585e-05,
+      "loss": 0.4236,
+      "step": 1191
+    },
+    {
+      "epoch": 1.781763826606876,
+      "grad_norm": 0.33799855298589,
+      "learning_rate": 2.2563676633444076e-05,
+      "loss": 0.4344,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7832585949177877,
+      "grad_norm": 0.3014745840738359,
+      "learning_rate": 2.2535991140642304e-05,
+      "loss": 0.4071,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7847533632286996,
+      "grad_norm": 0.28222329939865787,
+      "learning_rate": 2.2508305647840532e-05,
+      "loss": 0.4375,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7862481315396113,
+      "grad_norm": 0.3135789712583663,
+      "learning_rate": 2.2480620155038764e-05,
+      "loss": 0.4409,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7877428998505231,
+      "grad_norm": 0.3153935942065999,
+      "learning_rate": 2.245293466223699e-05,
+      "loss": 0.4276,
+      "step": 1196
+    },
+    {
+      "epoch": 1.789237668161435,
+      "grad_norm": 0.3138018336375337,
+      "learning_rate": 2.2425249169435217e-05,
+      "loss": 0.4163,
+      "step": 1197
+    },
+    {
+      "epoch": 1.790732436472347,
+      "grad_norm": 0.3441305308254253,
+      "learning_rate": 2.2397563676633445e-05,
+      "loss": 0.4206,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7922272047832586,
+      "grad_norm": 0.3001718015311762,
+      "learning_rate": 2.2369878183831673e-05,
+      "loss": 0.4238,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7937219730941703,
+      "grad_norm": 0.30134590980598075,
+      "learning_rate": 2.23421926910299e-05,
+      "loss": 0.4241,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7952167414050821,
+      "grad_norm": 0.331456695073465,
+      "learning_rate": 2.231450719822813e-05,
+      "loss": 0.4434,
+      "step": 1201
+    },
+    {
+      "epoch": 1.796711509715994,
+      "grad_norm": 0.3246399429951608,
+      "learning_rate": 2.2286821705426357e-05,
+      "loss": 0.4452,
+      "step": 1202
+    },
+    {
+      "epoch": 1.798206278026906,
+      "grad_norm": 0.3182155758301577,
+      "learning_rate": 2.2259136212624586e-05,
+      "loss": 0.4396,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7997010463378178,
+      "grad_norm": 0.33849866278458374,
+      "learning_rate": 2.2231450719822814e-05,
+      "loss": 0.423,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8011958146487295,
+      "grad_norm": 0.26874380572922596,
+      "learning_rate": 2.2203765227021042e-05,
+      "loss": 0.4236,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8026905829596411,
+      "grad_norm": 0.3104927895787378,
+      "learning_rate": 2.217607973421927e-05,
+      "loss": 0.426,
+      "step": 1206
+    },
+    {
+      "epoch": 1.804185351270553,
+      "grad_norm": 0.36514197391832903,
+      "learning_rate": 2.21483942414175e-05,
+      "loss": 0.4192,
+      "step": 1207
+    },
+    {
+      "epoch": 1.805680119581465,
+      "grad_norm": 0.2824421579410868,
+      "learning_rate": 2.2120708748615727e-05,
+      "loss": 0.4444,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8071748878923768,
+      "grad_norm": 0.31676336066705496,
+      "learning_rate": 2.2093023255813955e-05,
+      "loss": 0.4266,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8086696562032885,
+      "grad_norm": 0.2850025006279979,
+      "learning_rate": 2.2065337763012183e-05,
+      "loss": 0.438,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8101644245142003,
+      "grad_norm": 0.26490152900738845,
+      "learning_rate": 2.203765227021041e-05,
+      "loss": 0.4206,
+      "step": 1211
+    },
+    {
+      "epoch": 1.811659192825112,
+      "grad_norm": 0.32409404838649036,
+      "learning_rate": 2.200996677740864e-05,
+      "loss": 0.4099,
+      "step": 1212
+    },
+    {
+      "epoch": 1.813153961136024,
+      "grad_norm": 0.33138595157540335,
+      "learning_rate": 2.1982281284606868e-05,
+      "loss": 0.4172,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8146487294469358,
+      "grad_norm": 0.2845447930808825,
+      "learning_rate": 2.1954595791805096e-05,
+      "loss": 0.4074,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8161434977578477,
+      "grad_norm": 0.32135050850782043,
+      "learning_rate": 2.1926910299003324e-05,
+      "loss": 0.4237,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8176382660687593,
+      "grad_norm": 0.3143199638725208,
+      "learning_rate": 2.1899224806201552e-05,
+      "loss": 0.4224,
+      "step": 1216
+    },
+    {
+      "epoch": 1.819133034379671,
+      "grad_norm": 0.3325267320606801,
+      "learning_rate": 2.187153931339978e-05,
+      "loss": 0.4288,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8206278026905829,
+      "grad_norm": 0.30221017165816416,
+      "learning_rate": 2.1843853820598005e-05,
+      "loss": 0.4387,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8221225710014948,
+      "grad_norm": 0.2971684827555225,
+      "learning_rate": 2.1816168327796237e-05,
+      "loss": 0.4395,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8236173393124067,
+      "grad_norm": 0.3121042323174605,
+      "learning_rate": 2.1788482834994465e-05,
+      "loss": 0.448,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8251121076233185,
+      "grad_norm": 0.3404827123916544,
+      "learning_rate": 2.1760797342192693e-05,
+      "loss": 0.4376,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8266068759342302,
+      "grad_norm": 0.29340055096668693,
+      "learning_rate": 2.1733111849390918e-05,
+      "loss": 0.404,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8281016442451419,
+      "grad_norm": 0.3129131290696851,
+      "learning_rate": 2.170542635658915e-05,
+      "loss": 0.4202,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8295964125560538,
+      "grad_norm": 0.29753471071962206,
+      "learning_rate": 2.1677740863787378e-05,
+      "loss": 0.4108,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8310911808669657,
+      "grad_norm": 0.28125184427188965,
+      "learning_rate": 2.1650055370985606e-05,
+      "loss": 0.4067,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8325859491778775,
+      "grad_norm": 0.2835268244226538,
+      "learning_rate": 2.162236987818383e-05,
+      "loss": 0.4331,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8340807174887892,
+      "grad_norm": 0.3110004073058745,
+      "learning_rate": 2.1594684385382062e-05,
+      "loss": 0.4114,
+      "step": 1227
+    },
+    {
+      "epoch": 1.835575485799701,
+      "grad_norm": 0.29574566739251057,
+      "learning_rate": 2.156699889258029e-05,
+      "loss": 0.4331,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8370702541106128,
+      "grad_norm": 0.2771685369731435,
+      "learning_rate": 2.153931339977852e-05,
+      "loss": 0.4322,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8385650224215246,
+      "grad_norm": 0.32338557078619995,
+      "learning_rate": 2.1511627906976744e-05,
+      "loss": 0.4211,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8400597907324365,
+      "grad_norm": 0.30071568491722983,
+      "learning_rate": 2.1483942414174972e-05,
+      "loss": 0.4218,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8415545590433484,
+      "grad_norm": 0.29444101233444975,
+      "learning_rate": 2.1456256921373203e-05,
+      "loss": 0.4219,
+      "step": 1232
+    },
+    {
+      "epoch": 1.84304932735426,
+      "grad_norm": 0.7720093468942147,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.4495,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8445440956651717,
+      "grad_norm": 0.3906078256316968,
+      "learning_rate": 2.1400885935769656e-05,
+      "loss": 0.4459,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8460388639760836,
+      "grad_norm": 0.29792937841705525,
+      "learning_rate": 2.1373200442967885e-05,
+      "loss": 0.4102,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8475336322869955,
+      "grad_norm": 0.4073709059150026,
+      "learning_rate": 2.1345514950166116e-05,
+      "loss": 0.4272,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8490284005979074,
+      "grad_norm": 0.30315240537097765,
+      "learning_rate": 2.131782945736434e-05,
+      "loss": 0.4154,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8505231689088193,
+      "grad_norm": 0.35564047374028507,
+      "learning_rate": 2.129014396456257e-05,
+      "loss": 0.4371,
+      "step": 1238
+    },
+    {
+      "epoch": 1.852017937219731,
+      "grad_norm": 0.35406795796552387,
+      "learning_rate": 2.1262458471760797e-05,
+      "loss": 0.4284,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8535127055306426,
+      "grad_norm": 0.29415030589124785,
+      "learning_rate": 2.123477297895903e-05,
+      "loss": 0.4096,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8550074738415545,
+      "grad_norm": 0.368921965325846,
+      "learning_rate": 2.1207087486157254e-05,
+      "loss": 0.424,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8565022421524664,
+      "grad_norm": 0.28113268855539814,
+      "learning_rate": 2.1179401993355482e-05,
+      "loss": 0.4281,
+      "step": 1242
+    },
+    {
+      "epoch": 1.8579970104633783,
+      "grad_norm": 0.3317445942236754,
+      "learning_rate": 2.115171650055371e-05,
+      "loss": 0.437,
+      "step": 1243
+    },
+    {
+      "epoch": 1.85949177877429,
+      "grad_norm": 0.3032302616454194,
+      "learning_rate": 2.1124031007751938e-05,
+      "loss": 0.4411,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8609865470852018,
+      "grad_norm": 0.3161297686335451,
+      "learning_rate": 2.1096345514950166e-05,
+      "loss": 0.4381,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8624813153961135,
+      "grad_norm": 0.2601926265625027,
+      "learning_rate": 2.1068660022148395e-05,
+      "loss": 0.4214,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8639760837070254,
+      "grad_norm": 0.3161871168272268,
+      "learning_rate": 2.1040974529346623e-05,
+      "loss": 0.4223,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8654708520179373,
+      "grad_norm": 0.27336388450097937,
+      "learning_rate": 2.101328903654485e-05,
+      "loss": 0.4344,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8669656203288492,
+      "grad_norm": 0.3078155334847862,
+      "learning_rate": 2.098560354374308e-05,
+      "loss": 0.421,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8684603886397608,
+      "grad_norm": 0.26977036789853037,
+      "learning_rate": 2.0957918050941307e-05,
+      "loss": 0.4329,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8699551569506725,
+      "grad_norm": 0.29376421247064605,
+      "learning_rate": 2.0930232558139536e-05,
+      "loss": 0.4274,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8714499252615844,
+      "grad_norm": 0.2438286855754908,
+      "learning_rate": 2.0902547065337764e-05,
+      "loss": 0.4277,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8729446935724963,
+      "grad_norm": 0.34749858655469,
+      "learning_rate": 2.0874861572535992e-05,
+      "loss": 0.437,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8744394618834082,
+      "grad_norm": 0.28647931866656207,
+      "learning_rate": 2.084717607973422e-05,
+      "loss": 0.4291,
+      "step": 1254
+    },
+    {
+      "epoch": 1.87593423019432,
+      "grad_norm": 0.2569153092957249,
+      "learning_rate": 2.081949058693245e-05,
+      "loss": 0.4345,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8774289985052317,
+      "grad_norm": 0.26778557039858464,
+      "learning_rate": 2.0791805094130677e-05,
+      "loss": 0.4307,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8789237668161434,
+      "grad_norm": 0.3378897650847328,
+      "learning_rate": 2.0764119601328905e-05,
+      "loss": 0.4336,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8804185351270553,
+      "grad_norm": 0.25638690169440376,
+      "learning_rate": 2.0736434108527133e-05,
+      "loss": 0.4192,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8819133034379671,
+      "grad_norm": 0.33463531451802997,
+      "learning_rate": 2.070874861572536e-05,
+      "loss": 0.4304,
+      "step": 1259
+    },
+    {
+      "epoch": 1.883408071748879,
+      "grad_norm": 0.2795291330178031,
+      "learning_rate": 2.068106312292359e-05,
+      "loss": 0.4196,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8849028400597907,
+      "grad_norm": 0.31501557962885773,
+      "learning_rate": 2.0653377630121818e-05,
+      "loss": 0.4406,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8863976083707026,
+      "grad_norm": 0.2900319288588721,
+      "learning_rate": 2.0625692137320046e-05,
+      "loss": 0.4332,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8878923766816142,
+      "grad_norm": 0.30300974700330935,
+      "learning_rate": 2.0598006644518274e-05,
+      "loss": 0.4167,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8893871449925261,
+      "grad_norm": 0.282281462358467,
+      "learning_rate": 2.0570321151716502e-05,
+      "loss": 0.4398,
+      "step": 1264
+    },
+    {
+      "epoch": 1.890881913303438,
+      "grad_norm": 0.286932805031066,
+      "learning_rate": 2.054263565891473e-05,
+      "loss": 0.4524,
+      "step": 1265
+    },
+    {
+      "epoch": 1.89237668161435,
+      "grad_norm": 0.35105236866709255,
+      "learning_rate": 2.051495016611296e-05,
+      "loss": 0.437,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8938714499252616,
+      "grad_norm": 0.25378042688866215,
+      "learning_rate": 2.0487264673311183e-05,
+      "loss": 0.4128,
+      "step": 1267
+    },
+    {
+      "epoch": 1.8953662182361732,
+      "grad_norm": 0.2923469919996209,
+      "learning_rate": 2.0459579180509415e-05,
+      "loss": 0.4265,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8968609865470851,
+      "grad_norm": 0.29794325797991983,
+      "learning_rate": 2.0431893687707643e-05,
+      "loss": 0.4267,
+      "step": 1269
+    },
+    {
+      "epoch": 1.898355754857997,
+      "grad_norm": 0.31545129259823007,
+      "learning_rate": 2.040420819490587e-05,
+      "loss": 0.4263,
+      "step": 1270
+    },
+    {
+      "epoch": 1.899850523168909,
+      "grad_norm": 0.37460777553401664,
+      "learning_rate": 2.0376522702104096e-05,
+      "loss": 0.4281,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9013452914798208,
+      "grad_norm": 0.273962588180347,
+      "learning_rate": 2.0348837209302328e-05,
+      "loss": 0.4268,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9028400597907325,
+      "grad_norm": 0.33395467331659856,
+      "learning_rate": 2.0321151716500556e-05,
+      "loss": 0.4258,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9043348281016441,
+      "grad_norm": 0.34804181603672746,
+      "learning_rate": 2.029346622369878e-05,
+      "loss": 0.4201,
+      "step": 1274
+    },
+    {
+      "epoch": 1.905829596412556,
+      "grad_norm": 0.398655451869521,
+      "learning_rate": 2.026578073089701e-05,
+      "loss": 0.4306,
+      "step": 1275
+    },
+    {
+      "epoch": 1.907324364723468,
+      "grad_norm": 0.34780619907912347,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 0.4343,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9088191330343798,
+      "grad_norm": 0.31107320317526965,
+      "learning_rate": 2.021040974529347e-05,
+      "loss": 0.4105,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9103139013452914,
+      "grad_norm": 0.33955584765219027,
+      "learning_rate": 2.0182724252491694e-05,
+      "loss": 0.4544,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9118086696562033,
+      "grad_norm": 0.2954837926947831,
+      "learning_rate": 2.0155038759689922e-05,
+      "loss": 0.4324,
+      "step": 1279
+    },
+    {
+      "epoch": 1.913303437967115,
+      "grad_norm": 0.27837453738607404,
+      "learning_rate": 2.0127353266888153e-05,
+      "loss": 0.4395,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9147982062780269,
+      "grad_norm": 0.3522244526153378,
+      "learning_rate": 2.009966777408638e-05,
+      "loss": 0.4278,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9162929745889388,
+      "grad_norm": 0.28667590690977696,
+      "learning_rate": 2.0071982281284606e-05,
+      "loss": 0.4341,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9177877428998507,
+      "grad_norm": 0.33644263844423106,
+      "learning_rate": 2.0044296788482834e-05,
+      "loss": 0.4021,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9192825112107623,
+      "grad_norm": 0.3222948807838148,
+      "learning_rate": 2.0016611295681066e-05,
+      "loss": 0.4175,
+      "step": 1284
+    },
+    {
+      "epoch": 1.920777279521674,
+      "grad_norm": 0.2837932977925209,
+      "learning_rate": 1.998892580287929e-05,
+      "loss": 0.4193,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9222720478325859,
+      "grad_norm": 0.4111837016447267,
+      "learning_rate": 1.996124031007752e-05,
+      "loss": 0.4492,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9237668161434978,
+      "grad_norm": 0.2685275361392796,
+      "learning_rate": 1.9933554817275747e-05,
+      "loss": 0.4168,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9252615844544096,
+      "grad_norm": 0.4381873215915197,
+      "learning_rate": 1.990586932447398e-05,
+      "loss": 0.4272,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9267563527653215,
+      "grad_norm": 0.2886001323682502,
+      "learning_rate": 1.9878183831672204e-05,
+      "loss": 0.4333,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9282511210762332,
+      "grad_norm": 0.35060827263475736,
+      "learning_rate": 1.9850498338870432e-05,
+      "loss": 0.4297,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9297458893871449,
+      "grad_norm": 0.3606478325314803,
+      "learning_rate": 1.982281284606866e-05,
+      "loss": 0.4315,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9312406576980568,
+      "grad_norm": 0.33662587277942857,
+      "learning_rate": 1.979512735326689e-05,
+      "loss": 0.4021,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9327354260089686,
+      "grad_norm": 0.34881143476369697,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.4107,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9342301943198805,
+      "grad_norm": 0.32423781085949344,
+      "learning_rate": 1.9739756367663345e-05,
+      "loss": 0.4208,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9357249626307922,
+      "grad_norm": 0.31406802278105134,
+      "learning_rate": 1.9712070874861573e-05,
+      "loss": 0.4078,
+      "step": 1295
+    },
+    {
+      "epoch": 1.937219730941704,
+      "grad_norm": 0.3200538864191855,
+      "learning_rate": 1.9684385382059804e-05,
+      "loss": 0.4133,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9387144992526157,
+      "grad_norm": 0.3105935232737285,
+      "learning_rate": 1.965669988925803e-05,
+      "loss": 0.4273,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9402092675635276,
+      "grad_norm": 0.2991061877971967,
+      "learning_rate": 1.9629014396456257e-05,
+      "loss": 0.4391,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9417040358744395,
+      "grad_norm": 0.29168088041012424,
+      "learning_rate": 1.9601328903654486e-05,
+      "loss": 0.4335,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9431988041853514,
+      "grad_norm": 0.34775094932709094,
+      "learning_rate": 1.9573643410852714e-05,
+      "loss": 0.4189,
+      "step": 1300
+    },
+    {
+      "epoch": 1.944693572496263,
+      "grad_norm": 0.3033520892481037,
+      "learning_rate": 1.9545957918050942e-05,
+      "loss": 0.4316,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9461883408071747,
+      "grad_norm": 0.29763882794743257,
+      "learning_rate": 1.951827242524917e-05,
+      "loss": 0.4483,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9476831091180866,
+      "grad_norm": 0.3223954278660888,
+      "learning_rate": 1.94905869324474e-05,
+      "loss": 0.4321,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9491778774289985,
+      "grad_norm": 0.27855904480221716,
+      "learning_rate": 1.9462901439645627e-05,
+      "loss": 0.4196,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9506726457399104,
+      "grad_norm": 0.3340023556163215,
+      "learning_rate": 1.9435215946843855e-05,
+      "loss": 0.4255,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9521674140508223,
+      "grad_norm": 0.2956898828081276,
+      "learning_rate": 1.9407530454042083e-05,
+      "loss": 0.4143,
+      "step": 1306
+    },
+    {
+      "epoch": 1.953662182361734,
+      "grad_norm": 0.25944209331279766,
+      "learning_rate": 1.937984496124031e-05,
+      "loss": 0.4231,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9551569506726456,
+      "grad_norm": 0.32537545165946047,
+      "learning_rate": 1.935215946843854e-05,
+      "loss": 0.4255,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9566517189835575,
+      "grad_norm": 0.28758524749860553,
+      "learning_rate": 1.9324473975636768e-05,
+      "loss": 0.43,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9581464872944694,
+      "grad_norm": 0.3068636041787949,
+      "learning_rate": 1.9296788482834996e-05,
+      "loss": 0.4256,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9596412556053813,
+      "grad_norm": 0.3071216861845236,
+      "learning_rate": 1.9269102990033224e-05,
+      "loss": 0.4171,
+      "step": 1311
+    },
+    {
+      "epoch": 1.961136023916293,
+      "grad_norm": 0.32046760855604384,
+      "learning_rate": 1.9241417497231452e-05,
+      "loss": 0.44,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9626307922272048,
+      "grad_norm": 0.274769748155572,
+      "learning_rate": 1.921373200442968e-05,
+      "loss": 0.4522,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9641255605381165,
+      "grad_norm": 0.2833840267146538,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 0.4354,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9656203288490284,
+      "grad_norm": 0.28321401495521753,
+      "learning_rate": 1.9158361018826133e-05,
+      "loss": 0.4176,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9671150971599403,
+      "grad_norm": 0.29908741008902456,
+      "learning_rate": 1.9130675526024365e-05,
+      "loss": 0.4332,
+      "step": 1316
+    },
+    {
+      "epoch": 1.9686098654708521,
+      "grad_norm": 0.2678745895945535,
+      "learning_rate": 1.9102990033222593e-05,
+      "loss": 0.4339,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9701046337817638,
+      "grad_norm": 0.3596177140738104,
+      "learning_rate": 1.907530454042082e-05,
+      "loss": 0.4111,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9715994020926755,
+      "grad_norm": 0.30886672081603833,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.433,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9730941704035874,
+      "grad_norm": 0.3072596571714735,
+      "learning_rate": 1.9019933554817278e-05,
+      "loss": 0.4504,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9745889387144993,
+      "grad_norm": 0.3174280409010404,
+      "learning_rate": 1.8992248062015506e-05,
+      "loss": 0.4522,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9760837070254111,
+      "grad_norm": 0.3147819887538839,
+      "learning_rate": 1.8964562569213734e-05,
+      "loss": 0.4284,
+      "step": 1322
+    },
+    {
+      "epoch": 1.977578475336323,
+      "grad_norm": 0.30884330148447586,
+      "learning_rate": 1.893687707641196e-05,
+      "loss": 0.4041,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9790732436472347,
+      "grad_norm": 0.30260967083261414,
+      "learning_rate": 1.890919158361019e-05,
+      "loss": 0.4229,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9805680119581464,
+      "grad_norm": 0.3419573653240846,
+      "learning_rate": 1.888150609080842e-05,
+      "loss": 0.4173,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9820627802690582,
+      "grad_norm": 0.278423815410247,
+      "learning_rate": 1.8853820598006647e-05,
+      "loss": 0.417,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9835575485799701,
+      "grad_norm": 0.3009156022211277,
+      "learning_rate": 1.882613510520487e-05,
+      "loss": 0.4393,
+      "step": 1327
+    },
+    {
+      "epoch": 1.985052316890882,
+      "grad_norm": 0.31079221594589374,
+      "learning_rate": 1.8798449612403103e-05,
+      "loss": 0.4195,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9865470852017937,
+      "grad_norm": 0.36829460747941645,
+      "learning_rate": 1.877076411960133e-05,
+      "loss": 0.4483,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9880418535127056,
+      "grad_norm": 0.30228527470809224,
+      "learning_rate": 1.8743078626799556e-05,
+      "loss": 0.4264,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9895366218236172,
+      "grad_norm": 0.28410299133965794,
+      "learning_rate": 1.8715393133997784e-05,
+      "loss": 0.4038,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9910313901345291,
+      "grad_norm": 0.36966799291894603,
+      "learning_rate": 1.8687707641196016e-05,
+      "loss": 0.4239,
+      "step": 1332
+    },
+    {
+      "epoch": 1.992526158445441,
+      "grad_norm": 0.33987659610090876,
+      "learning_rate": 1.8660022148394244e-05,
+      "loss": 0.4436,
+      "step": 1333
+    },
+    {
+      "epoch": 1.994020926756353,
+      "grad_norm": 0.2765621205637022,
+      "learning_rate": 1.863233665559247e-05,
+      "loss": 0.4327,
+      "step": 1334
+    },
+    {
+      "epoch": 1.9955156950672646,
+      "grad_norm": 0.35626001258350964,
+      "learning_rate": 1.8604651162790697e-05,
+      "loss": 0.4263,
+      "step": 1335
+    },
+    {
+      "epoch": 1.9970104633781762,
+      "grad_norm": 0.3149869744229574,
+      "learning_rate": 1.857696566998893e-05,
+      "loss": 0.4364,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9985052316890881,
+      "grad_norm": 0.29692860656013026,
+      "learning_rate": 1.8549280177187157e-05,
+      "loss": 0.4223,
+      "step": 1337
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3397467995751796,
+      "learning_rate": 1.8521594684385382e-05,
+      "loss": 0.4199,
+      "step": 1338
+    },
+    {
+      "epoch": 2.001494768310912,
+      "grad_norm": 0.3674862971169319,
+      "learning_rate": 1.849390919158361e-05,
+      "loss": 0.3534,
+      "step": 1339
+    },
+    {
+      "epoch": 2.0029895366218238,
+      "grad_norm": 0.2902496843863615,
+      "learning_rate": 1.8466223698781838e-05,
+      "loss": 0.3593,
+      "step": 1340
+    },
+    {
+      "epoch": 2.004484304932735,
+      "grad_norm": 0.2862488595738825,
+      "learning_rate": 1.8438538205980066e-05,
+      "loss": 0.3138,
+      "step": 1341
+    },
+    {
+      "epoch": 2.005979073243647,
+      "grad_norm": 0.39705238585246155,
+      "learning_rate": 1.8410852713178295e-05,
+      "loss": 0.3445,
+      "step": 1342
+    },
+    {
+      "epoch": 2.007473841554559,
+      "grad_norm": 0.29082488871169365,
+      "learning_rate": 1.8383167220376523e-05,
+      "loss": 0.354,
+      "step": 1343
+    },
+    {
+      "epoch": 2.008968609865471,
+      "grad_norm": 0.33702937193520144,
+      "learning_rate": 1.835548172757475e-05,
+      "loss": 0.3474,
+      "step": 1344
+    },
+    {
+      "epoch": 2.0104633781763828,
+      "grad_norm": 0.3293935792554188,
+      "learning_rate": 1.832779623477298e-05,
+      "loss": 0.3464,
+      "step": 1345
+    },
+    {
+      "epoch": 2.0119581464872947,
+      "grad_norm": 0.2963277547449406,
+      "learning_rate": 1.8300110741971207e-05,
+      "loss": 0.3587,
+      "step": 1346
+    },
+    {
+      "epoch": 2.013452914798206,
+      "grad_norm": 0.31698243786758273,
+      "learning_rate": 1.8272425249169436e-05,
+      "loss": 0.3416,
+      "step": 1347
+    },
+    {
+      "epoch": 2.014947683109118,
+      "grad_norm": 0.35163627632269195,
+      "learning_rate": 1.8244739756367664e-05,
+      "loss": 0.3632,
+      "step": 1348
+    },
+    {
+      "epoch": 2.01644245142003,
+      "grad_norm": 0.270710120355098,
+      "learning_rate": 1.8217054263565892e-05,
+      "loss": 0.353,
+      "step": 1349
+    },
+    {
+      "epoch": 2.0179372197309418,
+      "grad_norm": 0.28040727617527056,
+      "learning_rate": 1.818936877076412e-05,
+      "loss": 0.3371,
+      "step": 1350
+    },
+    {
+      "epoch": 2.0194319880418536,
+      "grad_norm": 0.3204809542114785,
+      "learning_rate": 1.816168327796235e-05,
+      "loss": 0.3481,
+      "step": 1351
+    },
+    {
+      "epoch": 2.0209267563527655,
+      "grad_norm": 0.30005789054858845,
+      "learning_rate": 1.8133997785160577e-05,
+      "loss": 0.3352,
+      "step": 1352
+    },
+    {
+      "epoch": 2.022421524663677,
+      "grad_norm": 0.28400248698635155,
+      "learning_rate": 1.8106312292358805e-05,
+      "loss": 0.3433,
+      "step": 1353
+    },
+    {
+      "epoch": 2.023916292974589,
+      "grad_norm": 0.2759910803642707,
+      "learning_rate": 1.8078626799557033e-05,
+      "loss": 0.3353,
+      "step": 1354
+    },
+    {
+      "epoch": 2.0254110612855007,
+      "grad_norm": 0.327752947605851,
+      "learning_rate": 1.805094130675526e-05,
+      "loss": 0.353,
+      "step": 1355
+    },
+    {
+      "epoch": 2.0269058295964126,
+      "grad_norm": 0.34755046215803087,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 0.3575,
+      "step": 1356
+    },
+    {
+      "epoch": 2.0284005979073245,
+      "grad_norm": 0.27690127370770135,
+      "learning_rate": 1.7995570321151718e-05,
+      "loss": 0.3398,
+      "step": 1357
+    },
+    {
+      "epoch": 2.029895366218236,
+      "grad_norm": 0.3066530648255945,
+      "learning_rate": 1.7967884828349946e-05,
+      "loss": 0.3401,
+      "step": 1358
+    },
+    {
+      "epoch": 2.031390134529148,
+      "grad_norm": 0.29504977222265,
+      "learning_rate": 1.7940199335548174e-05,
+      "loss": 0.3431,
+      "step": 1359
+    },
+    {
+      "epoch": 2.0328849028400597,
+      "grad_norm": 0.3086144110464687,
+      "learning_rate": 1.7912513842746402e-05,
+      "loss": 0.3579,
+      "step": 1360
+    },
+    {
+      "epoch": 2.0343796711509716,
+      "grad_norm": 0.29209834757616826,
+      "learning_rate": 1.788482834994463e-05,
+      "loss": 0.3544,
+      "step": 1361
+    },
+    {
+      "epoch": 2.0358744394618835,
+      "grad_norm": 0.32473937453367985,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.3503,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0373692077727954,
+      "grad_norm": 0.2872110353329131,
+      "learning_rate": 1.7829457364341087e-05,
+      "loss": 0.3425,
+      "step": 1363
+    },
+    {
+      "epoch": 2.038863976083707,
+      "grad_norm": 0.31262209249317086,
+      "learning_rate": 1.7801771871539315e-05,
+      "loss": 0.3375,
+      "step": 1364
+    },
+    {
+      "epoch": 2.0403587443946187,
+      "grad_norm": 0.24911042354752044,
+      "learning_rate": 1.7774086378737543e-05,
+      "loss": 0.3444,
+      "step": 1365
+    },
+    {
+      "epoch": 2.0418535127055306,
+      "grad_norm": 0.2599412046320163,
+      "learning_rate": 1.774640088593577e-05,
+      "loss": 0.3406,
+      "step": 1366
+    },
+    {
+      "epoch": 2.0433482810164425,
+      "grad_norm": 0.2887859002076743,
+      "learning_rate": 1.7718715393134e-05,
+      "loss": 0.3465,
+      "step": 1367
+    },
+    {
+      "epoch": 2.0448430493273544,
+      "grad_norm": 0.2578169379882819,
+      "learning_rate": 1.7691029900332228e-05,
+      "loss": 0.3498,
+      "step": 1368
+    },
+    {
+      "epoch": 2.0463378176382663,
+      "grad_norm": 0.27827491435362073,
+      "learning_rate": 1.7663344407530456e-05,
+      "loss": 0.3498,
+      "step": 1369
+    },
+    {
+      "epoch": 2.0478325859491777,
+      "grad_norm": 0.2653946175090302,
+      "learning_rate": 1.7635658914728684e-05,
+      "loss": 0.3425,
+      "step": 1370
+    },
+    {
+      "epoch": 2.0493273542600896,
+      "grad_norm": 0.27259244083181394,
+      "learning_rate": 1.760797342192691e-05,
+      "loss": 0.3538,
+      "step": 1371
+    },
+    {
+      "epoch": 2.0508221225710015,
+      "grad_norm": 0.26354172940394305,
+      "learning_rate": 1.758028792912514e-05,
+      "loss": 0.3478,
+      "step": 1372
+    },
+    {
+      "epoch": 2.0523168908819134,
+      "grad_norm": 0.250362178604589,
+      "learning_rate": 1.755260243632337e-05,
+      "loss": 0.3282,
+      "step": 1373
+    },
+    {
+      "epoch": 2.0538116591928253,
+      "grad_norm": 0.28319952124417885,
+      "learning_rate": 1.7524916943521597e-05,
+      "loss": 0.3353,
+      "step": 1374
+    },
+    {
+      "epoch": 2.0553064275037367,
+      "grad_norm": 0.28520024894578044,
+      "learning_rate": 1.749723145071982e-05,
+      "loss": 0.3486,
+      "step": 1375
+    },
+    {
+      "epoch": 2.0568011958146486,
+      "grad_norm": 0.27340936154959455,
+      "learning_rate": 1.746954595791805e-05,
+      "loss": 0.3749,
+      "step": 1376
+    },
+    {
+      "epoch": 2.0582959641255605,
+      "grad_norm": 0.31638095987285925,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 0.3388,
+      "step": 1377
+    },
+    {
+      "epoch": 2.0597907324364724,
+      "grad_norm": 0.27850393964320486,
+      "learning_rate": 1.741417497231451e-05,
+      "loss": 0.3454,
+      "step": 1378
+    },
+    {
+      "epoch": 2.0612855007473843,
+      "grad_norm": 0.2682799836050719,
+      "learning_rate": 1.7386489479512734e-05,
+      "loss": 0.345,
+      "step": 1379
+    },
+    {
+      "epoch": 2.062780269058296,
+      "grad_norm": 0.30010193641170696,
+      "learning_rate": 1.7358803986710963e-05,
+      "loss": 0.3442,
+      "step": 1380
+    },
+    {
+      "epoch": 2.0642750373692076,
+      "grad_norm": 0.30146518473150224,
+      "learning_rate": 1.7331118493909194e-05,
+      "loss": 0.345,
+      "step": 1381
+    },
+    {
+      "epoch": 2.0657698056801195,
+      "grad_norm": 0.2692672568798684,
+      "learning_rate": 1.730343300110742e-05,
+      "loss": 0.3498,
+      "step": 1382
+    },
+    {
+      "epoch": 2.0672645739910314,
+      "grad_norm": 0.29171622337399805,
+      "learning_rate": 1.7275747508305647e-05,
+      "loss": 0.341,
+      "step": 1383
+    },
+    {
+      "epoch": 2.0687593423019432,
+      "grad_norm": 0.2559187627510791,
+      "learning_rate": 1.7248062015503875e-05,
+      "loss": 0.344,
+      "step": 1384
+    },
+    {
+      "epoch": 2.070254110612855,
+      "grad_norm": 0.2836996618203931,
+      "learning_rate": 1.7220376522702107e-05,
+      "loss": 0.3481,
+      "step": 1385
+    },
+    {
+      "epoch": 2.071748878923767,
+      "grad_norm": 0.2518439006172323,
+      "learning_rate": 1.7192691029900332e-05,
+      "loss": 0.3448,
+      "step": 1386
+    },
+    {
+      "epoch": 2.0732436472346785,
+      "grad_norm": 0.32796526614000937,
+      "learning_rate": 1.716500553709856e-05,
+      "loss": 0.3644,
+      "step": 1387
+    },
+    {
+      "epoch": 2.0747384155455904,
+      "grad_norm": 0.2506512687187237,
+      "learning_rate": 1.7137320044296788e-05,
+      "loss": 0.3294,
+      "step": 1388
+    },
+    {
+      "epoch": 2.0762331838565022,
+      "grad_norm": 0.2476923733130635,
+      "learning_rate": 1.710963455149502e-05,
+      "loss": 0.3414,
+      "step": 1389
+    },
+    {
+      "epoch": 2.077727952167414,
+      "grad_norm": 0.2698090876142066,
+      "learning_rate": 1.7081949058693245e-05,
+      "loss": 0.3485,
+      "step": 1390
+    },
+    {
+      "epoch": 2.079222720478326,
+      "grad_norm": 0.2614412929112908,
+      "learning_rate": 1.7054263565891473e-05,
+      "loss": 0.3599,
+      "step": 1391
+    },
+    {
+      "epoch": 2.0807174887892375,
+      "grad_norm": 0.2500109164163158,
+      "learning_rate": 1.70265780730897e-05,
+      "loss": 0.3435,
+      "step": 1392
+    },
+    {
+      "epoch": 2.0822122571001493,
+      "grad_norm": 0.2730725741164814,
+      "learning_rate": 1.6998892580287933e-05,
+      "loss": 0.3409,
+      "step": 1393
+    },
+    {
+      "epoch": 2.0837070254110612,
+      "grad_norm": 0.24796616506080948,
+      "learning_rate": 1.6971207087486157e-05,
+      "loss": 0.3477,
+      "step": 1394
+    },
+    {
+      "epoch": 2.085201793721973,
+      "grad_norm": 0.2574162789871206,
+      "learning_rate": 1.6943521594684386e-05,
+      "loss": 0.3428,
+      "step": 1395
+    },
+    {
+      "epoch": 2.086696562032885,
+      "grad_norm": 0.24099495696714113,
+      "learning_rate": 1.6915836101882614e-05,
+      "loss": 0.3493,
+      "step": 1396
+    },
+    {
+      "epoch": 2.088191330343797,
+      "grad_norm": 0.23669798879563633,
+      "learning_rate": 1.6888150609080842e-05,
+      "loss": 0.3263,
+      "step": 1397
+    },
+    {
+      "epoch": 2.0896860986547083,
+      "grad_norm": 0.2534300159070352,
+      "learning_rate": 1.686046511627907e-05,
+      "loss": 0.3483,
+      "step": 1398
+    },
+    {
+      "epoch": 2.09118086696562,
+      "grad_norm": 0.26144727440452237,
+      "learning_rate": 1.6832779623477298e-05,
+      "loss": 0.3618,
+      "step": 1399
+    },
+    {
+      "epoch": 2.092675635276532,
+      "grad_norm": 0.305124465052305,
+      "learning_rate": 1.6805094130675526e-05,
+      "loss": 0.3434,
+      "step": 1400
+    },
+    {
+      "epoch": 2.094170403587444,
+      "grad_norm": 0.23713445291508636,
+      "learning_rate": 1.6777408637873755e-05,
+      "loss": 0.3255,
+      "step": 1401
+    },
+    {
+      "epoch": 2.095665171898356,
+      "grad_norm": 0.26074376707700503,
+      "learning_rate": 1.6749723145071983e-05,
+      "loss": 0.3564,
+      "step": 1402
+    },
+    {
+      "epoch": 2.0971599402092678,
+      "grad_norm": 0.26660004039789087,
+      "learning_rate": 1.672203765227021e-05,
+      "loss": 0.3515,
+      "step": 1403
+    },
+    {
+      "epoch": 2.098654708520179,
+      "grad_norm": 0.2609759062198156,
+      "learning_rate": 1.669435215946844e-05,
+      "loss": 0.3523,
+      "step": 1404
+    },
+    {
+      "epoch": 2.100149476831091,
+      "grad_norm": 0.24873824708738737,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3301,
+      "step": 1405
+    },
+    {
+      "epoch": 2.101644245142003,
+      "grad_norm": 0.2746764056825652,
+      "learning_rate": 1.6638981173864896e-05,
+      "loss": 0.3373,
+      "step": 1406
+    },
+    {
+      "epoch": 2.103139013452915,
+      "grad_norm": 0.28678208029368607,
+      "learning_rate": 1.6611295681063124e-05,
+      "loss": 0.3413,
+      "step": 1407
+    },
+    {
+      "epoch": 2.1046337817638268,
+      "grad_norm": 0.2422517646134207,
+      "learning_rate": 1.6583610188261352e-05,
+      "loss": 0.3311,
+      "step": 1408
+    },
+    {
+      "epoch": 2.106128550074738,
+      "grad_norm": 0.27223830612459543,
+      "learning_rate": 1.655592469545958e-05,
+      "loss": 0.3427,
+      "step": 1409
+    },
+    {
+      "epoch": 2.10762331838565,
+      "grad_norm": 0.3222846591149712,
+      "learning_rate": 1.652823920265781e-05,
+      "loss": 0.349,
+      "step": 1410
+    },
+    {
+      "epoch": 2.109118086696562,
+      "grad_norm": 0.26899812331828105,
+      "learning_rate": 1.6500553709856037e-05,
+      "loss": 0.3298,
+      "step": 1411
+    },
+    {
+      "epoch": 2.110612855007474,
+      "grad_norm": 0.2630871145936935,
+      "learning_rate": 1.647286821705426e-05,
+      "loss": 0.3469,
+      "step": 1412
+    },
+    {
+      "epoch": 2.1121076233183858,
+      "grad_norm": 0.31101184541759913,
+      "learning_rate": 1.6445182724252493e-05,
+      "loss": 0.3511,
+      "step": 1413
+    },
+    {
+      "epoch": 2.1136023916292976,
+      "grad_norm": 0.2677888740441868,
+      "learning_rate": 1.641749723145072e-05,
+      "loss": 0.362,
+      "step": 1414
+    },
+    {
+      "epoch": 2.115097159940209,
+      "grad_norm": 0.3040682619989834,
+      "learning_rate": 1.638981173864895e-05,
+      "loss": 0.3377,
+      "step": 1415
+    },
+    {
+      "epoch": 2.116591928251121,
+      "grad_norm": 0.30930159520774986,
+      "learning_rate": 1.6362126245847174e-05,
+      "loss": 0.3611,
+      "step": 1416
+    },
+    {
+      "epoch": 2.118086696562033,
+      "grad_norm": 0.2632722530646921,
+      "learning_rate": 1.6334440753045406e-05,
+      "loss": 0.3614,
+      "step": 1417
+    },
+    {
+      "epoch": 2.1195814648729447,
+      "grad_norm": 0.3102659522040007,
+      "learning_rate": 1.6306755260243634e-05,
+      "loss": 0.3502,
+      "step": 1418
+    },
+    {
+      "epoch": 2.1210762331838566,
+      "grad_norm": 0.6918235377340805,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.3511,
+      "step": 1419
+    },
+    {
+      "epoch": 2.1225710014947685,
+      "grad_norm": 0.2633076916715839,
+      "learning_rate": 1.6251384274640087e-05,
+      "loss": 0.348,
+      "step": 1420
+    },
+    {
+      "epoch": 2.12406576980568,
+      "grad_norm": 0.33491820205601136,
+      "learning_rate": 1.622369878183832e-05,
+      "loss": 0.3305,
+      "step": 1421
+    },
+    {
+      "epoch": 2.125560538116592,
+      "grad_norm": 0.2443431568573533,
+      "learning_rate": 1.6196013289036547e-05,
+      "loss": 0.3376,
+      "step": 1422
+    },
+    {
+      "epoch": 2.1270553064275037,
+      "grad_norm": 0.2789295032581053,
+      "learning_rate": 1.6168327796234775e-05,
+      "loss": 0.3591,
+      "step": 1423
+    },
+    {
+      "epoch": 2.1285500747384156,
+      "grad_norm": 0.3072180119664886,
+      "learning_rate": 1.6140642303433e-05,
+      "loss": 0.3381,
+      "step": 1424
+    },
+    {
+      "epoch": 2.1300448430493275,
+      "grad_norm": 0.2931558402414626,
+      "learning_rate": 1.611295681063123e-05,
+      "loss": 0.3457,
+      "step": 1425
+    },
+    {
+      "epoch": 2.131539611360239,
+      "grad_norm": 0.2920486884549725,
+      "learning_rate": 1.608527131782946e-05,
+      "loss": 0.3582,
+      "step": 1426
+    },
+    {
+      "epoch": 2.133034379671151,
+      "grad_norm": 0.32010190212113254,
+      "learning_rate": 1.6057585825027684e-05,
+      "loss": 0.3561,
+      "step": 1427
+    },
+    {
+      "epoch": 2.1345291479820627,
+      "grad_norm": 0.26629698619607495,
+      "learning_rate": 1.6029900332225913e-05,
+      "loss": 0.3398,
+      "step": 1428
+    },
+    {
+      "epoch": 2.1360239162929746,
+      "grad_norm": 0.28518587554227476,
+      "learning_rate": 1.6002214839424144e-05,
+      "loss": 0.3482,
+      "step": 1429
+    },
+    {
+      "epoch": 2.1375186846038865,
+      "grad_norm": 0.30492646563480935,
+      "learning_rate": 1.5974529346622372e-05,
+      "loss": 0.3608,
+      "step": 1430
+    },
+    {
+      "epoch": 2.1390134529147984,
+      "grad_norm": 0.2609622955735452,
+      "learning_rate": 1.5946843853820597e-05,
+      "loss": 0.3458,
+      "step": 1431
+    },
+    {
+      "epoch": 2.14050822122571,
+      "grad_norm": 0.28023307051434004,
+      "learning_rate": 1.5919158361018825e-05,
+      "loss": 0.3555,
+      "step": 1432
+    },
+    {
+      "epoch": 2.1420029895366217,
+      "grad_norm": 0.28093417143138943,
+      "learning_rate": 1.5891472868217057e-05,
+      "loss": 0.3358,
+      "step": 1433
+    },
+    {
+      "epoch": 2.1434977578475336,
+      "grad_norm": 0.22672255772841898,
+      "learning_rate": 1.5863787375415285e-05,
+      "loss": 0.3419,
+      "step": 1434
+    },
+    {
+      "epoch": 2.1449925261584455,
+      "grad_norm": 0.2913750353307331,
+      "learning_rate": 1.583610188261351e-05,
+      "loss": 0.3204,
+      "step": 1435
+    },
+    {
+      "epoch": 2.1464872944693574,
+      "grad_norm": 0.24764198222861447,
+      "learning_rate": 1.5808416389811738e-05,
+      "loss": 0.3333,
+      "step": 1436
+    },
+    {
+      "epoch": 2.1479820627802693,
+      "grad_norm": 0.27582790069314234,
+      "learning_rate": 1.578073089700997e-05,
+      "loss": 0.3525,
+      "step": 1437
+    },
+    {
+      "epoch": 2.1494768310911807,
+      "grad_norm": 0.32472892036303097,
+      "learning_rate": 1.5753045404208195e-05,
+      "loss": 0.3455,
+      "step": 1438
+    },
+    {
+      "epoch": 2.1509715994020926,
+      "grad_norm": 0.3008220120001433,
+      "learning_rate": 1.5725359911406423e-05,
+      "loss": 0.3324,
+      "step": 1439
+    },
+    {
+      "epoch": 2.1524663677130045,
+      "grad_norm": 0.2663368999923895,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 0.34,
+      "step": 1440
+    },
+    {
+      "epoch": 2.1539611360239164,
+      "grad_norm": 0.2900904011732388,
+      "learning_rate": 1.5669988925802882e-05,
+      "loss": 0.3296,
+      "step": 1441
+    },
+    {
+      "epoch": 2.1554559043348283,
+      "grad_norm": 0.32094953381856944,
+      "learning_rate": 1.5642303433001107e-05,
+      "loss": 0.3467,
+      "step": 1442
+    },
+    {
+      "epoch": 2.15695067264574,
+      "grad_norm": 0.25518958348180326,
+      "learning_rate": 1.5614617940199335e-05,
+      "loss": 0.3623,
+      "step": 1443
+    },
+    {
+      "epoch": 2.1584454409566516,
+      "grad_norm": 0.2602089628152503,
+      "learning_rate": 1.5586932447397564e-05,
+      "loss": 0.3386,
+      "step": 1444
+    },
+    {
+      "epoch": 2.1599402092675635,
+      "grad_norm": 0.2924657263866658,
+      "learning_rate": 1.5559246954595795e-05,
+      "loss": 0.3453,
+      "step": 1445
+    },
+    {
+      "epoch": 2.1614349775784754,
+      "grad_norm": 0.2733756136759275,
+      "learning_rate": 1.553156146179402e-05,
+      "loss": 0.3435,
+      "step": 1446
+    },
+    {
+      "epoch": 2.1629297458893872,
+      "grad_norm": 0.24713884581900084,
+      "learning_rate": 1.5503875968992248e-05,
+      "loss": 0.329,
+      "step": 1447
+    },
+    {
+      "epoch": 2.164424514200299,
+      "grad_norm": 0.27183219189434127,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 0.367,
+      "step": 1448
+    },
+    {
+      "epoch": 2.1659192825112106,
+      "grad_norm": 0.2803790673960383,
+      "learning_rate": 1.5448504983388705e-05,
+      "loss": 0.3651,
+      "step": 1449
+    },
+    {
+      "epoch": 2.1674140508221225,
+      "grad_norm": 0.2623376559318404,
+      "learning_rate": 1.5420819490586933e-05,
+      "loss": 0.3422,
+      "step": 1450
+    },
+    {
+      "epoch": 2.1689088191330343,
+      "grad_norm": 0.2593344217008581,
+      "learning_rate": 1.539313399778516e-05,
+      "loss": 0.3448,
+      "step": 1451
+    },
+    {
+      "epoch": 2.1704035874439462,
+      "grad_norm": 0.2476334779086849,
+      "learning_rate": 1.536544850498339e-05,
+      "loss": 0.3374,
+      "step": 1452
+    },
+    {
+      "epoch": 2.171898355754858,
+      "grad_norm": 0.24783841573164786,
+      "learning_rate": 1.5337763012181617e-05,
+      "loss": 0.3278,
+      "step": 1453
+    },
+    {
+      "epoch": 2.17339312406577,
+      "grad_norm": 0.24271962976241726,
+      "learning_rate": 1.5310077519379846e-05,
+      "loss": 0.3458,
+      "step": 1454
+    },
+    {
+      "epoch": 2.1748878923766815,
+      "grad_norm": 0.24590726630039964,
+      "learning_rate": 1.5282392026578074e-05,
+      "loss": 0.3591,
+      "step": 1455
+    },
+    {
+      "epoch": 2.1763826606875933,
+      "grad_norm": 0.2914639146703328,
+      "learning_rate": 1.5254706533776304e-05,
+      "loss": 0.3662,
+      "step": 1456
+    },
+    {
+      "epoch": 2.1778774289985052,
+      "grad_norm": 0.25180030039825335,
+      "learning_rate": 1.5227021040974529e-05,
+      "loss": 0.3537,
+      "step": 1457
+    },
+    {
+      "epoch": 2.179372197309417,
+      "grad_norm": 0.24714913486454587,
+      "learning_rate": 1.5199335548172758e-05,
+      "loss": 0.3501,
+      "step": 1458
+    },
+    {
+      "epoch": 2.180866965620329,
+      "grad_norm": 0.27461559484149256,
+      "learning_rate": 1.5171650055370987e-05,
+      "loss": 0.3693,
+      "step": 1459
+    },
+    {
+      "epoch": 2.1823617339312404,
+      "grad_norm": 0.24957157038871805,
+      "learning_rate": 1.5143964562569215e-05,
+      "loss": 0.3406,
+      "step": 1460
+    },
+    {
+      "epoch": 2.1838565022421523,
+      "grad_norm": 0.24902460521608702,
+      "learning_rate": 1.5116279069767441e-05,
+      "loss": 0.3564,
+      "step": 1461
+    },
+    {
+      "epoch": 2.185351270553064,
+      "grad_norm": 0.2550598474749505,
+      "learning_rate": 1.5088593576965671e-05,
+      "loss": 0.3521,
+      "step": 1462
+    },
+    {
+      "epoch": 2.186846038863976,
+      "grad_norm": 0.2544236088422744,
+      "learning_rate": 1.50609080841639e-05,
+      "loss": 0.3312,
+      "step": 1463
+    },
+    {
+      "epoch": 2.188340807174888,
+      "grad_norm": 0.24793096428072336,
+      "learning_rate": 1.5033222591362128e-05,
+      "loss": 0.342,
+      "step": 1464
+    },
+    {
+      "epoch": 2.1898355754858,
+      "grad_norm": 0.24421564136416882,
+      "learning_rate": 1.5005537098560354e-05,
+      "loss": 0.3403,
+      "step": 1465
+    },
+    {
+      "epoch": 2.1913303437967113,
+      "grad_norm": 0.24830103826000602,
+      "learning_rate": 1.4977851605758584e-05,
+      "loss": 0.3471,
+      "step": 1466
+    },
+    {
+      "epoch": 2.192825112107623,
+      "grad_norm": 0.2595955427692041,
+      "learning_rate": 1.4950166112956812e-05,
+      "loss": 0.3502,
+      "step": 1467
+    },
+    {
+      "epoch": 2.194319880418535,
+      "grad_norm": 0.253469631334733,
+      "learning_rate": 1.4922480620155039e-05,
+      "loss": 0.3467,
+      "step": 1468
+    },
+    {
+      "epoch": 2.195814648729447,
+      "grad_norm": 0.2681659819246609,
+      "learning_rate": 1.4894795127353267e-05,
+      "loss": 0.3378,
+      "step": 1469
+    },
+    {
+      "epoch": 2.197309417040359,
+      "grad_norm": 0.2765117833048143,
+      "learning_rate": 1.4867109634551497e-05,
+      "loss": 0.3537,
+      "step": 1470
+    },
+    {
+      "epoch": 2.1988041853512708,
+      "grad_norm": 0.2800827360063099,
+      "learning_rate": 1.4839424141749725e-05,
+      "loss": 0.3461,
+      "step": 1471
+    },
+    {
+      "epoch": 2.200298953662182,
+      "grad_norm": 0.29678630228946956,
+      "learning_rate": 1.4811738648947951e-05,
+      "loss": 0.3648,
+      "step": 1472
+    },
+    {
+      "epoch": 2.201793721973094,
+      "grad_norm": 0.29393535369824536,
+      "learning_rate": 1.478405315614618e-05,
+      "loss": 0.3519,
+      "step": 1473
+    },
+    {
+      "epoch": 2.203288490284006,
+      "grad_norm": 0.29890919649779796,
+      "learning_rate": 1.475636766334441e-05,
+      "loss": 0.3444,
+      "step": 1474
+    },
+    {
+      "epoch": 2.204783258594918,
+      "grad_norm": 0.27192565241735883,
+      "learning_rate": 1.4728682170542638e-05,
+      "loss": 0.3553,
+      "step": 1475
+    },
+    {
+      "epoch": 2.2062780269058297,
+      "grad_norm": 0.2909697565485119,
+      "learning_rate": 1.4700996677740864e-05,
+      "loss": 0.3545,
+      "step": 1476
+    },
+    {
+      "epoch": 2.2077727952167416,
+      "grad_norm": 0.24833688810200938,
+      "learning_rate": 1.4673311184939092e-05,
+      "loss": 0.3397,
+      "step": 1477
+    },
+    {
+      "epoch": 2.209267563527653,
+      "grad_norm": 0.2653552327840514,
+      "learning_rate": 1.464562569213732e-05,
+      "loss": 0.3675,
+      "step": 1478
+    },
+    {
+      "epoch": 2.210762331838565,
+      "grad_norm": 0.23926358297107725,
+      "learning_rate": 1.461794019933555e-05,
+      "loss": 0.3363,
+      "step": 1479
+    },
+    {
+      "epoch": 2.212257100149477,
+      "grad_norm": 0.28079485801361687,
+      "learning_rate": 1.4590254706533777e-05,
+      "loss": 0.3325,
+      "step": 1480
+    },
+    {
+      "epoch": 2.2137518684603887,
+      "grad_norm": 0.25651182998254357,
+      "learning_rate": 1.4562569213732005e-05,
+      "loss": 0.3548,
+      "step": 1481
+    },
+    {
+      "epoch": 2.2152466367713006,
+      "grad_norm": 0.2590913843035247,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 0.3635,
+      "step": 1482
+    },
+    {
+      "epoch": 2.216741405082212,
+      "grad_norm": 0.23758038418981253,
+      "learning_rate": 1.450719822812846e-05,
+      "loss": 0.3371,
+      "step": 1483
+    },
+    {
+      "epoch": 2.218236173393124,
+      "grad_norm": 0.24725424006649302,
+      "learning_rate": 1.447951273532669e-05,
+      "loss": 0.3449,
+      "step": 1484
+    },
+    {
+      "epoch": 2.219730941704036,
+      "grad_norm": 0.2564185032655142,
+      "learning_rate": 1.4451827242524918e-05,
+      "loss": 0.349,
+      "step": 1485
+    },
+    {
+      "epoch": 2.2212257100149477,
+      "grad_norm": 0.2437344788148016,
+      "learning_rate": 1.4424141749723146e-05,
+      "loss": 0.3626,
+      "step": 1486
+    },
+    {
+      "epoch": 2.2227204783258596,
+      "grad_norm": 0.24994413913481753,
+      "learning_rate": 1.4396456256921373e-05,
+      "loss": 0.3332,
+      "step": 1487
+    },
+    {
+      "epoch": 2.2242152466367715,
+      "grad_norm": 0.25202447121120325,
+      "learning_rate": 1.4368770764119603e-05,
+      "loss": 0.3496,
+      "step": 1488
+    },
+    {
+      "epoch": 2.225710014947683,
+      "grad_norm": 0.24595204827342781,
+      "learning_rate": 1.434108527131783e-05,
+      "loss": 0.3531,
+      "step": 1489
+    },
+    {
+      "epoch": 2.227204783258595,
+      "grad_norm": 0.22737707295108817,
+      "learning_rate": 1.4313399778516059e-05,
+      "loss": 0.343,
+      "step": 1490
+    },
+    {
+      "epoch": 2.2286995515695067,
+      "grad_norm": 0.2375750534850708,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.3568,
+      "step": 1491
+    },
+    {
+      "epoch": 2.2301943198804186,
+      "grad_norm": 0.2372583353661071,
+      "learning_rate": 1.4258028792912515e-05,
+      "loss": 0.3501,
+      "step": 1492
+    },
+    {
+      "epoch": 2.2316890881913305,
+      "grad_norm": 0.2512252909861587,
+      "learning_rate": 1.4230343300110744e-05,
+      "loss": 0.349,
+      "step": 1493
+    },
+    {
+      "epoch": 2.233183856502242,
+      "grad_norm": 0.22923446668705474,
+      "learning_rate": 1.420265780730897e-05,
+      "loss": 0.3387,
+      "step": 1494
+    },
+    {
+      "epoch": 2.234678624813154,
+      "grad_norm": 0.24460142180825262,
+      "learning_rate": 1.4174972314507198e-05,
+      "loss": 0.3461,
+      "step": 1495
+    },
+    {
+      "epoch": 2.2361733931240657,
+      "grad_norm": 0.25086114363799383,
+      "learning_rate": 1.4147286821705426e-05,
+      "loss": 0.3504,
+      "step": 1496
+    },
+    {
+      "epoch": 2.2376681614349776,
+      "grad_norm": 0.2581559132244819,
+      "learning_rate": 1.4119601328903656e-05,
+      "loss": 0.3636,
+      "step": 1497
+    },
+    {
+      "epoch": 2.2391629297458895,
+      "grad_norm": 0.26247010137075677,
+      "learning_rate": 1.4091915836101883e-05,
+      "loss": 0.3659,
+      "step": 1498
+    },
+    {
+      "epoch": 2.2406576980568014,
+      "grad_norm": 0.23045320104764996,
+      "learning_rate": 1.4064230343300111e-05,
+      "loss": 0.3491,
+      "step": 1499
+    },
+    {
+      "epoch": 2.242152466367713,
+      "grad_norm": 0.26324595374096876,
+      "learning_rate": 1.403654485049834e-05,
+      "loss": 0.352,
+      "step": 1500
+    },
+    {
+      "epoch": 2.2436472346786247,
+      "grad_norm": 0.25698407192361866,
+      "learning_rate": 1.4008859357696569e-05,
+      "loss": 0.3431,
+      "step": 1501
+    },
+    {
+      "epoch": 2.2451420029895366,
+      "grad_norm": 0.24456969533288542,
+      "learning_rate": 1.3981173864894796e-05,
+      "loss": 0.3442,
+      "step": 1502
+    },
+    {
+      "epoch": 2.2466367713004485,
+      "grad_norm": 0.23437103613744287,
+      "learning_rate": 1.3953488372093024e-05,
+      "loss": 0.3437,
+      "step": 1503
+    },
+    {
+      "epoch": 2.2481315396113604,
+      "grad_norm": 0.2853650080466829,
+      "learning_rate": 1.3925802879291252e-05,
+      "loss": 0.3463,
+      "step": 1504
+    },
+    {
+      "epoch": 2.2496263079222723,
+      "grad_norm": 0.2593515190183951,
+      "learning_rate": 1.3898117386489482e-05,
+      "loss": 0.3539,
+      "step": 1505
+    },
+    {
+      "epoch": 2.2511210762331837,
+      "grad_norm": 0.24117111301372715,
+      "learning_rate": 1.3870431893687708e-05,
+      "loss": 0.3454,
+      "step": 1506
+    },
+    {
+      "epoch": 2.2526158445440956,
+      "grad_norm": 0.2497873118456191,
+      "learning_rate": 1.3842746400885937e-05,
+      "loss": 0.3671,
+      "step": 1507
+    },
+    {
+      "epoch": 2.2541106128550075,
+      "grad_norm": 0.2763296801598134,
+      "learning_rate": 1.3815060908084165e-05,
+      "loss": 0.3481,
+      "step": 1508
+    },
+    {
+      "epoch": 2.2556053811659194,
+      "grad_norm": 0.2513998192078893,
+      "learning_rate": 1.3787375415282391e-05,
+      "loss": 0.3369,
+      "step": 1509
+    },
+    {
+      "epoch": 2.2571001494768312,
+      "grad_norm": 0.28365828106689067,
+      "learning_rate": 1.375968992248062e-05,
+      "loss": 0.3704,
+      "step": 1510
+    },
+    {
+      "epoch": 2.258594917787743,
+      "grad_norm": 0.24947157764737324,
+      "learning_rate": 1.373200442967885e-05,
+      "loss": 0.3479,
+      "step": 1511
+    },
+    {
+      "epoch": 2.2600896860986546,
+      "grad_norm": 0.2518922024874687,
+      "learning_rate": 1.3704318936877078e-05,
+      "loss": 0.3446,
+      "step": 1512
+    },
+    {
+      "epoch": 2.2615844544095665,
+      "grad_norm": 0.2562412704138893,
+      "learning_rate": 1.3676633444075304e-05,
+      "loss": 0.3494,
+      "step": 1513
+    },
+    {
+      "epoch": 2.2630792227204783,
+      "grad_norm": 0.2671745178357787,
+      "learning_rate": 1.3648947951273532e-05,
+      "loss": 0.3663,
+      "step": 1514
+    },
+    {
+      "epoch": 2.2645739910313902,
+      "grad_norm": 0.24905237024666987,
+      "learning_rate": 1.3621262458471762e-05,
+      "loss": 0.3337,
+      "step": 1515
+    },
+    {
+      "epoch": 2.266068759342302,
+      "grad_norm": 0.23899898384882745,
+      "learning_rate": 1.359357696566999e-05,
+      "loss": 0.3361,
+      "step": 1516
+    },
+    {
+      "epoch": 2.2675635276532136,
+      "grad_norm": 0.2575710094453896,
+      "learning_rate": 1.3565891472868217e-05,
+      "loss": 0.3474,
+      "step": 1517
+    },
+    {
+      "epoch": 2.2690582959641254,
+      "grad_norm": 0.2602471759141405,
+      "learning_rate": 1.3538205980066445e-05,
+      "loss": 0.3491,
+      "step": 1518
+    },
+    {
+      "epoch": 2.2705530642750373,
+      "grad_norm": 0.23208146264939947,
+      "learning_rate": 1.3510520487264675e-05,
+      "loss": 0.339,
+      "step": 1519
+    },
+    {
+      "epoch": 2.2720478325859492,
+      "grad_norm": 0.2474760273284901,
+      "learning_rate": 1.3482834994462903e-05,
+      "loss": 0.3487,
+      "step": 1520
+    },
+    {
+      "epoch": 2.273542600896861,
+      "grad_norm": 0.25153021538004405,
+      "learning_rate": 1.345514950166113e-05,
+      "loss": 0.3481,
+      "step": 1521
+    },
+    {
+      "epoch": 2.275037369207773,
+      "grad_norm": 0.24664779846571083,
+      "learning_rate": 1.3427464008859358e-05,
+      "loss": 0.3423,
+      "step": 1522
+    },
+    {
+      "epoch": 2.2765321375186844,
+      "grad_norm": 0.26716528779973436,
+      "learning_rate": 1.3399778516057588e-05,
+      "loss": 0.3497,
+      "step": 1523
+    },
+    {
+      "epoch": 2.2780269058295963,
+      "grad_norm": 0.252875693508043,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 0.3494,
+      "step": 1524
+    },
+    {
+      "epoch": 2.279521674140508,
+      "grad_norm": 0.2636237262337298,
+      "learning_rate": 1.3344407530454042e-05,
+      "loss": 0.3558,
+      "step": 1525
+    },
+    {
+      "epoch": 2.28101644245142,
+      "grad_norm": 0.24261960280494754,
+      "learning_rate": 1.331672203765227e-05,
+      "loss": 0.3283,
+      "step": 1526
+    },
+    {
+      "epoch": 2.282511210762332,
+      "grad_norm": 0.24572565493794724,
+      "learning_rate": 1.32890365448505e-05,
+      "loss": 0.3474,
+      "step": 1527
+    },
+    {
+      "epoch": 2.2840059790732434,
+      "grad_norm": 0.24263860731861045,
+      "learning_rate": 1.3261351052048725e-05,
+      "loss": 0.3364,
+      "step": 1528
+    },
+    {
+      "epoch": 2.2855007473841553,
+      "grad_norm": 0.3016598140134729,
+      "learning_rate": 1.3233665559246955e-05,
+      "loss": 0.3628,
+      "step": 1529
+    },
+    {
+      "epoch": 2.286995515695067,
+      "grad_norm": 0.2832991727106963,
+      "learning_rate": 1.3205980066445183e-05,
+      "loss": 0.3349,
+      "step": 1530
+    },
+    {
+      "epoch": 2.288490284005979,
+      "grad_norm": 0.24472489069106423,
+      "learning_rate": 1.3178294573643413e-05,
+      "loss": 0.3449,
+      "step": 1531
+    },
+    {
+      "epoch": 2.289985052316891,
+      "grad_norm": 0.294352580178817,
+      "learning_rate": 1.3150609080841638e-05,
+      "loss": 0.3504,
+      "step": 1532
+    },
+    {
+      "epoch": 2.291479820627803,
+      "grad_norm": 0.2557073616138859,
+      "learning_rate": 1.3122923588039868e-05,
+      "loss": 0.3418,
+      "step": 1533
+    },
+    {
+      "epoch": 2.2929745889387143,
+      "grad_norm": 0.24458460736852505,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 0.3604,
+      "step": 1534
+    },
+    {
+      "epoch": 2.294469357249626,
+      "grad_norm": 0.26905780587720435,
+      "learning_rate": 1.3067552602436323e-05,
+      "loss": 0.3509,
+      "step": 1535
+    },
+    {
+      "epoch": 2.295964125560538,
+      "grad_norm": 0.24709932041148133,
+      "learning_rate": 1.303986710963455e-05,
+      "loss": 0.3582,
+      "step": 1536
+    },
+    {
+      "epoch": 2.29745889387145,
+      "grad_norm": 0.2394559051270739,
+      "learning_rate": 1.301218161683278e-05,
+      "loss": 0.3613,
+      "step": 1537
+    },
+    {
+      "epoch": 2.298953662182362,
+      "grad_norm": 0.27184489454616173,
+      "learning_rate": 1.2984496124031009e-05,
+      "loss": 0.3404,
+      "step": 1538
+    },
+    {
+      "epoch": 2.3004484304932733,
+      "grad_norm": 0.24176097444794792,
+      "learning_rate": 1.2956810631229235e-05,
+      "loss": 0.3366,
+      "step": 1539
+    },
+    {
+      "epoch": 2.301943198804185,
+      "grad_norm": 0.24285205073069588,
+      "learning_rate": 1.2929125138427464e-05,
+      "loss": 0.34,
+      "step": 1540
+    },
+    {
+      "epoch": 2.303437967115097,
+      "grad_norm": 0.2681067412191364,
+      "learning_rate": 1.2901439645625693e-05,
+      "loss": 0.3513,
+      "step": 1541
+    },
+    {
+      "epoch": 2.304932735426009,
+      "grad_norm": 0.23647661968712572,
+      "learning_rate": 1.2873754152823922e-05,
+      "loss": 0.3565,
+      "step": 1542
+    },
+    {
+      "epoch": 2.306427503736921,
+      "grad_norm": 0.24622741933344566,
+      "learning_rate": 1.2846068660022148e-05,
+      "loss": 0.3548,
+      "step": 1543
+    },
+    {
+      "epoch": 2.3079222720478327,
+      "grad_norm": 0.22073993435518316,
+      "learning_rate": 1.2818383167220376e-05,
+      "loss": 0.3493,
+      "step": 1544
+    },
+    {
+      "epoch": 2.3094170403587446,
+      "grad_norm": 0.23507902952979745,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 0.3292,
+      "step": 1545
+    },
+    {
+      "epoch": 2.310911808669656,
+      "grad_norm": 0.2479790631401503,
+      "learning_rate": 1.2763012181616834e-05,
+      "loss": 0.3367,
+      "step": 1546
+    },
+    {
+      "epoch": 2.312406576980568,
+      "grad_norm": 0.2710484047751056,
+      "learning_rate": 1.2735326688815061e-05,
+      "loss": 0.3494,
+      "step": 1547
+    },
+    {
+      "epoch": 2.31390134529148,
+      "grad_norm": 0.257659163011388,
+      "learning_rate": 1.2707641196013289e-05,
+      "loss": 0.3376,
+      "step": 1548
+    },
+    {
+      "epoch": 2.3153961136023917,
+      "grad_norm": 0.254892855733212,
+      "learning_rate": 1.2679955703211519e-05,
+      "loss": 0.3236,
+      "step": 1549
+    },
+    {
+      "epoch": 2.3168908819133036,
+      "grad_norm": 0.2597506659664171,
+      "learning_rate": 1.2652270210409744e-05,
+      "loss": 0.3505,
+      "step": 1550
+    },
+    {
+      "epoch": 2.318385650224215,
+      "grad_norm": 0.2784415427891222,
+      "learning_rate": 1.2624584717607974e-05,
+      "loss": 0.3409,
+      "step": 1551
+    },
+    {
+      "epoch": 2.319880418535127,
+      "grad_norm": 0.240817340762689,
+      "learning_rate": 1.2596899224806202e-05,
+      "loss": 0.3265,
+      "step": 1552
+    },
+    {
+      "epoch": 2.321375186846039,
+      "grad_norm": 0.2326707627940437,
+      "learning_rate": 1.2569213732004432e-05,
+      "loss": 0.3326,
+      "step": 1553
+    },
+    {
+      "epoch": 2.3228699551569507,
+      "grad_norm": 0.24990784008828967,
+      "learning_rate": 1.2541528239202657e-05,
+      "loss": 0.3541,
+      "step": 1554
+    },
+    {
+      "epoch": 2.3243647234678626,
+      "grad_norm": 0.2623751697375492,
+      "learning_rate": 1.2513842746400887e-05,
+      "loss": 0.3342,
+      "step": 1555
+    },
+    {
+      "epoch": 2.3258594917787745,
+      "grad_norm": 0.26346965730214383,
+      "learning_rate": 1.2486157253599115e-05,
+      "loss": 0.355,
+      "step": 1556
+    },
+    {
+      "epoch": 2.327354260089686,
+      "grad_norm": 0.23842939778304062,
+      "learning_rate": 1.2458471760797343e-05,
+      "loss": 0.3412,
+      "step": 1557
+    },
+    {
+      "epoch": 2.328849028400598,
+      "grad_norm": 0.2317850494073918,
+      "learning_rate": 1.2430786267995571e-05,
+      "loss": 0.3527,
+      "step": 1558
+    },
+    {
+      "epoch": 2.3303437967115097,
+      "grad_norm": 0.2297656920744873,
+      "learning_rate": 1.24031007751938e-05,
+      "loss": 0.3615,
+      "step": 1559
+    },
+    {
+      "epoch": 2.3318385650224216,
+      "grad_norm": 0.28837433905122034,
+      "learning_rate": 1.2375415282392027e-05,
+      "loss": 0.3528,
+      "step": 1560
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.23532658987921196,
+      "learning_rate": 1.2347729789590256e-05,
+      "loss": 0.3367,
+      "step": 1561
+    },
+    {
+      "epoch": 2.334828101644245,
+      "grad_norm": 0.25001514185944085,
+      "learning_rate": 1.2320044296788484e-05,
+      "loss": 0.3481,
+      "step": 1562
+    },
+    {
+      "epoch": 2.336322869955157,
+      "grad_norm": 0.25581039981467246,
+      "learning_rate": 1.2292358803986712e-05,
+      "loss": 0.3391,
+      "step": 1563
+    },
+    {
+      "epoch": 2.3378176382660687,
+      "grad_norm": 0.22931062817892672,
+      "learning_rate": 1.2264673311184939e-05,
+      "loss": 0.3425,
+      "step": 1564
+    },
+    {
+      "epoch": 2.3393124065769806,
+      "grad_norm": 0.2498212080293515,
+      "learning_rate": 1.2236987818383168e-05,
+      "loss": 0.3592,
+      "step": 1565
+    },
+    {
+      "epoch": 2.3408071748878925,
+      "grad_norm": 0.2667976321299467,
+      "learning_rate": 1.2209302325581395e-05,
+      "loss": 0.3409,
+      "step": 1566
+    },
+    {
+      "epoch": 2.3423019431988044,
+      "grad_norm": 0.2320742859426796,
+      "learning_rate": 1.2181616832779625e-05,
+      "loss": 0.3407,
+      "step": 1567
+    },
+    {
+      "epoch": 2.343796711509716,
+      "grad_norm": 0.2433144070773857,
+      "learning_rate": 1.2153931339977851e-05,
+      "loss": 0.3399,
+      "step": 1568
+    },
+    {
+      "epoch": 2.3452914798206277,
+      "grad_norm": 0.29698674947170245,
+      "learning_rate": 1.2126245847176081e-05,
+      "loss": 0.3659,
+      "step": 1569
+    },
+    {
+      "epoch": 2.3467862481315396,
+      "grad_norm": 0.24232482166841285,
+      "learning_rate": 1.2098560354374308e-05,
+      "loss": 0.3336,
+      "step": 1570
+    },
+    {
+      "epoch": 2.3482810164424515,
+      "grad_norm": 0.2415043045541668,
+      "learning_rate": 1.2070874861572538e-05,
+      "loss": 0.3332,
+      "step": 1571
+    },
+    {
+      "epoch": 2.3497757847533634,
+      "grad_norm": 0.2536133594384661,
+      "learning_rate": 1.2043189368770764e-05,
+      "loss": 0.3308,
+      "step": 1572
+    },
+    {
+      "epoch": 2.351270553064275,
+      "grad_norm": 0.29097090851363483,
+      "learning_rate": 1.2015503875968994e-05,
+      "loss": 0.3315,
+      "step": 1573
+    },
+    {
+      "epoch": 2.3527653213751867,
+      "grad_norm": 0.2670468600509829,
+      "learning_rate": 1.198781838316722e-05,
+      "loss": 0.3326,
+      "step": 1574
+    },
+    {
+      "epoch": 2.3542600896860986,
+      "grad_norm": 0.24253470974734717,
+      "learning_rate": 1.196013289036545e-05,
+      "loss": 0.3382,
+      "step": 1575
+    },
+    {
+      "epoch": 2.3557548579970105,
+      "grad_norm": 0.24600855461829854,
+      "learning_rate": 1.1932447397563677e-05,
+      "loss": 0.3451,
+      "step": 1576
+    },
+    {
+      "epoch": 2.3572496263079223,
+      "grad_norm": 0.27048412923957,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.351,
+      "step": 1577
+    },
+    {
+      "epoch": 2.3587443946188342,
+      "grad_norm": 0.2846251633197542,
+      "learning_rate": 1.1877076411960133e-05,
+      "loss": 0.3485,
+      "step": 1578
+    },
+    {
+      "epoch": 2.360239162929746,
+      "grad_norm": 0.21882039183562454,
+      "learning_rate": 1.1849390919158361e-05,
+      "loss": 0.3226,
+      "step": 1579
+    },
+    {
+      "epoch": 2.3617339312406576,
+      "grad_norm": 0.25146829867694315,
+      "learning_rate": 1.182170542635659e-05,
+      "loss": 0.3478,
+      "step": 1580
+    },
+    {
+      "epoch": 2.3632286995515694,
+      "grad_norm": 0.2932128653486548,
+      "learning_rate": 1.1794019933554818e-05,
+      "loss": 0.3486,
+      "step": 1581
+    },
+    {
+      "epoch": 2.3647234678624813,
+      "grad_norm": 0.24760344527470252,
+      "learning_rate": 1.1766334440753046e-05,
+      "loss": 0.3372,
+      "step": 1582
+    },
+    {
+      "epoch": 2.366218236173393,
+      "grad_norm": 0.2931302292056,
+      "learning_rate": 1.1738648947951274e-05,
+      "loss": 0.3349,
+      "step": 1583
+    },
+    {
+      "epoch": 2.367713004484305,
+      "grad_norm": 0.25533651834106247,
+      "learning_rate": 1.1710963455149502e-05,
+      "loss": 0.338,
+      "step": 1584
+    },
+    {
+      "epoch": 2.3692077727952165,
+      "grad_norm": 0.23833194173318106,
+      "learning_rate": 1.168327796234773e-05,
+      "loss": 0.3317,
+      "step": 1585
+    },
+    {
+      "epoch": 2.3707025411061284,
+      "grad_norm": 0.2401449478465909,
+      "learning_rate": 1.1655592469545959e-05,
+      "loss": 0.3388,
+      "step": 1586
+    },
+    {
+      "epoch": 2.3721973094170403,
+      "grad_norm": 0.26530400855687103,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.3595,
+      "step": 1587
+    },
+    {
+      "epoch": 2.373692077727952,
+      "grad_norm": 0.2509575094283734,
+      "learning_rate": 1.1600221483942415e-05,
+      "loss": 0.343,
+      "step": 1588
+    },
+    {
+      "epoch": 2.375186846038864,
+      "grad_norm": 0.24429350918619236,
+      "learning_rate": 1.1572535991140643e-05,
+      "loss": 0.3389,
+      "step": 1589
+    },
+    {
+      "epoch": 2.376681614349776,
+      "grad_norm": 0.2271158715745915,
+      "learning_rate": 1.1544850498338872e-05,
+      "loss": 0.3401,
+      "step": 1590
+    },
+    {
+      "epoch": 2.3781763826606874,
+      "grad_norm": 0.2494368263535274,
+      "learning_rate": 1.15171650055371e-05,
+      "loss": 0.3288,
+      "step": 1591
+    },
+    {
+      "epoch": 2.3796711509715993,
+      "grad_norm": 0.23680875513011931,
+      "learning_rate": 1.1489479512735326e-05,
+      "loss": 0.3461,
+      "step": 1592
+    },
+    {
+      "epoch": 2.381165919282511,
+      "grad_norm": 0.23575470809125845,
+      "learning_rate": 1.1461794019933556e-05,
+      "loss": 0.3524,
+      "step": 1593
+    },
+    {
+      "epoch": 2.382660687593423,
+      "grad_norm": 0.24007491939369022,
+      "learning_rate": 1.1434108527131783e-05,
+      "loss": 0.3509,
+      "step": 1594
+    },
+    {
+      "epoch": 2.384155455904335,
+      "grad_norm": 0.23011281194655347,
+      "learning_rate": 1.1406423034330013e-05,
+      "loss": 0.3488,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3856502242152464,
+      "grad_norm": 0.24504481282510163,
+      "learning_rate": 1.1378737541528239e-05,
+      "loss": 0.3328,
+      "step": 1596
+    },
+    {
+      "epoch": 2.3871449925261583,
+      "grad_norm": 0.2448684732435368,
+      "learning_rate": 1.1351052048726469e-05,
+      "loss": 0.3561,
+      "step": 1597
+    },
+    {
+      "epoch": 2.38863976083707,
+      "grad_norm": 0.23896122386229993,
+      "learning_rate": 1.1323366555924695e-05,
+      "loss": 0.3296,
+      "step": 1598
+    },
+    {
+      "epoch": 2.390134529147982,
+      "grad_norm": 0.25205890106436907,
+      "learning_rate": 1.1295681063122925e-05,
+      "loss": 0.3398,
+      "step": 1599
+    },
+    {
+      "epoch": 2.391629297458894,
+      "grad_norm": 0.24334434216211379,
+      "learning_rate": 1.1267995570321152e-05,
+      "loss": 0.3504,
+      "step": 1600
+    },
+    {
+      "epoch": 2.393124065769806,
+      "grad_norm": 0.24895682112257264,
+      "learning_rate": 1.1240310077519382e-05,
+      "loss": 0.3547,
+      "step": 1601
+    },
+    {
+      "epoch": 2.3946188340807173,
+      "grad_norm": 0.24372620830115457,
+      "learning_rate": 1.1212624584717608e-05,
+      "loss": 0.3427,
+      "step": 1602
+    },
+    {
+      "epoch": 2.396113602391629,
+      "grad_norm": 0.2515968794018395,
+      "learning_rate": 1.1184939091915836e-05,
+      "loss": 0.3542,
+      "step": 1603
+    },
+    {
+      "epoch": 2.397608370702541,
+      "grad_norm": 0.23540141870970982,
+      "learning_rate": 1.1157253599114065e-05,
+      "loss": 0.3563,
+      "step": 1604
+    },
+    {
+      "epoch": 2.399103139013453,
+      "grad_norm": 0.23424803296294505,
+      "learning_rate": 1.1129568106312293e-05,
+      "loss": 0.3528,
+      "step": 1605
+    },
+    {
+      "epoch": 2.400597907324365,
+      "grad_norm": 0.2448113197903768,
+      "learning_rate": 1.1101882613510521e-05,
+      "loss": 0.3552,
+      "step": 1606
+    },
+    {
+      "epoch": 2.4020926756352763,
+      "grad_norm": 0.2572075971438983,
+      "learning_rate": 1.107419712070875e-05,
+      "loss": 0.345,
+      "step": 1607
+    },
+    {
+      "epoch": 2.403587443946188,
+      "grad_norm": 0.24014434574617038,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 0.3453,
+      "step": 1608
+    },
+    {
+      "epoch": 2.4050822122571,
+      "grad_norm": 0.24201392296377913,
+      "learning_rate": 1.1018826135105206e-05,
+      "loss": 0.3503,
+      "step": 1609
+    },
+    {
+      "epoch": 2.406576980568012,
+      "grad_norm": 0.23549397651843973,
+      "learning_rate": 1.0991140642303434e-05,
+      "loss": 0.3552,
+      "step": 1610
+    },
+    {
+      "epoch": 2.408071748878924,
+      "grad_norm": 0.22436126073968177,
+      "learning_rate": 1.0963455149501662e-05,
+      "loss": 0.3347,
+      "step": 1611
+    },
+    {
+      "epoch": 2.4095665171898357,
+      "grad_norm": 0.22624846540709587,
+      "learning_rate": 1.093576965669989e-05,
+      "loss": 0.3473,
+      "step": 1612
+    },
+    {
+      "epoch": 2.4110612855007476,
+      "grad_norm": 0.24425236698538555,
+      "learning_rate": 1.0908084163898118e-05,
+      "loss": 0.3436,
+      "step": 1613
+    },
+    {
+      "epoch": 2.412556053811659,
+      "grad_norm": 0.24464986867762614,
+      "learning_rate": 1.0880398671096347e-05,
+      "loss": 0.3364,
+      "step": 1614
+    },
+    {
+      "epoch": 2.414050822122571,
+      "grad_norm": 0.2403817482065184,
+      "learning_rate": 1.0852713178294575e-05,
+      "loss": 0.3465,
+      "step": 1615
+    },
+    {
+      "epoch": 2.415545590433483,
+      "grad_norm": 0.22549750420566125,
+      "learning_rate": 1.0825027685492803e-05,
+      "loss": 0.3584,
+      "step": 1616
+    },
+    {
+      "epoch": 2.4170403587443947,
+      "grad_norm": 0.24683508781399968,
+      "learning_rate": 1.0797342192691031e-05,
+      "loss": 0.3321,
+      "step": 1617
+    },
+    {
+      "epoch": 2.4185351270553066,
+      "grad_norm": 0.23515344348896475,
+      "learning_rate": 1.076965669988926e-05,
+      "loss": 0.3409,
+      "step": 1618
+    },
+    {
+      "epoch": 2.420029895366218,
+      "grad_norm": 0.22417927224383183,
+      "learning_rate": 1.0741971207087486e-05,
+      "loss": 0.3362,
+      "step": 1619
+    },
+    {
+      "epoch": 2.42152466367713,
+      "grad_norm": 0.23526774193578573,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.3392,
+      "step": 1620
+    },
+    {
+      "epoch": 2.423019431988042,
+      "grad_norm": 0.2515855869471128,
+      "learning_rate": 1.0686600221483942e-05,
+      "loss": 0.3383,
+      "step": 1621
+    },
+    {
+      "epoch": 2.4245142002989537,
+      "grad_norm": 0.23998594902613546,
+      "learning_rate": 1.065891472868217e-05,
+      "loss": 0.3415,
+      "step": 1622
+    },
+    {
+      "epoch": 2.4260089686098656,
+      "grad_norm": 0.257123194617504,
+      "learning_rate": 1.0631229235880399e-05,
+      "loss": 0.351,
+      "step": 1623
+    },
+    {
+      "epoch": 2.4275037369207775,
+      "grad_norm": 0.23718817003620718,
+      "learning_rate": 1.0603543743078627e-05,
+      "loss": 0.347,
+      "step": 1624
+    },
+    {
+      "epoch": 2.428998505231689,
+      "grad_norm": 0.22169651818829825,
+      "learning_rate": 1.0575858250276855e-05,
+      "loss": 0.3386,
+      "step": 1625
+    },
+    {
+      "epoch": 2.430493273542601,
+      "grad_norm": 0.2393421050434122,
+      "learning_rate": 1.0548172757475083e-05,
+      "loss": 0.3401,
+      "step": 1626
+    },
+    {
+      "epoch": 2.4319880418535127,
+      "grad_norm": 0.24131861284469105,
+      "learning_rate": 1.0520487264673311e-05,
+      "loss": 0.3219,
+      "step": 1627
+    },
+    {
+      "epoch": 2.4334828101644246,
+      "grad_norm": 0.25470311773785226,
+      "learning_rate": 1.049280177187154e-05,
+      "loss": 0.3604,
+      "step": 1628
+    },
+    {
+      "epoch": 2.4349775784753365,
+      "grad_norm": 0.23271882473804995,
+      "learning_rate": 1.0465116279069768e-05,
+      "loss": 0.354,
+      "step": 1629
+    },
+    {
+      "epoch": 2.436472346786248,
+      "grad_norm": 0.2408818743275427,
+      "learning_rate": 1.0437430786267996e-05,
+      "loss": 0.3488,
+      "step": 1630
+    },
+    {
+      "epoch": 2.43796711509716,
+      "grad_norm": 0.2533494809956386,
+      "learning_rate": 1.0409745293466224e-05,
+      "loss": 0.3392,
+      "step": 1631
+    },
+    {
+      "epoch": 2.4394618834080717,
+      "grad_norm": 0.23878779644760664,
+      "learning_rate": 1.0382059800664452e-05,
+      "loss": 0.3603,
+      "step": 1632
+    },
+    {
+      "epoch": 2.4409566517189836,
+      "grad_norm": 0.22625712318920632,
+      "learning_rate": 1.035437430786268e-05,
+      "loss": 0.3381,
+      "step": 1633
+    },
+    {
+      "epoch": 2.4424514200298955,
+      "grad_norm": 0.2390655453512499,
+      "learning_rate": 1.0326688815060909e-05,
+      "loss": 0.3518,
+      "step": 1634
+    },
+    {
+      "epoch": 2.4439461883408073,
+      "grad_norm": 0.2423953593676019,
+      "learning_rate": 1.0299003322259137e-05,
+      "loss": 0.3465,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4454409566517192,
+      "grad_norm": 0.2408675422014972,
+      "learning_rate": 1.0271317829457365e-05,
+      "loss": 0.3435,
+      "step": 1636
+    },
+    {
+      "epoch": 2.4469357249626307,
+      "grad_norm": 0.23969880558847836,
+      "learning_rate": 1.0243632336655592e-05,
+      "loss": 0.3481,
+      "step": 1637
+    },
+    {
+      "epoch": 2.4484304932735426,
+      "grad_norm": 0.2349558312451453,
+      "learning_rate": 1.0215946843853822e-05,
+      "loss": 0.3439,
+      "step": 1638
+    },
+    {
+      "epoch": 2.4499252615844545,
+      "grad_norm": 0.231721980470865,
+      "learning_rate": 1.0188261351052048e-05,
+      "loss": 0.3368,
+      "step": 1639
+    },
+    {
+      "epoch": 2.4514200298953663,
+      "grad_norm": 0.22484341394011767,
+      "learning_rate": 1.0160575858250278e-05,
+      "loss": 0.3349,
+      "step": 1640
+    },
+    {
+      "epoch": 2.452914798206278,
+      "grad_norm": 0.23376210011486934,
+      "learning_rate": 1.0132890365448504e-05,
+      "loss": 0.3485,
+      "step": 1641
+    },
+    {
+      "epoch": 2.4544095665171897,
+      "grad_norm": 0.22896782729558493,
+      "learning_rate": 1.0105204872646734e-05,
+      "loss": 0.3476,
+      "step": 1642
+    },
+    {
+      "epoch": 2.4559043348281016,
+      "grad_norm": 0.23311411078812286,
+      "learning_rate": 1.0077519379844961e-05,
+      "loss": 0.3353,
+      "step": 1643
+    },
+    {
+      "epoch": 2.4573991031390134,
+      "grad_norm": 0.2724021743909134,
+      "learning_rate": 1.004983388704319e-05,
+      "loss": 0.3503,
+      "step": 1644
+    },
+    {
+      "epoch": 2.4588938714499253,
+      "grad_norm": 0.2401559806270789,
+      "learning_rate": 1.0022148394241417e-05,
+      "loss": 0.3572,
+      "step": 1645
+    },
+    {
+      "epoch": 2.460388639760837,
+      "grad_norm": 0.2286340746097179,
+      "learning_rate": 9.994462901439645e-06,
+      "loss": 0.3565,
+      "step": 1646
+    },
+    {
+      "epoch": 2.461883408071749,
+      "grad_norm": 0.24077096416541097,
+      "learning_rate": 9.966777408637874e-06,
+      "loss": 0.3444,
+      "step": 1647
+    },
+    {
+      "epoch": 2.4633781763826605,
+      "grad_norm": 0.25837639888821085,
+      "learning_rate": 9.939091915836102e-06,
+      "loss": 0.3562,
+      "step": 1648
+    },
+    {
+      "epoch": 2.4648729446935724,
+      "grad_norm": 0.2647164168182032,
+      "learning_rate": 9.91140642303433e-06,
+      "loss": 0.3577,
+      "step": 1649
+    },
+    {
+      "epoch": 2.4663677130044843,
+      "grad_norm": 0.22940932953967916,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 0.3461,
+      "step": 1650
+    },
+    {
+      "epoch": 2.467862481315396,
+      "grad_norm": 0.23406609471055229,
+      "learning_rate": 9.856035437430786e-06,
+      "loss": 0.3456,
+      "step": 1651
+    },
+    {
+      "epoch": 2.469357249626308,
+      "grad_norm": 0.2369380641108295,
+      "learning_rate": 9.828349944629015e-06,
+      "loss": 0.3435,
+      "step": 1652
+    },
+    {
+      "epoch": 2.4708520179372195,
+      "grad_norm": 0.24869939315864018,
+      "learning_rate": 9.800664451827243e-06,
+      "loss": 0.3525,
+      "step": 1653
+    },
+    {
+      "epoch": 2.4723467862481314,
+      "grad_norm": 0.2296791662317892,
+      "learning_rate": 9.772978959025471e-06,
+      "loss": 0.3379,
+      "step": 1654
+    },
+    {
+      "epoch": 2.4738415545590433,
+      "grad_norm": 0.24200440592701786,
+      "learning_rate": 9.7452934662237e-06,
+      "loss": 0.3421,
+      "step": 1655
+    },
+    {
+      "epoch": 2.475336322869955,
+      "grad_norm": 0.24974555127302797,
+      "learning_rate": 9.717607973421927e-06,
+      "loss": 0.3413,
+      "step": 1656
+    },
+    {
+      "epoch": 2.476831091180867,
+      "grad_norm": 0.23021376892365408,
+      "learning_rate": 9.689922480620156e-06,
+      "loss": 0.3547,
+      "step": 1657
+    },
+    {
+      "epoch": 2.478325859491779,
+      "grad_norm": 0.228355935345127,
+      "learning_rate": 9.662236987818384e-06,
+      "loss": 0.3408,
+      "step": 1658
+    },
+    {
+      "epoch": 2.4798206278026904,
+      "grad_norm": 0.23289772690976976,
+      "learning_rate": 9.634551495016612e-06,
+      "loss": 0.3587,
+      "step": 1659
+    },
+    {
+      "epoch": 2.4813153961136023,
+      "grad_norm": 0.2392026403014523,
+      "learning_rate": 9.60686600221484e-06,
+      "loss": 0.3335,
+      "step": 1660
+    },
+    {
+      "epoch": 2.482810164424514,
+      "grad_norm": 0.21592765238618347,
+      "learning_rate": 9.579180509413067e-06,
+      "loss": 0.3383,
+      "step": 1661
+    },
+    {
+      "epoch": 2.484304932735426,
+      "grad_norm": 0.2282675899240291,
+      "learning_rate": 9.551495016611297e-06,
+      "loss": 0.3563,
+      "step": 1662
+    },
+    {
+      "epoch": 2.485799701046338,
+      "grad_norm": 0.24383191281810568,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.3562,
+      "step": 1663
+    },
+    {
+      "epoch": 2.4872944693572494,
+      "grad_norm": 0.2375854384235552,
+      "learning_rate": 9.496124031007753e-06,
+      "loss": 0.3589,
+      "step": 1664
+    },
+    {
+      "epoch": 2.4887892376681613,
+      "grad_norm": 0.22770423862412123,
+      "learning_rate": 9.46843853820598e-06,
+      "loss": 0.3423,
+      "step": 1665
+    },
+    {
+      "epoch": 2.490284005979073,
+      "grad_norm": 0.23273420282035892,
+      "learning_rate": 9.44075304540421e-06,
+      "loss": 0.3311,
+      "step": 1666
+    },
+    {
+      "epoch": 2.491778774289985,
+      "grad_norm": 0.2318469476200431,
+      "learning_rate": 9.413067552602436e-06,
+      "loss": 0.353,
+      "step": 1667
+    },
+    {
+      "epoch": 2.493273542600897,
+      "grad_norm": 0.23617552196693867,
+      "learning_rate": 9.385382059800666e-06,
+      "loss": 0.354,
+      "step": 1668
+    },
+    {
+      "epoch": 2.494768310911809,
+      "grad_norm": 0.22034632579286184,
+      "learning_rate": 9.357696566998892e-06,
+      "loss": 0.348,
+      "step": 1669
+    },
+    {
+      "epoch": 2.4962630792227207,
+      "grad_norm": 0.2599189248321901,
+      "learning_rate": 9.330011074197122e-06,
+      "loss": 0.3459,
+      "step": 1670
+    },
+    {
+      "epoch": 2.497757847533632,
+      "grad_norm": 0.23629450968229948,
+      "learning_rate": 9.302325581395349e-06,
+      "loss": 0.3474,
+      "step": 1671
+    },
+    {
+      "epoch": 2.499252615844544,
+      "grad_norm": 0.23320878126262856,
+      "learning_rate": 9.274640088593579e-06,
+      "loss": 0.3503,
+      "step": 1672
+    },
+    {
+      "epoch": 2.500747384155456,
+      "grad_norm": 0.2418453903148267,
+      "learning_rate": 9.246954595791805e-06,
+      "loss": 0.3492,
+      "step": 1673
+    },
+    {
+      "epoch": 2.502242152466368,
+      "grad_norm": 0.2331374048266432,
+      "learning_rate": 9.219269102990033e-06,
+      "loss": 0.3481,
+      "step": 1674
+    },
+    {
+      "epoch": 2.5037369207772793,
+      "grad_norm": 0.23299412935863792,
+      "learning_rate": 9.191583610188261e-06,
+      "loss": 0.3354,
+      "step": 1675
+    },
+    {
+      "epoch": 2.505231689088191,
+      "grad_norm": 0.2211111354443292,
+      "learning_rate": 9.16389811738649e-06,
+      "loss": 0.3491,
+      "step": 1676
+    },
+    {
+      "epoch": 2.506726457399103,
+      "grad_norm": 0.26640162111591276,
+      "learning_rate": 9.136212624584718e-06,
+      "loss": 0.3484,
+      "step": 1677
+    },
+    {
+      "epoch": 2.508221225710015,
+      "grad_norm": 0.24483344834847384,
+      "learning_rate": 9.108527131782946e-06,
+      "loss": 0.3495,
+      "step": 1678
+    },
+    {
+      "epoch": 2.509715994020927,
+      "grad_norm": 0.23218642970681416,
+      "learning_rate": 9.080841638981174e-06,
+      "loss": 0.3472,
+      "step": 1679
+    },
+    {
+      "epoch": 2.5112107623318387,
+      "grad_norm": 0.26030723084224144,
+      "learning_rate": 9.053156146179402e-06,
+      "loss": 0.358,
+      "step": 1680
+    },
+    {
+      "epoch": 2.5127055306427506,
+      "grad_norm": 0.22503596496700046,
+      "learning_rate": 9.02547065337763e-06,
+      "loss": 0.3409,
+      "step": 1681
+    },
+    {
+      "epoch": 2.514200298953662,
+      "grad_norm": 0.23868557413588828,
+      "learning_rate": 8.997785160575859e-06,
+      "loss": 0.3373,
+      "step": 1682
+    },
+    {
+      "epoch": 2.515695067264574,
+      "grad_norm": 0.24870337625820704,
+      "learning_rate": 8.970099667774087e-06,
+      "loss": 0.3466,
+      "step": 1683
+    },
+    {
+      "epoch": 2.517189835575486,
+      "grad_norm": 0.2352643954434784,
+      "learning_rate": 8.942414174972315e-06,
+      "loss": 0.3481,
+      "step": 1684
+    },
+    {
+      "epoch": 2.5186846038863977,
+      "grad_norm": 0.22880009548826435,
+      "learning_rate": 8.914728682170543e-06,
+      "loss": 0.3304,
+      "step": 1685
+    },
+    {
+      "epoch": 2.520179372197309,
+      "grad_norm": 0.26497721861863055,
+      "learning_rate": 8.887043189368772e-06,
+      "loss": 0.3491,
+      "step": 1686
+    },
+    {
+      "epoch": 2.521674140508221,
+      "grad_norm": 0.24691323024635092,
+      "learning_rate": 8.859357696567e-06,
+      "loss": 0.3572,
+      "step": 1687
+    },
+    {
+      "epoch": 2.523168908819133,
+      "grad_norm": 0.22469962869247703,
+      "learning_rate": 8.831672203765228e-06,
+      "loss": 0.3268,
+      "step": 1688
+    },
+    {
+      "epoch": 2.524663677130045,
+      "grad_norm": 0.2542213888247696,
+      "learning_rate": 8.803986710963454e-06,
+      "loss": 0.3554,
+      "step": 1689
+    },
+    {
+      "epoch": 2.5261584454409567,
+      "grad_norm": 0.23371880441504678,
+      "learning_rate": 8.776301218161684e-06,
+      "loss": 0.3392,
+      "step": 1690
+    },
+    {
+      "epoch": 2.5276532137518686,
+      "grad_norm": 0.22496336386431415,
+      "learning_rate": 8.74861572535991e-06,
+      "loss": 0.3477,
+      "step": 1691
+    },
+    {
+      "epoch": 2.5291479820627805,
+      "grad_norm": 0.2479384250918033,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 0.3522,
+      "step": 1692
+    },
+    {
+      "epoch": 2.5306427503736924,
+      "grad_norm": 0.25335471006719107,
+      "learning_rate": 8.693244739756367e-06,
+      "loss": 0.3632,
+      "step": 1693
+    },
+    {
+      "epoch": 2.532137518684604,
+      "grad_norm": 0.24431291219362206,
+      "learning_rate": 8.665559246954597e-06,
+      "loss": 0.3536,
+      "step": 1694
+    },
+    {
+      "epoch": 2.5336322869955157,
+      "grad_norm": 0.24100932992418175,
+      "learning_rate": 8.637873754152824e-06,
+      "loss": 0.3485,
+      "step": 1695
+    },
+    {
+      "epoch": 2.5351270553064276,
+      "grad_norm": 0.24294618662909934,
+      "learning_rate": 8.610188261351053e-06,
+      "loss": 0.3527,
+      "step": 1696
+    },
+    {
+      "epoch": 2.5366218236173395,
+      "grad_norm": 0.22680321284372829,
+      "learning_rate": 8.58250276854928e-06,
+      "loss": 0.3472,
+      "step": 1697
+    },
+    {
+      "epoch": 2.538116591928251,
+      "grad_norm": 0.23498209529762668,
+      "learning_rate": 8.55481727574751e-06,
+      "loss": 0.3718,
+      "step": 1698
+    },
+    {
+      "epoch": 2.539611360239163,
+      "grad_norm": 0.25553954810454305,
+      "learning_rate": 8.527131782945736e-06,
+      "loss": 0.3676,
+      "step": 1699
+    },
+    {
+      "epoch": 2.5411061285500747,
+      "grad_norm": 0.23342945199753548,
+      "learning_rate": 8.499446290143966e-06,
+      "loss": 0.3371,
+      "step": 1700
+    },
+    {
+      "epoch": 2.5426008968609866,
+      "grad_norm": 0.25720037762097436,
+      "learning_rate": 8.471760797342193e-06,
+      "loss": 0.3443,
+      "step": 1701
+    },
+    {
+      "epoch": 2.5440956651718984,
+      "grad_norm": 0.22035459307361033,
+      "learning_rate": 8.444075304540421e-06,
+      "loss": 0.338,
+      "step": 1702
+    },
+    {
+      "epoch": 2.5455904334828103,
+      "grad_norm": 0.21549788941639528,
+      "learning_rate": 8.416389811738649e-06,
+      "loss": 0.3532,
+      "step": 1703
+    },
+    {
+      "epoch": 2.547085201793722,
+      "grad_norm": 0.21848186478437615,
+      "learning_rate": 8.388704318936877e-06,
+      "loss": 0.3357,
+      "step": 1704
+    },
+    {
+      "epoch": 2.5485799701046337,
+      "grad_norm": 0.23327402527750138,
+      "learning_rate": 8.361018826135106e-06,
+      "loss": 0.3388,
+      "step": 1705
+    },
+    {
+      "epoch": 2.5500747384155455,
+      "grad_norm": 0.22832265867859164,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.348,
+      "step": 1706
+    },
+    {
+      "epoch": 2.5515695067264574,
+      "grad_norm": 0.2238511445359443,
+      "learning_rate": 8.305647840531562e-06,
+      "loss": 0.3589,
+      "step": 1707
+    },
+    {
+      "epoch": 2.5530642750373693,
+      "grad_norm": 0.22354613970584744,
+      "learning_rate": 8.27796234772979e-06,
+      "loss": 0.3532,
+      "step": 1708
+    },
+    {
+      "epoch": 2.5545590433482808,
+      "grad_norm": 0.23099840562051846,
+      "learning_rate": 8.250276854928018e-06,
+      "loss": 0.3449,
+      "step": 1709
+    },
+    {
+      "epoch": 2.5560538116591927,
+      "grad_norm": 0.23187113553314434,
+      "learning_rate": 8.222591362126247e-06,
+      "loss": 0.3473,
+      "step": 1710
+    },
+    {
+      "epoch": 2.5575485799701045,
+      "grad_norm": 0.227085868954321,
+      "learning_rate": 8.194905869324475e-06,
+      "loss": 0.3504,
+      "step": 1711
+    },
+    {
+      "epoch": 2.5590433482810164,
+      "grad_norm": 0.25167137857961797,
+      "learning_rate": 8.167220376522703e-06,
+      "loss": 0.355,
+      "step": 1712
+    },
+    {
+      "epoch": 2.5605381165919283,
+      "grad_norm": 0.22007096232520068,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.3355,
+      "step": 1713
+    },
+    {
+      "epoch": 2.56203288490284,
+      "grad_norm": 0.2404386808895198,
+      "learning_rate": 8.11184939091916e-06,
+      "loss": 0.3638,
+      "step": 1714
+    },
+    {
+      "epoch": 2.563527653213752,
+      "grad_norm": 0.22216463420547736,
+      "learning_rate": 8.084163898117387e-06,
+      "loss": 0.3325,
+      "step": 1715
+    },
+    {
+      "epoch": 2.5650224215246635,
+      "grad_norm": 0.2321223548683117,
+      "learning_rate": 8.056478405315616e-06,
+      "loss": 0.3621,
+      "step": 1716
+    },
+    {
+      "epoch": 2.5665171898355754,
+      "grad_norm": 0.21206418601337038,
+      "learning_rate": 8.028792912513842e-06,
+      "loss": 0.3397,
+      "step": 1717
+    },
+    {
+      "epoch": 2.5680119581464873,
+      "grad_norm": 0.25426899217214066,
+      "learning_rate": 8.001107419712072e-06,
+      "loss": 0.359,
+      "step": 1718
+    },
+    {
+      "epoch": 2.569506726457399,
+      "grad_norm": 0.22764983890827897,
+      "learning_rate": 7.973421926910299e-06,
+      "loss": 0.3489,
+      "step": 1719
+    },
+    {
+      "epoch": 2.5710014947683106,
+      "grad_norm": 0.23257641315770505,
+      "learning_rate": 7.945736434108528e-06,
+      "loss": 0.3528,
+      "step": 1720
+    },
+    {
+      "epoch": 2.5724962630792225,
+      "grad_norm": 0.2218114496187336,
+      "learning_rate": 7.918050941306755e-06,
+      "loss": 0.3366,
+      "step": 1721
+    },
+    {
+      "epoch": 2.5739910313901344,
+      "grad_norm": 0.2333763103192322,
+      "learning_rate": 7.890365448504985e-06,
+      "loss": 0.345,
+      "step": 1722
+    },
+    {
+      "epoch": 2.5754857997010463,
+      "grad_norm": 0.2326483742613842,
+      "learning_rate": 7.862679955703211e-06,
+      "loss": 0.3396,
+      "step": 1723
+    },
+    {
+      "epoch": 2.576980568011958,
+      "grad_norm": 0.22730864989395216,
+      "learning_rate": 7.834994462901441e-06,
+      "loss": 0.331,
+      "step": 1724
+    },
+    {
+      "epoch": 2.57847533632287,
+      "grad_norm": 0.226403796193293,
+      "learning_rate": 7.807308970099668e-06,
+      "loss": 0.3377,
+      "step": 1725
+    },
+    {
+      "epoch": 2.579970104633782,
+      "grad_norm": 0.24141517721082018,
+      "learning_rate": 7.779623477297898e-06,
+      "loss": 0.3509,
+      "step": 1726
+    },
+    {
+      "epoch": 2.581464872944694,
+      "grad_norm": 0.24091332061237314,
+      "learning_rate": 7.751937984496124e-06,
+      "loss": 0.3431,
+      "step": 1727
+    },
+    {
+      "epoch": 2.5829596412556053,
+      "grad_norm": 0.2613120028653145,
+      "learning_rate": 7.724252491694352e-06,
+      "loss": 0.3517,
+      "step": 1728
+    },
+    {
+      "epoch": 2.584454409566517,
+      "grad_norm": 0.23966939769363316,
+      "learning_rate": 7.69656699889258e-06,
+      "loss": 0.3418,
+      "step": 1729
+    },
+    {
+      "epoch": 2.585949177877429,
+      "grad_norm": 0.22433234473802238,
+      "learning_rate": 7.668881506090809e-06,
+      "loss": 0.3488,
+      "step": 1730
+    },
+    {
+      "epoch": 2.587443946188341,
+      "grad_norm": 0.23466731344225947,
+      "learning_rate": 7.641196013289037e-06,
+      "loss": 0.3406,
+      "step": 1731
+    },
+    {
+      "epoch": 2.5889387144992524,
+      "grad_norm": 0.23566456562606097,
+      "learning_rate": 7.613510520487264e-06,
+      "loss": 0.3516,
+      "step": 1732
+    },
+    {
+      "epoch": 2.5904334828101643,
+      "grad_norm": 0.22634571454516944,
+      "learning_rate": 7.585825027685493e-06,
+      "loss": 0.3423,
+      "step": 1733
+    },
+    {
+      "epoch": 2.591928251121076,
+      "grad_norm": 0.2536709784548154,
+      "learning_rate": 7.558139534883721e-06,
+      "loss": 0.347,
+      "step": 1734
+    },
+    {
+      "epoch": 2.593423019431988,
+      "grad_norm": 0.23815412776518058,
+      "learning_rate": 7.53045404208195e-06,
+      "loss": 0.3371,
+      "step": 1735
+    },
+    {
+      "epoch": 2.5949177877429,
+      "grad_norm": 0.23093557969321732,
+      "learning_rate": 7.502768549280177e-06,
+      "loss": 0.3475,
+      "step": 1736
+    },
+    {
+      "epoch": 2.596412556053812,
+      "grad_norm": 0.22118450215516994,
+      "learning_rate": 7.475083056478406e-06,
+      "loss": 0.3443,
+      "step": 1737
+    },
+    {
+      "epoch": 2.5979073243647237,
+      "grad_norm": 0.23422853045992287,
+      "learning_rate": 7.447397563676633e-06,
+      "loss": 0.3546,
+      "step": 1738
+    },
+    {
+      "epoch": 2.599402092675635,
+      "grad_norm": 0.2459011577523459,
+      "learning_rate": 7.4197120708748625e-06,
+      "loss": 0.3478,
+      "step": 1739
+    },
+    {
+      "epoch": 2.600896860986547,
+      "grad_norm": 0.24860793453574415,
+      "learning_rate": 7.39202657807309e-06,
+      "loss": 0.336,
+      "step": 1740
+    },
+    {
+      "epoch": 2.602391629297459,
+      "grad_norm": 0.25188203080823973,
+      "learning_rate": 7.364341085271319e-06,
+      "loss": 0.3442,
+      "step": 1741
+    },
+    {
+      "epoch": 2.603886397608371,
+      "grad_norm": 0.206879483808265,
+      "learning_rate": 7.336655592469546e-06,
+      "loss": 0.3228,
+      "step": 1742
+    },
+    {
+      "epoch": 2.6053811659192823,
+      "grad_norm": 0.23284013894085195,
+      "learning_rate": 7.308970099667775e-06,
+      "loss": 0.3407,
+      "step": 1743
+    },
+    {
+      "epoch": 2.606875934230194,
+      "grad_norm": 0.24929844716911245,
+      "learning_rate": 7.281284606866003e-06,
+      "loss": 0.3504,
+      "step": 1744
+    },
+    {
+      "epoch": 2.608370702541106,
+      "grad_norm": 0.22583057412992175,
+      "learning_rate": 7.25359911406423e-06,
+      "loss": 0.3298,
+      "step": 1745
+    },
+    {
+      "epoch": 2.609865470852018,
+      "grad_norm": 0.2329889209038992,
+      "learning_rate": 7.225913621262459e-06,
+      "loss": 0.328,
+      "step": 1746
+    },
+    {
+      "epoch": 2.61136023916293,
+      "grad_norm": 0.24702388605782574,
+      "learning_rate": 7.198228128460686e-06,
+      "loss": 0.3396,
+      "step": 1747
+    },
+    {
+      "epoch": 2.6128550074738417,
+      "grad_norm": 0.2297075298688945,
+      "learning_rate": 7.170542635658915e-06,
+      "loss": 0.3366,
+      "step": 1748
+    },
+    {
+      "epoch": 2.6143497757847536,
+      "grad_norm": 0.23204020219890423,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.347,
+      "step": 1749
+    },
+    {
+      "epoch": 2.615844544095665,
+      "grad_norm": 0.22240754327027865,
+      "learning_rate": 7.115171650055372e-06,
+      "loss": 0.3426,
+      "step": 1750
+    },
+    {
+      "epoch": 2.617339312406577,
+      "grad_norm": 0.216440083883795,
+      "learning_rate": 7.087486157253599e-06,
+      "loss": 0.3453,
+      "step": 1751
+    },
+    {
+      "epoch": 2.618834080717489,
+      "grad_norm": 0.2238171571214644,
+      "learning_rate": 7.059800664451828e-06,
+      "loss": 0.3422,
+      "step": 1752
+    },
+    {
+      "epoch": 2.6203288490284007,
+      "grad_norm": 0.2291954604708445,
+      "learning_rate": 7.0321151716500555e-06,
+      "loss": 0.3452,
+      "step": 1753
+    },
+    {
+      "epoch": 2.621823617339312,
+      "grad_norm": 0.2538392539670155,
+      "learning_rate": 7.0044296788482845e-06,
+      "loss": 0.3612,
+      "step": 1754
+    },
+    {
+      "epoch": 2.623318385650224,
+      "grad_norm": 0.24098835347886044,
+      "learning_rate": 6.976744186046512e-06,
+      "loss": 0.3437,
+      "step": 1755
+    },
+    {
+      "epoch": 2.624813153961136,
+      "grad_norm": 0.2389697532590079,
+      "learning_rate": 6.949058693244741e-06,
+      "loss": 0.3446,
+      "step": 1756
+    },
+    {
+      "epoch": 2.626307922272048,
+      "grad_norm": 0.22342612830356817,
+      "learning_rate": 6.921373200442968e-06,
+      "loss": 0.3326,
+      "step": 1757
+    },
+    {
+      "epoch": 2.6278026905829597,
+      "grad_norm": 0.24930892055741222,
+      "learning_rate": 6.893687707641196e-06,
+      "loss": 0.3571,
+      "step": 1758
+    },
+    {
+      "epoch": 2.6292974588938716,
+      "grad_norm": 0.23617213355131875,
+      "learning_rate": 6.866002214839425e-06,
+      "loss": 0.3518,
+      "step": 1759
+    },
+    {
+      "epoch": 2.6307922272047835,
+      "grad_norm": 0.22910357833499392,
+      "learning_rate": 6.838316722037652e-06,
+      "loss": 0.356,
+      "step": 1760
+    },
+    {
+      "epoch": 2.6322869955156953,
+      "grad_norm": 0.22168805661174296,
+      "learning_rate": 6.810631229235881e-06,
+      "loss": 0.3429,
+      "step": 1761
+    },
+    {
+      "epoch": 2.633781763826607,
+      "grad_norm": 0.22516682631778526,
+      "learning_rate": 6.782945736434108e-06,
+      "loss": 0.3567,
+      "step": 1762
+    },
+    {
+      "epoch": 2.6352765321375187,
+      "grad_norm": 0.21817493499545296,
+      "learning_rate": 6.7552602436323374e-06,
+      "loss": 0.3377,
+      "step": 1763
+    },
+    {
+      "epoch": 2.6367713004484306,
+      "grad_norm": 0.22285314061324607,
+      "learning_rate": 6.727574750830565e-06,
+      "loss": 0.3502,
+      "step": 1764
+    },
+    {
+      "epoch": 2.6382660687593424,
+      "grad_norm": 0.22421719432449908,
+      "learning_rate": 6.699889258028794e-06,
+      "loss": 0.3418,
+      "step": 1765
+    },
+    {
+      "epoch": 2.639760837070254,
+      "grad_norm": 0.22815365665436424,
+      "learning_rate": 6.672203765227021e-06,
+      "loss": 0.3417,
+      "step": 1766
+    },
+    {
+      "epoch": 2.6412556053811658,
+      "grad_norm": 0.2359180134082138,
+      "learning_rate": 6.64451827242525e-06,
+      "loss": 0.358,
+      "step": 1767
+    },
+    {
+      "epoch": 2.6427503736920777,
+      "grad_norm": 0.21945140914396835,
+      "learning_rate": 6.6168327796234776e-06,
+      "loss": 0.3414,
+      "step": 1768
+    },
+    {
+      "epoch": 2.6442451420029895,
+      "grad_norm": 0.23896795226337073,
+      "learning_rate": 6.589147286821707e-06,
+      "loss": 0.3427,
+      "step": 1769
+    },
+    {
+      "epoch": 2.6457399103139014,
+      "grad_norm": 0.22891996692142982,
+      "learning_rate": 6.561461794019934e-06,
+      "loss": 0.3436,
+      "step": 1770
+    },
+    {
+      "epoch": 2.6472346786248133,
+      "grad_norm": 0.23761138232610027,
+      "learning_rate": 6.533776301218161e-06,
+      "loss": 0.3341,
+      "step": 1771
+    },
+    {
+      "epoch": 2.648729446935725,
+      "grad_norm": 0.24213795585355222,
+      "learning_rate": 6.50609080841639e-06,
+      "loss": 0.3521,
+      "step": 1772
+    },
+    {
+      "epoch": 2.6502242152466366,
+      "grad_norm": 0.2232389401736531,
+      "learning_rate": 6.478405315614618e-06,
+      "loss": 0.3363,
+      "step": 1773
+    },
+    {
+      "epoch": 2.6517189835575485,
+      "grad_norm": 0.21501573521189474,
+      "learning_rate": 6.450719822812847e-06,
+      "loss": 0.3207,
+      "step": 1774
+    },
+    {
+      "epoch": 2.6532137518684604,
+      "grad_norm": 0.2333400751542196,
+      "learning_rate": 6.423034330011074e-06,
+      "loss": 0.3678,
+      "step": 1775
+    },
+    {
+      "epoch": 2.6547085201793723,
+      "grad_norm": 0.23618346696846812,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 0.3391,
+      "step": 1776
+    },
+    {
+      "epoch": 2.6562032884902838,
+      "grad_norm": 0.2314023615183173,
+      "learning_rate": 6.3676633444075305e-06,
+      "loss": 0.3464,
+      "step": 1777
+    },
+    {
+      "epoch": 2.6576980568011956,
+      "grad_norm": 0.2352988668770188,
+      "learning_rate": 6.3399778516057595e-06,
+      "loss": 0.3527,
+      "step": 1778
+    },
+    {
+      "epoch": 2.6591928251121075,
+      "grad_norm": 0.22703400501274704,
+      "learning_rate": 6.312292358803987e-06,
+      "loss": 0.3466,
+      "step": 1779
+    },
+    {
+      "epoch": 2.6606875934230194,
+      "grad_norm": 0.2443188350324432,
+      "learning_rate": 6.284606866002216e-06,
+      "loss": 0.3463,
+      "step": 1780
+    },
+    {
+      "epoch": 2.6621823617339313,
+      "grad_norm": 0.2393782878863728,
+      "learning_rate": 6.256921373200443e-06,
+      "loss": 0.3533,
+      "step": 1781
+    },
+    {
+      "epoch": 2.663677130044843,
+      "grad_norm": 0.23046812072119735,
+      "learning_rate": 6.2292358803986714e-06,
+      "loss": 0.3503,
+      "step": 1782
+    },
+    {
+      "epoch": 2.665171898355755,
+      "grad_norm": 0.2144259910660163,
+      "learning_rate": 6.2015503875969e-06,
+      "loss": 0.3428,
+      "step": 1783
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.2536036964120041,
+      "learning_rate": 6.173864894795128e-06,
+      "loss": 0.3611,
+      "step": 1784
+    },
+    {
+      "epoch": 2.6681614349775784,
+      "grad_norm": 0.22400784280588265,
+      "learning_rate": 6.146179401993356e-06,
+      "loss": 0.3539,
+      "step": 1785
+    },
+    {
+      "epoch": 2.6696562032884903,
+      "grad_norm": 0.21747018993892098,
+      "learning_rate": 6.118493909191584e-06,
+      "loss": 0.3442,
+      "step": 1786
+    },
+    {
+      "epoch": 2.671150971599402,
+      "grad_norm": 0.22220318171419962,
+      "learning_rate": 6.090808416389812e-06,
+      "loss": 0.3467,
+      "step": 1787
+    },
+    {
+      "epoch": 2.672645739910314,
+      "grad_norm": 0.22127874338015066,
+      "learning_rate": 6.063122923588041e-06,
+      "loss": 0.3337,
+      "step": 1788
+    },
+    {
+      "epoch": 2.6741405082212255,
+      "grad_norm": 0.2159045118825795,
+      "learning_rate": 6.035437430786269e-06,
+      "loss": 0.3362,
+      "step": 1789
+    },
+    {
+      "epoch": 2.6756352765321374,
+      "grad_norm": 0.21672740435341065,
+      "learning_rate": 6.007751937984497e-06,
+      "loss": 0.3423,
+      "step": 1790
+    },
+    {
+      "epoch": 2.6771300448430493,
+      "grad_norm": 0.22029322071862198,
+      "learning_rate": 5.980066445182725e-06,
+      "loss": 0.3296,
+      "step": 1791
+    },
+    {
+      "epoch": 2.678624813153961,
+      "grad_norm": 0.21614754465411762,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 0.3517,
+      "step": 1792
+    },
+    {
+      "epoch": 2.680119581464873,
+      "grad_norm": 0.22264371583511636,
+      "learning_rate": 5.924695459579181e-06,
+      "loss": 0.3321,
+      "step": 1793
+    },
+    {
+      "epoch": 2.681614349775785,
+      "grad_norm": 0.21471674275192468,
+      "learning_rate": 5.897009966777409e-06,
+      "loss": 0.3481,
+      "step": 1794
+    },
+    {
+      "epoch": 2.683109118086697,
+      "grad_norm": 0.21170372316762678,
+      "learning_rate": 5.869324473975637e-06,
+      "loss": 0.3495,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6846038863976083,
+      "grad_norm": 0.23844996328908794,
+      "learning_rate": 5.841638981173865e-06,
+      "loss": 0.3324,
+      "step": 1796
+    },
+    {
+      "epoch": 2.68609865470852,
+      "grad_norm": 0.21227175105076126,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.3398,
+      "step": 1797
+    },
+    {
+      "epoch": 2.687593423019432,
+      "grad_norm": 0.2214191685474946,
+      "learning_rate": 5.786267995570322e-06,
+      "loss": 0.3303,
+      "step": 1798
+    },
+    {
+      "epoch": 2.689088191330344,
+      "grad_norm": 0.22482498436569123,
+      "learning_rate": 5.75858250276855e-06,
+      "loss": 0.3363,
+      "step": 1799
+    },
+    {
+      "epoch": 2.6905829596412554,
+      "grad_norm": 0.2192943029262021,
+      "learning_rate": 5.730897009966778e-06,
+      "loss": 0.3424,
+      "step": 1800
+    },
+    {
+      "epoch": 2.6920777279521673,
+      "grad_norm": 0.24186839478870298,
+      "learning_rate": 5.703211517165006e-06,
+      "loss": 0.3608,
+      "step": 1801
+    },
+    {
+      "epoch": 2.693572496263079,
+      "grad_norm": 0.2258121722756959,
+      "learning_rate": 5.6755260243632345e-06,
+      "loss": 0.3493,
+      "step": 1802
+    },
+    {
+      "epoch": 2.695067264573991,
+      "grad_norm": 0.2046833577493049,
+      "learning_rate": 5.647840531561463e-06,
+      "loss": 0.3341,
+      "step": 1803
+    },
+    {
+      "epoch": 2.696562032884903,
+      "grad_norm": 0.2249001062998578,
+      "learning_rate": 5.620155038759691e-06,
+      "loss": 0.3284,
+      "step": 1804
+    },
+    {
+      "epoch": 2.698056801195815,
+      "grad_norm": 0.2341562029996612,
+      "learning_rate": 5.592469545957918e-06,
+      "loss": 0.3429,
+      "step": 1805
+    },
+    {
+      "epoch": 2.6995515695067267,
+      "grad_norm": 0.22652306797698712,
+      "learning_rate": 5.5647840531561464e-06,
+      "loss": 0.3412,
+      "step": 1806
+    },
+    {
+      "epoch": 2.701046337817638,
+      "grad_norm": 0.2308476990731487,
+      "learning_rate": 5.537098560354375e-06,
+      "loss": 0.3392,
+      "step": 1807
+    },
+    {
+      "epoch": 2.70254110612855,
+      "grad_norm": 0.22107899157710142,
+      "learning_rate": 5.509413067552603e-06,
+      "loss": 0.3437,
+      "step": 1808
+    },
+    {
+      "epoch": 2.704035874439462,
+      "grad_norm": 0.23188990303294915,
+      "learning_rate": 5.481727574750831e-06,
+      "loss": 0.3462,
+      "step": 1809
+    },
+    {
+      "epoch": 2.705530642750374,
+      "grad_norm": 0.24938800287922946,
+      "learning_rate": 5.454042081949059e-06,
+      "loss": 0.342,
+      "step": 1810
+    },
+    {
+      "epoch": 2.7070254110612852,
+      "grad_norm": 0.2613518285042722,
+      "learning_rate": 5.426356589147287e-06,
+      "loss": 0.3407,
+      "step": 1811
+    },
+    {
+      "epoch": 2.708520179372197,
+      "grad_norm": 0.23654908576244107,
+      "learning_rate": 5.398671096345516e-06,
+      "loss": 0.35,
+      "step": 1812
+    },
+    {
+      "epoch": 2.710014947683109,
+      "grad_norm": 0.23757294143269403,
+      "learning_rate": 5.370985603543743e-06,
+      "loss": 0.3366,
+      "step": 1813
+    },
+    {
+      "epoch": 2.711509715994021,
+      "grad_norm": 0.21143957746329659,
+      "learning_rate": 5.343300110741971e-06,
+      "loss": 0.3636,
+      "step": 1814
+    },
+    {
+      "epoch": 2.713004484304933,
+      "grad_norm": 0.20533386657001457,
+      "learning_rate": 5.315614617940199e-06,
+      "loss": 0.3258,
+      "step": 1815
+    },
+    {
+      "epoch": 2.7144992526158447,
+      "grad_norm": 0.21431498187644712,
+      "learning_rate": 5.2879291251384275e-06,
+      "loss": 0.3298,
+      "step": 1816
+    },
+    {
+      "epoch": 2.7159940209267566,
+      "grad_norm": 0.2624099763379685,
+      "learning_rate": 5.260243632336656e-06,
+      "loss": 0.3495,
+      "step": 1817
+    },
+    {
+      "epoch": 2.717488789237668,
+      "grad_norm": 0.21632112757387492,
+      "learning_rate": 5.232558139534884e-06,
+      "loss": 0.3489,
+      "step": 1818
+    },
+    {
+      "epoch": 2.71898355754858,
+      "grad_norm": 0.21417849487231005,
+      "learning_rate": 5.204872646733112e-06,
+      "loss": 0.3344,
+      "step": 1819
+    },
+    {
+      "epoch": 2.720478325859492,
+      "grad_norm": 0.2212096341978614,
+      "learning_rate": 5.17718715393134e-06,
+      "loss": 0.3633,
+      "step": 1820
+    },
+    {
+      "epoch": 2.7219730941704037,
+      "grad_norm": 0.22416229067052926,
+      "learning_rate": 5.1495016611295685e-06,
+      "loss": 0.3608,
+      "step": 1821
+    },
+    {
+      "epoch": 2.7234678624813156,
+      "grad_norm": 0.2145578919369725,
+      "learning_rate": 5.121816168327796e-06,
+      "loss": 0.3409,
+      "step": 1822
+    },
+    {
+      "epoch": 2.724962630792227,
+      "grad_norm": 0.2131898528765146,
+      "learning_rate": 5.094130675526024e-06,
+      "loss": 0.3448,
+      "step": 1823
+    },
+    {
+      "epoch": 2.726457399103139,
+      "grad_norm": 0.22156663918995964,
+      "learning_rate": 5.066445182724252e-06,
+      "loss": 0.3394,
+      "step": 1824
+    },
+    {
+      "epoch": 2.7279521674140508,
+      "grad_norm": 0.22660184925674876,
+      "learning_rate": 5.0387596899224804e-06,
+      "loss": 0.3629,
+      "step": 1825
+    },
+    {
+      "epoch": 2.7294469357249627,
+      "grad_norm": 0.22133930750995479,
+      "learning_rate": 5.011074197120709e-06,
+      "loss": 0.341,
+      "step": 1826
+    },
+    {
+      "epoch": 2.7309417040358746,
+      "grad_norm": 0.22985645594210896,
+      "learning_rate": 4.983388704318937e-06,
+      "loss": 0.3423,
+      "step": 1827
+    },
+    {
+      "epoch": 2.7324364723467864,
+      "grad_norm": 0.19675774189931972,
+      "learning_rate": 4.955703211517165e-06,
+      "loss": 0.3372,
+      "step": 1828
+    },
+    {
+      "epoch": 2.7339312406576983,
+      "grad_norm": 0.21425264009522388,
+      "learning_rate": 4.928017718715393e-06,
+      "loss": 0.34,
+      "step": 1829
+    },
+    {
+      "epoch": 2.7354260089686098,
+      "grad_norm": 0.22232224375455825,
+      "learning_rate": 4.900332225913621e-06,
+      "loss": 0.3509,
+      "step": 1830
+    },
+    {
+      "epoch": 2.7369207772795217,
+      "grad_norm": 0.22124036117543916,
+      "learning_rate": 4.87264673311185e-06,
+      "loss": 0.3244,
+      "step": 1831
+    },
+    {
+      "epoch": 2.7384155455904335,
+      "grad_norm": 0.2175828250900761,
+      "learning_rate": 4.844961240310078e-06,
+      "loss": 0.3608,
+      "step": 1832
+    },
+    {
+      "epoch": 2.7399103139013454,
+      "grad_norm": 0.23042614801181316,
+      "learning_rate": 4.817275747508306e-06,
+      "loss": 0.3547,
+      "step": 1833
+    },
+    {
+      "epoch": 2.741405082212257,
+      "grad_norm": 0.21168844072564844,
+      "learning_rate": 4.789590254706533e-06,
+      "loss": 0.3334,
+      "step": 1834
+    },
+    {
+      "epoch": 2.7428998505231688,
+      "grad_norm": 0.20764803137887092,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.3363,
+      "step": 1835
+    },
+    {
+      "epoch": 2.7443946188340806,
+      "grad_norm": 0.21746850150509336,
+      "learning_rate": 4.73421926910299e-06,
+      "loss": 0.3391,
+      "step": 1836
+    },
+    {
+      "epoch": 2.7458893871449925,
+      "grad_norm": 0.22892184620193504,
+      "learning_rate": 4.706533776301218e-06,
+      "loss": 0.3452,
+      "step": 1837
+    },
+    {
+      "epoch": 2.7473841554559044,
+      "grad_norm": 0.2355644554392448,
+      "learning_rate": 4.678848283499446e-06,
+      "loss": 0.3401,
+      "step": 1838
+    },
+    {
+      "epoch": 2.7488789237668163,
+      "grad_norm": 0.22400364430704886,
+      "learning_rate": 4.651162790697674e-06,
+      "loss": 0.338,
+      "step": 1839
+    },
+    {
+      "epoch": 2.750373692077728,
+      "grad_norm": 0.21699457129549654,
+      "learning_rate": 4.6234772978959025e-06,
+      "loss": 0.338,
+      "step": 1840
+    },
+    {
+      "epoch": 2.7518684603886396,
+      "grad_norm": 0.21628560834198626,
+      "learning_rate": 4.595791805094131e-06,
+      "loss": 0.3458,
+      "step": 1841
+    },
+    {
+      "epoch": 2.7533632286995515,
+      "grad_norm": 0.2194846768198985,
+      "learning_rate": 4.568106312292359e-06,
+      "loss": 0.3371,
+      "step": 1842
+    },
+    {
+      "epoch": 2.7548579970104634,
+      "grad_norm": 0.21205470770504115,
+      "learning_rate": 4.540420819490587e-06,
+      "loss": 0.3454,
+      "step": 1843
+    },
+    {
+      "epoch": 2.7563527653213753,
+      "grad_norm": 0.23496366885403291,
+      "learning_rate": 4.512735326688815e-06,
+      "loss": 0.3457,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7578475336322867,
+      "grad_norm": 0.24977909409559299,
+      "learning_rate": 4.4850498338870435e-06,
+      "loss": 0.3592,
+      "step": 1845
+    },
+    {
+      "epoch": 2.7593423019431986,
+      "grad_norm": 0.23091818074767417,
+      "learning_rate": 4.457364341085272e-06,
+      "loss": 0.3408,
+      "step": 1846
+    },
+    {
+      "epoch": 2.7608370702541105,
+      "grad_norm": 0.2208446622611314,
+      "learning_rate": 4.4296788482835e-06,
+      "loss": 0.3277,
+      "step": 1847
+    },
+    {
+      "epoch": 2.7623318385650224,
+      "grad_norm": 0.21849992438203533,
+      "learning_rate": 4.401993355481727e-06,
+      "loss": 0.3467,
+      "step": 1848
+    },
+    {
+      "epoch": 2.7638266068759343,
+      "grad_norm": 0.2409319388153108,
+      "learning_rate": 4.374307862679955e-06,
+      "loss": 0.3437,
+      "step": 1849
+    },
+    {
+      "epoch": 2.765321375186846,
+      "grad_norm": 0.2061675967285374,
+      "learning_rate": 4.346622369878184e-06,
+      "loss": 0.34,
+      "step": 1850
+    },
+    {
+      "epoch": 2.766816143497758,
+      "grad_norm": 0.2079552897285853,
+      "learning_rate": 4.318936877076412e-06,
+      "loss": 0.3513,
+      "step": 1851
+    },
+    {
+      "epoch": 2.7683109118086695,
+      "grad_norm": 0.22664209669172278,
+      "learning_rate": 4.29125138427464e-06,
+      "loss": 0.3473,
+      "step": 1852
+    },
+    {
+      "epoch": 2.7698056801195814,
+      "grad_norm": 0.2395815026526975,
+      "learning_rate": 4.263565891472868e-06,
+      "loss": 0.3414,
+      "step": 1853
+    },
+    {
+      "epoch": 2.7713004484304933,
+      "grad_norm": 0.2249967475532358,
+      "learning_rate": 4.235880398671096e-06,
+      "loss": 0.3513,
+      "step": 1854
+    },
+    {
+      "epoch": 2.772795216741405,
+      "grad_norm": 0.2128907910399229,
+      "learning_rate": 4.2081949058693246e-06,
+      "loss": 0.3437,
+      "step": 1855
+    },
+    {
+      "epoch": 2.774289985052317,
+      "grad_norm": 0.20647431006736722,
+      "learning_rate": 4.180509413067553e-06,
+      "loss": 0.3309,
+      "step": 1856
+    },
+    {
+      "epoch": 2.7757847533632285,
+      "grad_norm": 0.2380914250689674,
+      "learning_rate": 4.152823920265781e-06,
+      "loss": 0.3577,
+      "step": 1857
+    },
+    {
+      "epoch": 2.7772795216741404,
+      "grad_norm": 0.21004335796822707,
+      "learning_rate": 4.125138427464009e-06,
+      "loss": 0.3303,
+      "step": 1858
+    },
+    {
+      "epoch": 2.7787742899850523,
+      "grad_norm": 0.22555962142776617,
+      "learning_rate": 4.097452934662237e-06,
+      "loss": 0.3515,
+      "step": 1859
+    },
+    {
+      "epoch": 2.780269058295964,
+      "grad_norm": 0.21048879701838388,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 0.334,
+      "step": 1860
+    },
+    {
+      "epoch": 2.781763826606876,
+      "grad_norm": 0.23066461986699155,
+      "learning_rate": 4.042081949058694e-06,
+      "loss": 0.3546,
+      "step": 1861
+    },
+    {
+      "epoch": 2.783258594917788,
+      "grad_norm": 0.22402040158499376,
+      "learning_rate": 4.014396456256921e-06,
+      "loss": 0.3539,
+      "step": 1862
+    },
+    {
+      "epoch": 2.7847533632287,
+      "grad_norm": 0.21724826287735982,
+      "learning_rate": 3.986710963455149e-06,
+      "loss": 0.3428,
+      "step": 1863
+    },
+    {
+      "epoch": 2.7862481315396113,
+      "grad_norm": 0.22424604870817222,
+      "learning_rate": 3.9590254706533775e-06,
+      "loss": 0.3226,
+      "step": 1864
+    },
+    {
+      "epoch": 2.787742899850523,
+      "grad_norm": 0.20750615183773335,
+      "learning_rate": 3.931339977851606e-06,
+      "loss": 0.3413,
+      "step": 1865
+    },
+    {
+      "epoch": 2.789237668161435,
+      "grad_norm": 0.22582275646140496,
+      "learning_rate": 3.903654485049834e-06,
+      "loss": 0.3588,
+      "step": 1866
+    },
+    {
+      "epoch": 2.790732436472347,
+      "grad_norm": 0.21540081359698546,
+      "learning_rate": 3.875968992248062e-06,
+      "loss": 0.3387,
+      "step": 1867
+    },
+    {
+      "epoch": 2.7922272047832584,
+      "grad_norm": 0.23328466365888145,
+      "learning_rate": 3.84828349944629e-06,
+      "loss": 0.3316,
+      "step": 1868
+    },
+    {
+      "epoch": 2.7937219730941703,
+      "grad_norm": 0.23346437874529052,
+      "learning_rate": 3.8205980066445185e-06,
+      "loss": 0.3338,
+      "step": 1869
+    },
+    {
+      "epoch": 2.795216741405082,
+      "grad_norm": 0.2215072158295504,
+      "learning_rate": 3.7929125138427466e-06,
+      "loss": 0.3402,
+      "step": 1870
+    },
+    {
+      "epoch": 2.796711509715994,
+      "grad_norm": 0.23345519719875976,
+      "learning_rate": 3.765227021040975e-06,
+      "loss": 0.3345,
+      "step": 1871
+    },
+    {
+      "epoch": 2.798206278026906,
+      "grad_norm": 0.22435466373548252,
+      "learning_rate": 3.737541528239203e-06,
+      "loss": 0.3464,
+      "step": 1872
+    },
+    {
+      "epoch": 2.799701046337818,
+      "grad_norm": 0.21717244855454707,
+      "learning_rate": 3.7098560354374312e-06,
+      "loss": 0.3287,
+      "step": 1873
+    },
+    {
+      "epoch": 2.8011958146487297,
+      "grad_norm": 0.226671395277339,
+      "learning_rate": 3.6821705426356594e-06,
+      "loss": 0.3475,
+      "step": 1874
+    },
+    {
+      "epoch": 2.802690582959641,
+      "grad_norm": 0.22236776284993368,
+      "learning_rate": 3.6544850498338876e-06,
+      "loss": 0.3417,
+      "step": 1875
+    },
+    {
+      "epoch": 2.804185351270553,
+      "grad_norm": 0.21996227689300382,
+      "learning_rate": 3.626799557032115e-06,
+      "loss": 0.3328,
+      "step": 1876
+    },
+    {
+      "epoch": 2.805680119581465,
+      "grad_norm": 0.22066118104510704,
+      "learning_rate": 3.599114064230343e-06,
+      "loss": 0.3446,
+      "step": 1877
+    },
+    {
+      "epoch": 2.807174887892377,
+      "grad_norm": 0.21570552716644853,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.3291,
+      "step": 1878
+    },
+    {
+      "epoch": 2.8086696562032882,
+      "grad_norm": 0.2152071115415961,
+      "learning_rate": 3.5437430786267996e-06,
+      "loss": 0.3586,
+      "step": 1879
+    },
+    {
+      "epoch": 2.8101644245142,
+      "grad_norm": 0.20802111653624747,
+      "learning_rate": 3.5160575858250277e-06,
+      "loss": 0.3504,
+      "step": 1880
+    },
+    {
+      "epoch": 2.811659192825112,
+      "grad_norm": 0.22072556850686775,
+      "learning_rate": 3.488372093023256e-06,
+      "loss": 0.3535,
+      "step": 1881
+    },
+    {
+      "epoch": 2.813153961136024,
+      "grad_norm": 0.21846489758856738,
+      "learning_rate": 3.460686600221484e-06,
+      "loss": 0.354,
+      "step": 1882
+    },
+    {
+      "epoch": 2.814648729446936,
+      "grad_norm": 0.21756597393393506,
+      "learning_rate": 3.4330011074197123e-06,
+      "loss": 0.3399,
+      "step": 1883
+    },
+    {
+      "epoch": 2.8161434977578477,
+      "grad_norm": 0.2118335951665565,
+      "learning_rate": 3.4053156146179405e-06,
+      "loss": 0.3382,
+      "step": 1884
+    },
+    {
+      "epoch": 2.8176382660687596,
+      "grad_norm": 0.20295995899869512,
+      "learning_rate": 3.3776301218161687e-06,
+      "loss": 0.3462,
+      "step": 1885
+    },
+    {
+      "epoch": 2.819133034379671,
+      "grad_norm": 0.20859531629005412,
+      "learning_rate": 3.349944629014397e-06,
+      "loss": 0.3304,
+      "step": 1886
+    },
+    {
+      "epoch": 2.820627802690583,
+      "grad_norm": 0.20790705977561386,
+      "learning_rate": 3.322259136212625e-06,
+      "loss": 0.3413,
+      "step": 1887
+    },
+    {
+      "epoch": 2.8221225710014948,
+      "grad_norm": 0.20462294140553627,
+      "learning_rate": 3.2945736434108533e-06,
+      "loss": 0.3567,
+      "step": 1888
+    },
+    {
+      "epoch": 2.8236173393124067,
+      "grad_norm": 0.22099349592605477,
+      "learning_rate": 3.2668881506090807e-06,
+      "loss": 0.3561,
+      "step": 1889
+    },
+    {
+      "epoch": 2.8251121076233185,
+      "grad_norm": 0.20124109255366243,
+      "learning_rate": 3.239202657807309e-06,
+      "loss": 0.3238,
+      "step": 1890
+    },
+    {
+      "epoch": 2.82660687593423,
+      "grad_norm": 0.2104809441709824,
+      "learning_rate": 3.211517165005537e-06,
+      "loss": 0.3403,
+      "step": 1891
+    },
+    {
+      "epoch": 2.828101644245142,
+      "grad_norm": 0.20453341126669386,
+      "learning_rate": 3.1838316722037652e-06,
+      "loss": 0.3372,
+      "step": 1892
+    },
+    {
+      "epoch": 2.8295964125560538,
+      "grad_norm": 0.21910531422941892,
+      "learning_rate": 3.1561461794019934e-06,
+      "loss": 0.3381,
+      "step": 1893
+    },
+    {
+      "epoch": 2.8310911808669657,
+      "grad_norm": 0.2321345277116509,
+      "learning_rate": 3.1284606866002216e-06,
+      "loss": 0.3457,
+      "step": 1894
+    },
+    {
+      "epoch": 2.8325859491778775,
+      "grad_norm": 0.229462549981798,
+      "learning_rate": 3.10077519379845e-06,
+      "loss": 0.3554,
+      "step": 1895
+    },
+    {
+      "epoch": 2.8340807174887894,
+      "grad_norm": 0.21988700916207501,
+      "learning_rate": 3.073089700996678e-06,
+      "loss": 0.33,
+      "step": 1896
+    },
+    {
+      "epoch": 2.8355754857997013,
+      "grad_norm": 0.20332832662642894,
+      "learning_rate": 3.045404208194906e-06,
+      "loss": 0.3303,
+      "step": 1897
+    },
+    {
+      "epoch": 2.8370702541106128,
+      "grad_norm": 0.22081997010346774,
+      "learning_rate": 3.0177187153931344e-06,
+      "loss": 0.3394,
+      "step": 1898
+    },
+    {
+      "epoch": 2.8385650224215246,
+      "grad_norm": 0.23355011362146388,
+      "learning_rate": 2.9900332225913626e-06,
+      "loss": 0.354,
+      "step": 1899
+    },
+    {
+      "epoch": 2.8400597907324365,
+      "grad_norm": 0.22173797881602894,
+      "learning_rate": 2.9623477297895904e-06,
+      "loss": 0.3485,
+      "step": 1900
+    },
+    {
+      "epoch": 2.8415545590433484,
+      "grad_norm": 0.22312920231959732,
+      "learning_rate": 2.9346622369878186e-06,
+      "loss": 0.3308,
+      "step": 1901
+    },
+    {
+      "epoch": 2.84304932735426,
+      "grad_norm": 0.22178292758815155,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 0.337,
+      "step": 1902
+    },
+    {
+      "epoch": 2.8445440956651717,
+      "grad_norm": 0.20437054260978404,
+      "learning_rate": 2.879291251384275e-06,
+      "loss": 0.3416,
+      "step": 1903
+    },
+    {
+      "epoch": 2.8460388639760836,
+      "grad_norm": 0.216615161632175,
+      "learning_rate": 2.851605758582503e-06,
+      "loss": 0.3432,
+      "step": 1904
+    },
+    {
+      "epoch": 2.8475336322869955,
+      "grad_norm": 0.214408851482875,
+      "learning_rate": 2.8239202657807313e-06,
+      "loss": 0.3359,
+      "step": 1905
+    },
+    {
+      "epoch": 2.8490284005979074,
+      "grad_norm": 0.20478230817759818,
+      "learning_rate": 2.796234772978959e-06,
+      "loss": 0.3387,
+      "step": 1906
+    },
+    {
+      "epoch": 2.8505231689088193,
+      "grad_norm": 0.2210045918528149,
+      "learning_rate": 2.7685492801771873e-06,
+      "loss": 0.3574,
+      "step": 1907
+    },
+    {
+      "epoch": 2.852017937219731,
+      "grad_norm": 0.21759572566745378,
+      "learning_rate": 2.7408637873754155e-06,
+      "loss": 0.3335,
+      "step": 1908
+    },
+    {
+      "epoch": 2.8535127055306426,
+      "grad_norm": 0.2134864426176502,
+      "learning_rate": 2.7131782945736437e-06,
+      "loss": 0.3489,
+      "step": 1909
+    },
+    {
+      "epoch": 2.8550074738415545,
+      "grad_norm": 0.20690726935267378,
+      "learning_rate": 2.6854928017718715e-06,
+      "loss": 0.3347,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8565022421524664,
+      "grad_norm": 0.21394452692242466,
+      "learning_rate": 2.6578073089700997e-06,
+      "loss": 0.3368,
+      "step": 1911
+    },
+    {
+      "epoch": 2.8579970104633783,
+      "grad_norm": 0.210952419852488,
+      "learning_rate": 2.630121816168328e-06,
+      "loss": 0.3565,
+      "step": 1912
+    },
+    {
+      "epoch": 2.8594917787742897,
+      "grad_norm": 0.20876658412271523,
+      "learning_rate": 2.602436323366556e-06,
+      "loss": 0.3373,
+      "step": 1913
+    },
+    {
+      "epoch": 2.8609865470852016,
+      "grad_norm": 0.2045972128259521,
+      "learning_rate": 2.5747508305647842e-06,
+      "loss": 0.3407,
+      "step": 1914
+    },
+    {
+      "epoch": 2.8624813153961135,
+      "grad_norm": 0.22114881901645006,
+      "learning_rate": 2.547065337763012e-06,
+      "loss": 0.3435,
+      "step": 1915
+    },
+    {
+      "epoch": 2.8639760837070254,
+      "grad_norm": 0.23528775077188038,
+      "learning_rate": 2.5193798449612402e-06,
+      "loss": 0.332,
+      "step": 1916
+    },
+    {
+      "epoch": 2.8654708520179373,
+      "grad_norm": 0.19789114963880858,
+      "learning_rate": 2.4916943521594684e-06,
+      "loss": 0.3255,
+      "step": 1917
+    },
+    {
+      "epoch": 2.866965620328849,
+      "grad_norm": 0.19783888733120641,
+      "learning_rate": 2.4640088593576966e-06,
+      "loss": 0.3508,
+      "step": 1918
+    },
+    {
+      "epoch": 2.868460388639761,
+      "grad_norm": 0.2110092878689821,
+      "learning_rate": 2.436323366555925e-06,
+      "loss": 0.3412,
+      "step": 1919
+    },
+    {
+      "epoch": 2.8699551569506725,
+      "grad_norm": 0.21654575150876104,
+      "learning_rate": 2.408637873754153e-06,
+      "loss": 0.3329,
+      "step": 1920
+    },
+    {
+      "epoch": 2.8714499252615844,
+      "grad_norm": 0.20329400157583838,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.3549,
+      "step": 1921
+    },
+    {
+      "epoch": 2.8729446935724963,
+      "grad_norm": 0.20421400973828532,
+      "learning_rate": 2.353266888150609e-06,
+      "loss": 0.3493,
+      "step": 1922
+    },
+    {
+      "epoch": 2.874439461883408,
+      "grad_norm": 0.21782304713055226,
+      "learning_rate": 2.325581395348837e-06,
+      "loss": 0.3342,
+      "step": 1923
+    },
+    {
+      "epoch": 2.87593423019432,
+      "grad_norm": 0.23937439880901415,
+      "learning_rate": 2.2978959025470653e-06,
+      "loss": 0.3667,
+      "step": 1924
+    },
+    {
+      "epoch": 2.8774289985052315,
+      "grad_norm": 0.22184212968855252,
+      "learning_rate": 2.2702104097452935e-06,
+      "loss": 0.3441,
+      "step": 1925
+    },
+    {
+      "epoch": 2.8789237668161434,
+      "grad_norm": 0.22755534480575992,
+      "learning_rate": 2.2425249169435217e-06,
+      "loss": 0.3529,
+      "step": 1926
+    },
+    {
+      "epoch": 2.8804185351270553,
+      "grad_norm": 0.20214239233471787,
+      "learning_rate": 2.21483942414175e-06,
+      "loss": 0.3326,
+      "step": 1927
+    },
+    {
+      "epoch": 2.881913303437967,
+      "grad_norm": 0.20895406429162833,
+      "learning_rate": 2.1871539313399777e-06,
+      "loss": 0.344,
+      "step": 1928
+    },
+    {
+      "epoch": 2.883408071748879,
+      "grad_norm": 0.21056311035683462,
+      "learning_rate": 2.159468438538206e-06,
+      "loss": 0.3512,
+      "step": 1929
+    },
+    {
+      "epoch": 2.884902840059791,
+      "grad_norm": 0.20288428431055278,
+      "learning_rate": 2.131782945736434e-06,
+      "loss": 0.3352,
+      "step": 1930
+    },
+    {
+      "epoch": 2.886397608370703,
+      "grad_norm": 0.21942143092332136,
+      "learning_rate": 2.1040974529346623e-06,
+      "loss": 0.3456,
+      "step": 1931
+    },
+    {
+      "epoch": 2.8878923766816142,
+      "grad_norm": 0.21057652171140478,
+      "learning_rate": 2.0764119601328905e-06,
+      "loss": 0.337,
+      "step": 1932
+    },
+    {
+      "epoch": 2.889387144992526,
+      "grad_norm": 0.20695056515673813,
+      "learning_rate": 2.0487264673311187e-06,
+      "loss": 0.3302,
+      "step": 1933
+    },
+    {
+      "epoch": 2.890881913303438,
+      "grad_norm": 0.20988845919306717,
+      "learning_rate": 2.021040974529347e-06,
+      "loss": 0.357,
+      "step": 1934
+    },
+    {
+      "epoch": 2.89237668161435,
+      "grad_norm": 0.21503345922379283,
+      "learning_rate": 1.9933554817275746e-06,
+      "loss": 0.3487,
+      "step": 1935
+    },
+    {
+      "epoch": 2.8938714499252614,
+      "grad_norm": 0.20192465453104996,
+      "learning_rate": 1.965669988925803e-06,
+      "loss": 0.3343,
+      "step": 1936
+    },
+    {
+      "epoch": 2.8953662182361732,
+      "grad_norm": 0.21010225375027156,
+      "learning_rate": 1.937984496124031e-06,
+      "loss": 0.3577,
+      "step": 1937
+    },
+    {
+      "epoch": 2.896860986547085,
+      "grad_norm": 0.2028005247448409,
+      "learning_rate": 1.9102990033222592e-06,
+      "loss": 0.3434,
+      "step": 1938
+    },
+    {
+      "epoch": 2.898355754857997,
+      "grad_norm": 0.20525393372928627,
+      "learning_rate": 1.8826135105204874e-06,
+      "loss": 0.3323,
+      "step": 1939
+    },
+    {
+      "epoch": 2.899850523168909,
+      "grad_norm": 0.21015347161556747,
+      "learning_rate": 1.8549280177187156e-06,
+      "loss": 0.3491,
+      "step": 1940
+    },
+    {
+      "epoch": 2.901345291479821,
+      "grad_norm": 0.21249454184380948,
+      "learning_rate": 1.8272425249169438e-06,
+      "loss": 0.3487,
+      "step": 1941
+    },
+    {
+      "epoch": 2.9028400597907327,
+      "grad_norm": 0.20710518955244947,
+      "learning_rate": 1.7995570321151716e-06,
+      "loss": 0.346,
+      "step": 1942
+    },
+    {
+      "epoch": 2.904334828101644,
+      "grad_norm": 0.19732684726522856,
+      "learning_rate": 1.7718715393133998e-06,
+      "loss": 0.3593,
+      "step": 1943
+    },
+    {
+      "epoch": 2.905829596412556,
+      "grad_norm": 0.203610759345602,
+      "learning_rate": 1.744186046511628e-06,
+      "loss": 0.325,
+      "step": 1944
+    },
+    {
+      "epoch": 2.907324364723468,
+      "grad_norm": 0.2202184574549953,
+      "learning_rate": 1.7165005537098562e-06,
+      "loss": 0.3548,
+      "step": 1945
+    },
+    {
+      "epoch": 2.90881913303438,
+      "grad_norm": 0.2214764363711459,
+      "learning_rate": 1.6888150609080844e-06,
+      "loss": 0.3261,
+      "step": 1946
+    },
+    {
+      "epoch": 2.910313901345291,
+      "grad_norm": 0.20963214091490842,
+      "learning_rate": 1.6611295681063126e-06,
+      "loss": 0.3299,
+      "step": 1947
+    },
+    {
+      "epoch": 2.911808669656203,
+      "grad_norm": 0.23357177877077664,
+      "learning_rate": 1.6334440753045403e-06,
+      "loss": 0.3606,
+      "step": 1948
+    },
+    {
+      "epoch": 2.913303437967115,
+      "grad_norm": 0.21887111014985486,
+      "learning_rate": 1.6057585825027685e-06,
+      "loss": 0.3494,
+      "step": 1949
+    },
+    {
+      "epoch": 2.914798206278027,
+      "grad_norm": 0.218895785054277,
+      "learning_rate": 1.5780730897009967e-06,
+      "loss": 0.3506,
+      "step": 1950
+    },
+    {
+      "epoch": 2.9162929745889388,
+      "grad_norm": 0.21345075908147337,
+      "learning_rate": 1.550387596899225e-06,
+      "loss": 0.3416,
+      "step": 1951
+    },
+    {
+      "epoch": 2.9177877428998507,
+      "grad_norm": 0.20652652580767397,
+      "learning_rate": 1.522702104097453e-06,
+      "loss": 0.332,
+      "step": 1952
+    },
+    {
+      "epoch": 2.9192825112107625,
+      "grad_norm": 0.19541288421423722,
+      "learning_rate": 1.4950166112956813e-06,
+      "loss": 0.342,
+      "step": 1953
+    },
+    {
+      "epoch": 2.920777279521674,
+      "grad_norm": 0.21630205710153558,
+      "learning_rate": 1.4673311184939093e-06,
+      "loss": 0.3596,
+      "step": 1954
+    },
+    {
+      "epoch": 2.922272047832586,
+      "grad_norm": 0.20736893022083003,
+      "learning_rate": 1.4396456256921375e-06,
+      "loss": 0.3459,
+      "step": 1955
+    },
+    {
+      "epoch": 2.9237668161434978,
+      "grad_norm": 0.2173471911727096,
+      "learning_rate": 1.4119601328903657e-06,
+      "loss": 0.3518,
+      "step": 1956
+    },
+    {
+      "epoch": 2.9252615844544096,
+      "grad_norm": 0.19316355826112722,
+      "learning_rate": 1.3842746400885937e-06,
+      "loss": 0.3326,
+      "step": 1957
+    },
+    {
+      "epoch": 2.9267563527653215,
+      "grad_norm": 0.2056324476755035,
+      "learning_rate": 1.3565891472868218e-06,
+      "loss": 0.3405,
+      "step": 1958
+    },
+    {
+      "epoch": 2.928251121076233,
+      "grad_norm": 0.2127352964144356,
+      "learning_rate": 1.3289036544850498e-06,
+      "loss": 0.3225,
+      "step": 1959
+    },
+    {
+      "epoch": 2.929745889387145,
+      "grad_norm": 0.20645462547346227,
+      "learning_rate": 1.301218161683278e-06,
+      "loss": 0.343,
+      "step": 1960
+    },
+    {
+      "epoch": 2.9312406576980568,
+      "grad_norm": 0.21447679593141114,
+      "learning_rate": 1.273532668881506e-06,
+      "loss": 0.3334,
+      "step": 1961
+    },
+    {
+      "epoch": 2.9327354260089686,
+      "grad_norm": 0.21026723696101599,
+      "learning_rate": 1.2458471760797342e-06,
+      "loss": 0.354,
+      "step": 1962
+    },
+    {
+      "epoch": 2.9342301943198805,
+      "grad_norm": 0.22936909655279183,
+      "learning_rate": 1.2181616832779624e-06,
+      "loss": 0.3413,
+      "step": 1963
+    },
+    {
+      "epoch": 2.9357249626307924,
+      "grad_norm": 0.2121186813838369,
+      "learning_rate": 1.1904761904761904e-06,
+      "loss": 0.3468,
+      "step": 1964
+    },
+    {
+      "epoch": 2.9372197309417043,
+      "grad_norm": 0.20471950552933696,
+      "learning_rate": 1.1627906976744186e-06,
+      "loss": 0.3358,
+      "step": 1965
+    },
+    {
+      "epoch": 2.9387144992526157,
+      "grad_norm": 0.2238442433371041,
+      "learning_rate": 1.1351052048726468e-06,
+      "loss": 0.3641,
+      "step": 1966
+    },
+    {
+      "epoch": 2.9402092675635276,
+      "grad_norm": 0.20257208978782898,
+      "learning_rate": 1.107419712070875e-06,
+      "loss": 0.3524,
+      "step": 1967
+    },
+    {
+      "epoch": 2.9417040358744395,
+      "grad_norm": 0.21984271862251512,
+      "learning_rate": 1.079734219269103e-06,
+      "loss": 0.3537,
+      "step": 1968
+    },
+    {
+      "epoch": 2.9431988041853514,
+      "grad_norm": 0.1963726637813834,
+      "learning_rate": 1.0520487264673311e-06,
+      "loss": 0.3334,
+      "step": 1969
+    },
+    {
+      "epoch": 2.944693572496263,
+      "grad_norm": 0.20450812036308044,
+      "learning_rate": 1.0243632336655593e-06,
+      "loss": 0.3423,
+      "step": 1970
+    },
+    {
+      "epoch": 2.9461883408071747,
+      "grad_norm": 0.21160273308017824,
+      "learning_rate": 9.966777408637873e-07,
+      "loss": 0.3418,
+      "step": 1971
+    },
+    {
+      "epoch": 2.9476831091180866,
+      "grad_norm": 0.20275027125614634,
+      "learning_rate": 9.689922480620155e-07,
+      "loss": 0.3426,
+      "step": 1972
+    },
+    {
+      "epoch": 2.9491778774289985,
+      "grad_norm": 0.20372675098901297,
+      "learning_rate": 9.413067552602437e-07,
+      "loss": 0.3454,
+      "step": 1973
+    },
+    {
+      "epoch": 2.9506726457399104,
+      "grad_norm": 0.20411210929594867,
+      "learning_rate": 9.136212624584719e-07,
+      "loss": 0.349,
+      "step": 1974
+    },
+    {
+      "epoch": 2.9521674140508223,
+      "grad_norm": 0.21650377624319586,
+      "learning_rate": 8.859357696566999e-07,
+      "loss": 0.3613,
+      "step": 1975
+    },
+    {
+      "epoch": 2.953662182361734,
+      "grad_norm": 0.2262110264527186,
+      "learning_rate": 8.582502768549281e-07,
+      "loss": 0.3528,
+      "step": 1976
+    },
+    {
+      "epoch": 2.9551569506726456,
+      "grad_norm": 0.2242812495020997,
+      "learning_rate": 8.305647840531563e-07,
+      "loss": 0.3669,
+      "step": 1977
+    },
+    {
+      "epoch": 2.9566517189835575,
+      "grad_norm": 0.20933279538468128,
+      "learning_rate": 8.028792912513843e-07,
+      "loss": 0.3492,
+      "step": 1978
+    },
+    {
+      "epoch": 2.9581464872944694,
+      "grad_norm": 0.22086437964510128,
+      "learning_rate": 7.751937984496125e-07,
+      "loss": 0.3474,
+      "step": 1979
+    },
+    {
+      "epoch": 2.9596412556053813,
+      "grad_norm": 0.20156989435950562,
+      "learning_rate": 7.475083056478406e-07,
+      "loss": 0.3388,
+      "step": 1980
+    },
+    {
+      "epoch": 2.9611360239162927,
+      "grad_norm": 0.20875757044554535,
+      "learning_rate": 7.198228128460687e-07,
+      "loss": 0.3499,
+      "step": 1981
+    },
+    {
+      "epoch": 2.9626307922272046,
+      "grad_norm": 0.19376329314096624,
+      "learning_rate": 6.921373200442968e-07,
+      "loss": 0.3434,
+      "step": 1982
+    },
+    {
+      "epoch": 2.9641255605381165,
+      "grad_norm": 0.21197349390156894,
+      "learning_rate": 6.644518272425249e-07,
+      "loss": 0.346,
+      "step": 1983
+    },
+    {
+      "epoch": 2.9656203288490284,
+      "grad_norm": 0.20422476834860726,
+      "learning_rate": 6.36766334440753e-07,
+      "loss": 0.3265,
+      "step": 1984
+    },
+    {
+      "epoch": 2.9671150971599403,
+      "grad_norm": 0.20573236864229327,
+      "learning_rate": 6.090808416389812e-07,
+      "loss": 0.3678,
+      "step": 1985
+    },
+    {
+      "epoch": 2.968609865470852,
+      "grad_norm": 0.20926881378152945,
+      "learning_rate": 5.813953488372093e-07,
+      "loss": 0.3416,
+      "step": 1986
+    },
+    {
+      "epoch": 2.970104633781764,
+      "grad_norm": 0.20835516384428782,
+      "learning_rate": 5.537098560354375e-07,
+      "loss": 0.3505,
+      "step": 1987
+    },
+    {
+      "epoch": 2.9715994020926755,
+      "grad_norm": 0.1952899730484928,
+      "learning_rate": 5.260243632336656e-07,
+      "loss": 0.3416,
+      "step": 1988
+    },
+    {
+      "epoch": 2.9730941704035874,
+      "grad_norm": 0.2044183807910725,
+      "learning_rate": 4.983388704318937e-07,
+      "loss": 0.345,
+      "step": 1989
+    },
+    {
+      "epoch": 2.9745889387144993,
+      "grad_norm": 0.21812634874418335,
+      "learning_rate": 4.7065337763012186e-07,
+      "loss": 0.3321,
+      "step": 1990
+    },
+    {
+      "epoch": 2.976083707025411,
+      "grad_norm": 0.20992580301557814,
+      "learning_rate": 4.4296788482834994e-07,
+      "loss": 0.3451,
+      "step": 1991
+    },
+    {
+      "epoch": 2.977578475336323,
+      "grad_norm": 0.2023197773170547,
+      "learning_rate": 4.1528239202657814e-07,
+      "loss": 0.3451,
+      "step": 1992
+    },
+    {
+      "epoch": 2.9790732436472345,
+      "grad_norm": 0.20037635087772215,
+      "learning_rate": 3.8759689922480623e-07,
+      "loss": 0.3428,
+      "step": 1993
+    },
+    {
+      "epoch": 2.9805680119581464,
+      "grad_norm": 0.2121075249995569,
+      "learning_rate": 3.5991140642303437e-07,
+      "loss": 0.3333,
+      "step": 1994
+    },
+    {
+      "epoch": 2.9820627802690582,
+      "grad_norm": 0.1954452218488001,
+      "learning_rate": 3.3222591362126246e-07,
+      "loss": 0.3324,
+      "step": 1995
+    },
+    {
+      "epoch": 2.98355754857997,
+      "grad_norm": 0.20115245050675856,
+      "learning_rate": 3.045404208194906e-07,
+      "loss": 0.3492,
+      "step": 1996
+    },
+    {
+      "epoch": 2.985052316890882,
+      "grad_norm": 0.20402753428718695,
+      "learning_rate": 2.7685492801771874e-07,
+      "loss": 0.349,
+      "step": 1997
+    },
+    {
+      "epoch": 2.986547085201794,
+      "grad_norm": 0.21440948473923,
+      "learning_rate": 2.4916943521594683e-07,
+      "loss": 0.3519,
+      "step": 1998
+    },
+    {
+      "epoch": 2.988041853512706,
+      "grad_norm": 0.2246951004987662,
+      "learning_rate": 2.2148394241417497e-07,
+      "loss": 0.3459,
+      "step": 1999
+    },
+    {
+      "epoch": 2.9895366218236172,
+      "grad_norm": 0.2087919360930793,
+      "learning_rate": 1.9379844961240311e-07,
+      "loss": 0.3517,
+      "step": 2000
+    },
+    {
+      "epoch": 2.991031390134529,
+      "grad_norm": 0.21186085280501327,
+      "learning_rate": 1.6611295681063123e-07,
+      "loss": 0.3487,
+      "step": 2001
+    },
+    {
+      "epoch": 2.992526158445441,
+      "grad_norm": 0.20808981927546513,
+      "learning_rate": 1.3842746400885937e-07,
+      "loss": 0.3522,
+      "step": 2002
+    },
+    {
+      "epoch": 2.994020926756353,
+      "grad_norm": 0.20254426866913747,
+      "learning_rate": 1.1074197120708749e-07,
+      "loss": 0.3401,
+      "step": 2003
+    },
+    {
+      "epoch": 2.9955156950672643,
+      "grad_norm": 0.21780513289100528,
+      "learning_rate": 8.305647840531561e-08,
+      "loss": 0.3415,
+      "step": 2004
+    },
+    {
+      "epoch": 2.9970104633781762,
+      "grad_norm": 0.21729041476880495,
+      "learning_rate": 5.537098560354374e-08,
+      "loss": 0.329,
+      "step": 2005
+    },
+    {
+      "epoch": 2.998505231689088,
+      "grad_norm": 0.20062315926869276,
+      "learning_rate": 2.768549280177187e-08,
+      "loss": 0.3358,
+      "step": 2006
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.20956701781560585,
+      "learning_rate": 0.0,
+      "loss": 0.3532,
+      "step": 2007
+    },
+    {
+      "epoch": 3.0,
+      "step": 2007,
+      "total_flos": 1.7126160710573752e+18,
+      "train_loss": 0.5120905198502077,
+      "train_runtime": 117090.8784,
+      "train_samples_per_second": 0.274,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2007,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7126160710573752e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}