diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6964827620516392,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8465005159378052,
+      "learning_rate": 6.622516556291391e-07,
+      "loss": 3.3059,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9715436697006226,
+      "learning_rate": 1.3245033112582782e-06,
+      "loss": 3.7826,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6590728759765625,
+      "learning_rate": 1.9867549668874175e-06,
+      "loss": 3.5184,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.264482259750366,
+      "learning_rate": 2.6490066225165563e-06,
+      "loss": 4.0008,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6213676929473877,
+      "learning_rate": 3.3112582781456956e-06,
+      "loss": 3.4438,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2400014400482178,
+      "learning_rate": 3.973509933774835e-06,
+      "loss": 3.8402,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1491827964782715,
+      "learning_rate": 4.635761589403974e-06,
+      "loss": 3.6056,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4942920207977295,
+      "learning_rate": 5.298013245033113e-06,
+      "loss": 3.8327,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3179354667663574,
+      "learning_rate": 5.9602649006622515e-06,
+      "loss": 3.6016,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2473526000976562,
+      "learning_rate": 6.622516556291391e-06,
+      "loss": 3.1156,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.013592481613159,
+      "learning_rate": 7.28476821192053e-06,
+      "loss": 3.741,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3654344081878662,
+      "learning_rate": 7.94701986754967e-06,
+      "loss": 3.1595,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4851272106170654,
+      "learning_rate": 8.609271523178809e-06,
+      "loss": 3.309,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2894127368927002,
+      "learning_rate": 9.271523178807948e-06,
+      "loss": 3.41,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9651306867599487,
+      "learning_rate": 9.933774834437086e-06,
+      "loss": 3.0962,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1612597703933716,
+      "learning_rate": 1.0596026490066225e-05,
+      "loss": 3.2608,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.13624906539917,
+      "learning_rate": 1.1258278145695366e-05,
+      "loss": 3.4395,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0920512676239014,
+      "learning_rate": 1.1920529801324503e-05,
+      "loss": 3.1039,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1041334867477417,
+      "learning_rate": 1.2582781456953644e-05,
+      "loss": 3.04,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2223341464996338,
+      "learning_rate": 1.3245033112582782e-05,
+      "loss": 3.3573,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.423256516456604,
+      "learning_rate": 1.390728476821192e-05,
+      "loss": 3.2344,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.019239902496338,
+      "learning_rate": 1.456953642384106e-05,
+      "loss": 2.961,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8667553067207336,
+      "learning_rate": 1.5231788079470199e-05,
+      "loss": 3.0434,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9331719875335693,
+      "learning_rate": 1.589403973509934e-05,
+      "loss": 2.9039,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8311322331428528,
+      "learning_rate": 1.655629139072848e-05,
+      "loss": 2.9542,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9529271125793457,
+      "learning_rate": 1.7218543046357617e-05,
+      "loss": 3.0846,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0235741138458252,
+      "learning_rate": 1.7880794701986755e-05,
+      "loss": 2.9234,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0178136825561523,
+      "learning_rate": 1.8543046357615895e-05,
+      "loss": 2.9683,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9590712785720825,
+      "learning_rate": 1.9205298013245036e-05,
+      "loss": 3.083,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9728456735610962,
+      "learning_rate": 1.9867549668874173e-05,
+      "loss": 2.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.054449439048767,
+      "learning_rate": 2.052980132450331e-05,
+      "loss": 3.2505,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.025812029838562,
+      "learning_rate": 2.119205298013245e-05,
+      "loss": 3.1406,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1379340887069702,
+      "learning_rate": 2.185430463576159e-05,
+      "loss": 3.055,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1058827638626099,
+      "learning_rate": 2.2516556291390732e-05,
+      "loss": 2.9731,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8949688076972961,
+      "learning_rate": 2.317880794701987e-05,
+      "loss": 2.7599,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.061889886856079,
+      "learning_rate": 2.3841059602649006e-05,
+      "loss": 2.9647,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2604342699050903,
+      "learning_rate": 2.4503311258278147e-05,
+      "loss": 3.2488,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0576081275939941,
+      "learning_rate": 2.5165562913907287e-05,
+      "loss": 2.8317,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9627434611320496,
+      "learning_rate": 2.5827814569536424e-05,
+      "loss": 3.057,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.138532042503357,
+      "learning_rate": 2.6490066225165565e-05,
+      "loss": 3.0145,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2501425743103027,
+      "learning_rate": 2.7152317880794702e-05,
+      "loss": 3.0757,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1380525827407837,
+      "learning_rate": 2.781456953642384e-05,
+      "loss": 3.0791,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8773428797721863,
+      "learning_rate": 2.8476821192052983e-05,
+      "loss": 2.73,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.869491457939148,
+      "learning_rate": 2.913907284768212e-05,
+      "loss": 2.6191,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9522172212600708,
+      "learning_rate": 2.980132450331126e-05,
+      "loss": 2.8169,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9050353765487671,
+      "learning_rate": 3.0463576158940398e-05,
+      "loss": 2.7318,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9091339111328125,
+      "learning_rate": 3.1125827814569535e-05,
+      "loss": 2.8713,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9983739852905273,
+      "learning_rate": 3.178807947019868e-05,
+      "loss": 2.9996,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9158467054367065,
+      "learning_rate": 3.2450331125827816e-05,
+      "loss": 2.9113,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.071158528327942,
+      "learning_rate": 3.311258278145696e-05,
+      "loss": 3.0376,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9090975522994995,
+      "learning_rate": 3.37748344370861e-05,
+      "loss": 2.8224,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9709640741348267,
+      "learning_rate": 3.4437086092715235e-05,
+      "loss": 2.832,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0022692680358887,
+      "learning_rate": 3.509933774834437e-05,
+      "loss": 2.908,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9489349126815796,
+      "learning_rate": 3.576158940397351e-05,
+      "loss": 3.0759,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0647804737091064,
+      "learning_rate": 3.6423841059602646e-05,
+      "loss": 2.6719,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9491623640060425,
+      "learning_rate": 3.708609271523179e-05,
+      "loss": 2.8885,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8797650933265686,
+      "learning_rate": 3.774834437086093e-05,
+      "loss": 2.762,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0328469276428223,
+      "learning_rate": 3.841059602649007e-05,
+      "loss": 2.9797,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0455362796783447,
+      "learning_rate": 3.907284768211921e-05,
+      "loss": 3.0064,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.973304808139801,
+      "learning_rate": 3.9735099337748346e-05,
+      "loss": 2.6846,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.053493857383728,
+      "learning_rate": 4.039735099337749e-05,
+      "loss": 2.8232,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0137115716934204,
+      "learning_rate": 4.105960264900662e-05,
+      "loss": 2.7637,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8970140814781189,
+      "learning_rate": 4.1721854304635764e-05,
+      "loss": 2.8644,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0503255128860474,
+      "learning_rate": 4.23841059602649e-05,
+      "loss": 2.7927,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9206955432891846,
+      "learning_rate": 4.304635761589404e-05,
+      "loss": 2.8839,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9703933596611023,
+      "learning_rate": 4.370860927152318e-05,
+      "loss": 2.8651,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9761046767234802,
+      "learning_rate": 4.437086092715232e-05,
+      "loss": 2.6735,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9341240525245667,
+      "learning_rate": 4.5033112582781463e-05,
+      "loss": 2.8067,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8800592422485352,
+      "learning_rate": 4.56953642384106e-05,
+      "loss": 2.7855,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9529682993888855,
+      "learning_rate": 4.635761589403974e-05,
+      "loss": 2.8796,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1102032661437988,
+      "learning_rate": 4.7019867549668875e-05,
+      "loss": 2.9459,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.97463458776474,
+      "learning_rate": 4.768211920529801e-05,
+      "loss": 2.7726,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.059367299079895,
+      "learning_rate": 4.8344370860927156e-05,
+      "loss": 3.025,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0793780088424683,
+      "learning_rate": 4.900662251655629e-05,
+      "loss": 2.9883,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0120257139205933,
+      "learning_rate": 4.966887417218543e-05,
+      "loss": 2.702,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.836804986000061,
+      "learning_rate": 5.0331125827814574e-05,
+      "loss": 2.6563,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1675254106521606,
+      "learning_rate": 5.099337748344372e-05,
+      "loss": 2.7923,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1154323816299438,
+      "learning_rate": 5.165562913907285e-05,
+      "loss": 2.9183,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0326870679855347,
+      "learning_rate": 5.231788079470199e-05,
+      "loss": 2.7754,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0615193843841553,
+      "learning_rate": 5.298013245033113e-05,
+      "loss": 2.8237,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.096045732498169,
+      "learning_rate": 5.364238410596027e-05,
+      "loss": 2.6867,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.99045729637146,
+      "learning_rate": 5.4304635761589404e-05,
+      "loss": 2.9822,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0876522064208984,
+      "learning_rate": 5.496688741721855e-05,
+      "loss": 2.9194,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8541275858879089,
+      "learning_rate": 5.562913907284768e-05,
+      "loss": 2.793,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1198952198028564,
+      "learning_rate": 5.629139072847682e-05,
+      "loss": 2.9692,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0689620971679688,
+      "learning_rate": 5.6953642384105966e-05,
+      "loss": 2.7703,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.114166259765625,
+      "learning_rate": 5.76158940397351e-05,
+      "loss": 2.8391,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9224385023117065,
+      "learning_rate": 5.827814569536424e-05,
+      "loss": 2.7175,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0710177421569824,
+      "learning_rate": 5.8940397350993385e-05,
+      "loss": 2.9163,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9974583983421326,
+      "learning_rate": 5.960264900662252e-05,
+      "loss": 2.7971,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0250120162963867,
+      "learning_rate": 6.026490066225165e-05,
+      "loss": 2.6168,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9977954626083374,
+      "learning_rate": 6.0927152317880796e-05,
+      "loss": 2.5418,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9334400296211243,
+      "learning_rate": 6.158940397350994e-05,
+      "loss": 2.727,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8602076172828674,
+      "learning_rate": 6.225165562913907e-05,
+      "loss": 2.6817,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0017945766448975,
+      "learning_rate": 6.291390728476821e-05,
+      "loss": 2.8375,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9383439421653748,
+      "learning_rate": 6.357615894039736e-05,
+      "loss": 2.7961,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1213310956954956,
+      "learning_rate": 6.423841059602649e-05,
+      "loss": 2.9757,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0118752717971802,
+      "learning_rate": 6.490066225165563e-05,
+      "loss": 2.7794,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.931228518486023,
+      "learning_rate": 6.556291390728478e-05,
+      "loss": 2.8484,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8528026938438416,
+      "learning_rate": 6.622516556291392e-05,
+      "loss": 2.6649,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.046879529953003,
+      "learning_rate": 6.688741721854305e-05,
+      "loss": 2.7008,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9575727581977844,
+      "learning_rate": 6.75496688741722e-05,
+      "loss": 2.8019,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9638158679008484,
+      "learning_rate": 6.821192052980133e-05,
+      "loss": 2.6091,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9592703580856323,
+      "learning_rate": 6.887417218543047e-05,
+      "loss": 2.7826,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.076717734336853,
+      "learning_rate": 6.95364238410596e-05,
+      "loss": 3.0087,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.886050820350647,
+      "learning_rate": 7.019867549668874e-05,
+      "loss": 2.8504,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9256192445755005,
+      "learning_rate": 7.086092715231787e-05,
+      "loss": 2.7476,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.977173924446106,
+      "learning_rate": 7.152317880794702e-05,
+      "loss": 2.929,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0920604467391968,
+      "learning_rate": 7.218543046357616e-05,
+      "loss": 2.9099,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0105801820755005,
+      "learning_rate": 7.284768211920529e-05,
+      "loss": 2.9241,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9057108759880066,
+      "learning_rate": 7.350993377483444e-05,
+      "loss": 2.578,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8995856046676636,
+      "learning_rate": 7.417218543046358e-05,
+      "loss": 2.8721,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.010469675064087,
+      "learning_rate": 7.483443708609272e-05,
+      "loss": 2.9527,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8397200703620911,
+      "learning_rate": 7.549668874172185e-05,
+      "loss": 2.8898,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9744095206260681,
+      "learning_rate": 7.6158940397351e-05,
+      "loss": 2.8472,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8169723749160767,
+      "learning_rate": 7.682119205298014e-05,
+      "loss": 2.9016,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7275145649909973,
+      "learning_rate": 7.748344370860927e-05,
+      "loss": 2.6558,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8188539743423462,
+      "learning_rate": 7.814569536423842e-05,
+      "loss": 2.6763,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6780896782875061,
+      "learning_rate": 7.880794701986756e-05,
+      "loss": 2.6888,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9559083580970764,
+      "learning_rate": 7.947019867549669e-05,
+      "loss": 2.7511,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7042127251625061,
+      "learning_rate": 8.013245033112584e-05,
+      "loss": 2.741,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0874255895614624,
+      "learning_rate": 8.079470198675498e-05,
+      "loss": 2.771,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7785601019859314,
+      "learning_rate": 8.145695364238411e-05,
+      "loss": 2.5553,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9149391651153564,
+      "learning_rate": 8.211920529801324e-05,
+      "loss": 2.8121,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8360904455184937,
+      "learning_rate": 8.278145695364238e-05,
+      "loss": 2.6646,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9527228474617004,
+      "learning_rate": 8.344370860927153e-05,
+      "loss": 2.8205,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0456985235214233,
+      "learning_rate": 8.410596026490066e-05,
+      "loss": 2.6169,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8443238139152527,
+      "learning_rate": 8.47682119205298e-05,
+      "loss": 2.7697,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8376434445381165,
+      "learning_rate": 8.543046357615895e-05,
+      "loss": 2.8327,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9154295325279236,
+      "learning_rate": 8.609271523178808e-05,
+      "loss": 2.695,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0350686311721802,
+      "learning_rate": 8.675496688741722e-05,
+      "loss": 2.8694,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1221824884414673,
+      "learning_rate": 8.741721854304636e-05,
+      "loss": 2.8344,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9455471634864807,
+      "learning_rate": 8.80794701986755e-05,
+      "loss": 2.7086,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7347345352172852,
+      "learning_rate": 8.874172185430464e-05,
+      "loss": 2.6523,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8115938901901245,
+      "learning_rate": 8.940397350993378e-05,
+      "loss": 2.7553,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0158753395080566,
+      "learning_rate": 9.006622516556293e-05,
+      "loss": 2.7497,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8055458068847656,
+      "learning_rate": 9.072847682119206e-05,
+      "loss": 2.5587,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9538119435310364,
+      "learning_rate": 9.13907284768212e-05,
+      "loss": 2.7344,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8980190753936768,
+      "learning_rate": 9.205298013245035e-05,
+      "loss": 2.9033,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9373752474784851,
+      "learning_rate": 9.271523178807948e-05,
+      "loss": 2.7375,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.724567711353302,
+      "learning_rate": 9.337748344370862e-05,
+      "loss": 2.7507,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9333958029747009,
+      "learning_rate": 9.403973509933775e-05,
+      "loss": 2.7951,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.81438148021698,
+      "learning_rate": 9.47019867549669e-05,
+      "loss": 2.5708,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.844967782497406,
+      "learning_rate": 9.536423841059602e-05,
+      "loss": 2.6766,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8641153573989868,
+      "learning_rate": 9.602649006622517e-05,
+      "loss": 2.8005,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8951542973518372,
+      "learning_rate": 9.668874172185431e-05,
+      "loss": 2.5282,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8378608822822571,
+      "learning_rate": 9.735099337748344e-05,
+      "loss": 2.9217,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7680026888847351,
+      "learning_rate": 9.801324503311259e-05,
+      "loss": 2.7549,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9696276187896729,
+      "learning_rate": 9.867549668874173e-05,
+      "loss": 2.6386,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.781838059425354,
+      "learning_rate": 9.933774834437086e-05,
+      "loss": 2.6572,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8659349083900452,
+      "learning_rate": 0.0001,
+      "loss": 2.7133,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8928912281990051,
+      "learning_rate": 9.999998961351283e-05,
+      "loss": 2.8301,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.934087872505188,
+      "learning_rate": 9.999995845405561e-05,
+      "loss": 2.8315,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9130839705467224,
+      "learning_rate": 9.999990652164128e-05,
+      "loss": 2.7198,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.843783438205719,
+      "learning_rate": 9.999983381629146e-05,
+      "loss": 2.8028,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0018725395202637,
+      "learning_rate": 9.99997403380363e-05,
+      "loss": 2.8903,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8355838656425476,
+      "learning_rate": 9.999962608691468e-05,
+      "loss": 2.5834,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8316748738288879,
+      "learning_rate": 9.999949106297403e-05,
+      "loss": 2.6615,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.829491376876831,
+      "learning_rate": 9.999933526627048e-05,
+      "loss": 2.5355,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9862785339355469,
+      "learning_rate": 9.999915869686875e-05,
+      "loss": 2.6536,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7654109001159668,
+      "learning_rate": 9.999896135484217e-05,
+      "loss": 2.7126,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9111725091934204,
+      "learning_rate": 9.999874324027277e-05,
+      "loss": 2.7498,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.930557906627655,
+      "learning_rate": 9.999850435325113e-05,
+      "loss": 2.9114,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8987687826156616,
+      "learning_rate": 9.999824469387652e-05,
+      "loss": 2.6973,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8717585206031799,
+      "learning_rate": 9.99979642622568e-05,
+      "loss": 2.8706,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9925357103347778,
+      "learning_rate": 9.99976630585085e-05,
+      "loss": 2.6998,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8843782544136047,
+      "learning_rate": 9.999734108275676e-05,
+      "loss": 2.9575,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7779749035835266,
+      "learning_rate": 9.999699833513531e-05,
+      "loss": 2.6043,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8790146708488464,
+      "learning_rate": 9.99966348157866e-05,
+      "loss": 2.8257,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8007485866546631,
+      "learning_rate": 9.999625052486161e-05,
+      "loss": 2.7505,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0786597728729248,
+      "learning_rate": 9.999584546252003e-05,
+      "loss": 2.9278,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8645161986351013,
+      "learning_rate": 9.999541962893014e-05,
+      "loss": 2.8234,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8798420429229736,
+      "learning_rate": 9.999497302426883e-05,
+      "loss": 2.7924,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8675272464752197,
+      "learning_rate": 9.99945056487217e-05,
+      "loss": 2.673,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9363759160041809,
+      "learning_rate": 9.999401750248287e-05,
+      "loss": 2.7721,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.817405104637146,
+      "learning_rate": 9.999350858575518e-05,
+      "loss": 2.785,
+      "step": 176
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9421597719192505,
+      "learning_rate": 9.999297889875005e-05,
+      "loss": 2.8091,
+      "step": 177
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7078314423561096,
+      "learning_rate": 9.999242844168754e-05,
+      "loss": 2.6964,
+      "step": 178
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8067195415496826,
+      "learning_rate": 9.999185721479635e-05,
+      "loss": 2.7226,
+      "step": 179
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8672513961791992,
+      "learning_rate": 9.99912652183138e-05,
+      "loss": 2.8741,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8595795035362244,
+      "learning_rate": 9.999065245248585e-05,
+      "loss": 2.6998,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.812088668346405,
+      "learning_rate": 9.999001891756707e-05,
+      "loss": 2.7476,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7425401210784912,
+      "learning_rate": 9.998936461382066e-05,
+      "loss": 2.7981,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8452320098876953,
+      "learning_rate": 9.998868954151847e-05,
+      "loss": 2.6615,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8758852481842041,
+      "learning_rate": 9.998799370094097e-05,
+      "loss": 2.7954,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8468999862670898,
+      "learning_rate": 9.998727709237722e-05,
+      "loss": 2.8161,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.923998236656189,
+      "learning_rate": 9.998653971612497e-05,
+      "loss": 2.8769,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9320216774940491,
+      "learning_rate": 9.998578157249059e-05,
+      "loss": 2.6246,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7383199334144592,
+      "learning_rate": 9.9985002661789e-05,
+      "loss": 2.6324,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9730477333068848,
+      "learning_rate": 9.998420298434388e-05,
+      "loss": 2.7908,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0173331499099731,
+      "learning_rate": 9.99833825404874e-05,
+      "loss": 2.8669,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8507503271102905,
+      "learning_rate": 9.998254133056045e-05,
+      "loss": 2.7045,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9344580173492432,
+      "learning_rate": 9.998167935491251e-05,
+      "loss": 2.8493,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8826327919960022,
+      "learning_rate": 9.998079661390169e-05,
+      "loss": 2.7399,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0255029201507568,
+      "learning_rate": 9.997989310789475e-05,
+      "loss": 2.6824,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8689097166061401,
+      "learning_rate": 9.997896883726706e-05,
+      "loss": 2.8631,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9297754764556885,
+      "learning_rate": 9.997802380240259e-05,
+      "loss": 2.8237,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7634474039077759,
+      "learning_rate": 9.997705800369402e-05,
+      "loss": 2.6348,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8396182060241699,
+      "learning_rate": 9.997607144154253e-05,
+      "loss": 2.8023,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0354857444763184,
+      "learning_rate": 9.997506411635805e-05,
+      "loss": 2.745,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8294510841369629,
+      "learning_rate": 9.997403602855903e-05,
+      "loss": 2.5586,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8469874262809753,
+      "learning_rate": 9.997298717857266e-05,
+      "loss": 2.6223,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9981289505958557,
+      "learning_rate": 9.997191756683466e-05,
+      "loss": 2.703,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9857164621353149,
+      "learning_rate": 9.997082719378942e-05,
+      "loss": 2.7876,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8563696146011353,
+      "learning_rate": 9.996971605988994e-05,
+      "loss": 2.7564,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8454917073249817,
+      "learning_rate": 9.996858416559785e-05,
+      "loss": 2.8257,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9124134182929993,
+      "learning_rate": 9.996743151138342e-05,
+      "loss": 2.7383,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6618683934211731,
+      "learning_rate": 9.996625809772551e-05,
+      "loss": 2.5695,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7349535226821899,
+      "learning_rate": 9.996506392511165e-05,
+      "loss": 2.6056,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8854600191116333,
+      "learning_rate": 9.996384899403793e-05,
+      "loss": 2.7825,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9201147556304932,
+      "learning_rate": 9.996261330500916e-05,
+      "loss": 2.7961,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8609025478363037,
+      "learning_rate": 9.996135685853866e-05,
+      "loss": 2.7295,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9264471530914307,
+      "learning_rate": 9.99600796551485e-05,
+      "loss": 2.7816,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8366650938987732,
+      "learning_rate": 9.995878169536925e-05,
+      "loss": 2.7115,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8762885332107544,
+      "learning_rate": 9.995746297974018e-05,
+      "loss": 2.7398,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8016989827156067,
+      "learning_rate": 9.995612350880917e-05,
+      "loss": 2.801,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8133795261383057,
+      "learning_rate": 9.995476328313269e-05,
+      "loss": 2.5766,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7489718198776245,
+      "learning_rate": 9.995338230327589e-05,
+      "loss": 2.6348,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9690249562263489,
+      "learning_rate": 9.99519805698125e-05,
+      "loss": 2.7175,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7692612409591675,
+      "learning_rate": 9.995055808332486e-05,
+      "loss": 2.5221,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7513999342918396,
+      "learning_rate": 9.9949114844404e-05,
+      "loss": 2.8638,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9739452004432678,
+      "learning_rate": 9.99476508536495e-05,
+      "loss": 2.5738,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9342159628868103,
+      "learning_rate": 9.994616611166958e-05,
+      "loss": 2.8495,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8598986864089966,
+      "learning_rate": 9.994466061908112e-05,
+      "loss": 2.659,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.007734775543213,
+      "learning_rate": 9.994313437650956e-05,
+      "loss": 2.8778,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8156813979148865,
+      "learning_rate": 9.9941587384589e-05,
+      "loss": 2.6719,
+      "step": 226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.889741837978363,
+      "learning_rate": 9.994001964396218e-05,
+      "loss": 2.7015,
+      "step": 227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9016849994659424,
+      "learning_rate": 9.993843115528041e-05,
+      "loss": 2.9616,
+      "step": 228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8703715801239014,
+      "learning_rate": 9.993682191920362e-05,
+      "loss": 2.6917,
+      "step": 229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8929210305213928,
+      "learning_rate": 9.993519193640043e-05,
+      "loss": 2.6865,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.739349901676178,
+      "learning_rate": 9.9933541207548e-05,
+      "loss": 2.5295,
+      "step": 231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7522986531257629,
+      "learning_rate": 9.993186973333215e-05,
+      "loss": 2.6763,
+      "step": 232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7804232835769653,
+      "learning_rate": 9.993017751444732e-05,
+      "loss": 2.5456,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.896909236907959,
+      "learning_rate": 9.992846455159653e-05,
+      "loss": 2.7084,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8710148930549622,
+      "learning_rate": 9.992673084549147e-05,
+      "loss": 2.7376,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7927027940750122,
+      "learning_rate": 9.992497639685241e-05,
+      "loss": 2.587,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8940434455871582,
+      "learning_rate": 9.992320120640826e-05,
+      "loss": 2.6464,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7628607749938965,
+      "learning_rate": 9.992140527489655e-05,
+      "loss": 2.4599,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8806402087211609,
+      "learning_rate": 9.99195886030634e-05,
+      "loss": 2.7733,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8961281180381775,
+      "learning_rate": 9.991775119166359e-05,
+      "loss": 2.7053,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7889245748519897,
+      "learning_rate": 9.991589304146046e-05,
+      "loss": 2.7942,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8384144306182861,
+      "learning_rate": 9.991401415322601e-05,
+      "loss": 2.6484,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8577688932418823,
+      "learning_rate": 9.991211452774084e-05,
+      "loss": 2.7185,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.774991512298584,
+      "learning_rate": 9.991019416579414e-05,
+      "loss": 2.5045,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8736111521720886,
+      "learning_rate": 9.99082530681838e-05,
+      "loss": 2.9223,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7778720855712891,
+      "learning_rate": 9.990629123571624e-05,
+      "loss": 2.6362,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7742685675621033,
+      "learning_rate": 9.99043086692065e-05,
+      "loss": 2.6949,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7513969540596008,
+      "learning_rate": 9.990230536947827e-05,
+      "loss": 2.8282,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7694494724273682,
+      "learning_rate": 9.990028133736386e-05,
+      "loss": 2.7614,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8768795728683472,
+      "learning_rate": 9.989823657370415e-05,
+      "loss": 2.7261,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7335816621780396,
+      "learning_rate": 9.989617107934868e-05,
+      "loss": 2.5521,
+      "step": 251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8202827572822571,
+      "learning_rate": 9.989408485515555e-05,
+      "loss": 2.705,
+      "step": 252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7582573890686035,
+      "learning_rate": 9.98919779019915e-05,
+      "loss": 2.5687,
+      "step": 253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9693670868873596,
+      "learning_rate": 9.988985022073192e-05,
+      "loss": 2.6554,
+      "step": 254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8855671882629395,
+      "learning_rate": 9.988770181226075e-05,
+      "loss": 2.8819,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9269035458564758,
+      "learning_rate": 9.988553267747056e-05,
+      "loss": 2.8489,
+      "step": 256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.937470555305481,
+      "learning_rate": 9.988334281726257e-05,
+      "loss": 2.8375,
+      "step": 257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8407672643661499,
+      "learning_rate": 9.988113223254654e-05,
+      "loss": 2.7291,
+      "step": 258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8289220929145813,
+      "learning_rate": 9.98789009242409e-05,
+      "loss": 2.6686,
+      "step": 259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.827402651309967,
+      "learning_rate": 9.987664889327266e-05,
+      "loss": 2.6866,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9115194082260132,
+      "learning_rate": 9.987437614057744e-05,
+      "loss": 2.4584,
+      "step": 261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.937545657157898,
+      "learning_rate": 9.98720826670995e-05,
+      "loss": 2.9154,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7565220594406128,
+      "learning_rate": 9.986976847379168e-05,
+      "loss": 2.6595,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.973150908946991,
+      "learning_rate": 9.986743356161543e-05,
+      "loss": 2.678,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7280756831169128,
+      "learning_rate": 9.986507793154078e-05,
+      "loss": 2.5601,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8045365214347839,
+      "learning_rate": 9.986270158454645e-05,
+      "loss": 2.6674,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8584613800048828,
+      "learning_rate": 9.986030452161968e-05,
+      "loss": 2.8085,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8294486999511719,
+      "learning_rate": 9.985788674375637e-05,
+      "loss": 2.8551,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7818707227706909,
+      "learning_rate": 9.985544825196102e-05,
+      "loss": 2.7435,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9071153402328491,
+      "learning_rate": 9.98529890472467e-05,
+      "loss": 2.807,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7957189679145813,
+      "learning_rate": 9.98505091306351e-05,
+      "loss": 2.7899,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8975158333778381,
+      "learning_rate": 9.984800850315655e-05,
+      "loss": 2.7777,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.915144145488739,
+      "learning_rate": 9.984548716584995e-05,
+      "loss": 2.6418,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7521941065788269,
+      "learning_rate": 9.984294511976283e-05,
+      "loss": 2.7099,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8055357933044434,
+      "learning_rate": 9.984038236595129e-05,
+      "loss": 2.6501,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8220183253288269,
+      "learning_rate": 9.983779890548005e-05,
+      "loss": 2.7531,
+      "step": 276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7943485975265503,
+      "learning_rate": 9.983519473942245e-05,
+      "loss": 2.6919,
+      "step": 277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7502560019493103,
+      "learning_rate": 9.983256986886038e-05,
+      "loss": 2.3672,
+      "step": 278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8013080358505249,
+      "learning_rate": 9.98299242948844e-05,
+      "loss": 2.6279,
+      "step": 279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8688856363296509,
+      "learning_rate": 9.982725801859362e-05,
+      "loss": 2.7523,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7903696298599243,
+      "learning_rate": 9.98245710410958e-05,
+      "loss": 2.7453,
+      "step": 281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.858173668384552,
+      "learning_rate": 9.982186336350722e-05,
+      "loss": 2.6647,
+      "step": 282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8307147026062012,
+      "learning_rate": 9.981913498695285e-05,
+      "loss": 2.5604,
+      "step": 283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7315438389778137,
+      "learning_rate": 9.981638591256622e-05,
+      "loss": 2.7369,
+      "step": 284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8580665588378906,
+      "learning_rate": 9.981361614148945e-05,
+      "loss": 2.696,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7809886336326599,
+      "learning_rate": 9.981082567487325e-05,
+      "loss": 2.5879,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7830343842506409,
+      "learning_rate": 9.980801451387697e-05,
+      "loss": 2.5484,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8183286786079407,
+      "learning_rate": 9.980518265966852e-05,
+      "loss": 2.7612,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8343474268913269,
+      "learning_rate": 9.980233011342442e-05,
+      "loss": 2.5901,
+      "step": 289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9522347450256348,
+      "learning_rate": 9.979945687632982e-05,
+      "loss": 2.8261,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8049488067626953,
+      "learning_rate": 9.979656294957837e-05,
+      "loss": 2.7218,
+      "step": 291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.857140839099884,
+      "learning_rate": 9.979364833437243e-05,
+      "loss": 2.7545,
+      "step": 292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.899485170841217,
+      "learning_rate": 9.97907130319229e-05,
+      "loss": 2.7352,
+      "step": 293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7595685124397278,
+      "learning_rate": 9.978775704344928e-05,
+      "loss": 2.6128,
+      "step": 294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8028084635734558,
+      "learning_rate": 9.978478037017961e-05,
+      "loss": 2.6504,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.721229076385498,
+      "learning_rate": 9.978178301335066e-05,
+      "loss": 2.6998,
+      "step": 296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8368968367576599,
+      "learning_rate": 9.977876497420766e-05,
+      "loss": 2.6211,
+      "step": 297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8401256799697876,
+      "learning_rate": 9.977572625400448e-05,
+      "loss": 2.5657,
+      "step": 298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7477930784225464,
+      "learning_rate": 9.977266685400361e-05,
+      "loss": 2.5437,
+      "step": 299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7372848391532898,
+      "learning_rate": 9.97695867754761e-05,
+      "loss": 2.573,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6785719394683838,
+      "learning_rate": 9.97664860197016e-05,
+      "loss": 2.542,
+      "step": 301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8921648859977722,
+      "learning_rate": 9.976336458796832e-05,
+      "loss": 2.7424,
+      "step": 302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8670565485954285,
+      "learning_rate": 9.976022248157313e-05,
+      "loss": 2.6454,
+      "step": 303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9507176280021667,
+      "learning_rate": 9.975705970182141e-05,
+      "loss": 2.7046,
+      "step": 304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9055517911911011,
+      "learning_rate": 9.975387625002719e-05,
+      "loss": 2.7399,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8196343183517456,
+      "learning_rate": 9.975067212751308e-05,
+      "loss": 2.7075,
+      "step": 306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8114293217658997,
+      "learning_rate": 9.974744733561021e-05,
+      "loss": 2.4954,
+      "step": 307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8591504693031311,
+      "learning_rate": 9.97442018756584e-05,
+      "loss": 2.8986,
+      "step": 308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6892009973526001,
+      "learning_rate": 9.974093574900601e-05,
+      "loss": 2.6031,
+      "step": 309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7819831371307373,
+      "learning_rate": 9.973764895700994e-05,
+      "loss": 2.6815,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8065125346183777,
+      "learning_rate": 9.973434150103576e-05,
+      "loss": 2.6724,
+      "step": 311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9093921184539795,
+      "learning_rate": 9.973101338245753e-05,
+      "loss": 2.8612,
+      "step": 312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7525083422660828,
+      "learning_rate": 9.972766460265804e-05,
+      "loss": 2.598,
+      "step": 313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8567810654640198,
+      "learning_rate": 9.972429516302848e-05,
+      "loss": 2.6601,
+      "step": 314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7696500420570374,
+      "learning_rate": 9.972090506496876e-05,
+      "loss": 2.6938,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7607791423797607,
+      "learning_rate": 9.971749430988733e-05,
+      "loss": 2.7679,
+      "step": 316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.020056962966919,
+      "learning_rate": 9.97140628992012e-05,
+      "loss": 2.7033,
+      "step": 317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7067403197288513,
+      "learning_rate": 9.9710610834336e-05,
+      "loss": 2.6961,
+      "step": 318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7647778391838074,
+      "learning_rate": 9.970713811672591e-05,
+      "loss": 2.7684,
+      "step": 319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8393383026123047,
+      "learning_rate": 9.970364474781372e-05,
+      "loss": 2.6111,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7869490385055542,
+      "learning_rate": 9.970013072905078e-05,
+      "loss": 2.7934,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8329804539680481,
+      "learning_rate": 9.969659606189701e-05,
+      "loss": 2.6876,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8282068371772766,
+      "learning_rate": 9.969304074782093e-05,
+      "loss": 2.6275,
+      "step": 323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9450337290763855,
+      "learning_rate": 9.968946478829963e-05,
+      "loss": 2.7459,
+      "step": 324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7967404723167419,
+      "learning_rate": 9.968586818481877e-05,
+      "loss": 2.6018,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8883475661277771,
+      "learning_rate": 9.96822509388726e-05,
+      "loss": 2.7468,
+      "step": 326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8554742336273193,
+      "learning_rate": 9.967861305196393e-05,
+      "loss": 2.6533,
+      "step": 327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8726681470870972,
+      "learning_rate": 9.967495452560416e-05,
+      "loss": 2.7442,
+      "step": 328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.923151969909668,
+      "learning_rate": 9.967127536131326e-05,
+      "loss": 2.5924,
+      "step": 329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8226321935653687,
+      "learning_rate": 9.966757556061976e-05,
+      "loss": 2.9735,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8337855339050293,
+      "learning_rate": 9.96638551250608e-05,
+      "loss": 2.8677,
+      "step": 331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6853298544883728,
+      "learning_rate": 9.966011405618208e-05,
+      "loss": 2.5116,
+      "step": 332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9352625608444214,
+      "learning_rate": 9.965635235553784e-05,
+      "loss": 2.7025,
+      "step": 333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7430493235588074,
+      "learning_rate": 9.965257002469089e-05,
+      "loss": 2.6504,
+      "step": 334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8830888271331787,
+      "learning_rate": 9.964876706521268e-05,
+      "loss": 2.6294,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7388954758644104,
+      "learning_rate": 9.964494347868316e-05,
+      "loss": 2.6125,
+      "step": 336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7738839983940125,
+      "learning_rate": 9.96410992666909e-05,
+      "loss": 2.6884,
+      "step": 337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9271988272666931,
+      "learning_rate": 9.963723443083298e-05,
+      "loss": 2.7873,
+      "step": 338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7353416085243225,
+      "learning_rate": 9.96333489727151e-05,
+      "loss": 2.6941,
+      "step": 339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7990018725395203,
+      "learning_rate": 9.962944289395153e-05,
+      "loss": 2.7083,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8863693475723267,
+      "learning_rate": 9.962551619616505e-05,
+      "loss": 2.9436,
+      "step": 341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6994494795799255,
+      "learning_rate": 9.962156888098708e-05,
+      "loss": 2.7438,
+      "step": 342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9128270149230957,
+      "learning_rate": 9.961760095005753e-05,
+      "loss": 2.6318,
+      "step": 343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7746107578277588,
+      "learning_rate": 9.961361240502495e-05,
+      "loss": 2.5779,
+      "step": 344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9767360091209412,
+      "learning_rate": 9.96096032475464e-05,
+      "loss": 2.6333,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7274709343910217,
+      "learning_rate": 9.960557347928754e-05,
+      "loss": 2.7426,
+      "step": 346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7930894494056702,
+      "learning_rate": 9.960152310192255e-05,
+      "loss": 2.6029,
+      "step": 347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8602296710014343,
+      "learning_rate": 9.95974521171342e-05,
+      "loss": 2.7584,
+      "step": 348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7529052495956421,
+      "learning_rate": 9.959336052661386e-05,
+      "loss": 2.6912,
+      "step": 349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.79721599817276,
+      "learning_rate": 9.958924833206137e-05,
+      "loss": 2.6802,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8383491039276123,
+      "learning_rate": 9.958511553518521e-05,
+      "loss": 2.7049,
+      "step": 351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7134909629821777,
+      "learning_rate": 9.958096213770239e-05,
+      "loss": 2.5222,
+      "step": 352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8162690997123718,
+      "learning_rate": 9.957678814133844e-05,
+      "loss": 2.6294,
+      "step": 353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8344608545303345,
+      "learning_rate": 9.957259354782753e-05,
+      "loss": 2.8939,
+      "step": 354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7988254427909851,
+      "learning_rate": 9.956837835891232e-05,
+      "loss": 2.713,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8969252705574036,
+      "learning_rate": 9.956414257634408e-05,
+      "loss": 2.5782,
+      "step": 356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8966469764709473,
+      "learning_rate": 9.955988620188258e-05,
+      "loss": 2.5406,
+      "step": 357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8078041672706604,
+      "learning_rate": 9.955560923729617e-05,
+      "loss": 2.602,
+      "step": 358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7947049140930176,
+      "learning_rate": 9.955131168436175e-05,
+      "loss": 2.839,
+      "step": 359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7592992782592773,
+      "learning_rate": 9.954699354486481e-05,
+      "loss": 2.5064,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8197783827781677,
+      "learning_rate": 9.954265482059932e-05,
+      "loss": 2.7911,
+      "step": 361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8161508440971375,
+      "learning_rate": 9.953829551336788e-05,
+      "loss": 2.5422,
+      "step": 362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7685711979866028,
+      "learning_rate": 9.95339156249816e-05,
+      "loss": 2.4845,
+      "step": 363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7320737838745117,
+      "learning_rate": 9.952951515726013e-05,
+      "loss": 2.4413,
+      "step": 364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.689552366733551,
+      "learning_rate": 9.95250941120317e-05,
+      "loss": 2.3715,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8327648043632507,
+      "learning_rate": 9.952065249113307e-05,
+      "loss": 2.4651,
+      "step": 366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8683032989501953,
+      "learning_rate": 9.951619029640954e-05,
+      "loss": 2.8514,
+      "step": 367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8329164385795593,
+      "learning_rate": 9.951170752971501e-05,
+      "loss": 2.7054,
+      "step": 368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7724478840827942,
+      "learning_rate": 9.950720419291184e-05,
+      "loss": 2.3897,
+      "step": 369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9404852986335754,
+      "learning_rate": 9.950268028787102e-05,
+      "loss": 2.6536,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.824918270111084,
+      "learning_rate": 9.949813581647203e-05,
+      "loss": 2.6706,
+      "step": 371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8168641924858093,
+      "learning_rate": 9.949357078060291e-05,
+      "loss": 2.5127,
+      "step": 372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.829546332359314,
+      "learning_rate": 9.948898518216028e-05,
+      "loss": 2.68,
+      "step": 373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8319979310035706,
+      "learning_rate": 9.948437902304922e-05,
+      "loss": 2.521,
+      "step": 374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.818394660949707,
+      "learning_rate": 9.947975230518345e-05,
+      "loss": 2.6686,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0310651063919067,
+      "learning_rate": 9.947510503048516e-05,
+      "loss": 2.9114,
+      "step": 376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7248244285583496,
+      "learning_rate": 9.947043720088511e-05,
+      "loss": 2.4616,
+      "step": 377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8024706244468689,
+      "learning_rate": 9.946574881832259e-05,
+      "loss": 2.7584,
+      "step": 378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6976470947265625,
+      "learning_rate": 9.946103988474542e-05,
+      "loss": 2.569,
+      "step": 379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7425373196601868,
+      "learning_rate": 9.945631040211001e-05,
+      "loss": 2.4392,
+      "step": 380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8619625568389893,
+      "learning_rate": 9.945156037238122e-05,
+      "loss": 2.6985,
+      "step": 381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8687835931777954,
+      "learning_rate": 9.944678979753255e-05,
+      "loss": 2.6899,
+      "step": 382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8138022422790527,
+      "learning_rate": 9.944199867954593e-05,
+      "loss": 2.6538,
+      "step": 383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8898317813873291,
+      "learning_rate": 9.94371870204119e-05,
+      "loss": 2.762,
+      "step": 384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8393351435661316,
+      "learning_rate": 9.94323548221295e-05,
+      "loss": 2.6527,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8877520561218262,
+      "learning_rate": 9.942750208670633e-05,
+      "loss": 2.8081,
+      "step": 386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8737496137619019,
+      "learning_rate": 9.942262881615848e-05,
+      "loss": 2.5237,
+      "step": 387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7914659976959229,
+      "learning_rate": 9.941773501251061e-05,
+      "loss": 2.7301,
+      "step": 388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8276338577270508,
+      "learning_rate": 9.94128206777959e-05,
+      "loss": 2.6864,
+      "step": 389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8775596022605896,
+      "learning_rate": 9.940788581405606e-05,
+      "loss": 2.616,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8196595907211304,
+      "learning_rate": 9.940293042334131e-05,
+      "loss": 2.7994,
+      "step": 391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7924444675445557,
+      "learning_rate": 9.939795450771043e-05,
+      "loss": 2.598,
+      "step": 392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0316113233566284,
+      "learning_rate": 9.939295806923071e-05,
+      "loss": 2.8257,
+      "step": 393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8323584794998169,
+      "learning_rate": 9.938794110997795e-05,
+      "loss": 2.639,
+      "step": 394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6941940188407898,
+      "learning_rate": 9.938290363203651e-05,
+      "loss": 2.4029,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7764073610305786,
+      "learning_rate": 9.937784563749926e-05,
+      "loss": 2.5625,
+      "step": 396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8002656102180481,
+      "learning_rate": 9.937276712846757e-05,
+      "loss": 2.936,
+      "step": 397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7883482575416565,
+      "learning_rate": 9.936766810705138e-05,
+      "loss": 2.6994,
+      "step": 398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7349295020103455,
+      "learning_rate": 9.936254857536912e-05,
+      "loss": 2.7544,
+      "step": 399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7450819611549377,
+      "learning_rate": 9.935740853554773e-05,
+      "loss": 2.5475,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8077352046966553,
+      "learning_rate": 9.935224798972273e-05,
+      "loss": 2.7203,
+      "step": 401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9065093398094177,
+      "learning_rate": 9.934706694003808e-05,
+      "loss": 2.743,
+      "step": 402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7674577832221985,
+      "learning_rate": 9.934186538864629e-05,
+      "loss": 2.4251,
+      "step": 403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9564427137374878,
+      "learning_rate": 9.933664333770844e-05,
+      "loss": 2.7572,
+      "step": 404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8135406970977783,
+      "learning_rate": 9.933140078939404e-05,
+      "loss": 2.7079,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8982585668563843,
+      "learning_rate": 9.932613774588117e-05,
+      "loss": 2.669,
+      "step": 406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8759541511535645,
+      "learning_rate": 9.93208542093564e-05,
+      "loss": 2.8098,
+      "step": 407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9763906002044678,
+      "learning_rate": 9.931555018201484e-05,
+      "loss": 2.6328,
+      "step": 408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8444701433181763,
+      "learning_rate": 9.93102256660601e-05,
+      "loss": 2.5238,
+      "step": 409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9019443392753601,
+      "learning_rate": 9.93048806637043e-05,
+      "loss": 2.8322,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8568253517150879,
+      "learning_rate": 9.929951517716804e-05,
+      "loss": 2.7088,
+      "step": 411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8701837658882141,
+      "learning_rate": 9.929412920868051e-05,
+      "loss": 2.716,
+      "step": 412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8999288082122803,
+      "learning_rate": 9.928872276047933e-05,
+      "loss": 2.4605,
+      "step": 413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8750538229942322,
+      "learning_rate": 9.928329583481067e-05,
+      "loss": 2.7552,
+      "step": 414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8071305751800537,
+      "learning_rate": 9.927784843392921e-05,
+      "loss": 2.4979,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7555310726165771,
+      "learning_rate": 9.92723805600981e-05,
+      "loss": 2.4109,
+      "step": 416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7860155701637268,
+      "learning_rate": 9.926689221558904e-05,
+      "loss": 2.4743,
+      "step": 417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8705489635467529,
+      "learning_rate": 9.926138340268221e-05,
+      "loss": 2.5176,
+      "step": 418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9023087024688721,
+      "learning_rate": 9.92558541236663e-05,
+      "loss": 2.7436,
+      "step": 419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9064571261405945,
+      "learning_rate": 9.925030438083849e-05,
+      "loss": 2.6743,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8368931412696838,
+      "learning_rate": 9.924473417650446e-05,
+      "loss": 2.4735,
+      "step": 421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8718896508216858,
+      "learning_rate": 9.923914351297845e-05,
+      "loss": 2.5572,
+      "step": 422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7399803996086121,
+      "learning_rate": 9.923353239258313e-05,
+      "loss": 2.4494,
+      "step": 423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8385767340660095,
+      "learning_rate": 9.922790081764967e-05,
+      "loss": 2.6536,
+      "step": 424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8154153823852539,
+      "learning_rate": 9.922224879051782e-05,
+      "loss": 2.6535,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.721518337726593,
+      "learning_rate": 9.921657631353572e-05,
+      "loss": 2.5068,
+      "step": 426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6836869120597839,
+      "learning_rate": 9.921088338906004e-05,
+      "loss": 2.6774,
+      "step": 427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7791640758514404,
+      "learning_rate": 9.920517001945602e-05,
+      "loss": 2.5462,
+      "step": 428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8162728548049927,
+      "learning_rate": 9.919943620709727e-05,
+      "loss": 2.6862,
+      "step": 429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8810282945632935,
+      "learning_rate": 9.919368195436601e-05,
+      "loss": 2.6322,
+      "step": 430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8009904026985168,
+      "learning_rate": 9.918790726365286e-05,
+      "loss": 2.9101,
+      "step": 431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6784713864326477,
+      "learning_rate": 9.918211213735699e-05,
+      "loss": 2.4451,
+      "step": 432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7634642124176025,
+      "learning_rate": 9.917629657788603e-05,
+      "loss": 2.3801,
+      "step": 433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7026737332344055,
+      "learning_rate": 9.917046058765612e-05,
+      "loss": 2.5363,
+      "step": 434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8839834332466125,
+      "learning_rate": 9.916460416909187e-05,
+      "loss": 2.7754,
+      "step": 435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8632910847663879,
+      "learning_rate": 9.915872732462639e-05,
+      "loss": 2.8152,
+      "step": 436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8932034969329834,
+      "learning_rate": 9.915283005670126e-05,
+      "loss": 2.6962,
+      "step": 437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7760161757469177,
+      "learning_rate": 9.914691236776657e-05,
+      "loss": 2.7129,
+      "step": 438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8482321500778198,
+      "learning_rate": 9.914097426028088e-05,
+      "loss": 2.4466,
+      "step": 439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7343797087669373,
+      "learning_rate": 9.913501573671122e-05,
+      "loss": 2.5819,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8916664123535156,
+      "learning_rate": 9.912903679953311e-05,
+      "loss": 2.6591,
+      "step": 441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9987776279449463,
+      "learning_rate": 9.912303745123059e-05,
+      "loss": 2.6127,
+      "step": 442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.006719946861267,
+      "learning_rate": 9.911701769429611e-05,
+      "loss": 2.7387,
+      "step": 443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9053053855895996,
+      "learning_rate": 9.911097753123064e-05,
+      "loss": 2.6412,
+      "step": 444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0155466794967651,
+      "learning_rate": 9.910491696454365e-05,
+      "loss": 2.8149,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.887499213218689,
+      "learning_rate": 9.909883599675303e-05,
+      "loss": 2.5725,
+      "step": 446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8541029095649719,
+      "learning_rate": 9.90927346303852e-05,
+      "loss": 2.8406,
+      "step": 447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9062759280204773,
+      "learning_rate": 9.908661286797502e-05,
+      "loss": 2.6437,
+      "step": 448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8206595182418823,
+      "learning_rate": 9.908047071206582e-05,
+      "loss": 2.4447,
+      "step": 449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7528868317604065,
+      "learning_rate": 9.907430816520944e-05,
+      "loss": 2.3723,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7756978869438171,
+      "learning_rate": 9.906812522996615e-05,
+      "loss": 2.741,
+      "step": 451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8813048005104065,
+      "learning_rate": 9.906192190890472e-05,
+      "loss": 2.6291,
+      "step": 452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8096169829368591,
+      "learning_rate": 9.905569820460238e-05,
+      "loss": 2.5103,
+      "step": 453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7809406518936157,
+      "learning_rate": 9.904945411964481e-05,
+      "loss": 2.4597,
+      "step": 454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9555474519729614,
+      "learning_rate": 9.90431896566262e-05,
+      "loss": 2.8689,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0121536254882812,
+      "learning_rate": 9.903690481814916e-05,
+      "loss": 2.7397,
+      "step": 456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7492336630821228,
+      "learning_rate": 9.903059960682481e-05,
+      "loss": 2.6305,
+      "step": 457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8646408915519714,
+      "learning_rate": 9.902427402527268e-05,
+      "loss": 2.6508,
+      "step": 458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8946760892868042,
+      "learning_rate": 9.90179280761208e-05,
+      "loss": 2.549,
+      "step": 459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0213191509246826,
+      "learning_rate": 9.901156176200566e-05,
+      "loss": 2.8957,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9659153819084167,
+      "learning_rate": 9.900517508557223e-05,
+      "loss": 2.6606,
+      "step": 461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9098890423774719,
+      "learning_rate": 9.899876804947388e-05,
+      "loss": 2.8194,
+      "step": 462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8891843557357788,
+      "learning_rate": 9.899234065637247e-05,
+      "loss": 2.6815,
+      "step": 463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9320822358131409,
+      "learning_rate": 9.898589290893836e-05,
+      "loss": 2.6761,
+      "step": 464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8484837412834167,
+      "learning_rate": 9.897942480985029e-05,
+      "loss": 2.6326,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7779737114906311,
+      "learning_rate": 9.897293636179551e-05,
+      "loss": 2.6753,
+      "step": 466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7977185249328613,
+      "learning_rate": 9.896642756746973e-05,
+      "loss": 2.5886,
+      "step": 467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7587082982063293,
+      "learning_rate": 9.895989842957703e-05,
+      "loss": 2.7967,
+      "step": 468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6483625769615173,
+      "learning_rate": 9.895334895083006e-05,
+      "loss": 2.3075,
+      "step": 469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7477641105651855,
+      "learning_rate": 9.894677913394983e-05,
+      "loss": 2.7634,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7287459373474121,
+      "learning_rate": 9.894018898166584e-05,
+      "loss": 2.6134,
+      "step": 471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7630150318145752,
+      "learning_rate": 9.893357849671605e-05,
+      "loss": 2.4481,
+      "step": 472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.806857705116272,
+      "learning_rate": 9.892694768184684e-05,
+      "loss": 2.6888,
+      "step": 473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8435704112052917,
+      "learning_rate": 9.892029653981303e-05,
+      "loss": 2.5365,
+      "step": 474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8922801613807678,
+      "learning_rate": 9.89136250733779e-05,
+      "loss": 2.6923,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9545143246650696,
+      "learning_rate": 9.89069332853132e-05,
+      "loss": 2.7206,
+      "step": 476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7804939150810242,
+      "learning_rate": 9.890022117839906e-05,
+      "loss": 2.4251,
+      "step": 477
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7924111485481262,
+      "learning_rate": 9.889348875542413e-05,
+      "loss": 2.5708,
+      "step": 478
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7692276835441589,
+      "learning_rate": 9.88867360191854e-05,
+      "loss": 2.3966,
+      "step": 479
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8817324042320251,
+      "learning_rate": 9.887996297248844e-05,
+      "loss": 2.5312,
+      "step": 480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8083483576774597,
+      "learning_rate": 9.887316961814712e-05,
+      "loss": 2.5101,
+      "step": 481
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8889942169189453,
+      "learning_rate": 9.886635595898382e-05,
+      "loss": 2.5571,
+      "step": 482
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8455393314361572,
+      "learning_rate": 9.88595219978293e-05,
+      "loss": 2.494,
+      "step": 483
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8857603073120117,
+      "learning_rate": 9.885266773752286e-05,
+      "loss": 2.4952,
+      "step": 484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8819442391395569,
+      "learning_rate": 9.884579318091214e-05,
+      "loss": 2.6347,
+      "step": 485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8141264319419861,
+      "learning_rate": 9.883889833085322e-05,
+      "loss": 2.5474,
+      "step": 486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8088395595550537,
+      "learning_rate": 9.883198319021066e-05,
+      "loss": 2.4982,
+      "step": 487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8304647207260132,
+      "learning_rate": 9.88250477618574e-05,
+      "loss": 2.5598,
+      "step": 488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8097513914108276,
+      "learning_rate": 9.881809204867485e-05,
+      "loss": 2.4064,
+      "step": 489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7212309241294861,
+      "learning_rate": 9.88111160535528e-05,
+      "loss": 2.4212,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6799522638320923,
+      "learning_rate": 9.880411977938952e-05,
+      "loss": 2.4958,
+      "step": 491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7942051291465759,
+      "learning_rate": 9.879710322909165e-05,
+      "loss": 2.7001,
+      "step": 492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8030378222465515,
+      "learning_rate": 9.879006640557431e-05,
+      "loss": 2.502,
+      "step": 493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7573240399360657,
+      "learning_rate": 9.8783009311761e-05,
+      "loss": 2.438,
+      "step": 494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7588917016983032,
+      "learning_rate": 9.877593195058367e-05,
+      "loss": 2.5639,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8095919489860535,
+      "learning_rate": 9.876883432498265e-05,
+      "loss": 2.6628,
+      "step": 496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8059120774269104,
+      "learning_rate": 9.876171643790675e-05,
+      "loss": 2.4055,
+      "step": 497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9393121600151062,
+      "learning_rate": 9.875457829231313e-05,
+      "loss": 2.7048,
+      "step": 498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.730537474155426,
+      "learning_rate": 9.874741989116743e-05,
+      "loss": 2.4618,
+      "step": 499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8142967224121094,
+      "learning_rate": 9.874024123744367e-05,
+      "loss": 2.6862,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7797869443893433,
+      "learning_rate": 9.873304233412426e-05,
+      "loss": 2.4399,
+      "step": 501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9106787443161011,
+      "learning_rate": 9.87258231842001e-05,
+      "loss": 2.8232,
+      "step": 502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7860686779022217,
+      "learning_rate": 9.871858379067041e-05,
+      "loss": 2.5431,
+      "step": 503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7612290382385254,
+      "learning_rate": 9.87113241565429e-05,
+      "loss": 2.5759,
+      "step": 504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8477287888526917,
+      "learning_rate": 9.870404428483362e-05,
+      "loss": 2.7125,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8533612489700317,
+      "learning_rate": 9.86967441785671e-05,
+      "loss": 2.5665,
+      "step": 506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8847472071647644,
+      "learning_rate": 9.868942384077621e-05,
+      "loss": 2.5987,
+      "step": 507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.771891176700592,
+      "learning_rate": 9.868208327450226e-05,
+      "loss": 2.4868,
+      "step": 508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7635993957519531,
+      "learning_rate": 9.867472248279497e-05,
+      "loss": 2.4985,
+      "step": 509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8201240301132202,
+      "learning_rate": 9.866734146871242e-05,
+      "loss": 2.4618,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8513836860656738,
+      "learning_rate": 9.865994023532115e-05,
+      "loss": 2.6808,
+      "step": 511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8396844267845154,
+      "learning_rate": 9.865251878569607e-05,
+      "loss": 2.4539,
+      "step": 512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8808681964874268,
+      "learning_rate": 9.86450771229205e-05,
+      "loss": 2.6926,
+      "step": 513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.031578779220581,
+      "learning_rate": 9.863761525008612e-05,
+      "loss": 2.8921,
+      "step": 514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8537565469741821,
+      "learning_rate": 9.863013317029305e-05,
+      "loss": 2.6498,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9759238958358765,
+      "learning_rate": 9.862263088664981e-05,
+      "loss": 2.5966,
+      "step": 516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8199980854988098,
+      "learning_rate": 9.861510840227329e-05,
+      "loss": 2.7307,
+      "step": 517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8716624975204468,
+      "learning_rate": 9.860756572028873e-05,
+      "loss": 2.9306,
+      "step": 518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7007376551628113,
+      "learning_rate": 9.860000284382987e-05,
+      "loss": 2.4597,
+      "step": 519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9326262474060059,
+      "learning_rate": 9.859241977603875e-05,
+      "loss": 2.7057,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8084246516227722,
+      "learning_rate": 9.858481652006581e-05,
+      "loss": 2.7191,
+      "step": 521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7747828960418701,
+      "learning_rate": 9.857719307906995e-05,
+      "loss": 2.6666,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8678902983665466,
+      "learning_rate": 9.856954945621834e-05,
+      "loss": 2.5272,
+      "step": 523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8224223852157593,
+      "learning_rate": 9.856188565468664e-05,
+      "loss": 2.6295,
+      "step": 524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7473587393760681,
+      "learning_rate": 9.855420167765882e-05,
+      "loss": 2.6095,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7718611359596252,
+      "learning_rate": 9.854649752832728e-05,
+      "loss": 2.3885,
+      "step": 526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.879264771938324,
+      "learning_rate": 9.853877320989277e-05,
+      "loss": 2.6725,
+      "step": 527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6908172369003296,
+      "learning_rate": 9.853102872556443e-05,
+      "loss": 2.5639,
+      "step": 528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9951748847961426,
+      "learning_rate": 9.85232640785598e-05,
+      "loss": 2.5916,
+      "step": 529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9338299036026001,
+      "learning_rate": 9.851547927210477e-05,
+      "loss": 2.6414,
+      "step": 530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7971630096435547,
+      "learning_rate": 9.850767430943358e-05,
+      "loss": 2.7777,
+      "step": 531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7246478796005249,
+      "learning_rate": 9.84998491937889e-05,
+      "loss": 2.4613,
+      "step": 532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9147344827651978,
+      "learning_rate": 9.849200392842176e-05,
+      "loss": 2.5972,
+      "step": 533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0179131031036377,
+      "learning_rate": 9.848413851659152e-05,
+      "loss": 2.7454,
+      "step": 534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8567588329315186,
+      "learning_rate": 9.847625296156598e-05,
+      "loss": 2.5307,
+      "step": 535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8043563365936279,
+      "learning_rate": 9.846834726662124e-05,
+      "loss": 2.5642,
+      "step": 536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8981109857559204,
+      "learning_rate": 9.84604214350418e-05,
+      "loss": 2.6313,
+      "step": 537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7680789232254028,
+      "learning_rate": 9.845247547012054e-05,
+      "loss": 2.4995,
+      "step": 538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9186317920684814,
+      "learning_rate": 9.844450937515864e-05,
+      "loss": 2.6727,
+      "step": 539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7494387626647949,
+      "learning_rate": 9.843652315346574e-05,
+      "loss": 2.6276,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8913398385047913,
+      "learning_rate": 9.842851680835977e-05,
+      "loss": 2.5588,
+      "step": 541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7759923934936523,
+      "learning_rate": 9.842049034316705e-05,
+      "loss": 2.6583,
+      "step": 542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8929766416549683,
+      "learning_rate": 9.841244376122223e-05,
+      "loss": 2.4762,
+      "step": 543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7863863706588745,
+      "learning_rate": 9.840437706586836e-05,
+      "loss": 2.5423,
+      "step": 544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8556399345397949,
+      "learning_rate": 9.839629026045682e-05,
+      "loss": 2.6239,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8204936385154724,
+      "learning_rate": 9.838818334834735e-05,
+      "loss": 2.5346,
+      "step": 546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8093518018722534,
+      "learning_rate": 9.838005633290804e-05,
+      "loss": 2.4062,
+      "step": 547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.849304735660553,
+      "learning_rate": 9.837190921751535e-05,
+      "loss": 2.5108,
+      "step": 548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7636126279830933,
+      "learning_rate": 9.836374200555405e-05,
+      "loss": 2.5399,
+      "step": 549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8015439510345459,
+      "learning_rate": 9.83555547004173e-05,
+      "loss": 2.6545,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7884140014648438,
+      "learning_rate": 9.83473473055066e-05,
+      "loss": 2.4885,
+      "step": 551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9432367086410522,
+      "learning_rate": 9.833911982423179e-05,
+      "loss": 2.6253,
+      "step": 552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9360383152961731,
+      "learning_rate": 9.833087226001105e-05,
+      "loss": 2.6967,
+      "step": 553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8199325203895569,
+      "learning_rate": 9.832260461627089e-05,
+      "loss": 2.4797,
+      "step": 554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8625393509864807,
+      "learning_rate": 9.831431689644621e-05,
+      "loss": 2.7208,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8158754706382751,
+      "learning_rate": 9.830600910398023e-05,
+      "loss": 2.5712,
+      "step": 556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8487555980682373,
+      "learning_rate": 9.829768124232446e-05,
+      "loss": 2.74,
+      "step": 557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7938382029533386,
+      "learning_rate": 9.828933331493881e-05,
+      "loss": 2.4722,
+      "step": 558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8773726224899292,
+      "learning_rate": 9.828096532529152e-05,
+      "loss": 2.3106,
+      "step": 559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7719241380691528,
+      "learning_rate": 9.827257727685912e-05,
+      "loss": 2.6555,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8159040808677673,
+      "learning_rate": 9.826416917312654e-05,
+      "loss": 2.5329,
+      "step": 561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8229182958602905,
+      "learning_rate": 9.825574101758697e-05,
+      "loss": 2.4891,
+      "step": 562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8250560760498047,
+      "learning_rate": 9.824729281374201e-05,
+      "loss": 2.5347,
+      "step": 563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9243699312210083,
+      "learning_rate": 9.823882456510149e-05,
+      "loss": 2.592,
+      "step": 564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7569814920425415,
+      "learning_rate": 9.823033627518367e-05,
+      "loss": 2.6381,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8083333373069763,
+      "learning_rate": 9.822182794751509e-05,
+      "loss": 2.6116,
+      "step": 566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8862336874008179,
+      "learning_rate": 9.82132995856306e-05,
+      "loss": 2.4314,
+      "step": 567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9001768827438354,
+      "learning_rate": 9.820475119307337e-05,
+      "loss": 2.7729,
+      "step": 568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8941066861152649,
+      "learning_rate": 9.819618277339495e-05,
+      "loss": 2.4904,
+      "step": 569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.964531660079956,
+      "learning_rate": 9.818759433015516e-05,
+      "loss": 2.7135,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9375636577606201,
+      "learning_rate": 9.817898586692214e-05,
+      "loss": 2.6623,
+      "step": 571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7890806794166565,
+      "learning_rate": 9.817035738727234e-05,
+      "loss": 2.4595,
+      "step": 572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.811140775680542,
+      "learning_rate": 9.816170889479058e-05,
+      "loss": 2.5288,
+      "step": 573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.91343092918396,
+      "learning_rate": 9.815304039306995e-05,
+      "loss": 2.6651,
+      "step": 574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.931253969669342,
+      "learning_rate": 9.814435188571184e-05,
+      "loss": 2.7396,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8641043305397034,
+      "learning_rate": 9.813564337632599e-05,
+      "loss": 2.538,
+      "step": 576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8326981663703918,
+      "learning_rate": 9.812691486853042e-05,
+      "loss": 2.6228,
+      "step": 577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8766738176345825,
+      "learning_rate": 9.81181663659515e-05,
+      "loss": 2.4112,
+      "step": 578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8376889824867249,
+      "learning_rate": 9.810939787222386e-05,
+      "loss": 2.273,
+      "step": 579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9325894713401794,
+      "learning_rate": 9.810060939099045e-05,
+      "loss": 2.8498,
+      "step": 580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.85801762342453,
+      "learning_rate": 9.809180092590254e-05,
+      "loss": 2.5871,
+      "step": 581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9565813541412354,
+      "learning_rate": 9.808297248061966e-05,
+      "loss": 2.5834,
+      "step": 582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7715215086936951,
+      "learning_rate": 9.807412405880972e-05,
+      "loss": 2.4791,
+      "step": 583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9487622380256653,
+      "learning_rate": 9.806525566414883e-05,
+      "loss": 2.7606,
+      "step": 584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8899068832397461,
+      "learning_rate": 9.805636730032149e-05,
+      "loss": 2.6684,
+      "step": 585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.922802746295929,
+      "learning_rate": 9.804745897102041e-05,
+      "loss": 2.5697,
+      "step": 586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8120271563529968,
+      "learning_rate": 9.80385306799467e-05,
+      "loss": 2.5061,
+      "step": 587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.848046600818634,
+      "learning_rate": 9.802958243080966e-05,
+      "loss": 2.65,
+      "step": 588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8484374284744263,
+      "learning_rate": 9.802061422732692e-05,
+      "loss": 2.5112,
+      "step": 589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9390422105789185,
+      "learning_rate": 9.801162607322444e-05,
+      "loss": 2.5327,
+      "step": 590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9485512375831604,
+      "learning_rate": 9.800261797223641e-05,
+      "loss": 2.6107,
+      "step": 591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0079869031906128,
+      "learning_rate": 9.799358992810533e-05,
+      "loss": 2.696,
+      "step": 592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9294269680976868,
+      "learning_rate": 9.7984541944582e-05,
+      "loss": 2.5018,
+      "step": 593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8038244843482971,
+      "learning_rate": 9.797547402542549e-05,
+      "loss": 2.4804,
+      "step": 594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9279298186302185,
+      "learning_rate": 9.796638617440314e-05,
+      "loss": 2.4988,
+      "step": 595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8789168000221252,
+      "learning_rate": 9.795727839529058e-05,
+      "loss": 2.6111,
+      "step": 596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7688995599746704,
+      "learning_rate": 9.794815069187176e-05,
+      "loss": 2.6335,
+      "step": 597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7985723614692688,
+      "learning_rate": 9.793900306793882e-05,
+      "loss": 2.5446,
+      "step": 598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7922737002372742,
+      "learning_rate": 9.792983552729225e-05,
+      "loss": 2.5358,
+      "step": 599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7974783778190613,
+      "learning_rate": 9.79206480737408e-05,
+      "loss": 2.5939,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7665889859199524,
+      "learning_rate": 9.791144071110149e-05,
+      "loss": 2.5671,
+      "step": 601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9134510159492493,
+      "learning_rate": 9.790221344319958e-05,
+      "loss": 2.6896,
+      "step": 602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9396125674247742,
+      "learning_rate": 9.789296627386865e-05,
+      "loss": 2.7384,
+      "step": 603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8175870180130005,
+      "learning_rate": 9.78836992069505e-05,
+      "loss": 2.5994,
+      "step": 604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.866368293762207,
+      "learning_rate": 9.787441224629525e-05,
+      "loss": 2.6615,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8564729690551758,
+      "learning_rate": 9.786510539576124e-05,
+      "loss": 2.7554,
+      "step": 606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8464537262916565,
+      "learning_rate": 9.785577865921509e-05,
+      "loss": 2.6275,
+      "step": 607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8628912568092346,
+      "learning_rate": 9.784643204053167e-05,
+      "loss": 2.4489,
+      "step": 608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.911533534526825,
+      "learning_rate": 9.783706554359416e-05,
+      "loss": 2.6346,
+      "step": 609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9320758581161499,
+      "learning_rate": 9.78276791722939e-05,
+      "loss": 2.6102,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0003029108047485,
+      "learning_rate": 9.78182729305306e-05,
+      "loss": 2.6623,
+      "step": 611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8281003832817078,
+      "learning_rate": 9.780884682221215e-05,
+      "loss": 2.6449,
+      "step": 612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9109224081039429,
+      "learning_rate": 9.779940085125471e-05,
+      "loss": 2.5409,
+      "step": 613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7955072522163391,
+      "learning_rate": 9.778993502158271e-05,
+      "loss": 2.5369,
+      "step": 614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8612529635429382,
+      "learning_rate": 9.778044933712883e-05,
+      "loss": 2.4815,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7752905488014221,
+      "learning_rate": 9.777094380183396e-05,
+      "loss": 2.5578,
+      "step": 616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8656027913093567,
+      "learning_rate": 9.776141841964728e-05,
+      "loss": 2.5509,
+      "step": 617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9025648236274719,
+      "learning_rate": 9.77518731945262e-05,
+      "loss": 2.6159,
+      "step": 618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7324679493904114,
+      "learning_rate": 9.774230813043638e-05,
+      "loss": 2.4396,
+      "step": 619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.805213212966919,
+      "learning_rate": 9.77327232313517e-05,
+      "loss": 2.5829,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7210841774940491,
+      "learning_rate": 9.772311850125431e-05,
+      "loss": 2.3443,
+      "step": 621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7831017971038818,
+      "learning_rate": 9.771349394413458e-05,
+      "loss": 2.6018,
+      "step": 622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7572575211524963,
+      "learning_rate": 9.770384956399112e-05,
+      "loss": 2.4386,
+      "step": 623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9449385404586792,
+      "learning_rate": 9.76941853648308e-05,
+      "loss": 2.4336,
+      "step": 624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.800679087638855,
+      "learning_rate": 9.768450135066868e-05,
+      "loss": 2.4859,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7943803668022156,
+      "learning_rate": 9.767479752552808e-05,
+      "loss": 2.5869,
+      "step": 626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.826631486415863,
+      "learning_rate": 9.766507389344057e-05,
+      "loss": 2.4083,
+      "step": 627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8961843848228455,
+      "learning_rate": 9.76553304584459e-05,
+      "loss": 2.5688,
+      "step": 628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7377244234085083,
+      "learning_rate": 9.764556722459206e-05,
+      "loss": 2.51,
+      "step": 629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7931573987007141,
+      "learning_rate": 9.763578419593532e-05,
+      "loss": 2.4611,
+      "step": 630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8119449615478516,
+      "learning_rate": 9.762598137654009e-05,
+      "loss": 2.6067,
+      "step": 631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7726951837539673,
+      "learning_rate": 9.761615877047907e-05,
+      "loss": 2.4436,
+      "step": 632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8500878214836121,
+      "learning_rate": 9.760631638183315e-05,
+      "loss": 2.5177,
+      "step": 633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.969874918460846,
+      "learning_rate": 9.759645421469146e-05,
+      "loss": 2.6097,
+      "step": 634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8367302417755127,
+      "learning_rate": 9.75865722731513e-05,
+      "loss": 2.6787,
+      "step": 635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8564853072166443,
+      "learning_rate": 9.757667056131821e-05,
+      "loss": 2.7243,
+      "step": 636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8722653985023499,
+      "learning_rate": 9.756674908330598e-05,
+      "loss": 2.4775,
+      "step": 637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8727867603302002,
+      "learning_rate": 9.755680784323657e-05,
+      "loss": 2.4932,
+      "step": 638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.851186990737915,
+      "learning_rate": 9.754684684524016e-05,
+      "loss": 2.6851,
+      "step": 639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9135060906410217,
+      "learning_rate": 9.753686609345516e-05,
+      "loss": 2.6079,
+      "step": 640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8025040030479431,
+      "learning_rate": 9.752686559202815e-05,
+      "loss": 2.6485,
+      "step": 641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8136054277420044,
+      "learning_rate": 9.751684534511392e-05,
+      "loss": 2.4585,
+      "step": 642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8092338442802429,
+      "learning_rate": 9.75068053568755e-05,
+      "loss": 2.5357,
+      "step": 643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8788452744483948,
+      "learning_rate": 9.74967456314841e-05,
+      "loss": 2.6177,
+      "step": 644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8854349851608276,
+      "learning_rate": 9.748666617311912e-05,
+      "loss": 2.738,
+      "step": 645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.963509202003479,
+      "learning_rate": 9.747656698596816e-05,
+      "loss": 2.622,
+      "step": 646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8236713409423828,
+      "learning_rate": 9.746644807422702e-05,
+      "loss": 2.4663,
+      "step": 647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8729363083839417,
+      "learning_rate": 9.745630944209971e-05,
+      "loss": 2.4499,
+      "step": 648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8630483746528625,
+      "learning_rate": 9.744615109379843e-05,
+      "loss": 2.5845,
+      "step": 649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8970080018043518,
+      "learning_rate": 9.743597303354355e-05,
+      "loss": 2.6337,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8775519728660583,
+      "learning_rate": 9.742577526556363e-05,
+      "loss": 2.705,
+      "step": 651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7180139422416687,
+      "learning_rate": 9.741555779409545e-05,
+      "loss": 2.2397,
+      "step": 652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8728063702583313,
+      "learning_rate": 9.740532062338394e-05,
+      "loss": 2.7192,
+      "step": 653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7323952317237854,
+      "learning_rate": 9.739506375768224e-05,
+      "loss": 2.337,
+      "step": 654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7435590624809265,
+      "learning_rate": 9.738478720125165e-05,
+      "loss": 2.3323,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7636069655418396,
+      "learning_rate": 9.737449095836169e-05,
+      "loss": 2.5429,
+      "step": 656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8566385507583618,
+      "learning_rate": 9.736417503329001e-05,
+      "loss": 2.4715,
+      "step": 657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7937480807304382,
+      "learning_rate": 9.735383943032244e-05,
+      "loss": 2.5833,
+      "step": 658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9081477522850037,
+      "learning_rate": 9.734348415375305e-05,
+      "loss": 2.6781,
+      "step": 659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8089854121208191,
+      "learning_rate": 9.733310920788401e-05,
+      "loss": 2.4107,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9076220989227295,
+      "learning_rate": 9.732271459702569e-05,
+      "loss": 2.8088,
+      "step": 661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7662338018417358,
+      "learning_rate": 9.731230032549663e-05,
+      "loss": 2.4892,
+      "step": 662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8634386658668518,
+      "learning_rate": 9.730186639762353e-05,
+      "loss": 2.6197,
+      "step": 663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8576626777648926,
+      "learning_rate": 9.72914128177413e-05,
+      "loss": 2.5535,
+      "step": 664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9263651371002197,
+      "learning_rate": 9.728093959019294e-05,
+      "loss": 2.7611,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8200981020927429,
+      "learning_rate": 9.727044671932968e-05,
+      "loss": 2.3712,
+      "step": 666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.980901300907135,
+      "learning_rate": 9.725993420951086e-05,
+      "loss": 2.5217,
+      "step": 667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8675286173820496,
+      "learning_rate": 9.7249402065104e-05,
+      "loss": 2.405,
+      "step": 668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8116831183433533,
+      "learning_rate": 9.723885029048481e-05,
+      "loss": 2.6801,
+      "step": 669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8333988785743713,
+      "learning_rate": 9.72282788900371e-05,
+      "loss": 2.3584,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0178238153457642,
+      "learning_rate": 9.721768786815286e-05,
+      "loss": 2.6897,
+      "step": 671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8240295648574829,
+      "learning_rate": 9.720707722923225e-05,
+      "loss": 2.404,
+      "step": 672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.812338650226593,
+      "learning_rate": 9.719644697768353e-05,
+      "loss": 2.6146,
+      "step": 673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8723152875900269,
+      "learning_rate": 9.718579711792315e-05,
+      "loss": 2.4352,
+      "step": 674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8809595704078674,
+      "learning_rate": 9.717512765437571e-05,
+      "loss": 2.6343,
+      "step": 675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8982685804367065,
+      "learning_rate": 9.716443859147392e-05,
+      "loss": 2.6872,
+      "step": 676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8772156834602356,
+      "learning_rate": 9.715372993365868e-05,
+      "loss": 2.5171,
+      "step": 677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.837755560874939,
+      "learning_rate": 9.714300168537897e-05,
+      "loss": 2.4678,
+      "step": 678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7710533738136292,
+      "learning_rate": 9.713225385109197e-05,
+      "loss": 2.4434,
+      "step": 679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9214946627616882,
+      "learning_rate": 9.712148643526294e-05,
+      "loss": 2.4139,
+      "step": 680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0227956771850586,
+      "learning_rate": 9.711069944236532e-05,
+      "loss": 2.5425,
+      "step": 681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9563687443733215,
+      "learning_rate": 9.70998928768807e-05,
+      "loss": 2.5767,
+      "step": 682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.129826545715332,
+      "learning_rate": 9.708906674329872e-05,
+      "loss": 2.9332,
+      "step": 683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8897944092750549,
+      "learning_rate": 9.707822104611721e-05,
+      "loss": 2.6602,
+      "step": 684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8710518479347229,
+      "learning_rate": 9.706735578984215e-05,
+      "loss": 2.5553,
+      "step": 685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9633013606071472,
+      "learning_rate": 9.70564709789876e-05,
+      "loss": 2.5662,
+      "step": 686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8311755061149597,
+      "learning_rate": 9.704556661807573e-05,
+      "loss": 2.6726,
+      "step": 687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7835612893104553,
+      "learning_rate": 9.703464271163689e-05,
+      "loss": 2.5764,
+      "step": 688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.945631742477417,
+      "learning_rate": 9.70236992642095e-05,
+      "loss": 2.6301,
+      "step": 689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7845411896705627,
+      "learning_rate": 9.701273628034015e-05,
+      "loss": 2.7705,
+      "step": 690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9557234048843384,
+      "learning_rate": 9.700175376458348e-05,
+      "loss": 2.5626,
+      "step": 691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8278346061706543,
+      "learning_rate": 9.69907517215023e-05,
+      "loss": 2.4655,
+      "step": 692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8081557154655457,
+      "learning_rate": 9.697973015566752e-05,
+      "loss": 2.3697,
+      "step": 693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8579469919204712,
+      "learning_rate": 9.696868907165813e-05,
+      "loss": 2.5245,
+      "step": 694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8345386981964111,
+      "learning_rate": 9.695762847406127e-05,
+      "loss": 2.4396,
+      "step": 695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9190948605537415,
+      "learning_rate": 9.694654836747216e-05,
+      "loss": 2.5501,
+      "step": 696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9420356154441833,
+      "learning_rate": 9.693544875649416e-05,
+      "loss": 2.5066,
+      "step": 697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8013123869895935,
+      "learning_rate": 9.69243296457387e-05,
+      "loss": 2.5356,
+      "step": 698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8184309005737305,
+      "learning_rate": 9.691319103982527e-05,
+      "loss": 2.5957,
+      "step": 699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.022031545639038,
+      "learning_rate": 9.69020329433816e-05,
+      "loss": 2.8988,
+      "step": 700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9064924120903015,
+      "learning_rate": 9.689085536104334e-05,
+      "loss": 2.5065,
+      "step": 701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9356926083564758,
+      "learning_rate": 9.687965829745438e-05,
+      "loss": 2.4153,
+      "step": 702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8587701320648193,
+      "learning_rate": 9.686844175726661e-05,
+      "loss": 2.5206,
+      "step": 703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8800788521766663,
+      "learning_rate": 9.685720574514008e-05,
+      "loss": 2.5057,
+      "step": 704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7496289014816284,
+      "learning_rate": 9.684595026574288e-05,
+      "loss": 2.4448,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8469542860984802,
+      "learning_rate": 9.683467532375121e-05,
+      "loss": 2.4796,
+      "step": 706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8564528822898865,
+      "learning_rate": 9.682338092384935e-05,
+      "loss": 2.5557,
+      "step": 707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7172966003417969,
+      "learning_rate": 9.681206707072966e-05,
+      "loss": 2.39,
+      "step": 708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.844623863697052,
+      "learning_rate": 9.68007337690926e-05,
+      "loss": 2.4081,
+      "step": 709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9841399192810059,
+      "learning_rate": 9.678938102364667e-05,
+      "loss": 2.7453,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8438463807106018,
+      "learning_rate": 9.677800883910852e-05,
+      "loss": 2.4746,
+      "step": 711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8266069889068604,
+      "learning_rate": 9.67666172202028e-05,
+      "loss": 2.5717,
+      "step": 712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8954083919525146,
+      "learning_rate": 9.675520617166227e-05,
+      "loss": 2.5884,
+      "step": 713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8748486042022705,
+      "learning_rate": 9.674377569822778e-05,
+      "loss": 2.4401,
+      "step": 714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.793143093585968,
+      "learning_rate": 9.673232580464819e-05,
+      "loss": 2.6018,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8197920322418213,
+      "learning_rate": 9.67208564956805e-05,
+      "loss": 2.5166,
+      "step": 716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0339230298995972,
+      "learning_rate": 9.670936777608975e-05,
+      "loss": 2.5271,
+      "step": 717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8470713496208191,
+      "learning_rate": 9.669785965064899e-05,
+      "loss": 2.6493,
+      "step": 718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7828465104103088,
+      "learning_rate": 9.668633212413942e-05,
+      "loss": 2.437,
+      "step": 719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7941486835479736,
+      "learning_rate": 9.667478520135025e-05,
+      "loss": 2.4226,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7724357843399048,
+      "learning_rate": 9.666321888707874e-05,
+      "loss": 2.2573,
+      "step": 721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7300595045089722,
+      "learning_rate": 9.665163318613026e-05,
+      "loss": 2.4728,
+      "step": 722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8689148426055908,
+      "learning_rate": 9.66400281033182e-05,
+      "loss": 2.5707,
+      "step": 723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8709267377853394,
+      "learning_rate": 9.662840364346396e-05,
+      "loss": 2.7022,
+      "step": 724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9303498864173889,
+      "learning_rate": 9.661675981139707e-05,
+      "loss": 2.4126,
+      "step": 725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8274940848350525,
+      "learning_rate": 9.660509661195505e-05,
+      "loss": 2.5632,
+      "step": 726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8323959112167358,
+      "learning_rate": 9.659341404998348e-05,
+      "loss": 2.582,
+      "step": 727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9584940075874329,
+      "learning_rate": 9.658171213033603e-05,
+      "loss": 2.6593,
+      "step": 728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8095754384994507,
+      "learning_rate": 9.656999085787434e-05,
+      "loss": 2.5547,
+      "step": 729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8836039900779724,
+      "learning_rate": 9.655825023746812e-05,
+      "loss": 2.4262,
+      "step": 730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9213065505027771,
+      "learning_rate": 9.654649027399516e-05,
+      "loss": 2.6227,
+      "step": 731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9351211786270142,
+      "learning_rate": 9.653471097234119e-05,
+      "loss": 2.5129,
+      "step": 732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.86330246925354,
+      "learning_rate": 9.65229123374001e-05,
+      "loss": 2.4141,
+      "step": 733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7810982465744019,
+      "learning_rate": 9.651109437407369e-05,
+      "loss": 2.5093,
+      "step": 734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8625754117965698,
+      "learning_rate": 9.649925708727185e-05,
+      "loss": 2.6302,
+      "step": 735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9204151630401611,
+      "learning_rate": 9.648740048191254e-05,
+      "loss": 2.468,
+      "step": 736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8967533707618713,
+      "learning_rate": 9.647552456292165e-05,
+      "loss": 2.6723,
+      "step": 737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9060606956481934,
+      "learning_rate": 9.646362933523317e-05,
+      "loss": 2.5017,
+      "step": 738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9275190830230713,
+      "learning_rate": 9.645171480378908e-05,
+      "loss": 2.3794,
+      "step": 739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8264337182044983,
+      "learning_rate": 9.643978097353937e-05,
+      "loss": 2.2715,
+      "step": 740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9493167400360107,
+      "learning_rate": 9.642782784944206e-05,
+      "loss": 2.5221,
+      "step": 741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8093529939651489,
+      "learning_rate": 9.641585543646322e-05,
+      "loss": 2.4926,
+      "step": 742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.840020477771759,
+      "learning_rate": 9.640386373957688e-05,
+      "loss": 2.4638,
+      "step": 743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8975186944007874,
+      "learning_rate": 9.639185276376513e-05,
+      "loss": 2.6565,
+      "step": 744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8403806090354919,
+      "learning_rate": 9.6379822514018e-05,
+      "loss": 2.4144,
+      "step": 745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8126936554908752,
+      "learning_rate": 9.636777299533359e-05,
+      "loss": 2.6467,
+      "step": 746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9501340985298157,
+      "learning_rate": 9.635570421271803e-05,
+      "loss": 2.6686,
+      "step": 747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.850297749042511,
+      "learning_rate": 9.634361617118532e-05,
+      "loss": 2.475,
+      "step": 748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7798040509223938,
+      "learning_rate": 9.633150887575764e-05,
+      "loss": 2.4367,
+      "step": 749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9985675811767578,
+      "learning_rate": 9.631938233146504e-05,
+      "loss": 2.3772,
+      "step": 750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8391686081886292,
+      "learning_rate": 9.63072365433456e-05,
+      "loss": 2.5381,
+      "step": 751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8428494334220886,
+      "learning_rate": 9.629507151644542e-05,
+      "loss": 2.5493,
+      "step": 752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8951964974403381,
+      "learning_rate": 9.628288725581858e-05,
+      "loss": 2.7116,
+      "step": 753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8796242475509644,
+      "learning_rate": 9.627068376652712e-05,
+      "loss": 2.4386,
+      "step": 754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.810857892036438,
+      "learning_rate": 9.625846105364113e-05,
+      "loss": 2.5598,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8799995183944702,
+      "learning_rate": 9.624621912223864e-05,
+      "loss": 2.6602,
+      "step": 756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8351497054100037,
+      "learning_rate": 9.623395797740564e-05,
+      "loss": 2.5776,
+      "step": 757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8841829895973206,
+      "learning_rate": 9.62216776242362e-05,
+      "loss": 2.5041,
+      "step": 758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9105660915374756,
+      "learning_rate": 9.620937806783226e-05,
+      "loss": 2.6163,
+      "step": 759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7271343469619751,
+      "learning_rate": 9.619705931330382e-05,
+      "loss": 2.5229,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9027129411697388,
+      "learning_rate": 9.618472136576882e-05,
+      "loss": 2.6388,
+      "step": 761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8056872487068176,
+      "learning_rate": 9.617236423035314e-05,
+      "loss": 2.5266,
+      "step": 762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8042852878570557,
+      "learning_rate": 9.615998791219072e-05,
+      "loss": 2.3966,
+      "step": 763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7782443165779114,
+      "learning_rate": 9.61475924164234e-05,
+      "loss": 2.3041,
+      "step": 764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8327529430389404,
+      "learning_rate": 9.613517774820098e-05,
+      "loss": 2.4143,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9447781443595886,
+      "learning_rate": 9.612274391268128e-05,
+      "loss": 2.5807,
+      "step": 766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8174476027488708,
+      "learning_rate": 9.611029091503005e-05,
+      "loss": 2.516,
+      "step": 767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0648137331008911,
+      "learning_rate": 9.6097818760421e-05,
+      "loss": 2.7869,
+      "step": 768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8991522192955017,
+      "learning_rate": 9.608532745403584e-05,
+      "loss": 2.5163,
+      "step": 769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8338145613670349,
+      "learning_rate": 9.607281700106414e-05,
+      "loss": 2.2953,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9511364102363586,
+      "learning_rate": 9.606028740670354e-05,
+      "loss": 2.5325,
+      "step": 771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8175514340400696,
+      "learning_rate": 9.604773867615954e-05,
+      "loss": 2.6717,
+      "step": 772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0112950801849365,
+      "learning_rate": 9.603517081464564e-05,
+      "loss": 2.7084,
+      "step": 773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7411983013153076,
+      "learning_rate": 9.60225838273833e-05,
+      "loss": 2.3901,
+      "step": 774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8197214007377625,
+      "learning_rate": 9.600997771960188e-05,
+      "loss": 2.4592,
+      "step": 775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9533175826072693,
+      "learning_rate": 9.599735249653872e-05,
+      "loss": 2.4463,
+      "step": 776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9318834543228149,
+      "learning_rate": 9.598470816343907e-05,
+      "loss": 2.6103,
+      "step": 777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8868743777275085,
+      "learning_rate": 9.597204472555615e-05,
+      "loss": 2.4081,
+      "step": 778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7808035016059875,
+      "learning_rate": 9.595936218815111e-05,
+      "loss": 2.5651,
+      "step": 779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7980866432189941,
+      "learning_rate": 9.594666055649304e-05,
+      "loss": 2.4865,
+      "step": 780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8971570134162903,
+      "learning_rate": 9.593393983585892e-05,
+      "loss": 2.5587,
+      "step": 781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.792402446269989,
+      "learning_rate": 9.592120003153373e-05,
+      "loss": 2.4907,
+      "step": 782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7143792510032654,
+      "learning_rate": 9.590844114881032e-05,
+      "loss": 2.5947,
+      "step": 783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.872031033039093,
+      "learning_rate": 9.589566319298949e-05,
+      "loss": 2.6744,
+      "step": 784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8474131226539612,
+      "learning_rate": 9.588286616937996e-05,
+      "loss": 2.4586,
+      "step": 785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8205456733703613,
+      "learning_rate": 9.587005008329839e-05,
+      "loss": 2.6144,
+      "step": 786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7831835746765137,
+      "learning_rate": 9.585721494006935e-05,
+      "loss": 2.4825,
+      "step": 787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9346792101860046,
+      "learning_rate": 9.58443607450253e-05,
+      "loss": 2.6398,
+      "step": 788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7935715913772583,
+      "learning_rate": 9.583148750350665e-05,
+      "loss": 2.5888,
+      "step": 789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.830780565738678,
+      "learning_rate": 9.58185952208617e-05,
+      "loss": 2.3966,
+      "step": 790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9217642545700073,
+      "learning_rate": 9.580568390244669e-05,
+      "loss": 2.3978,
+      "step": 791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8393696546554565,
+      "learning_rate": 9.579275355362574e-05,
+      "loss": 2.4813,
+      "step": 792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9995031952857971,
+      "learning_rate": 9.577980417977086e-05,
+      "loss": 2.6132,
+      "step": 793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.892956018447876,
+      "learning_rate": 9.576683578626205e-05,
+      "loss": 2.3532,
+      "step": 794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8786612749099731,
+      "learning_rate": 9.57538483784871e-05,
+      "loss": 2.6084,
+      "step": 795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9275163412094116,
+      "learning_rate": 9.574084196184175e-05,
+      "loss": 2.5752,
+      "step": 796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8300831317901611,
+      "learning_rate": 9.572781654172967e-05,
+      "loss": 2.4248,
+      "step": 797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8611246943473816,
+      "learning_rate": 9.571477212356239e-05,
+      "loss": 2.5123,
+      "step": 798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9518934488296509,
+      "learning_rate": 9.570170871275932e-05,
+      "loss": 2.6475,
+      "step": 799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7589889764785767,
+      "learning_rate": 9.56886263147478e-05,
+      "loss": 2.4233,
+      "step": 800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9097497463226318,
+      "learning_rate": 9.5675524934963e-05,
+      "loss": 2.636,
+      "step": 801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8767960667610168,
+      "learning_rate": 9.566240457884805e-05,
+      "loss": 2.56,
+      "step": 802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8318452835083008,
+      "learning_rate": 9.564926525185392e-05,
+      "loss": 2.6699,
+      "step": 803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.980240523815155,
+      "learning_rate": 9.563610695943945e-05,
+      "loss": 2.6884,
+      "step": 804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8870283365249634,
+      "learning_rate": 9.562292970707138e-05,
+      "loss": 2.4095,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.893090546131134,
+      "learning_rate": 9.560973350022436e-05,
+      "loss": 2.605,
+      "step": 806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.110241413116455,
+      "learning_rate": 9.559651834438083e-05,
+      "loss": 2.702,
+      "step": 807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7626559138298035,
+      "learning_rate": 9.558328424503118e-05,
+      "loss": 2.51,
+      "step": 808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7921006083488464,
+      "learning_rate": 9.557003120767363e-05,
+      "loss": 2.3669,
+      "step": 809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8864078521728516,
+      "learning_rate": 9.555675923781429e-05,
+      "loss": 2.3243,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7603578567504883,
+      "learning_rate": 9.554346834096713e-05,
+      "loss": 2.4609,
+      "step": 811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8214945197105408,
+      "learning_rate": 9.553015852265395e-05,
+      "loss": 2.3925,
+      "step": 812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8361346125602722,
+      "learning_rate": 9.551682978840448e-05,
+      "loss": 2.6193,
+      "step": 813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8879817128181458,
+      "learning_rate": 9.550348214375625e-05,
+      "loss": 2.7049,
+      "step": 814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7941809892654419,
+      "learning_rate": 9.549011559425465e-05,
+      "loss": 2.393,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.034500002861023,
+      "learning_rate": 9.547673014545295e-05,
+      "loss": 2.7093,
+      "step": 816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8976350426673889,
+      "learning_rate": 9.546332580291228e-05,
+      "loss": 2.8013,
+      "step": 817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9602030515670776,
+      "learning_rate": 9.544990257220159e-05,
+      "loss": 2.7109,
+      "step": 818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8363465070724487,
+      "learning_rate": 9.54364604588977e-05,
+      "loss": 2.4781,
+      "step": 819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8406535387039185,
+      "learning_rate": 9.542299946858523e-05,
+      "loss": 2.686,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.988840639591217,
+      "learning_rate": 9.54095196068567e-05,
+      "loss": 2.6869,
+      "step": 821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9482634663581848,
+      "learning_rate": 9.539602087931246e-05,
+      "loss": 2.6711,
+      "step": 822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8609281778335571,
+      "learning_rate": 9.538250329156064e-05,
+      "loss": 2.421,
+      "step": 823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9972183704376221,
+      "learning_rate": 9.53689668492173e-05,
+      "loss": 2.4293,
+      "step": 824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6926944851875305,
+      "learning_rate": 9.535541155790626e-05,
+      "loss": 2.3612,
+      "step": 825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7985569834709167,
+      "learning_rate": 9.534183742325918e-05,
+      "loss": 2.4615,
+      "step": 826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.884579598903656,
+      "learning_rate": 9.532824445091558e-05,
+      "loss": 2.5927,
+      "step": 827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9123578667640686,
+      "learning_rate": 9.53146326465228e-05,
+      "loss": 2.5826,
+      "step": 828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.848058819770813,
+      "learning_rate": 9.530100201573597e-05,
+      "loss": 2.4503,
+      "step": 829
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6764326691627502,
+      "learning_rate": 9.528735256421809e-05,
+      "loss": 2.34,
+      "step": 830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0358911752700806,
+      "learning_rate": 9.527368429763994e-05,
+      "loss": 2.6063,
+      "step": 831
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8172800540924072,
+      "learning_rate": 9.525999722168011e-05,
+      "loss": 2.5198,
+      "step": 832
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7579197883605957,
+      "learning_rate": 9.524629134202507e-05,
+      "loss": 2.4502,
+      "step": 833
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7847051620483398,
+      "learning_rate": 9.523256666436902e-05,
+      "loss": 2.5417,
+      "step": 834
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8176968097686768,
+      "learning_rate": 9.521882319441402e-05,
+      "loss": 2.6626,
+      "step": 835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8821489214897156,
+      "learning_rate": 9.520506093786994e-05,
+      "loss": 2.6296,
+      "step": 836
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7884302139282227,
+      "learning_rate": 9.519127990045441e-05,
+      "loss": 2.4921,
+      "step": 837
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9690931439399719,
+      "learning_rate": 9.517748008789292e-05,
+      "loss": 2.8136,
+      "step": 838
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9310109615325928,
+      "learning_rate": 9.516366150591871e-05,
+      "loss": 2.5605,
+      "step": 839
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7614098191261292,
+      "learning_rate": 9.514982416027287e-05,
+      "loss": 2.4479,
+      "step": 840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8594020009040833,
+      "learning_rate": 9.513596805670425e-05,
+      "loss": 2.5495,
+      "step": 841
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7793421745300293,
+      "learning_rate": 9.512209320096946e-05,
+      "loss": 2.5705,
+      "step": 842
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8242400288581848,
+      "learning_rate": 9.5108199598833e-05,
+      "loss": 2.5479,
+      "step": 843
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8235220909118652,
+      "learning_rate": 9.509428725606704e-05,
+      "loss": 2.3669,
+      "step": 844
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8361103534698486,
+      "learning_rate": 9.508035617845163e-05,
+      "loss": 2.5808,
+      "step": 845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7369599342346191,
+      "learning_rate": 9.506640637177457e-05,
+      "loss": 2.3542,
+      "step": 846
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7609150409698486,
+      "learning_rate": 9.505243784183145e-05,
+      "loss": 2.2528,
+      "step": 847
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7985784411430359,
+      "learning_rate": 9.503845059442559e-05,
+      "loss": 2.6028,
+      "step": 848
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8320029377937317,
+      "learning_rate": 9.502444463536815e-05,
+      "loss": 2.4414,
+      "step": 849
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8459235429763794,
+      "learning_rate": 9.501041997047803e-05,
+      "loss": 2.3267,
+      "step": 850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8740043044090271,
+      "learning_rate": 9.499637660558191e-05,
+      "loss": 2.4507,
+      "step": 851
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8699366450309753,
+      "learning_rate": 9.498231454651425e-05,
+      "loss": 2.4738,
+      "step": 852
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9228084087371826,
+      "learning_rate": 9.496823379911726e-05,
+      "loss": 2.6663,
+      "step": 853
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8658406138420105,
+      "learning_rate": 9.49541343692409e-05,
+      "loss": 2.6661,
+      "step": 854
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8771939873695374,
+      "learning_rate": 9.494001626274295e-05,
+      "loss": 2.698,
+      "step": 855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8918392658233643,
+      "learning_rate": 9.492587948548887e-05,
+      "loss": 2.6501,
+      "step": 856
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8132167458534241,
+      "learning_rate": 9.491172404335197e-05,
+      "loss": 2.4009,
+      "step": 857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.842668890953064,
+      "learning_rate": 9.489754994221321e-05,
+      "loss": 2.5852,
+      "step": 858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8682898879051208,
+      "learning_rate": 9.488335718796138e-05,
+      "loss": 2.4827,
+      "step": 859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9704009294509888,
+      "learning_rate": 9.4869145786493e-05,
+      "loss": 2.7836,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.038689136505127,
+      "learning_rate": 9.485491574371233e-05,
+      "loss": 2.613,
+      "step": 861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7846161127090454,
+      "learning_rate": 9.484066706553136e-05,
+      "loss": 2.5457,
+      "step": 862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8110592365264893,
+      "learning_rate": 9.482639975786985e-05,
+      "loss": 2.5826,
+      "step": 863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7480689883232117,
+      "learning_rate": 9.481211382665528e-05,
+      "loss": 2.3785,
+      "step": 864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8808575868606567,
+      "learning_rate": 9.479780927782289e-05,
+      "loss": 2.7185,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8465996980667114,
+      "learning_rate": 9.478348611731562e-05,
+      "loss": 2.561,
+      "step": 866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.78768390417099,
+      "learning_rate": 9.47691443510842e-05,
+      "loss": 2.3462,
+      "step": 867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6276262402534485,
+      "learning_rate": 9.475478398508701e-05,
+      "loss": 2.4951,
+      "step": 868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8047231435775757,
+      "learning_rate": 9.474040502529022e-05,
+      "loss": 2.3482,
+      "step": 869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7850313782691956,
+      "learning_rate": 9.472600747766771e-05,
+      "loss": 2.4704,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7781289219856262,
+      "learning_rate": 9.471159134820107e-05,
+      "loss": 2.5146,
+      "step": 871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9193789958953857,
+      "learning_rate": 9.469715664287961e-05,
+      "loss": 2.4903,
+      "step": 872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9298500418663025,
+      "learning_rate": 9.468270336770037e-05,
+      "loss": 2.595,
+      "step": 873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8443459868431091,
+      "learning_rate": 9.466823152866813e-05,
+      "loss": 2.5182,
+      "step": 874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8444223403930664,
+      "learning_rate": 9.465374113179531e-05,
+      "loss": 2.6845,
+      "step": 875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9759942889213562,
+      "learning_rate": 9.463923218310212e-05,
+      "loss": 2.708,
+      "step": 876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9447507262229919,
+      "learning_rate": 9.46247046886164e-05,
+      "loss": 2.6129,
+      "step": 877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8106656074523926,
+      "learning_rate": 9.461015865437375e-05,
+      "loss": 2.6191,
+      "step": 878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9456419348716736,
+      "learning_rate": 9.45955940864175e-05,
+      "loss": 2.6838,
+      "step": 879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8303319811820984,
+      "learning_rate": 9.458101099079857e-05,
+      "loss": 2.4344,
+      "step": 880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.950612485408783,
+      "learning_rate": 9.456640937357568e-05,
+      "loss": 2.7616,
+      "step": 881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7732024788856506,
+      "learning_rate": 9.455178924081522e-05,
+      "loss": 2.4238,
+      "step": 882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8276551365852356,
+      "learning_rate": 9.453715059859125e-05,
+      "loss": 2.603,
+      "step": 883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8684731125831604,
+      "learning_rate": 9.452249345298555e-05,
+      "loss": 2.5918,
+      "step": 884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9662395119667053,
+      "learning_rate": 9.450781781008753e-05,
+      "loss": 2.6982,
+      "step": 885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8765259385108948,
+      "learning_rate": 9.449312367599435e-05,
+      "loss": 2.5693,
+      "step": 886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8437281250953674,
+      "learning_rate": 9.447841105681084e-05,
+      "loss": 2.6621,
+      "step": 887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.766871988773346,
+      "learning_rate": 9.446367995864946e-05,
+      "loss": 2.3755,
+      "step": 888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8586955666542053,
+      "learning_rate": 9.444893038763042e-05,
+      "loss": 2.3746,
+      "step": 889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7787945866584778,
+      "learning_rate": 9.443416234988156e-05,
+      "loss": 2.5312,
+      "step": 890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8924590945243835,
+      "learning_rate": 9.44193758515384e-05,
+      "loss": 2.597,
+      "step": 891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9893290996551514,
+      "learning_rate": 9.440457089874414e-05,
+      "loss": 2.5071,
+      "step": 892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8814725875854492,
+      "learning_rate": 9.438974749764961e-05,
+      "loss": 2.4408,
+      "step": 893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9844701290130615,
+      "learning_rate": 9.437490565441336e-05,
+      "loss": 2.5038,
+      "step": 894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9421939253807068,
+      "learning_rate": 9.436004537520157e-05,
+      "loss": 2.6053,
+      "step": 895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7952485680580139,
+      "learning_rate": 9.434516666618807e-05,
+      "loss": 2.727,
+      "step": 896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8446345329284668,
+      "learning_rate": 9.433026953355437e-05,
+      "loss": 2.4626,
+      "step": 897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8816962242126465,
+      "learning_rate": 9.431535398348962e-05,
+      "loss": 2.5186,
+      "step": 898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9463314414024353,
+      "learning_rate": 9.430042002219065e-05,
+      "loss": 2.7589,
+      "step": 899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9740270972251892,
+      "learning_rate": 9.428546765586188e-05,
+      "loss": 2.6169,
+      "step": 900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8869308829307556,
+      "learning_rate": 9.427049689071545e-05,
+      "loss": 2.4491,
+      "step": 901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7530148029327393,
+      "learning_rate": 9.425550773297108e-05,
+      "loss": 2.3414,
+      "step": 902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7882416248321533,
+      "learning_rate": 9.424050018885615e-05,
+      "loss": 2.55,
+      "step": 903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.830909252166748,
+      "learning_rate": 9.422547426460571e-05,
+      "loss": 2.4008,
+      "step": 904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.843914270401001,
+      "learning_rate": 9.421042996646241e-05,
+      "loss": 2.5754,
+      "step": 905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9130939841270447,
+      "learning_rate": 9.419536730067656e-05,
+      "loss": 2.4601,
+      "step": 906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8956150412559509,
+      "learning_rate": 9.418028627350606e-05,
+      "loss": 2.5112,
+      "step": 907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9193796515464783,
+      "learning_rate": 9.41651868912165e-05,
+      "loss": 2.6283,
+      "step": 908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8346453905105591,
+      "learning_rate": 9.415006916008105e-05,
+      "loss": 2.2356,
+      "step": 909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8787338137626648,
+      "learning_rate": 9.413493308638048e-05,
+      "loss": 2.6376,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8426021337509155,
+      "learning_rate": 9.411977867640328e-05,
+      "loss": 2.3543,
+      "step": 911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8989477753639221,
+      "learning_rate": 9.410460593644543e-05,
+      "loss": 2.5671,
+      "step": 912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7146804332733154,
+      "learning_rate": 9.408941487281064e-05,
+      "loss": 2.3545,
+      "step": 913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8000743389129639,
+      "learning_rate": 9.407420549181014e-05,
+      "loss": 2.416,
+      "step": 914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6681689620018005,
+      "learning_rate": 9.405897779976284e-05,
+      "loss": 2.2648,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9739978909492493,
+      "learning_rate": 9.404373180299523e-05,
+      "loss": 2.5621,
+      "step": 916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6980319619178772,
+      "learning_rate": 9.402846750784137e-05,
+      "loss": 2.4718,
+      "step": 917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0059524774551392,
+      "learning_rate": 9.4013184920643e-05,
+      "loss": 2.6842,
+      "step": 918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8616735339164734,
+      "learning_rate": 9.399788404774938e-05,
+      "loss": 2.4514,
+      "step": 919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8543574213981628,
+      "learning_rate": 9.398256489551744e-05,
+      "loss": 2.5394,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0317810773849487,
+      "learning_rate": 9.396722747031163e-05,
+      "loss": 2.6999,
+      "step": 921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.96778804063797,
+      "learning_rate": 9.395187177850404e-05,
+      "loss": 2.4032,
+      "step": 922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9467160105705261,
+      "learning_rate": 9.393649782647436e-05,
+      "loss": 2.6036,
+      "step": 923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9159897565841675,
+      "learning_rate": 9.39211056206098e-05,
+      "loss": 2.64,
+      "step": 924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9492524266242981,
+      "learning_rate": 9.390569516730525e-05,
+      "loss": 2.7583,
+      "step": 925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0030841827392578,
+      "learning_rate": 9.38902664729631e-05,
+      "loss": 2.4751,
+      "step": 926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8102176189422607,
+      "learning_rate": 9.387481954399334e-05,
+      "loss": 2.4437,
+      "step": 927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8322480916976929,
+      "learning_rate": 9.385935438681357e-05,
+      "loss": 2.6311,
+      "step": 928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8280768394470215,
+      "learning_rate": 9.384387100784891e-05,
+      "loss": 2.4861,
+      "step": 929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8847587704658508,
+      "learning_rate": 9.382836941353208e-05,
+      "loss": 2.6443,
+      "step": 930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.919711172580719,
+      "learning_rate": 9.381284961030338e-05,
+      "loss": 2.5238,
+      "step": 931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6651988625526428,
+      "learning_rate": 9.379731160461065e-05,
+      "loss": 2.2246,
+      "step": 932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8964563608169556,
+      "learning_rate": 9.37817554029093e-05,
+      "loss": 2.6937,
+      "step": 933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8492236733436584,
+      "learning_rate": 9.376618101166231e-05,
+      "loss": 2.5313,
+      "step": 934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7911650538444519,
+      "learning_rate": 9.37505884373402e-05,
+      "loss": 2.5269,
+      "step": 935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9214297533035278,
+      "learning_rate": 9.373497768642106e-05,
+      "loss": 2.6525,
+      "step": 936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8177318572998047,
+      "learning_rate": 9.37193487653905e-05,
+      "loss": 2.5084,
+      "step": 937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.007588267326355,
+      "learning_rate": 9.370370168074174e-05,
+      "loss": 2.466,
+      "step": 938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9075914621353149,
+      "learning_rate": 9.36880364389755e-05,
+      "loss": 2.5316,
+      "step": 939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9407746195793152,
+      "learning_rate": 9.367235304660005e-05,
+      "loss": 2.5966,
+      "step": 940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8306323885917664,
+      "learning_rate": 9.365665151013119e-05,
+      "loss": 2.4598,
+      "step": 941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9019886255264282,
+      "learning_rate": 9.364093183609229e-05,
+      "loss": 2.6141,
+      "step": 942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8965043425559998,
+      "learning_rate": 9.362519403101422e-05,
+      "loss": 2.6351,
+      "step": 943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9223279356956482,
+      "learning_rate": 9.360943810143542e-05,
+      "loss": 2.4849,
+      "step": 944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9433805346488953,
+      "learning_rate": 9.359366405390183e-05,
+      "loss": 2.6868,
+      "step": 945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7972301244735718,
+      "learning_rate": 9.357787189496694e-05,
+      "loss": 2.5145,
+      "step": 946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9764590859413147,
+      "learning_rate": 9.356206163119172e-05,
+      "loss": 2.5633,
+      "step": 947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8737182021141052,
+      "learning_rate": 9.354623326914474e-05,
+      "loss": 2.5008,
+      "step": 948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8596807718276978,
+      "learning_rate": 9.353038681540201e-05,
+      "loss": 2.4589,
+      "step": 949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.871933102607727,
+      "learning_rate": 9.351452227654708e-05,
+      "loss": 2.4158,
+      "step": 950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.821906566619873,
+      "learning_rate": 9.349863965917107e-05,
+      "loss": 2.7072,
+      "step": 951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9650446772575378,
+      "learning_rate": 9.348273896987253e-05,
+      "loss": 2.4062,
+      "step": 952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1025364398956299,
+      "learning_rate": 9.346682021525755e-05,
+      "loss": 2.5719,
+      "step": 953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8566573262214661,
+      "learning_rate": 9.345088340193974e-05,
+      "loss": 2.5346,
+      "step": 954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.929720938205719,
+      "learning_rate": 9.34349285365402e-05,
+      "loss": 2.3618,
+      "step": 955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8955583572387695,
+      "learning_rate": 9.341895562568752e-05,
+      "loss": 2.5537,
+      "step": 956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9748162627220154,
+      "learning_rate": 9.340296467601782e-05,
+      "loss": 2.4996,
+      "step": 957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9762101173400879,
+      "learning_rate": 9.338695569417467e-05,
+      "loss": 2.7555,
+      "step": 958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8479787707328796,
+      "learning_rate": 9.337092868680913e-05,
+      "loss": 2.4813,
+      "step": 959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8120766282081604,
+      "learning_rate": 9.335488366057982e-05,
+      "loss": 2.6262,
+      "step": 960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7780633568763733,
+      "learning_rate": 9.33388206221528e-05,
+      "loss": 2.5308,
+      "step": 961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9647269248962402,
+      "learning_rate": 9.332273957820156e-05,
+      "loss": 2.5904,
+      "step": 962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9312331080436707,
+      "learning_rate": 9.330664053540718e-05,
+      "loss": 2.5641,
+      "step": 963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8897035121917725,
+      "learning_rate": 9.329052350045812e-05,
+      "loss": 2.5313,
+      "step": 964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8122804164886475,
+      "learning_rate": 9.327438848005036e-05,
+      "loss": 2.4242,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9119512438774109,
+      "learning_rate": 9.325823548088737e-05,
+      "loss": 2.4559,
+      "step": 966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0928702354431152,
+      "learning_rate": 9.324206450968005e-05,
+      "loss": 2.6577,
+      "step": 967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9350084662437439,
+      "learning_rate": 9.322587557314678e-05,
+      "loss": 2.4781,
+      "step": 968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.014357328414917,
+      "learning_rate": 9.320966867801341e-05,
+      "loss": 2.4139,
+      "step": 969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9140161275863647,
+      "learning_rate": 9.319344383101327e-05,
+      "loss": 2.4852,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.024887204170227,
+      "learning_rate": 9.31772010388871e-05,
+      "loss": 2.7674,
+      "step": 971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9189098477363586,
+      "learning_rate": 9.316094030838313e-05,
+      "loss": 2.6592,
+      "step": 972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9161933660507202,
+      "learning_rate": 9.314466164625703e-05,
+      "loss": 2.435,
+      "step": 973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0341871976852417,
+      "learning_rate": 9.312836505927194e-05,
+      "loss": 2.5199,
+      "step": 974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8874825835227966,
+      "learning_rate": 9.31120505541984e-05,
+      "loss": 2.6712,
+      "step": 975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8985300064086914,
+      "learning_rate": 9.309571813781448e-05,
+      "loss": 2.3086,
+      "step": 976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8720566630363464,
+      "learning_rate": 9.307936781690558e-05,
+      "loss": 2.5829,
+      "step": 977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9364365339279175,
+      "learning_rate": 9.306299959826463e-05,
+      "loss": 2.6623,
+      "step": 978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9393642544746399,
+      "learning_rate": 9.304661348869195e-05,
+      "loss": 2.4638,
+      "step": 979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8153301477432251,
+      "learning_rate": 9.303020949499531e-05,
+      "loss": 2.2676,
+      "step": 980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9059273600578308,
+      "learning_rate": 9.30137876239899e-05,
+      "loss": 2.5445,
+      "step": 981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9256536960601807,
+      "learning_rate": 9.299734788249834e-05,
+      "loss": 2.4554,
+      "step": 982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9293302297592163,
+      "learning_rate": 9.298089027735068e-05,
+      "loss": 2.3912,
+      "step": 983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.881770670413971,
+      "learning_rate": 9.296441481538441e-05,
+      "loss": 2.5226,
+      "step": 984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9255176782608032,
+      "learning_rate": 9.294792150344438e-05,
+      "loss": 2.5611,
+      "step": 985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8549072742462158,
+      "learning_rate": 9.29314103483829e-05,
+      "loss": 2.4901,
+      "step": 986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8663883805274963,
+      "learning_rate": 9.291488135705971e-05,
+      "loss": 2.6376,
+      "step": 987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7615084052085876,
+      "learning_rate": 9.28983345363419e-05,
+      "loss": 2.3795,
+      "step": 988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9023071527481079,
+      "learning_rate": 9.288176989310404e-05,
+      "loss": 2.5392,
+      "step": 989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8277377486228943,
+      "learning_rate": 9.286518743422805e-05,
+      "loss": 2.346,
+      "step": 990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9475431442260742,
+      "learning_rate": 9.284858716660328e-05,
+      "loss": 2.6726,
+      "step": 991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8818802833557129,
+      "learning_rate": 9.283196909712646e-05,
+      "loss": 2.2485,
+      "step": 992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9061498641967773,
+      "learning_rate": 9.281533323270171e-05,
+      "loss": 2.3202,
+      "step": 993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9872395992279053,
+      "learning_rate": 9.27986795802406e-05,
+      "loss": 2.5707,
+      "step": 994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9309643507003784,
+      "learning_rate": 9.278200814666201e-05,
+      "loss": 2.4423,
+      "step": 995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9359556436538696,
+      "learning_rate": 9.276531893889225e-05,
+      "loss": 2.3096,
+      "step": 996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9526696801185608,
+      "learning_rate": 9.2748611963865e-05,
+      "loss": 2.5548,
+      "step": 997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.856442928314209,
+      "learning_rate": 9.273188722852138e-05,
+      "loss": 2.5596,
+      "step": 998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0660444498062134,
+      "learning_rate": 9.271514473980979e-05,
+      "loss": 2.6831,
+      "step": 999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9052556157112122,
+      "learning_rate": 9.269838450468608e-05,
+      "loss": 2.3991,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8943884968757629,
+      "learning_rate": 9.268160653011344e-05,
+      "loss": 2.4875,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9386797547340393,
+      "learning_rate": 9.266481082306244e-05,
+      "loss": 2.4033,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0029648542404175,
+      "learning_rate": 9.264799739051102e-05,
+      "loss": 2.5432,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9869605898857117,
+      "learning_rate": 9.263116623944448e-05,
+      "loss": 2.5666,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.792017936706543,
+      "learning_rate": 9.261431737685546e-05,
+      "loss": 2.2348,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8796958327293396,
+      "learning_rate": 9.259745080974402e-05,
+      "loss": 2.4769,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8809022307395935,
+      "learning_rate": 9.258056654511751e-05,
+      "loss": 2.4272,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7356773018836975,
+      "learning_rate": 9.256366458999065e-05,
+      "loss": 2.3428,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8494225144386292,
+      "learning_rate": 9.254674495138554e-05,
+      "loss": 2.3777,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8876265287399292,
+      "learning_rate": 9.252980763633159e-05,
+      "loss": 2.4677,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9220447540283203,
+      "learning_rate": 9.251285265186556e-05,
+      "loss": 2.6099,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8722644448280334,
+      "learning_rate": 9.249588000503159e-05,
+      "loss": 2.496,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8423951268196106,
+      "learning_rate": 9.24788897028811e-05,
+      "loss": 2.419,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7509445548057556,
+      "learning_rate": 9.246188175247288e-05,
+      "loss": 2.3515,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6867886781692505,
+      "learning_rate": 9.244485616087304e-05,
+      "loss": 2.3216,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8120637536048889,
+      "learning_rate": 9.242781293515503e-05,
+      "loss": 2.2259,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9875243306159973,
+      "learning_rate": 9.24107520823996e-05,
+      "loss": 2.7031,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8543577194213867,
+      "learning_rate": 9.239367360969489e-05,
+      "loss": 2.3726,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8599010109901428,
+      "learning_rate": 9.237657752413626e-05,
+      "loss": 2.4518,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8669360876083374,
+      "learning_rate": 9.23594638328265e-05,
+      "loss": 2.5562,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8919858336448669,
+      "learning_rate": 9.234233254287559e-05,
+      "loss": 2.7564,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9792090654373169,
+      "learning_rate": 9.232518366140093e-05,
+      "loss": 2.4883,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8052428960800171,
+      "learning_rate": 9.230801719552717e-05,
+      "loss": 2.2951,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8317286372184753,
+      "learning_rate": 9.229083315238629e-05,
+      "loss": 2.3103,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9874253869056702,
+      "learning_rate": 9.227363153911754e-05,
+      "loss": 2.6029,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7766485810279846,
+      "learning_rate": 9.225641236286753e-05,
+      "loss": 2.2478,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8950728178024292,
+      "learning_rate": 9.22391756307901e-05,
+      "loss": 2.5281,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8016728758811951,
+      "learning_rate": 9.222192135004644e-05,
+      "loss": 2.3754,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.814794659614563,
+      "learning_rate": 9.220464952780498e-05,
+      "loss": 2.3512,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8678061962127686,
+      "learning_rate": 9.218736017124148e-05,
+      "loss": 2.4466,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8539865612983704,
+      "learning_rate": 9.217005328753894e-05,
+      "loss": 2.3827,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8633853197097778,
+      "learning_rate": 9.21527288838877e-05,
+      "loss": 2.5785,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7764385342597961,
+      "learning_rate": 9.213538696748534e-05,
+      "loss": 2.4505,
+      "step": 1033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9363821744918823,
+      "learning_rate": 9.211802754553672e-05,
+      "loss": 2.4268,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9726277589797974,
+      "learning_rate": 9.210065062525398e-05,
+      "loss": 2.609,
+      "step": 1035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7196357846260071,
+      "learning_rate": 9.208325621385652e-05,
+      "loss": 2.3731,
+      "step": 1036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8246290683746338,
+      "learning_rate": 9.206584431857101e-05,
+      "loss": 2.4908,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9217603802680969,
+      "learning_rate": 9.20484149466314e-05,
+      "loss": 2.5659,
+      "step": 1038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8578044176101685,
+      "learning_rate": 9.203096810527888e-05,
+      "loss": 2.4808,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8699897527694702,
+      "learning_rate": 9.20135038017619e-05,
+      "loss": 2.3864,
+      "step": 1040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8162538409233093,
+      "learning_rate": 9.19960220433362e-05,
+      "loss": 2.4027,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8621248006820679,
+      "learning_rate": 9.19785228372647e-05,
+      "loss": 2.5931,
+      "step": 1042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8855344653129578,
+      "learning_rate": 9.196100619081764e-05,
+      "loss": 2.4308,
+      "step": 1043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8291260004043579,
+      "learning_rate": 9.194347211127246e-05,
+      "loss": 2.5386,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0021919012069702,
+      "learning_rate": 9.192592060591387e-05,
+      "loss": 2.5567,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9060608148574829,
+      "learning_rate": 9.190835168203382e-05,
+      "loss": 2.3288,
+      "step": 1046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9935576915740967,
+      "learning_rate": 9.189076534693146e-05,
+      "loss": 2.649,
+      "step": 1047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9028780460357666,
+      "learning_rate": 9.187316160791322e-05,
+      "loss": 2.5309,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.022945523262024,
+      "learning_rate": 9.185554047229273e-05,
+      "loss": 2.5774,
+      "step": 1049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9208734631538391,
+      "learning_rate": 9.183790194739086e-05,
+      "loss": 2.5721,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9340197443962097,
+      "learning_rate": 9.182024604053572e-05,
+      "loss": 2.509,
+      "step": 1051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9187864065170288,
+      "learning_rate": 9.180257275906259e-05,
+      "loss": 2.5169,
+      "step": 1052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8688451647758484,
+      "learning_rate": 9.178488211031402e-05,
+      "loss": 2.5453,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8868038654327393,
+      "learning_rate": 9.176717410163978e-05,
+      "loss": 2.3754,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7854194045066833,
+      "learning_rate": 9.174944874039678e-05,
+      "loss": 2.4251,
+      "step": 1055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8835120797157288,
+      "learning_rate": 9.173170603394923e-05,
+      "loss": 2.3694,
+      "step": 1056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8619648218154907,
+      "learning_rate": 9.17139459896685e-05,
+      "loss": 2.3382,
+      "step": 1057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9625738859176636,
+      "learning_rate": 9.169616861493316e-05,
+      "loss": 2.5523,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9585489630699158,
+      "learning_rate": 9.167837391712899e-05,
+      "loss": 2.7438,
+      "step": 1059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9888617992401123,
+      "learning_rate": 9.166056190364896e-05,
+      "loss": 2.5043,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9654977917671204,
+      "learning_rate": 9.164273258189325e-05,
+      "loss": 2.8235,
+      "step": 1061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8183463215827942,
+      "learning_rate": 9.162488595926923e-05,
+      "loss": 2.4726,
+      "step": 1062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9576370120048523,
+      "learning_rate": 9.160702204319143e-05,
+      "loss": 2.7495,
+      "step": 1063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8773769736289978,
+      "learning_rate": 9.158914084108159e-05,
+      "loss": 2.4684,
+      "step": 1064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9947913289070129,
+      "learning_rate": 9.157124236036864e-05,
+      "loss": 2.6337,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0025322437286377,
+      "learning_rate": 9.155332660848865e-05,
+      "loss": 2.6835,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8898522257804871,
+      "learning_rate": 9.153539359288487e-05,
+      "loss": 2.647,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9381152987480164,
+      "learning_rate": 9.151744332100782e-05,
+      "loss": 2.6797,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9054981470108032,
+      "learning_rate": 9.149947580031503e-05,
+      "loss": 2.4181,
+      "step": 1069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8213066458702087,
+      "learning_rate": 9.148149103827131e-05,
+      "loss": 2.3769,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9516487121582031,
+      "learning_rate": 9.146348904234861e-05,
+      "loss": 2.5512,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9731013774871826,
+      "learning_rate": 9.1445469820026e-05,
+      "loss": 2.6495,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9506499767303467,
+      "learning_rate": 9.142743337878977e-05,
+      "loss": 2.6963,
+      "step": 1073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.794478178024292,
+      "learning_rate": 9.14093797261333e-05,
+      "loss": 2.4706,
+      "step": 1074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9115039110183716,
+      "learning_rate": 9.139130886955718e-05,
+      "loss": 2.5146,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.965015709400177,
+      "learning_rate": 9.137322081656911e-05,
+      "loss": 2.5628,
+      "step": 1076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9145756363868713,
+      "learning_rate": 9.135511557468392e-05,
+      "loss": 2.2908,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0827358961105347,
+      "learning_rate": 9.133699315142363e-05,
+      "loss": 2.6058,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9285373091697693,
+      "learning_rate": 9.131885355431738e-05,
+      "loss": 2.5317,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8579372763633728,
+      "learning_rate": 9.130069679090141e-05,
+      "loss": 2.4054,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0522046089172363,
+      "learning_rate": 9.128252286871912e-05,
+      "loss": 2.6825,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9709950685501099,
+      "learning_rate": 9.126433179532108e-05,
+      "loss": 2.4924,
+      "step": 1082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.974951446056366,
+      "learning_rate": 9.12461235782649e-05,
+      "loss": 2.406,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9844278693199158,
+      "learning_rate": 9.122789822511537e-05,
+      "loss": 2.474,
+      "step": 1084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9462881088256836,
+      "learning_rate": 9.12096557434444e-05,
+      "loss": 2.4048,
+      "step": 1085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9415122866630554,
+      "learning_rate": 9.119139614083097e-05,
+      "loss": 2.5365,
+      "step": 1086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9685835838317871,
+      "learning_rate": 9.117311942486122e-05,
+      "loss": 2.6016,
+      "step": 1087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9535225033760071,
+      "learning_rate": 9.115482560312843e-05,
+      "loss": 2.3265,
+      "step": 1088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8494290709495544,
+      "learning_rate": 9.113651468323287e-05,
+      "loss": 2.2863,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9062087535858154,
+      "learning_rate": 9.111818667278203e-05,
+      "loss": 2.5857,
+      "step": 1090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.850395679473877,
+      "learning_rate": 9.109984157939042e-05,
+      "loss": 2.2044,
+      "step": 1091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9424256086349487,
+      "learning_rate": 9.108147941067972e-05,
+      "loss": 2.5653,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8327465653419495,
+      "learning_rate": 9.106310017427865e-05,
+      "loss": 2.4289,
+      "step": 1093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8099372386932373,
+      "learning_rate": 9.104470387782307e-05,
+      "loss": 2.3482,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8970416784286499,
+      "learning_rate": 9.102629052895584e-05,
+      "loss": 2.5226,
+      "step": 1095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7897661924362183,
+      "learning_rate": 9.1007860135327e-05,
+      "loss": 2.4414,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8243594765663147,
+      "learning_rate": 9.09894127045936e-05,
+      "loss": 2.4651,
+      "step": 1097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8980042934417725,
+      "learning_rate": 9.097094824441984e-05,
+      "loss": 2.4888,
+      "step": 1098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7792158722877502,
+      "learning_rate": 9.095246676247693e-05,
+      "loss": 2.3139,
+      "step": 1099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8563408851623535,
+      "learning_rate": 9.093396826644318e-05,
+      "loss": 2.5715,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8901385068893433,
+      "learning_rate": 9.091545276400397e-05,
+      "loss": 2.5369,
+      "step": 1101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8428673148155212,
+      "learning_rate": 9.089692026285173e-05,
+      "loss": 2.4167,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8971716165542603,
+      "learning_rate": 9.087837077068597e-05,
+      "loss": 2.5039,
+      "step": 1103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.886114239692688,
+      "learning_rate": 9.085980429521328e-05,
+      "loss": 2.4393,
+      "step": 1104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8074424862861633,
+      "learning_rate": 9.084122084414723e-05,
+      "loss": 2.5128,
+      "step": 1105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9695908427238464,
+      "learning_rate": 9.082262042520851e-05,
+      "loss": 2.5492,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8067036271095276,
+      "learning_rate": 9.080400304612487e-05,
+      "loss": 2.4789,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9244163036346436,
+      "learning_rate": 9.078536871463103e-05,
+      "loss": 2.5178,
+      "step": 1108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8792081475257874,
+      "learning_rate": 9.076671743846883e-05,
+      "loss": 2.4326,
+      "step": 1109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8852723836898804,
+      "learning_rate": 9.07480492253871e-05,
+      "loss": 2.2175,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8586215972900391,
+      "learning_rate": 9.072936408314176e-05,
+      "loss": 2.4153,
+      "step": 1111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9502882957458496,
+      "learning_rate": 9.071066201949569e-05,
+      "loss": 2.3679,
+      "step": 1112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8658397793769836,
+      "learning_rate": 9.069194304221886e-05,
+      "loss": 2.5857,
+      "step": 1113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8643784523010254,
+      "learning_rate": 9.067320715908824e-05,
+      "loss": 2.4697,
+      "step": 1114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8049342036247253,
+      "learning_rate": 9.065445437788783e-05,
+      "loss": 2.483,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9818577170372009,
+      "learning_rate": 9.063568470640867e-05,
+      "loss": 2.3772,
+      "step": 1116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0456651449203491,
+      "learning_rate": 9.061689815244876e-05,
+      "loss": 2.4417,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8706830739974976,
+      "learning_rate": 9.05980947238132e-05,
+      "loss": 2.6157,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0073730945587158,
+      "learning_rate": 9.0579274428314e-05,
+      "loss": 2.6269,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0647164583206177,
+      "learning_rate": 9.056043727377028e-05,
+      "loss": 2.6279,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8685073256492615,
+      "learning_rate": 9.054158326800808e-05,
+      "loss": 2.3641,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8698757886886597,
+      "learning_rate": 9.05227124188605e-05,
+      "loss": 2.5636,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9103327989578247,
+      "learning_rate": 9.050382473416758e-05,
+      "loss": 2.5255,
+      "step": 1123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.970798909664154,
+      "learning_rate": 9.048492022177643e-05,
+      "loss": 2.5965,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9833136200904846,
+      "learning_rate": 9.046599888954108e-05,
+      "loss": 2.4772,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.990537166595459,
+      "learning_rate": 9.044706074532259e-05,
+      "loss": 2.4164,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9494500160217285,
+      "learning_rate": 9.042810579698898e-05,
+      "loss": 2.6508,
+      "step": 1127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8141872882843018,
+      "learning_rate": 9.040913405241527e-05,
+      "loss": 2.4638,
+      "step": 1128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9225784540176392,
+      "learning_rate": 9.039014551948346e-05,
+      "loss": 2.3552,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9780427813529968,
+      "learning_rate": 9.037114020608251e-05,
+      "loss": 2.5607,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.030666470527649,
+      "learning_rate": 9.035211812010833e-05,
+      "loss": 2.4212,
+      "step": 1131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9744720458984375,
+      "learning_rate": 9.033307926946387e-05,
+      "loss": 2.4857,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.141334056854248,
+      "learning_rate": 9.031402366205898e-05,
+      "loss": 2.4911,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8100097179412842,
+      "learning_rate": 9.029495130581049e-05,
+      "loss": 2.4863,
+      "step": 1134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8299288153648376,
+      "learning_rate": 9.02758622086422e-05,
+      "loss": 2.3395,
+      "step": 1135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0275336503982544,
+      "learning_rate": 9.025675637848485e-05,
+      "loss": 2.4244,
+      "step": 1136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8657293915748596,
+      "learning_rate": 9.023763382327614e-05,
+      "loss": 2.4512,
+      "step": 1137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9719887375831604,
+      "learning_rate": 9.021849455096073e-05,
+      "loss": 2.6969,
+      "step": 1138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7690579891204834,
+      "learning_rate": 9.019933856949019e-05,
+      "loss": 2.4655,
+      "step": 1139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8043075799942017,
+      "learning_rate": 9.018016588682307e-05,
+      "loss": 2.3019,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9188481569290161,
+      "learning_rate": 9.016097651092483e-05,
+      "loss": 2.6083,
+      "step": 1141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.811289370059967,
+      "learning_rate": 9.01417704497679e-05,
+      "loss": 2.4611,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8761153817176819,
+      "learning_rate": 9.01225477113316e-05,
+      "loss": 2.3257,
+      "step": 1143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9558963179588318,
+      "learning_rate": 9.010330830360221e-05,
+      "loss": 2.6686,
+      "step": 1144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9471624493598938,
+      "learning_rate": 9.00840522345729e-05,
+      "loss": 2.4635,
+      "step": 1145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8171020150184631,
+      "learning_rate": 9.006477951224381e-05,
+      "loss": 2.4567,
+      "step": 1146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7581790089607239,
+      "learning_rate": 9.004549014462199e-05,
+      "loss": 2.5151,
+      "step": 1147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0500810146331787,
+      "learning_rate": 9.002618413972137e-05,
+      "loss": 2.8132,
+      "step": 1148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8393862843513489,
+      "learning_rate": 9.00068615055628e-05,
+      "loss": 2.3723,
+      "step": 1149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9680187702178955,
+      "learning_rate": 8.998752225017409e-05,
+      "loss": 2.5539,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8167526721954346,
+      "learning_rate": 8.996816638158986e-05,
+      "loss": 2.5251,
+      "step": 1151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9344148635864258,
+      "learning_rate": 8.994879390785176e-05,
+      "loss": 2.4844,
+      "step": 1152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0253651142120361,
+      "learning_rate": 8.992940483700821e-05,
+      "loss": 2.6635,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8864468932151794,
+      "learning_rate": 8.99099991771146e-05,
+      "loss": 2.3302,
+      "step": 1154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9903461933135986,
+      "learning_rate": 8.989057693623321e-05,
+      "loss": 2.6768,
+      "step": 1155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7636241316795349,
+      "learning_rate": 8.987113812243319e-05,
+      "loss": 2.3806,
+      "step": 1156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0741918087005615,
+      "learning_rate": 8.985168274379057e-05,
+      "loss": 2.4979,
+      "step": 1157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9200177788734436,
+      "learning_rate": 8.983221080838827e-05,
+      "loss": 2.6339,
+      "step": 1158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8899011611938477,
+      "learning_rate": 8.98127223243161e-05,
+      "loss": 2.612,
+      "step": 1159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9580606818199158,
+      "learning_rate": 8.979321729967074e-05,
+      "loss": 2.4566,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9352087378501892,
+      "learning_rate": 8.977369574255571e-05,
+      "loss": 2.3214,
+      "step": 1161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.931174635887146,
+      "learning_rate": 8.975415766108146e-05,
+      "loss": 2.6523,
+      "step": 1162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8986103534698486,
+      "learning_rate": 8.973460306336526e-05,
+      "loss": 2.5525,
+      "step": 1163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8644623756408691,
+      "learning_rate": 8.971503195753124e-05,
+      "loss": 2.3252,
+      "step": 1164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8920726180076599,
+      "learning_rate": 8.969544435171042e-05,
+      "loss": 2.4068,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9629347324371338,
+      "learning_rate": 8.967584025404063e-05,
+      "loss": 2.4767,
+      "step": 1166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9677354097366333,
+      "learning_rate": 8.965621967266662e-05,
+      "loss": 2.5798,
+      "step": 1167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8486101627349854,
+      "learning_rate": 8.963658261573991e-05,
+      "loss": 2.3634,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7743979692459106,
+      "learning_rate": 8.961692909141892e-05,
+      "loss": 2.3581,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8173879384994507,
+      "learning_rate": 8.959725910786888e-05,
+      "loss": 2.391,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.860262393951416,
+      "learning_rate": 8.957757267326188e-05,
+      "loss": 2.4532,
+      "step": 1171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9127695560455322,
+      "learning_rate": 8.955786979577685e-05,
+      "loss": 2.449,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8726258873939514,
+      "learning_rate": 8.953815048359951e-05,
+      "loss": 2.3197,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0369006395339966,
+      "learning_rate": 8.951841474492244e-05,
+      "loss": 2.4852,
+      "step": 1174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8455232977867126,
+      "learning_rate": 8.949866258794506e-05,
+      "loss": 2.3956,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.934421181678772,
+      "learning_rate": 8.947889402087358e-05,
+      "loss": 2.3669,
+      "step": 1176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9072303771972656,
+      "learning_rate": 8.945910905192103e-05,
+      "loss": 2.529,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8356785774230957,
+      "learning_rate": 8.943930768930728e-05,
+      "loss": 2.5501,
+      "step": 1178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.019317865371704,
+      "learning_rate": 8.941948994125897e-05,
+      "loss": 2.467,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7343520522117615,
+      "learning_rate": 8.93996558160096e-05,
+      "loss": 2.2714,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0315858125686646,
+      "learning_rate": 8.937980532179943e-05,
+      "loss": 2.4345,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8153426051139832,
+      "learning_rate": 8.935993846687552e-05,
+      "loss": 2.3441,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9352467060089111,
+      "learning_rate": 8.934005525949179e-05,
+      "loss": 2.4735,
+      "step": 1183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9458906650543213,
+      "learning_rate": 8.932015570790887e-05,
+      "loss": 2.4563,
+      "step": 1184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8895012140274048,
+      "learning_rate": 8.93002398203942e-05,
+      "loss": 2.2666,
+      "step": 1185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9538923501968384,
+      "learning_rate": 8.928030760522207e-05,
+      "loss": 2.513,
+      "step": 1186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9325218796730042,
+      "learning_rate": 8.92603590706735e-05,
+      "loss": 2.5391,
+      "step": 1187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8284532427787781,
+      "learning_rate": 8.924039422503626e-05,
+      "loss": 2.176,
+      "step": 1188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0173803567886353,
+      "learning_rate": 8.922041307660497e-05,
+      "loss": 2.4491,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0019792318344116,
+      "learning_rate": 8.920041563368098e-05,
+      "loss": 2.6786,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9906332492828369,
+      "learning_rate": 8.918040190457242e-05,
+      "loss": 2.5025,
+      "step": 1191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8936331272125244,
+      "learning_rate": 8.916037189759416e-05,
+      "loss": 2.3254,
+      "step": 1192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8801282644271851,
+      "learning_rate": 8.91403256210679e-05,
+      "loss": 2.41,
+      "step": 1193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9023463726043701,
+      "learning_rate": 8.912026308332201e-05,
+      "loss": 2.4279,
+      "step": 1194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.012753963470459,
+      "learning_rate": 8.91001842926917e-05,
+      "loss": 2.4381,
+      "step": 1195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9439823627471924,
+      "learning_rate": 8.908008925751888e-05,
+      "loss": 2.4221,
+      "step": 1196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8599189519882202,
+      "learning_rate": 8.90599779861522e-05,
+      "loss": 2.3373,
+      "step": 1197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9668784141540527,
+      "learning_rate": 8.903985048694712e-05,
+      "loss": 2.5529,
+      "step": 1198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9312662482261658,
+      "learning_rate": 8.901970676826576e-05,
+      "loss": 2.4649,
+      "step": 1199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8912866115570068,
+      "learning_rate": 8.899954683847704e-05,
+      "loss": 2.3248,
+      "step": 1200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.962126612663269,
+      "learning_rate": 8.89793707059566e-05,
+      "loss": 2.4365,
+      "step": 1201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9224323630332947,
+      "learning_rate": 8.89591783790868e-05,
+      "loss": 2.1344,
+      "step": 1202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9377404451370239,
+      "learning_rate": 8.893896986625673e-05,
+      "loss": 2.5669,
+      "step": 1203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9715995788574219,
+      "learning_rate": 8.891874517586221e-05,
+      "loss": 2.694,
+      "step": 1204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0126399993896484,
+      "learning_rate": 8.889850431630577e-05,
+      "loss": 2.4282,
+      "step": 1205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0099446773529053,
+      "learning_rate": 8.887824729599668e-05,
+      "loss": 2.6477,
+      "step": 1206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9335098266601562,
+      "learning_rate": 8.88579741233509e-05,
+      "loss": 2.4883,
+      "step": 1207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9435465931892395,
+      "learning_rate": 8.883768480679114e-05,
+      "loss": 2.3982,
+      "step": 1208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8381285667419434,
+      "learning_rate": 8.881737935474677e-05,
+      "loss": 2.5224,
+      "step": 1209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9300739169120789,
+      "learning_rate": 8.879705777565386e-05,
+      "loss": 2.311,
+      "step": 1210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.785260796546936,
+      "learning_rate": 8.877672007795524e-05,
+      "loss": 2.2148,
+      "step": 1211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8577236533164978,
+      "learning_rate": 8.875636627010037e-05,
+      "loss": 2.4244,
+      "step": 1212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8655155897140503,
+      "learning_rate": 8.873599636054544e-05,
+      "loss": 2.5623,
+      "step": 1213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8206347823143005,
+      "learning_rate": 8.871561035775336e-05,
+      "loss": 2.3771,
+      "step": 1214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7978485822677612,
+      "learning_rate": 8.869520827019364e-05,
+      "loss": 2.3987,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8523750305175781,
+      "learning_rate": 8.867479010634253e-05,
+      "loss": 2.6638,
+      "step": 1216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7599415183067322,
+      "learning_rate": 8.865435587468296e-05,
+      "loss": 2.3783,
+      "step": 1217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9095607995986938,
+      "learning_rate": 8.863390558370452e-05,
+      "loss": 2.4568,
+      "step": 1218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8383930325508118,
+      "learning_rate": 8.86134392419035e-05,
+      "loss": 2.4133,
+      "step": 1219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9924961924552917,
+      "learning_rate": 8.85929568577828e-05,
+      "loss": 2.4735,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8585272431373596,
+      "learning_rate": 8.857245843985204e-05,
+      "loss": 2.593,
+      "step": 1221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.97022944688797,
+      "learning_rate": 8.855194399662748e-05,
+      "loss": 2.7754,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7763843536376953,
+      "learning_rate": 8.853141353663203e-05,
+      "loss": 2.4352,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9691436290740967,
+      "learning_rate": 8.851086706839529e-05,
+      "loss": 2.4792,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8545634150505066,
+      "learning_rate": 8.849030460045345e-05,
+      "loss": 2.4046,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8784313201904297,
+      "learning_rate": 8.846972614134942e-05,
+      "loss": 2.3699,
+      "step": 1226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9085052013397217,
+      "learning_rate": 8.844913169963269e-05,
+      "loss": 2.5335,
+      "step": 1227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7833656072616577,
+      "learning_rate": 8.842852128385941e-05,
+      "loss": 2.169,
+      "step": 1228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8538340926170349,
+      "learning_rate": 8.84078949025924e-05,
+      "loss": 2.5127,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9951622486114502,
+      "learning_rate": 8.838725256440108e-05,
+      "loss": 2.6328,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9019765853881836,
+      "learning_rate": 8.836659427786149e-05,
+      "loss": 2.2056,
+      "step": 1231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9898159503936768,
+      "learning_rate": 8.834592005155631e-05,
+      "loss": 2.3833,
+      "step": 1232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0198637247085571,
+      "learning_rate": 8.832522989407485e-05,
+      "loss": 2.4335,
+      "step": 1233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9030969738960266,
+      "learning_rate": 8.830452381401304e-05,
+      "loss": 2.352,
+      "step": 1234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8404765725135803,
+      "learning_rate": 8.828380181997341e-05,
+      "loss": 2.4401,
+      "step": 1235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8854008316993713,
+      "learning_rate": 8.826306392056511e-05,
+      "loss": 2.3014,
+      "step": 1236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8738580942153931,
+      "learning_rate": 8.82423101244039e-05,
+      "loss": 2.3402,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9823501706123352,
+      "learning_rate": 8.822154044011214e-05,
+      "loss": 2.4474,
+      "step": 1238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8651440143585205,
+      "learning_rate": 8.82007548763188e-05,
+      "loss": 2.328,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.934158980846405,
+      "learning_rate": 8.817995344165942e-05,
+      "loss": 2.3368,
+      "step": 1240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1003262996673584,
+      "learning_rate": 8.815913614477615e-05,
+      "loss": 2.6825,
+      "step": 1241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8706275224685669,
+      "learning_rate": 8.813830299431778e-05,
+      "loss": 2.2062,
+      "step": 1242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8888944983482361,
+      "learning_rate": 8.811745399893958e-05,
+      "loss": 2.3942,
+      "step": 1243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8310203552246094,
+      "learning_rate": 8.80965891673035e-05,
+      "loss": 2.2394,
+      "step": 1244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8618929386138916,
+      "learning_rate": 8.807570850807803e-05,
+      "loss": 2.38,
+      "step": 1245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9461740255355835,
+      "learning_rate": 8.805481202993821e-05,
+      "loss": 2.4237,
+      "step": 1246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7501404285430908,
+      "learning_rate": 8.803389974156573e-05,
+      "loss": 2.4216,
+      "step": 1247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9186726808547974,
+      "learning_rate": 8.801297165164876e-05,
+      "loss": 2.3686,
+      "step": 1248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9284569025039673,
+      "learning_rate": 8.799202776888206e-05,
+      "loss": 2.4972,
+      "step": 1249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9179290533065796,
+      "learning_rate": 8.797106810196702e-05,
+      "loss": 2.3126,
+      "step": 1250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9057959318161011,
+      "learning_rate": 8.79500926596115e-05,
+      "loss": 2.2471,
+      "step": 1251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9998093247413635,
+      "learning_rate": 8.792910145052992e-05,
+      "loss": 2.2297,
+      "step": 1252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0257833003997803,
+      "learning_rate": 8.790809448344333e-05,
+      "loss": 2.4944,
+      "step": 1253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.936707615852356,
+      "learning_rate": 8.788707176707923e-05,
+      "loss": 2.5696,
+      "step": 1254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9692200422286987,
+      "learning_rate": 8.786603331017172e-05,
+      "loss": 2.484,
+      "step": 1255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9758712649345398,
+      "learning_rate": 8.784497912146145e-05,
+      "loss": 2.343,
+      "step": 1256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9709192514419556,
+      "learning_rate": 8.782390920969554e-05,
+      "loss": 2.5705,
+      "step": 1257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9986327290534973,
+      "learning_rate": 8.780282358362774e-05,
+      "loss": 2.4892,
+      "step": 1258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9084057211875916,
+      "learning_rate": 8.778172225201822e-05,
+      "loss": 2.6553,
+      "step": 1259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8762896656990051,
+      "learning_rate": 8.776060522363377e-05,
+      "loss": 2.5571,
+      "step": 1260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.006292700767517,
+      "learning_rate": 8.773947250724761e-05,
+      "loss": 2.5154,
+      "step": 1261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0106823444366455,
+      "learning_rate": 8.771832411163959e-05,
+      "loss": 2.6049,
+      "step": 1262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9188042879104614,
+      "learning_rate": 8.769716004559594e-05,
+      "loss": 2.544,
+      "step": 1263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9266941547393799,
+      "learning_rate": 8.767598031790953e-05,
+      "loss": 2.3695,
+      "step": 1264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8934957981109619,
+      "learning_rate": 8.765478493737967e-05,
+      "loss": 2.3516,
+      "step": 1265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.94706791639328,
+      "learning_rate": 8.763357391281218e-05,
+      "loss": 2.5135,
+      "step": 1266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9178247451782227,
+      "learning_rate": 8.761234725301935e-05,
+      "loss": 2.3471,
+      "step": 1267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9558742642402649,
+      "learning_rate": 8.759110496682002e-05,
+      "loss": 2.46,
+      "step": 1268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8093961477279663,
+      "learning_rate": 8.75698470630395e-05,
+      "loss": 2.3877,
+      "step": 1269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9310894012451172,
+      "learning_rate": 8.75485735505096e-05,
+      "loss": 2.2904,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8870509266853333,
+      "learning_rate": 8.752728443806859e-05,
+      "loss": 2.4087,
+      "step": 1271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9511570930480957,
+      "learning_rate": 8.750597973456123e-05,
+      "loss": 2.4415,
+      "step": 1272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8211982846260071,
+      "learning_rate": 8.748465944883877e-05,
+      "loss": 2.4292,
+      "step": 1273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.877618670463562,
+      "learning_rate": 8.746332358975891e-05,
+      "loss": 2.3004,
+      "step": 1274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0565955638885498,
+      "learning_rate": 8.744197216618584e-05,
+      "loss": 2.5232,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0309162139892578,
+      "learning_rate": 8.742060518699023e-05,
+      "loss": 2.5272,
+      "step": 1276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8899449110031128,
+      "learning_rate": 8.739922266104918e-05,
+      "loss": 2.4276,
+      "step": 1277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8462796211242676,
+      "learning_rate": 8.737782459724627e-05,
+      "loss": 2.4194,
+      "step": 1278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1118632555007935,
+      "learning_rate": 8.73564110044715e-05,
+      "loss": 2.5851,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7825729846954346,
+      "learning_rate": 8.733498189162138e-05,
+      "loss": 2.274,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8847973942756653,
+      "learning_rate": 8.731353726759883e-05,
+      "loss": 2.4869,
+      "step": 1281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8786745071411133,
+      "learning_rate": 8.729207714131324e-05,
+      "loss": 2.4299,
+      "step": 1282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9233691096305847,
+      "learning_rate": 8.72706015216804e-05,
+      "loss": 2.437,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0109699964523315,
+      "learning_rate": 8.724911041762256e-05,
+      "loss": 2.4546,
+      "step": 1284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8807622790336609,
+      "learning_rate": 8.722760383806841e-05,
+      "loss": 2.1743,
+      "step": 1285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.026498556137085,
+      "learning_rate": 8.720608179195307e-05,
+      "loss": 2.4339,
+      "step": 1286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8491703867912292,
+      "learning_rate": 8.718454428821806e-05,
+      "loss": 2.2832,
+      "step": 1287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9114482402801514,
+      "learning_rate": 8.716299133581136e-05,
+      "loss": 2.471,
+      "step": 1288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1746423244476318,
+      "learning_rate": 8.714142294368734e-05,
+      "loss": 2.6351,
+      "step": 1289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8985783457756042,
+      "learning_rate": 8.711983912080679e-05,
+      "loss": 2.4947,
+      "step": 1290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0163264274597168,
+      "learning_rate": 8.709823987613691e-05,
+      "loss": 2.4195,
+      "step": 1291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.937718391418457,
+      "learning_rate": 8.707662521865132e-05,
+      "loss": 2.5993,
+      "step": 1292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9024103879928589,
+      "learning_rate": 8.705499515733005e-05,
+      "loss": 2.5558,
+      "step": 1293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8637648224830627,
+      "learning_rate": 8.703334970115947e-05,
+      "loss": 2.4271,
+      "step": 1294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8713198900222778,
+      "learning_rate": 8.701168885913243e-05,
+      "loss": 2.3647,
+      "step": 1295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8685842156410217,
+      "learning_rate": 8.699001264024811e-05,
+      "loss": 2.4023,
+      "step": 1296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9419881701469421,
+      "learning_rate": 8.696832105351212e-05,
+      "loss": 2.4557,
+      "step": 1297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.986113429069519,
+      "learning_rate": 8.694661410793641e-05,
+      "loss": 2.6096,
+      "step": 1298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8952358961105347,
+      "learning_rate": 8.692489181253937e-05,
+      "loss": 2.5447,
+      "step": 1299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.836007297039032,
+      "learning_rate": 8.69031541763457e-05,
+      "loss": 2.4536,
+      "step": 1300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7221465706825256,
+      "learning_rate": 8.688140120838652e-05,
+      "loss": 2.3228,
+      "step": 1301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9126811027526855,
+      "learning_rate": 8.685963291769933e-05,
+      "loss": 2.5393,
+      "step": 1302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9124241471290588,
+      "learning_rate": 8.683784931332794e-05,
+      "loss": 2.6192,
+      "step": 1303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.923842191696167,
+      "learning_rate": 8.681605040432257e-05,
+      "loss": 2.5196,
+      "step": 1304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7877792119979858,
+      "learning_rate": 8.679423619973976e-05,
+      "loss": 2.444,
+      "step": 1305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1193318367004395,
+      "learning_rate": 8.677240670864249e-05,
+      "loss": 2.4068,
+      "step": 1306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.954078197479248,
+      "learning_rate": 8.675056194009996e-05,
+      "loss": 2.5363,
+      "step": 1307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8487145900726318,
+      "learning_rate": 8.672870190318781e-05,
+      "loss": 2.3671,
+      "step": 1308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8956751823425293,
+      "learning_rate": 8.670682660698801e-05,
+      "loss": 2.3762,
+      "step": 1309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8769878149032593,
+      "learning_rate": 8.668493606058886e-05,
+      "loss": 2.3119,
+      "step": 1310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9458174109458923,
+      "learning_rate": 8.666303027308499e-05,
+      "loss": 2.3648,
+      "step": 1311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0918606519699097,
+      "learning_rate": 8.664110925357736e-05,
+      "loss": 2.5894,
+      "step": 1312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9428215622901917,
+      "learning_rate": 8.661917301117328e-05,
+      "loss": 2.4353,
+      "step": 1313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8792629837989807,
+      "learning_rate": 8.659722155498634e-05,
+      "loss": 2.3754,
+      "step": 1314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9429607391357422,
+      "learning_rate": 8.657525489413651e-05,
+      "loss": 2.6701,
+      "step": 1315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7704954147338867,
+      "learning_rate": 8.655327303775007e-05,
+      "loss": 2.2996,
+      "step": 1316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.844926118850708,
+      "learning_rate": 8.653127599495951e-05,
+      "loss": 2.4638,
+      "step": 1317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9198400974273682,
+      "learning_rate": 8.650926377490381e-05,
+      "loss": 2.4675,
+      "step": 1318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8717504739761353,
+      "learning_rate": 8.648723638672806e-05,
+      "loss": 2.4731,
+      "step": 1319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8117406964302063,
+      "learning_rate": 8.646519383958382e-05,
+      "loss": 2.2429,
+      "step": 1320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9988822937011719,
+      "learning_rate": 8.644313614262887e-05,
+      "loss": 2.6359,
+      "step": 1321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8359012603759766,
+      "learning_rate": 8.642106330502724e-05,
+      "loss": 2.5576,
+      "step": 1322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.018871784210205,
+      "learning_rate": 8.639897533594934e-05,
+      "loss": 2.5495,
+      "step": 1323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8558893203735352,
+      "learning_rate": 8.63768722445718e-05,
+      "loss": 2.3541,
+      "step": 1324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0427305698394775,
+      "learning_rate": 8.635475404007758e-05,
+      "loss": 2.4013,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8323906064033508,
+      "learning_rate": 8.63326207316559e-05,
+      "loss": 2.3773,
+      "step": 1326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8837379217147827,
+      "learning_rate": 8.631047232850223e-05,
+      "loss": 2.4897,
+      "step": 1327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9127394556999207,
+      "learning_rate": 8.628830883981835e-05,
+      "loss": 2.4759,
+      "step": 1328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9428324699401855,
+      "learning_rate": 8.62661302748123e-05,
+      "loss": 2.4638,
+      "step": 1329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9634432196617126,
+      "learning_rate": 8.624393664269836e-05,
+      "loss": 2.6957,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8720947504043579,
+      "learning_rate": 8.62217279526971e-05,
+      "loss": 2.3717,
+      "step": 1331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8520815372467041,
+      "learning_rate": 8.619950421403532e-05,
+      "loss": 2.383,
+      "step": 1332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.048311471939087,
+      "learning_rate": 8.617726543594609e-05,
+      "loss": 2.5579,
+      "step": 1333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8008101582527161,
+      "learning_rate": 8.615501162766871e-05,
+      "loss": 2.3024,
+      "step": 1334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8426388502120972,
+      "learning_rate": 8.613274279844873e-05,
+      "loss": 2.2161,
+      "step": 1335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9289656281471252,
+      "learning_rate": 8.611045895753796e-05,
+      "loss": 2.2594,
+      "step": 1336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9178404808044434,
+      "learning_rate": 8.608816011419445e-05,
+      "loss": 2.2611,
+      "step": 1337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1265740394592285,
+      "learning_rate": 8.606584627768242e-05,
+      "loss": 2.5153,
+      "step": 1338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0409668684005737,
+      "learning_rate": 8.604351745727241e-05,
+      "loss": 2.5967,
+      "step": 1339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8417184352874756,
+      "learning_rate": 8.602117366224113e-05,
+      "loss": 2.3716,
+      "step": 1340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9766790270805359,
+      "learning_rate": 8.59988149018715e-05,
+      "loss": 2.5131,
+      "step": 1341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8678521513938904,
+      "learning_rate": 8.597644118545269e-05,
+      "loss": 2.4937,
+      "step": 1342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8407182693481445,
+      "learning_rate": 8.59540525222801e-05,
+      "loss": 2.4775,
+      "step": 1343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.926662027835846,
+      "learning_rate": 8.593164892165526e-05,
+      "loss": 2.2536,
+      "step": 1344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9972390532493591,
+      "learning_rate": 8.590923039288601e-05,
+      "loss": 2.403,
+      "step": 1345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.048362135887146,
+      "learning_rate": 8.58867969452863e-05,
+      "loss": 2.5234,
+      "step": 1346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9322709441184998,
+      "learning_rate": 8.586434858817635e-05,
+      "loss": 2.403,
+      "step": 1347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8918463587760925,
+      "learning_rate": 8.584188533088253e-05,
+      "loss": 2.3859,
+      "step": 1348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9062890410423279,
+      "learning_rate": 8.58194071827374e-05,
+      "loss": 2.446,
+      "step": 1349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.119522213935852,
+      "learning_rate": 8.579691415307973e-05,
+      "loss": 2.5026,
+      "step": 1350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0720216035842896,
+      "learning_rate": 8.577440625125447e-05,
+      "loss": 2.4907,
+      "step": 1351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9925363659858704,
+      "learning_rate": 8.575188348661275e-05,
+      "loss": 2.4733,
+      "step": 1352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9293316006660461,
+      "learning_rate": 8.572934586851184e-05,
+      "loss": 2.3468,
+      "step": 1353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9493741393089294,
+      "learning_rate": 8.570679340631522e-05,
+      "loss": 2.4145,
+      "step": 1354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8539941906929016,
+      "learning_rate": 8.568422610939252e-05,
+      "loss": 2.4664,
+      "step": 1355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8588550090789795,
+      "learning_rate": 8.566164398711955e-05,
+      "loss": 2.456,
+      "step": 1356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9095320701599121,
+      "learning_rate": 8.563904704887826e-05,
+      "loss": 2.2273,
+      "step": 1357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8710384368896484,
+      "learning_rate": 8.561643530405676e-05,
+      "loss": 2.343,
+      "step": 1358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8165090680122375,
+      "learning_rate": 8.559380876204932e-05,
+      "loss": 2.4341,
+      "step": 1359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7897449135780334,
+      "learning_rate": 8.557116743225633e-05,
+      "loss": 2.4482,
+      "step": 1360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9430137276649475,
+      "learning_rate": 8.554851132408438e-05,
+      "loss": 2.2562,
+      "step": 1361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0089588165283203,
+      "learning_rate": 8.552584044694613e-05,
+      "loss": 2.5534,
+      "step": 1362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8530250787734985,
+      "learning_rate": 8.550315481026043e-05,
+      "loss": 2.2048,
+      "step": 1363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0357913970947266,
+      "learning_rate": 8.548045442345225e-05,
+      "loss": 2.4439,
+      "step": 1364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9317297339439392,
+      "learning_rate": 8.545773929595268e-05,
+      "loss": 2.5152,
+      "step": 1365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9390052556991577,
+      "learning_rate": 8.54350094371989e-05,
+      "loss": 2.3238,
+      "step": 1366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.985398530960083,
+      "learning_rate": 8.541226485663429e-05,
+      "loss": 2.4807,
+      "step": 1367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8479779362678528,
+      "learning_rate": 8.53895055637083e-05,
+      "loss": 2.4679,
+      "step": 1368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8749542832374573,
+      "learning_rate": 8.536673156787645e-05,
+      "loss": 2.4432,
+      "step": 1369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7867797613143921,
+      "learning_rate": 8.534394287860046e-05,
+      "loss": 2.3606,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7601848840713501,
+      "learning_rate": 8.532113950534807e-05,
+      "loss": 2.2212,
+      "step": 1371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8882381319999695,
+      "learning_rate": 8.529832145759319e-05,
+      "loss": 2.4097,
+      "step": 1372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9234427213668823,
+      "learning_rate": 8.527548874481578e-05,
+      "loss": 2.5212,
+      "step": 1373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9820677042007446,
+      "learning_rate": 8.52526413765019e-05,
+      "loss": 2.5844,
+      "step": 1374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.838736891746521,
+      "learning_rate": 8.522977936214373e-05,
+      "loss": 2.3947,
+      "step": 1375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7949293851852417,
+      "learning_rate": 8.520690271123948e-05,
+      "loss": 2.1995,
+      "step": 1376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8908659815788269,
+      "learning_rate": 8.518401143329348e-05,
+      "loss": 2.4008,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.046911358833313,
+      "learning_rate": 8.516110553781615e-05,
+      "loss": 2.6181,
+      "step": 1378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.784465491771698,
+      "learning_rate": 8.513818503432393e-05,
+      "loss": 2.2847,
+      "step": 1379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8926481604576111,
+      "learning_rate": 8.511524993233939e-05,
+      "loss": 2.363,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9636290073394775,
+      "learning_rate": 8.509230024139113e-05,
+      "loss": 2.5346,
+      "step": 1381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0433545112609863,
+      "learning_rate": 8.50693359710138e-05,
+      "loss": 2.6249,
+      "step": 1382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9527367949485779,
+      "learning_rate": 8.504635713074814e-05,
+      "loss": 2.5234,
+      "step": 1383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9840210676193237,
+      "learning_rate": 8.50233637301409e-05,
+      "loss": 2.338,
+      "step": 1384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9387891292572021,
+      "learning_rate": 8.500035577874494e-05,
+      "loss": 2.4895,
+      "step": 1385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9152513146400452,
+      "learning_rate": 8.497733328611912e-05,
+      "loss": 2.661,
+      "step": 1386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9593742489814758,
+      "learning_rate": 8.495429626182834e-05,
+      "loss": 2.4732,
+      "step": 1387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8461337685585022,
+      "learning_rate": 8.493124471544358e-05,
+      "loss": 2.3296,
+      "step": 1388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9233540892601013,
+      "learning_rate": 8.49081786565418e-05,
+      "loss": 2.4693,
+      "step": 1389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8849382996559143,
+      "learning_rate": 8.488509809470602e-05,
+      "loss": 2.3098,
+      "step": 1390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9804632663726807,
+      "learning_rate": 8.486200303952527e-05,
+      "loss": 2.3908,
+      "step": 1391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9048060774803162,
+      "learning_rate": 8.483889350059461e-05,
+      "loss": 2.2079,
+      "step": 1392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0337049961090088,
+      "learning_rate": 8.481576948751512e-05,
+      "loss": 2.5141,
+      "step": 1393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8692469000816345,
+      "learning_rate": 8.479263100989392e-05,
+      "loss": 2.2939,
+      "step": 1394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9407290816307068,
+      "learning_rate": 8.476947807734407e-05,
+      "loss": 2.4221,
+      "step": 1395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9571899771690369,
+      "learning_rate": 8.47463106994847e-05,
+      "loss": 2.274,
+      "step": 1396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0597962141036987,
+      "learning_rate": 8.47231288859409e-05,
+      "loss": 2.4992,
+      "step": 1397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9570518732070923,
+      "learning_rate": 8.469993264634378e-05,
+      "loss": 2.5023,
+      "step": 1398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9299846887588501,
+      "learning_rate": 8.467672199033044e-05,
+      "loss": 2.5294,
+      "step": 1399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.926623523235321,
+      "learning_rate": 8.465349692754396e-05,
+      "loss": 2.4238,
+      "step": 1400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.981641948223114,
+      "learning_rate": 8.463025746763342e-05,
+      "loss": 2.3257,
+      "step": 1401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9025770425796509,
+      "learning_rate": 8.460700362025389e-05,
+      "loss": 2.4027,
+      "step": 1402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9181666374206543,
+      "learning_rate": 8.458373539506638e-05,
+      "loss": 2.3799,
+      "step": 1403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9037690758705139,
+      "learning_rate": 8.45604528017379e-05,
+      "loss": 2.4595,
+      "step": 1404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0120702981948853,
+      "learning_rate": 8.453715584994141e-05,
+      "loss": 2.2945,
+      "step": 1405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0788570642471313,
+      "learning_rate": 8.451384454935587e-05,
+      "loss": 2.5852,
+      "step": 1406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9411907196044922,
+      "learning_rate": 8.449051890966617e-05,
+      "loss": 2.3421,
+      "step": 1407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9842630624771118,
+      "learning_rate": 8.446717894056318e-05,
+      "loss": 2.4269,
+      "step": 1408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9808139204978943,
+      "learning_rate": 8.444382465174369e-05,
+      "loss": 2.3937,
+      "step": 1409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8482744693756104,
+      "learning_rate": 8.442045605291049e-05,
+      "loss": 2.3403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0291088819503784,
+      "learning_rate": 8.439707315377228e-05,
+      "loss": 2.5207,
+      "step": 1411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7998470067977905,
+      "learning_rate": 8.437367596404369e-05,
+      "loss": 2.3441,
+      "step": 1412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9823687076568604,
+      "learning_rate": 8.43502644934453e-05,
+      "loss": 2.5467,
+      "step": 1413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.963265061378479,
+      "learning_rate": 8.432683875170365e-05,
+      "loss": 2.5613,
+      "step": 1414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9774287939071655,
+      "learning_rate": 8.430339874855118e-05,
+      "loss": 2.5247,
+      "step": 1415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9005679488182068,
+      "learning_rate": 8.427994449372625e-05,
+      "loss": 2.4375,
+      "step": 1416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9155099391937256,
+      "learning_rate": 8.425647599697317e-05,
+      "loss": 2.6025,
+      "step": 1417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8719702959060669,
+      "learning_rate": 8.423299326804215e-05,
+      "loss": 2.3498,
+      "step": 1418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8664777278900146,
+      "learning_rate": 8.420949631668929e-05,
+      "loss": 2.367,
+      "step": 1419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0079071521759033,
+      "learning_rate": 8.418598515267663e-05,
+      "loss": 2.616,
+      "step": 1420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7984552979469299,
+      "learning_rate": 8.416245978577213e-05,
+      "loss": 2.1078,
+      "step": 1421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8963137269020081,
+      "learning_rate": 8.413892022574959e-05,
+      "loss": 2.4931,
+      "step": 1422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8914421200752258,
+      "learning_rate": 8.411536648238878e-05,
+      "loss": 2.3895,
+      "step": 1423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8427234292030334,
+      "learning_rate": 8.40917985654753e-05,
+      "loss": 2.3548,
+      "step": 1424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7873695492744446,
+      "learning_rate": 8.406821648480066e-05,
+      "loss": 2.15,
+      "step": 1425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8814392685890198,
+      "learning_rate": 8.404462025016229e-05,
+      "loss": 2.4973,
+      "step": 1426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.791801929473877,
+      "learning_rate": 8.402100987136344e-05,
+      "loss": 2.0897,
+      "step": 1427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8160238265991211,
+      "learning_rate": 8.399738535821329e-05,
+      "loss": 2.2563,
+      "step": 1428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8476583361625671,
+      "learning_rate": 8.397374672052683e-05,
+      "loss": 2.3409,
+      "step": 1429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0012695789337158,
+      "learning_rate": 8.395009396812501e-05,
+      "loss": 2.426,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0525578260421753,
+      "learning_rate": 8.392642711083454e-05,
+      "loss": 2.3694,
+      "step": 1431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.912517249584198,
+      "learning_rate": 8.390274615848808e-05,
+      "loss": 2.5228,
+      "step": 1432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8228232264518738,
+      "learning_rate": 8.387905112092406e-05,
+      "loss": 2.2044,
+      "step": 1433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9813492894172668,
+      "learning_rate": 8.385534200798686e-05,
+      "loss": 2.3165,
+      "step": 1434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0138626098632812,
+      "learning_rate": 8.383161882952663e-05,
+      "loss": 2.3502,
+      "step": 1435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.023398518562317,
+      "learning_rate": 8.380788159539939e-05,
+      "loss": 2.4472,
+      "step": 1436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9842982888221741,
+      "learning_rate": 8.378413031546702e-05,
+      "loss": 2.4613,
+      "step": 1437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0570728778839111,
+      "learning_rate": 8.376036499959717e-05,
+      "loss": 2.624,
+      "step": 1438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9541673064231873,
+      "learning_rate": 8.37365856576634e-05,
+      "loss": 2.3958,
+      "step": 1439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9733685851097107,
+      "learning_rate": 8.371279229954505e-05,
+      "loss": 2.6015,
+      "step": 1440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.88368821144104,
+      "learning_rate": 8.368898493512732e-05,
+      "loss": 2.3729,
+      "step": 1441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9597542881965637,
+      "learning_rate": 8.366516357430116e-05,
+      "loss": 2.2637,
+      "step": 1442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0069377422332764,
+      "learning_rate": 8.364132822696343e-05,
+      "loss": 2.5042,
+      "step": 1443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2083885669708252,
+      "learning_rate": 8.36174789030167e-05,
+      "loss": 2.6947,
+      "step": 1444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0174486637115479,
+      "learning_rate": 8.359361561236944e-05,
+      "loss": 2.3809,
+      "step": 1445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8031017184257507,
+      "learning_rate": 8.35697383649359e-05,
+      "loss": 2.481,
+      "step": 1446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8628156185150146,
+      "learning_rate": 8.354584717063602e-05,
+      "loss": 2.4611,
+      "step": 1447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8374050259590149,
+      "learning_rate": 8.352194203939571e-05,
+      "loss": 2.4647,
+      "step": 1448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9824017286300659,
+      "learning_rate": 8.349802298114655e-05,
+      "loss": 2.1456,
+      "step": 1449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7881754636764526,
+      "learning_rate": 8.347409000582595e-05,
+      "loss": 2.3322,
+      "step": 1450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9925867915153503,
+      "learning_rate": 8.345014312337706e-05,
+      "loss": 2.5439,
+      "step": 1451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.070968508720398,
+      "learning_rate": 8.342618234374889e-05,
+      "loss": 2.5754,
+      "step": 1452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9470415115356445,
+      "learning_rate": 8.340220767689613e-05,
+      "loss": 2.4319,
+      "step": 1453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9974439144134521,
+      "learning_rate": 8.33782191327793e-05,
+      "loss": 2.3889,
+      "step": 1454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0775346755981445,
+      "learning_rate": 8.335421672136468e-05,
+      "loss": 2.3978,
+      "step": 1455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1436883211135864,
+      "learning_rate": 8.333020045262427e-05,
+      "loss": 2.576,
+      "step": 1456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8908794522285461,
+      "learning_rate": 8.33061703365359e-05,
+      "loss": 2.4527,
+      "step": 1457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8172414302825928,
+      "learning_rate": 8.328212638308306e-05,
+      "loss": 2.2842,
+      "step": 1458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9501521587371826,
+      "learning_rate": 8.325806860225507e-05,
+      "loss": 2.2384,
+      "step": 1459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9726687669754028,
+      "learning_rate": 8.323399700404694e-05,
+      "loss": 2.4263,
+      "step": 1460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9996610879898071,
+      "learning_rate": 8.320991159845947e-05,
+      "loss": 2.4601,
+      "step": 1461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9289079308509827,
+      "learning_rate": 8.318581239549918e-05,
+      "loss": 2.3415,
+      "step": 1462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8809939622879028,
+      "learning_rate": 8.316169940517826e-05,
+      "loss": 2.471,
+      "step": 1463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.031815528869629,
+      "learning_rate": 8.313757263751473e-05,
+      "loss": 2.4692,
+      "step": 1464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9613248705863953,
+      "learning_rate": 8.311343210253226e-05,
+      "loss": 2.4178,
+      "step": 1465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8248059153556824,
+      "learning_rate": 8.308927781026029e-05,
+      "loss": 2.3307,
+      "step": 1466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.949759304523468,
+      "learning_rate": 8.306510977073392e-05,
+      "loss": 2.4047,
+      "step": 1467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0060029029846191,
+      "learning_rate": 8.304092799399399e-05,
+      "loss": 2.4158,
+      "step": 1468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.868643581867218,
+      "learning_rate": 8.301673249008708e-05,
+      "loss": 2.2753,
+      "step": 1469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8650954961776733,
+      "learning_rate": 8.299252326906541e-05,
+      "loss": 2.5876,
+      "step": 1470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8953088521957397,
+      "learning_rate": 8.296830034098695e-05,
+      "loss": 2.3993,
+      "step": 1471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0168226957321167,
+      "learning_rate": 8.294406371591533e-05,
+      "loss": 2.4307,
+      "step": 1472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8477437496185303,
+      "learning_rate": 8.29198134039199e-05,
+      "loss": 2.2247,
+      "step": 1473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9346543550491333,
+      "learning_rate": 8.289554941507566e-05,
+      "loss": 2.5248,
+      "step": 1474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.053155779838562,
+      "learning_rate": 8.287127175946336e-05,
+      "loss": 2.5135,
+      "step": 1475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8327992558479309,
+      "learning_rate": 8.284698044716935e-05,
+      "loss": 2.3653,
+      "step": 1476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9269668459892273,
+      "learning_rate": 8.282267548828568e-05,
+      "loss": 2.3869,
+      "step": 1477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.897818386554718,
+      "learning_rate": 8.279835689291008e-05,
+      "loss": 2.3615,
+      "step": 1478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9141201972961426,
+      "learning_rate": 8.277402467114595e-05,
+      "loss": 2.3508,
+      "step": 1479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8783124685287476,
+      "learning_rate": 8.274967883310234e-05,
+      "loss": 2.3497,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8458665609359741,
+      "learning_rate": 8.272531938889396e-05,
+      "loss": 2.5229,
+      "step": 1481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8869600892066956,
+      "learning_rate": 8.270094634864118e-05,
+      "loss": 2.3903,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8718274235725403,
+      "learning_rate": 8.267655972246998e-05,
+      "loss": 2.3601,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8808475136756897,
+      "learning_rate": 8.265215952051206e-05,
+      "loss": 2.44,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7231338620185852,
+      "learning_rate": 8.262774575290467e-05,
+      "loss": 2.2595,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0035008192062378,
+      "learning_rate": 8.260331842979076e-05,
+      "loss": 2.3557,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.91319739818573,
+      "learning_rate": 8.257887756131892e-05,
+      "loss": 2.4275,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9774386286735535,
+      "learning_rate": 8.255442315764331e-05,
+      "loss": 2.6651,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0150998830795288,
+      "learning_rate": 8.252995522892375e-05,
+      "loss": 2.3775,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8351843953132629,
+      "learning_rate": 8.250547378532567e-05,
+      "loss": 2.5,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8958282470703125,
+      "learning_rate": 8.248097883702011e-05,
+      "loss": 2.2757,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.876357913017273,
+      "learning_rate": 8.245647039418375e-05,
+      "loss": 2.3521,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8717342019081116,
+      "learning_rate": 8.243194846699887e-05,
+      "loss": 2.3348,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9814518094062805,
+      "learning_rate": 8.24074130656533e-05,
+      "loss": 2.5646,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9229702949523926,
+      "learning_rate": 8.238286420034049e-05,
+      "loss": 2.5632,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9203149080276489,
+      "learning_rate": 8.235830188125955e-05,
+      "loss": 2.3702,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8121488094329834,
+      "learning_rate": 8.233372611861512e-05,
+      "loss": 2.1517,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9292315244674683,
+      "learning_rate": 8.23091369226174e-05,
+      "loss": 2.4424,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0272107124328613,
+      "learning_rate": 8.228453430348222e-05,
+      "loss": 2.5152,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8490802049636841,
+      "learning_rate": 8.225991827143098e-05,
+      "loss": 2.3058,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9373388886451721,
+      "learning_rate": 8.223528883669065e-05,
+      "loss": 2.4686,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0922669172286987,
+      "learning_rate": 8.221064600949374e-05,
+      "loss": 2.5701,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.007336974143982,
+      "learning_rate": 8.218598980007837e-05,
+      "loss": 2.3931,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9471147060394287,
+      "learning_rate": 8.216132021868817e-05,
+      "loss": 2.2377,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7934107780456543,
+      "learning_rate": 8.213663727557237e-05,
+      "loss": 2.2511,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0819346904754639,
+      "learning_rate": 8.211194098098574e-05,
+      "loss": 2.371,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9092562198638916,
+      "learning_rate": 8.208723134518859e-05,
+      "loss": 2.3666,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8435171842575073,
+      "learning_rate": 8.206250837844674e-05,
+      "loss": 2.3675,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8781112432479858,
+      "learning_rate": 8.203777209103161e-05,
+      "loss": 2.2931,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0469880104064941,
+      "learning_rate": 8.201302249322011e-05,
+      "loss": 2.5547,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8628454804420471,
+      "learning_rate": 8.198825959529473e-05,
+      "loss": 2.2135,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9164783358573914,
+      "learning_rate": 8.196348340754338e-05,
+      "loss": 2.2829,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9868070483207703,
+      "learning_rate": 8.193869394025964e-05,
+      "loss": 2.3783,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9827185869216919,
+      "learning_rate": 8.191389120374249e-05,
+      "loss": 2.5015,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9773440361022949,
+      "learning_rate": 8.188907520829647e-05,
+      "loss": 2.6463,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8450796604156494,
+      "learning_rate": 8.186424596423161e-05,
+      "loss": 2.3453,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8871644139289856,
+      "learning_rate": 8.183940348186347e-05,
+      "loss": 2.323,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7731435298919678,
+      "learning_rate": 8.181454777151307e-05,
+      "loss": 2.3729,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0794804096221924,
+      "learning_rate": 8.178967884350699e-05,
+      "loss": 2.585,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.859815776348114,
+      "learning_rate": 8.176479670817725e-05,
+      "loss": 2.4392,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8893265128135681,
+      "learning_rate": 8.173990137586133e-05,
+      "loss": 2.4489,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8557251691818237,
+      "learning_rate": 8.17149928569023e-05,
+      "loss": 2.2361,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9627293944358826,
+      "learning_rate": 8.169007116164857e-05,
+      "loss": 2.3853,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9357194304466248,
+      "learning_rate": 8.166513630045414e-05,
+      "loss": 2.438,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9882711172103882,
+      "learning_rate": 8.164018828367843e-05,
+      "loss": 2.4298,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.822153627872467,
+      "learning_rate": 8.161522712168632e-05,
+      "loss": 2.1808,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9008060097694397,
+      "learning_rate": 8.159025282484816e-05,
+      "loss": 2.2287,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.027705430984497,
+      "learning_rate": 8.156526540353976e-05,
+      "loss": 2.4711,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9967756867408752,
+      "learning_rate": 8.154026486814239e-05,
+      "loss": 2.3868,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.050546646118164,
+      "learning_rate": 8.151525122904276e-05,
+      "loss": 2.4032,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9265164136886597,
+      "learning_rate": 8.1490224496633e-05,
+      "loss": 2.5126,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9791779518127441,
+      "learning_rate": 8.146518468131075e-05,
+      "loss": 2.5232,
+      "step": 1532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9770310521125793,
+      "learning_rate": 8.144013179347898e-05,
+      "loss": 2.5373,
+      "step": 1533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9917840361595154,
+      "learning_rate": 8.141506584354618e-05,
+      "loss": 2.5462,
+      "step": 1534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.063856601715088,
+      "learning_rate": 8.138998684192625e-05,
+      "loss": 2.5666,
+      "step": 1535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0337369441986084,
+      "learning_rate": 8.136489479903848e-05,
+      "loss": 2.4472,
+      "step": 1536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9763365983963013,
+      "learning_rate": 8.133978972530761e-05,
+      "loss": 2.4625,
+      "step": 1537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0077166557312012,
+      "learning_rate": 8.131467163116377e-05,
+      "loss": 2.3769,
+      "step": 1538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.026965856552124,
+      "learning_rate": 8.128954052704252e-05,
+      "loss": 2.4918,
+      "step": 1539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9736360311508179,
+      "learning_rate": 8.126439642338479e-05,
+      "loss": 2.2129,
+      "step": 1540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9614072442054749,
+      "learning_rate": 8.123923933063697e-05,
+      "loss": 2.3022,
+      "step": 1541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.958809494972229,
+      "learning_rate": 8.121406925925081e-05,
+      "loss": 2.5857,
+      "step": 1542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.074947476387024,
+      "learning_rate": 8.118888621968342e-05,
+      "loss": 2.5047,
+      "step": 1543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8737003207206726,
+      "learning_rate": 8.116369022239738e-05,
+      "loss": 2.5888,
+      "step": 1544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8335123062133789,
+      "learning_rate": 8.113848127786056e-05,
+      "loss": 2.2014,
+      "step": 1545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0281398296356201,
+      "learning_rate": 8.111325939654631e-05,
+      "loss": 2.6757,
+      "step": 1546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0703166723251343,
+      "learning_rate": 8.108802458893323e-05,
+      "loss": 2.4089,
+      "step": 1547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9576767086982727,
+      "learning_rate": 8.106277686550543e-05,
+      "loss": 2.2467,
+      "step": 1548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1322723627090454,
+      "learning_rate": 8.103751623675227e-05,
+      "loss": 2.5782,
+      "step": 1549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9306889176368713,
+      "learning_rate": 8.101224271316852e-05,
+      "loss": 2.5857,
+      "step": 1550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.055438756942749,
+      "learning_rate": 8.098695630525433e-05,
+      "loss": 2.4077,
+      "step": 1551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0312957763671875,
+      "learning_rate": 8.096165702351518e-05,
+      "loss": 2.3839,
+      "step": 1552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9521524906158447,
+      "learning_rate": 8.093634487846186e-05,
+      "loss": 2.4441,
+      "step": 1553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9686403274536133,
+      "learning_rate": 8.091101988061058e-05,
+      "loss": 2.3736,
+      "step": 1554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8817377090454102,
+      "learning_rate": 8.088568204048283e-05,
+      "loss": 2.2058,
+      "step": 1555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9367691874504089,
+      "learning_rate": 8.086033136860544e-05,
+      "loss": 2.339,
+      "step": 1556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8272611498832703,
+      "learning_rate": 8.083496787551063e-05,
+      "loss": 2.2718,
+      "step": 1557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9063159823417664,
+      "learning_rate": 8.080959157173587e-05,
+      "loss": 2.4131,
+      "step": 1558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8549443483352661,
+      "learning_rate": 8.0784202467824e-05,
+      "loss": 2.4492,
+      "step": 1559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0062401294708252,
+      "learning_rate": 8.075880057432317e-05,
+      "loss": 2.3982,
+      "step": 1560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1214648485183716,
+      "learning_rate": 8.073338590178682e-05,
+      "loss": 2.6155,
+      "step": 1561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9181610345840454,
+      "learning_rate": 8.070795846077373e-05,
+      "loss": 2.3173,
+      "step": 1562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.077675461769104,
+      "learning_rate": 8.068251826184794e-05,
+      "loss": 2.5408,
+      "step": 1563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9182681441307068,
+      "learning_rate": 8.065706531557888e-05,
+      "loss": 2.0756,
+      "step": 1564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8897507786750793,
+      "learning_rate": 8.063159963254118e-05,
+      "loss": 2.3082,
+      "step": 1565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.038323998451233,
+      "learning_rate": 8.06061212233148e-05,
+      "loss": 2.4941,
+      "step": 1566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9944480061531067,
+      "learning_rate": 8.0580630098485e-05,
+      "loss": 2.5625,
+      "step": 1567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8923318982124329,
+      "learning_rate": 8.05551262686423e-05,
+      "loss": 2.5473,
+      "step": 1568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8775807023048401,
+      "learning_rate": 8.052960974438251e-05,
+      "loss": 2.2249,
+      "step": 1569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9791279435157776,
+      "learning_rate": 8.050408053630671e-05,
+      "loss": 2.4105,
+      "step": 1570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8444132804870605,
+      "learning_rate": 8.047853865502126e-05,
+      "loss": 2.5006,
+      "step": 1571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8552398085594177,
+      "learning_rate": 8.045298411113778e-05,
+      "loss": 2.2266,
+      "step": 1572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.851406455039978,
+      "learning_rate": 8.042741691527313e-05,
+      "loss": 2.3015,
+      "step": 1573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8829087615013123,
+      "learning_rate": 8.040183707804945e-05,
+      "loss": 2.3797,
+      "step": 1574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9659262299537659,
+      "learning_rate": 8.037624461009415e-05,
+      "loss": 2.4004,
+      "step": 1575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9305048584938049,
+      "learning_rate": 8.03506395220398e-05,
+      "loss": 2.3955,
+      "step": 1576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.977260172367096,
+      "learning_rate": 8.032502182452437e-05,
+      "loss": 2.5368,
+      "step": 1577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.019663691520691,
+      "learning_rate": 8.029939152819089e-05,
+      "loss": 2.4676,
+      "step": 1578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0480598211288452,
+      "learning_rate": 8.027374864368775e-05,
+      "loss": 2.6509,
+      "step": 1579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8900166749954224,
+      "learning_rate": 8.024809318166854e-05,
+      "loss": 2.238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9190627336502075,
+      "learning_rate": 8.022242515279203e-05,
+      "loss": 2.3796,
+      "step": 1581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9604819416999817,
+      "learning_rate": 8.01967445677223e-05,
+      "loss": 2.374,
+      "step": 1582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9925508499145508,
+      "learning_rate": 8.017105143712853e-05,
+      "loss": 2.3021,
+      "step": 1583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0334984064102173,
+      "learning_rate": 8.014534577168521e-05,
+      "loss": 2.4076,
+      "step": 1584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.022028923034668,
+      "learning_rate": 8.011962758207199e-05,
+      "loss": 2.6099,
+      "step": 1585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.078336238861084,
+      "learning_rate": 8.009389687897375e-05,
+      "loss": 2.2794,
+      "step": 1586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.799945592880249,
+      "learning_rate": 8.006815367308056e-05,
+      "loss": 2.4669,
+      "step": 1587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1424075365066528,
+      "learning_rate": 8.004239797508764e-05,
+      "loss": 2.4094,
+      "step": 1588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9280880093574524,
+      "learning_rate": 8.001662979569548e-05,
+      "loss": 2.3132,
+      "step": 1589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.975516140460968,
+      "learning_rate": 7.999084914560969e-05,
+      "loss": 2.6215,
+      "step": 1590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1786541938781738,
+      "learning_rate": 7.99650560355411e-05,
+      "loss": 2.4746,
+      "step": 1591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9080499410629272,
+      "learning_rate": 7.99392504762057e-05,
+      "loss": 2.3907,
+      "step": 1592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.050630807876587,
+      "learning_rate": 7.991343247832466e-05,
+      "loss": 2.415,
+      "step": 1593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9351238012313843,
+      "learning_rate": 7.98876020526243e-05,
+      "loss": 2.4579,
+      "step": 1594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9871669411659241,
+      "learning_rate": 7.986175920983611e-05,
+      "loss": 2.3876,
+      "step": 1595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0169905424118042,
+      "learning_rate": 7.983590396069675e-05,
+      "loss": 2.5182,
+      "step": 1596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9110769033432007,
+      "learning_rate": 7.981003631594806e-05,
+      "loss": 2.3365,
+      "step": 1597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.875284731388092,
+      "learning_rate": 7.978415628633695e-05,
+      "loss": 2.2204,
+      "step": 1598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8939157128334045,
+      "learning_rate": 7.975826388261554e-05,
+      "loss": 2.4103,
+      "step": 1599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9853616952896118,
+      "learning_rate": 7.973235911554108e-05,
+      "loss": 2.5692,
+      "step": 1600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9268829226493835,
+      "learning_rate": 7.970644199587596e-05,
+      "loss": 2.3525,
+      "step": 1601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8710259199142456,
+      "learning_rate": 7.968051253438768e-05,
+      "loss": 2.19,
+      "step": 1602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0537748336791992,
+      "learning_rate": 7.965457074184888e-05,
+      "loss": 2.4712,
+      "step": 1603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9105566740036011,
+      "learning_rate": 7.962861662903733e-05,
+      "loss": 2.5011,
+      "step": 1604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9500436782836914,
+      "learning_rate": 7.960265020673592e-05,
+      "loss": 2.3596,
+      "step": 1605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9974099397659302,
+      "learning_rate": 7.957667148573263e-05,
+      "loss": 2.25,
+      "step": 1606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8677740097045898,
+      "learning_rate": 7.955068047682057e-05,
+      "loss": 2.4628,
+      "step": 1607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0048409700393677,
+      "learning_rate": 7.952467719079796e-05,
+      "loss": 2.4108,
+      "step": 1608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0863333940505981,
+      "learning_rate": 7.949866163846811e-05,
+      "loss": 2.5084,
+      "step": 1609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0699412822723389,
+      "learning_rate": 7.947263383063941e-05,
+      "loss": 2.4657,
+      "step": 1610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9292821884155273,
+      "learning_rate": 7.94465937781254e-05,
+      "loss": 2.2268,
+      "step": 1611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8992481827735901,
+      "learning_rate": 7.94205414917446e-05,
+      "loss": 2.2405,
+      "step": 1612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0061519145965576,
+      "learning_rate": 7.939447698232077e-05,
+      "loss": 2.3951,
+      "step": 1613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9798409938812256,
+      "learning_rate": 7.936840026068258e-05,
+      "loss": 2.4678,
+      "step": 1614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0307414531707764,
+      "learning_rate": 7.934231133766387e-05,
+      "loss": 2.2083,
+      "step": 1615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.008912444114685,
+      "learning_rate": 7.931621022410356e-05,
+      "loss": 2.339,
+      "step": 1616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.065599799156189,
+      "learning_rate": 7.929009693084559e-05,
+      "loss": 2.3649,
+      "step": 1617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.823039174079895,
+      "learning_rate": 7.926397146873897e-05,
+      "loss": 2.1857,
+      "step": 1618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9298186898231506,
+      "learning_rate": 7.923783384863775e-05,
+      "loss": 2.3464,
+      "step": 1619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8747079372406006,
+      "learning_rate": 7.921168408140109e-05,
+      "loss": 2.0845,
+      "step": 1620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.918271005153656,
+      "learning_rate": 7.918552217789314e-05,
+      "loss": 2.3711,
+      "step": 1621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9284515380859375,
+      "learning_rate": 7.915934814898312e-05,
+      "loss": 2.3659,
+      "step": 1622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9019615054130554,
+      "learning_rate": 7.913316200554525e-05,
+      "loss": 2.3964,
+      "step": 1623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9223300218582153,
+      "learning_rate": 7.910696375845885e-05,
+      "loss": 2.3233,
+      "step": 1624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0259531736373901,
+      "learning_rate": 7.908075341860823e-05,
+      "loss": 2.5053,
+      "step": 1625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9418337345123291,
+      "learning_rate": 7.905453099688269e-05,
+      "loss": 2.5233,
+      "step": 1626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9323391318321228,
+      "learning_rate": 7.902829650417662e-05,
+      "loss": 2.1136,
+      "step": 1627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1033257246017456,
+      "learning_rate": 7.900204995138935e-05,
+      "loss": 2.6354,
+      "step": 1628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8796046376228333,
+      "learning_rate": 7.89757913494253e-05,
+      "loss": 2.4251,
+      "step": 1629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9551258683204651,
+      "learning_rate": 7.894952070919383e-05,
+      "loss": 2.372,
+      "step": 1630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8407449722290039,
+      "learning_rate": 7.892323804160932e-05,
+      "loss": 2.2472,
+      "step": 1631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8407036662101746,
+      "learning_rate": 7.889694335759119e-05,
+      "loss": 2.164,
+      "step": 1632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0701960325241089,
+      "learning_rate": 7.887063666806378e-05,
+      "loss": 2.3797,
+      "step": 1633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8991835713386536,
+      "learning_rate": 7.884431798395646e-05,
+      "loss": 2.2957,
+      "step": 1634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0140999555587769,
+      "learning_rate": 7.881798731620361e-05,
+      "loss": 2.2481,
+      "step": 1635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9329394102096558,
+      "learning_rate": 7.87916446757445e-05,
+      "loss": 2.242,
+      "step": 1636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9739601612091064,
+      "learning_rate": 7.876529007352348e-05,
+      "loss": 2.3578,
+      "step": 1637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9564885497093201,
+      "learning_rate": 7.87389235204898e-05,
+      "loss": 2.4352,
+      "step": 1638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9778159260749817,
+      "learning_rate": 7.871254502759767e-05,
+      "loss": 2.1641,
+      "step": 1639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1343963146209717,
+      "learning_rate": 7.868615460580633e-05,
+      "loss": 2.4189,
+      "step": 1640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9649724960327148,
+      "learning_rate": 7.86597522660799e-05,
+      "loss": 2.32,
+      "step": 1641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0997480154037476,
+      "learning_rate": 7.863333801938748e-05,
+      "loss": 2.4274,
+      "step": 1642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.975153386592865,
+      "learning_rate": 7.860691187670315e-05,
+      "loss": 2.4466,
+      "step": 1643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9430555701255798,
+      "learning_rate": 7.858047384900589e-05,
+      "loss": 2.3311,
+      "step": 1644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9202498197555542,
+      "learning_rate": 7.855402394727959e-05,
+      "loss": 2.3672,
+      "step": 1645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9534029364585876,
+      "learning_rate": 7.852756218251318e-05,
+      "loss": 2.2137,
+      "step": 1646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9546650648117065,
+      "learning_rate": 7.85010885657004e-05,
+      "loss": 2.2562,
+      "step": 1647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0434573888778687,
+      "learning_rate": 7.847460310783999e-05,
+      "loss": 2.3069,
+      "step": 1648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1709288358688354,
+      "learning_rate": 7.844810581993557e-05,
+      "loss": 2.5331,
+      "step": 1649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8715768456459045,
+      "learning_rate": 7.842159671299569e-05,
+      "loss": 2.0688,
+      "step": 1650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9610153436660767,
+      "learning_rate": 7.839507579803382e-05,
+      "loss": 2.1094,
+      "step": 1651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9678767323493958,
+      "learning_rate": 7.836854308606834e-05,
+      "loss": 2.546,
+      "step": 1652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.962037205696106,
+      "learning_rate": 7.834199858812248e-05,
+      "loss": 2.4358,
+      "step": 1653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1131154298782349,
+      "learning_rate": 7.831544231522443e-05,
+      "loss": 2.5138,
+      "step": 1654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9640947580337524,
+      "learning_rate": 7.828887427840723e-05,
+      "loss": 2.3419,
+      "step": 1655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9368012547492981,
+      "learning_rate": 7.826229448870883e-05,
+      "loss": 2.5076,
+      "step": 1656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0151622295379639,
+      "learning_rate": 7.823570295717206e-05,
+      "loss": 2.4759,
+      "step": 1657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0502291917800903,
+      "learning_rate": 7.820909969484462e-05,
+      "loss": 2.3837,
+      "step": 1658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9294648170471191,
+      "learning_rate": 7.818248471277909e-05,
+      "loss": 2.3006,
+      "step": 1659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.924385130405426,
+      "learning_rate": 7.815585802203291e-05,
+      "loss": 2.3374,
+      "step": 1660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0526052713394165,
+      "learning_rate": 7.812921963366841e-05,
+      "loss": 2.6644,
+      "step": 1661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8251699209213257,
+      "learning_rate": 7.810256955875275e-05,
+      "loss": 2.3058,
+      "step": 1662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9224830269813538,
+      "learning_rate": 7.807590780835795e-05,
+      "loss": 2.3892,
+      "step": 1663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0357741117477417,
+      "learning_rate": 7.804923439356086e-05,
+      "loss": 2.525,
+      "step": 1664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0423439741134644,
+      "learning_rate": 7.802254932544327e-05,
+      "loss": 2.3406,
+      "step": 1665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9470587968826294,
+      "learning_rate": 7.799585261509169e-05,
+      "loss": 2.3737,
+      "step": 1666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8244056701660156,
+      "learning_rate": 7.796914427359754e-05,
+      "loss": 2.4296,
+      "step": 1667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.02104914188385,
+      "learning_rate": 7.794242431205707e-05,
+      "loss": 2.4904,
+      "step": 1668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9128826856613159,
+      "learning_rate": 7.791569274157132e-05,
+      "loss": 2.2568,
+      "step": 1669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9822860360145569,
+      "learning_rate": 7.788894957324618e-05,
+      "loss": 2.2926,
+      "step": 1670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9249267578125,
+      "learning_rate": 7.786219481819235e-05,
+      "loss": 2.3373,
+      "step": 1671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8880137801170349,
+      "learning_rate": 7.783542848752536e-05,
+      "loss": 2.3843,
+      "step": 1672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0186752080917358,
+      "learning_rate": 7.78086505923655e-05,
+      "loss": 2.3696,
+      "step": 1673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9560384154319763,
+      "learning_rate": 7.778186114383795e-05,
+      "loss": 2.2024,
+      "step": 1674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0898208618164062,
+      "learning_rate": 7.775506015307262e-05,
+      "loss": 2.3136,
+      "step": 1675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.075258493423462,
+      "learning_rate": 7.772824763120424e-05,
+      "loss": 2.4146,
+      "step": 1676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8624078631401062,
+      "learning_rate": 7.770142358937228e-05,
+      "loss": 2.1998,
+      "step": 1677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0495119094848633,
+      "learning_rate": 7.767458803872111e-05,
+      "loss": 2.5301,
+      "step": 1678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9451799392700195,
+      "learning_rate": 7.76477409903998e-05,
+      "loss": 2.4468,
+      "step": 1679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9552488327026367,
+      "learning_rate": 7.762088245556219e-05,
+      "loss": 2.3662,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8727592825889587,
+      "learning_rate": 7.759401244536691e-05,
+      "loss": 2.4058,
+      "step": 1681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9953304529190063,
+      "learning_rate": 7.756713097097739e-05,
+      "loss": 2.451,
+      "step": 1682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.066450834274292,
+      "learning_rate": 7.754023804356176e-05,
+      "loss": 2.66,
+      "step": 1683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0633366107940674,
+      "learning_rate": 7.751333367429295e-05,
+      "loss": 2.5229,
+      "step": 1684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8529301285743713,
+      "learning_rate": 7.748641787434865e-05,
+      "loss": 2.1595,
+      "step": 1685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9444707632064819,
+      "learning_rate": 7.745949065491128e-05,
+      "loss": 2.2415,
+      "step": 1686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0203944444656372,
+      "learning_rate": 7.7432552027168e-05,
+      "loss": 2.3591,
+      "step": 1687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9548871517181396,
+      "learning_rate": 7.740560200231073e-05,
+      "loss": 2.2299,
+      "step": 1688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9815899133682251,
+      "learning_rate": 7.73786405915361e-05,
+      "loss": 2.2496,
+      "step": 1689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9313283562660217,
+      "learning_rate": 7.73516678060455e-05,
+      "loss": 2.5009,
+      "step": 1690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0647212266921997,
+      "learning_rate": 7.732468365704502e-05,
+      "loss": 2.3737,
+      "step": 1691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.096922516822815,
+      "learning_rate": 7.729768815574547e-05,
+      "loss": 2.2951,
+      "step": 1692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1384544372558594,
+      "learning_rate": 7.727068131336241e-05,
+      "loss": 2.2196,
+      "step": 1693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9567999243736267,
+      "learning_rate": 7.724366314111608e-05,
+      "loss": 2.3588,
+      "step": 1694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0849462747573853,
+      "learning_rate": 7.721663365023142e-05,
+      "loss": 2.4306,
+      "step": 1695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0322370529174805,
+      "learning_rate": 7.718959285193812e-05,
+      "loss": 2.5054,
+      "step": 1696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1019150018692017,
+      "learning_rate": 7.71625407574705e-05,
+      "loss": 2.5258,
+      "step": 1697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9688391089439392,
+      "learning_rate": 7.713547737806763e-05,
+      "loss": 2.2402,
+      "step": 1698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1122242212295532,
+      "learning_rate": 7.710840272497326e-05,
+      "loss": 2.4363,
+      "step": 1699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9423515796661377,
+      "learning_rate": 7.70813168094358e-05,
+      "loss": 2.2356,
+      "step": 1700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.989656925201416,
+      "learning_rate": 7.705421964270834e-05,
+      "loss": 2.4774,
+      "step": 1701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1426745653152466,
+      "learning_rate": 7.702711123604867e-05,
+      "loss": 2.5183,
+      "step": 1702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9326692819595337,
+      "learning_rate": 7.69999916007192e-05,
+      "loss": 2.2893,
+      "step": 1703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9828354716300964,
+      "learning_rate": 7.697286074798708e-05,
+      "loss": 2.2449,
+      "step": 1704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0485544204711914,
+      "learning_rate": 7.694571868912408e-05,
+      "loss": 2.4795,
+      "step": 1705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.115509271621704,
+      "learning_rate": 7.691856543540661e-05,
+      "loss": 2.4799,
+      "step": 1706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.835258424282074,
+      "learning_rate": 7.689140099811574e-05,
+      "loss": 2.1941,
+      "step": 1707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0000500679016113,
+      "learning_rate": 7.68642253885372e-05,
+      "loss": 2.349,
+      "step": 1708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8784769177436829,
+      "learning_rate": 7.683703861796136e-05,
+      "loss": 2.3486,
+      "step": 1709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1318634748458862,
+      "learning_rate": 7.680984069768322e-05,
+      "loss": 2.4156,
+      "step": 1710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8953444361686707,
+      "learning_rate": 7.678263163900243e-05,
+      "loss": 2.3663,
+      "step": 1711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7773551344871521,
+      "learning_rate": 7.675541145322321e-05,
+      "loss": 2.1373,
+      "step": 1712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9479961395263672,
+      "learning_rate": 7.672818015165449e-05,
+      "loss": 2.4248,
+      "step": 1713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.008703589439392,
+      "learning_rate": 7.670093774560974e-05,
+      "loss": 2.2687,
+      "step": 1714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9986369013786316,
+      "learning_rate": 7.66736842464071e-05,
+      "loss": 2.3341,
+      "step": 1715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9651576280593872,
+      "learning_rate": 7.664641966536928e-05,
+      "loss": 2.2617,
+      "step": 1716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0687973499298096,
+      "learning_rate": 7.661914401382359e-05,
+      "loss": 2.2649,
+      "step": 1717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9479822516441345,
+      "learning_rate": 7.6591857303102e-05,
+      "loss": 2.4282,
+      "step": 1718
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0188570022583008,
+      "learning_rate": 7.6564559544541e-05,
+      "loss": 2.4346,
+      "step": 1719
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9700555801391602,
+      "learning_rate": 7.653725074948174e-05,
+      "loss": 2.2705,
+      "step": 1720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1296528577804565,
+      "learning_rate": 7.650993092926988e-05,
+      "loss": 2.4526,
+      "step": 1721
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8774310946464539,
+      "learning_rate": 7.648260009525569e-05,
+      "loss": 2.4718,
+      "step": 1722
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9145401120185852,
+      "learning_rate": 7.64552582587941e-05,
+      "loss": 2.1461,
+      "step": 1723
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0154246091842651,
+      "learning_rate": 7.642790543124444e-05,
+      "loss": 2.521,
+      "step": 1724
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9274670481681824,
+      "learning_rate": 7.640054162397074e-05,
+      "loss": 2.3038,
+      "step": 1725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.978484570980072,
+      "learning_rate": 7.637316684834159e-05,
+      "loss": 2.4821,
+      "step": 1726
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9475973844528198,
+      "learning_rate": 7.634578111573006e-05,
+      "loss": 2.2641,
+      "step": 1727
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9114319682121277,
+      "learning_rate": 7.631838443751379e-05,
+      "loss": 2.2765,
+      "step": 1728
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9779497385025024,
+      "learning_rate": 7.629097682507503e-05,
+      "loss": 2.3086,
+      "step": 1729
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.971633791923523,
+      "learning_rate": 7.626355828980054e-05,
+      "loss": 2.3454,
+      "step": 1730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0291191339492798,
+      "learning_rate": 7.623612884308158e-05,
+      "loss": 2.3156,
+      "step": 1731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1426750421524048,
+      "learning_rate": 7.620868849631397e-05,
+      "loss": 2.5646,
+      "step": 1732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9908868074417114,
+      "learning_rate": 7.61812372608981e-05,
+      "loss": 2.3487,
+      "step": 1733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1234275102615356,
+      "learning_rate": 7.615377514823882e-05,
+      "loss": 2.5744,
+      "step": 1734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9337838292121887,
+      "learning_rate": 7.612630216974553e-05,
+      "loss": 2.3193,
+      "step": 1735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.126224160194397,
+      "learning_rate": 7.609881833683215e-05,
+      "loss": 2.3314,
+      "step": 1736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9801412224769592,
+      "learning_rate": 7.607132366091709e-05,
+      "loss": 2.4754,
+      "step": 1737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0631413459777832,
+      "learning_rate": 7.604381815342326e-05,
+      "loss": 2.3085,
+      "step": 1738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8829604983329773,
+      "learning_rate": 7.601630182577809e-05,
+      "loss": 2.3546,
+      "step": 1739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9654369950294495,
+      "learning_rate": 7.598877468941352e-05,
+      "loss": 2.2068,
+      "step": 1740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0241409540176392,
+      "learning_rate": 7.596123675576594e-05,
+      "loss": 2.5773,
+      "step": 1741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9042023420333862,
+      "learning_rate": 7.593368803627625e-05,
+      "loss": 2.4095,
+      "step": 1742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2156932353973389,
+      "learning_rate": 7.590612854238983e-05,
+      "loss": 2.3651,
+      "step": 1743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0340590476989746,
+      "learning_rate": 7.587855828555655e-05,
+      "loss": 2.5495,
+      "step": 1744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.120925784111023,
+      "learning_rate": 7.58509772772307e-05,
+      "loss": 2.4524,
+      "step": 1745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0074207782745361,
+      "learning_rate": 7.58233855288711e-05,
+      "loss": 2.201,
+      "step": 1746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.888594388961792,
+      "learning_rate": 7.579578305194097e-05,
+      "loss": 2.2109,
+      "step": 1747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7927941083908081,
+      "learning_rate": 7.576816985790805e-05,
+      "loss": 2.2584,
+      "step": 1748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.912196159362793,
+      "learning_rate": 7.574054595824452e-05,
+      "loss": 2.3528,
+      "step": 1749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0093685388565063,
+      "learning_rate": 7.571291136442696e-05,
+      "loss": 2.3224,
+      "step": 1750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0923992395401,
+      "learning_rate": 7.56852660879364e-05,
+      "loss": 2.5051,
+      "step": 1751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0423188209533691,
+      "learning_rate": 7.56576101402584e-05,
+      "loss": 2.4773,
+      "step": 1752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9690961837768555,
+      "learning_rate": 7.562994353288282e-05,
+      "loss": 2.3725,
+      "step": 1753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9942700862884521,
+      "learning_rate": 7.560226627730407e-05,
+      "loss": 2.5143,
+      "step": 1754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9441214799880981,
+      "learning_rate": 7.557457838502089e-05,
+      "loss": 2.354,
+      "step": 1755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9925492405891418,
+      "learning_rate": 7.554687986753651e-05,
+      "loss": 2.3802,
+      "step": 1756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2071778774261475,
+      "learning_rate": 7.551917073635849e-05,
+      "loss": 2.4032,
+      "step": 1757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1086249351501465,
+      "learning_rate": 7.54914510029989e-05,
+      "loss": 2.307,
+      "step": 1758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0008466243743896,
+      "learning_rate": 7.546372067897416e-05,
+      "loss": 2.3567,
+      "step": 1759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9172022938728333,
+      "learning_rate": 7.543597977580508e-05,
+      "loss": 2.386,
+      "step": 1760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9578985571861267,
+      "learning_rate": 7.540822830501688e-05,
+      "loss": 2.3633,
+      "step": 1761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.008805513381958,
+      "learning_rate": 7.538046627813919e-05,
+      "loss": 2.5445,
+      "step": 1762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9858459830284119,
+      "learning_rate": 7.535269370670598e-05,
+      "loss": 2.4394,
+      "step": 1763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0243449211120605,
+      "learning_rate": 7.532491060225567e-05,
+      "loss": 2.556,
+      "step": 1764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.054688811302185,
+      "learning_rate": 7.529711697633097e-05,
+      "loss": 2.3328,
+      "step": 1765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8582972288131714,
+      "learning_rate": 7.526931284047903e-05,
+      "loss": 2.245,
+      "step": 1766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9144033193588257,
+      "learning_rate": 7.524149820625133e-05,
+      "loss": 2.1838,
+      "step": 1767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9844692945480347,
+      "learning_rate": 7.521367308520374e-05,
+      "loss": 2.2951,
+      "step": 1768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9576483964920044,
+      "learning_rate": 7.518583748889646e-05,
+      "loss": 2.3026,
+      "step": 1769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1517951488494873,
+      "learning_rate": 7.515799142889405e-05,
+      "loss": 2.4929,
+      "step": 1770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8902363777160645,
+      "learning_rate": 7.513013491676544e-05,
+      "loss": 2.429,
+      "step": 1771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9414577484130859,
+      "learning_rate": 7.510226796408384e-05,
+      "loss": 2.3442,
+      "step": 1772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.012868881225586,
+      "learning_rate": 7.50743905824269e-05,
+      "loss": 2.4226,
+      "step": 1773
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1457428932189941,
+      "learning_rate": 7.504650278337647e-05,
+      "loss": 2.3876,
+      "step": 1774
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9788988828659058,
+      "learning_rate": 7.501860457851886e-05,
+      "loss": 2.1157,
+      "step": 1775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.999226450920105,
+      "learning_rate": 7.499069597944462e-05,
+      "loss": 2.0496,
+      "step": 1776
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0303400754928589,
+      "learning_rate": 7.496277699774863e-05,
+      "loss": 2.5661,
+      "step": 1777
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0577235221862793,
+      "learning_rate": 7.493484764503013e-05,
+      "loss": 2.3873,
+      "step": 1778
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9700968265533447,
+      "learning_rate": 7.490690793289261e-05,
+      "loss": 2.2857,
+      "step": 1779
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9228655695915222,
+      "learning_rate": 7.487895787294388e-05,
+      "loss": 2.1987,
+      "step": 1780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0193064212799072,
+      "learning_rate": 7.48509974767961e-05,
+      "loss": 2.3916,
+      "step": 1781
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9101388454437256,
+      "learning_rate": 7.482302675606565e-05,
+      "loss": 2.3217,
+      "step": 1782
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9145199060440063,
+      "learning_rate": 7.479504572237321e-05,
+      "loss": 2.1168,
+      "step": 1783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9714588522911072,
+      "learning_rate": 7.476705438734382e-05,
+      "loss": 2.3367,
+      "step": 1784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8335629105567932,
+      "learning_rate": 7.473905276260669e-05,
+      "loss": 2.1418,
+      "step": 1785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9727364182472229,
+      "learning_rate": 7.47110408597954e-05,
+      "loss": 2.3412,
+      "step": 1786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0289790630340576,
+      "learning_rate": 7.468301869054776e-05,
+      "loss": 2.5406,
+      "step": 1787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9801586270332336,
+      "learning_rate": 7.465498626650582e-05,
+      "loss": 2.3414,
+      "step": 1788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.044488787651062,
+      "learning_rate": 7.462694359931593e-05,
+      "loss": 2.4094,
+      "step": 1789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9189467430114746,
+      "learning_rate": 7.459889070062867e-05,
+      "loss": 2.2104,
+      "step": 1790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0389673709869385,
+      "learning_rate": 7.457082758209892e-05,
+      "loss": 2.3476,
+      "step": 1791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0399079322814941,
+      "learning_rate": 7.454275425538572e-05,
+      "loss": 2.3829,
+      "step": 1792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8790834546089172,
+      "learning_rate": 7.451467073215245e-05,
+      "loss": 2.1827,
+      "step": 1793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9631142616271973,
+      "learning_rate": 7.448657702406664e-05,
+      "loss": 2.3577,
+      "step": 1794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.185268759727478,
+      "learning_rate": 7.44584731428001e-05,
+      "loss": 2.5392,
+      "step": 1795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8667134046554565,
+      "learning_rate": 7.443035910002885e-05,
+      "loss": 2.299,
+      "step": 1796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0530925989151,
+      "learning_rate": 7.440223490743312e-05,
+      "loss": 2.4703,
+      "step": 1797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9795252084732056,
+      "learning_rate": 7.437410057669742e-05,
+      "loss": 2.5102,
+      "step": 1798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0500110387802124,
+      "learning_rate": 7.434595611951037e-05,
+      "loss": 2.3285,
+      "step": 1799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9572514891624451,
+      "learning_rate": 7.43178015475649e-05,
+      "loss": 2.2936,
+      "step": 1800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8998790383338928,
+      "learning_rate": 7.428963687255805e-05,
+      "loss": 2.178,
+      "step": 1801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0332844257354736,
+      "learning_rate": 7.426146210619112e-05,
+      "loss": 2.3708,
+      "step": 1802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9197377562522888,
+      "learning_rate": 7.42332772601696e-05,
+      "loss": 2.2983,
+      "step": 1803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.106593132019043,
+      "learning_rate": 7.420508234620315e-05,
+      "loss": 2.3155,
+      "step": 1804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9009300470352173,
+      "learning_rate": 7.417687737600558e-05,
+      "loss": 2.2135,
+      "step": 1805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1325452327728271,
+      "learning_rate": 7.414866236129494e-05,
+      "loss": 2.4041,
+      "step": 1806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1101938486099243,
+      "learning_rate": 7.412043731379343e-05,
+      "loss": 2.3182,
+      "step": 1807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1332283020019531,
+      "learning_rate": 7.40922022452274e-05,
+      "loss": 2.2634,
+      "step": 1808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9621933698654175,
+      "learning_rate": 7.406395716732737e-05,
+      "loss": 2.1033,
+      "step": 1809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1238906383514404,
+      "learning_rate": 7.403570209182805e-05,
+      "loss": 2.3988,
+      "step": 1810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9589847922325134,
+      "learning_rate": 7.400743703046826e-05,
+      "loss": 2.3482,
+      "step": 1811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8727018237113953,
+      "learning_rate": 7.3979161994991e-05,
+      "loss": 2.1521,
+      "step": 1812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0051244497299194,
+      "learning_rate": 7.395087699714339e-05,
+      "loss": 2.2396,
+      "step": 1813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.162473440170288,
+      "learning_rate": 7.39225820486767e-05,
+      "loss": 2.4082,
+      "step": 1814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8432785272598267,
+      "learning_rate": 7.389427716134634e-05,
+      "loss": 2.2289,
+      "step": 1815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.052862286567688,
+      "learning_rate": 7.386596234691186e-05,
+      "loss": 2.3775,
+      "step": 1816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9011712074279785,
+      "learning_rate": 7.38376376171369e-05,
+      "loss": 2.2758,
+      "step": 1817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0116006135940552,
+      "learning_rate": 7.380930298378925e-05,
+      "loss": 2.443,
+      "step": 1818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8988835215568542,
+      "learning_rate": 7.378095845864077e-05,
+      "loss": 2.1778,
+      "step": 1819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9231415390968323,
+      "learning_rate": 7.375260405346749e-05,
+      "loss": 2.4313,
+      "step": 1820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9672991037368774,
+      "learning_rate": 7.372423978004952e-05,
+      "loss": 2.2927,
+      "step": 1821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1778262853622437,
+      "learning_rate": 7.369586565017106e-05,
+      "loss": 2.2631,
+      "step": 1822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9229978919029236,
+      "learning_rate": 7.36674816756204e-05,
+      "loss": 2.1571,
+      "step": 1823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0724033117294312,
+      "learning_rate": 7.363908786818996e-05,
+      "loss": 2.4812,
+      "step": 1824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9825274348258972,
+      "learning_rate": 7.361068423967616e-05,
+      "loss": 2.4882,
+      "step": 1825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0209020376205444,
+      "learning_rate": 7.358227080187963e-05,
+      "loss": 2.0878,
+      "step": 1826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9900394082069397,
+      "learning_rate": 7.355384756660496e-05,
+      "loss": 2.2618,
+      "step": 1827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9824588298797607,
+      "learning_rate": 7.352541454566086e-05,
+      "loss": 2.5273,
+      "step": 1828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0064198970794678,
+      "learning_rate": 7.349697175086009e-05,
+      "loss": 2.2903,
+      "step": 1829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0262513160705566,
+      "learning_rate": 7.34685191940195e-05,
+      "loss": 2.3089,
+      "step": 1830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0536631345748901,
+      "learning_rate": 7.344005688695996e-05,
+      "loss": 2.3251,
+      "step": 1831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1370151042938232,
+      "learning_rate": 7.341158484150639e-05,
+      "loss": 2.3745,
+      "step": 1832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9724609851837158,
+      "learning_rate": 7.338310306948781e-05,
+      "loss": 2.2782,
+      "step": 1833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1237895488739014,
+      "learning_rate": 7.335461158273721e-05,
+      "loss": 2.432,
+      "step": 1834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0730245113372803,
+      "learning_rate": 7.332611039309166e-05,
+      "loss": 2.5047,
+      "step": 1835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8537101149559021,
+      "learning_rate": 7.329759951239226e-05,
+      "loss": 2.2199,
+      "step": 1836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0017913579940796,
+      "learning_rate": 7.32690789524841e-05,
+      "loss": 2.1947,
+      "step": 1837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0268797874450684,
+      "learning_rate": 7.324054872521634e-05,
+      "loss": 2.3912,
+      "step": 1838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.177985429763794,
+      "learning_rate": 7.321200884244212e-05,
+      "loss": 2.3969,
+      "step": 1839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1149883270263672,
+      "learning_rate": 7.318345931601862e-05,
+      "loss": 2.3235,
+      "step": 1840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9933745861053467,
+      "learning_rate": 7.315490015780699e-05,
+      "loss": 2.391,
+      "step": 1841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.023175835609436,
+      "learning_rate": 7.312633137967243e-05,
+      "loss": 2.2734,
+      "step": 1842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.037279486656189,
+      "learning_rate": 7.309775299348408e-05,
+      "loss": 2.3447,
+      "step": 1843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1776492595672607,
+      "learning_rate": 7.306916501111512e-05,
+      "loss": 2.518,
+      "step": 1844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9285102486610413,
+      "learning_rate": 7.30405674444427e-05,
+      "loss": 2.1527,
+      "step": 1845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9894176721572876,
+      "learning_rate": 7.301196030534796e-05,
+      "loss": 2.4074,
+      "step": 1846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9558159112930298,
+      "learning_rate": 7.298334360571599e-05,
+      "loss": 2.4047,
+      "step": 1847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.047155499458313,
+      "learning_rate": 7.295471735743586e-05,
+      "loss": 2.4479,
+      "step": 1848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0158107280731201,
+      "learning_rate": 7.292608157240063e-05,
+      "loss": 2.3767,
+      "step": 1849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9218646287918091,
+      "learning_rate": 7.289743626250731e-05,
+      "loss": 2.211,
+      "step": 1850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9160715341567993,
+      "learning_rate": 7.286878143965689e-05,
+      "loss": 2.4634,
+      "step": 1851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8560699820518494,
+      "learning_rate": 7.284011711575423e-05,
+      "loss": 2.2871,
+      "step": 1852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9146595597267151,
+      "learning_rate": 7.281144330270824e-05,
+      "loss": 2.3682,
+      "step": 1853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8751729130744934,
+      "learning_rate": 7.27827600124317e-05,
+      "loss": 2.2389,
+      "step": 1854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9473608136177063,
+      "learning_rate": 7.275406725684137e-05,
+      "loss": 2.4411,
+      "step": 1855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0765432119369507,
+      "learning_rate": 7.272536504785794e-05,
+      "loss": 2.4631,
+      "step": 1856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9788899421691895,
+      "learning_rate": 7.2696653397406e-05,
+      "loss": 2.2689,
+      "step": 1857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.944067656993866,
+      "learning_rate": 7.266793231741408e-05,
+      "loss": 2.4057,
+      "step": 1858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0392214059829712,
+      "learning_rate": 7.263920181981461e-05,
+      "loss": 2.5401,
+      "step": 1859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0435481071472168,
+      "learning_rate": 7.261046191654398e-05,
+      "loss": 2.4703,
+      "step": 1860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9915140271186829,
+      "learning_rate": 7.258171261954245e-05,
+      "loss": 2.2865,
+      "step": 1861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8746885061264038,
+      "learning_rate": 7.255295394075417e-05,
+      "loss": 2.1706,
+      "step": 1862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1280632019042969,
+      "learning_rate": 7.25241858921272e-05,
+      "loss": 2.5916,
+      "step": 1863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1003673076629639,
+      "learning_rate": 7.24954084856135e-05,
+      "loss": 2.6791,
+      "step": 1864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0396860837936401,
+      "learning_rate": 7.246662173316894e-05,
+      "loss": 2.4586,
+      "step": 1865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0224331617355347,
+      "learning_rate": 7.243782564675325e-05,
+      "loss": 2.3271,
+      "step": 1866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9120011925697327,
+      "learning_rate": 7.240902023833e-05,
+      "loss": 2.4112,
+      "step": 1867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9646201729774475,
+      "learning_rate": 7.23802055198667e-05,
+      "loss": 2.2623,
+      "step": 1868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9280524253845215,
+      "learning_rate": 7.235138150333471e-05,
+      "loss": 2.2508,
+      "step": 1869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8653500080108643,
+      "learning_rate": 7.232254820070922e-05,
+      "loss": 2.1536,
+      "step": 1870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1189990043640137,
+      "learning_rate": 7.22937056239693e-05,
+      "loss": 2.618,
+      "step": 1871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0262922048568726,
+      "learning_rate": 7.226485378509788e-05,
+      "loss": 2.3422,
+      "step": 1872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0643494129180908,
+      "learning_rate": 7.223599269608172e-05,
+      "loss": 2.2168,
+      "step": 1873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9295349717140198,
+      "learning_rate": 7.220712236891144e-05,
+      "loss": 2.2115,
+      "step": 1874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0115851163864136,
+      "learning_rate": 7.217824281558148e-05,
+      "loss": 2.4658,
+      "step": 1875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0030776262283325,
+      "learning_rate": 7.214935404809015e-05,
+      "loss": 2.3486,
+      "step": 1876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.920015275478363,
+      "learning_rate": 7.212045607843953e-05,
+      "loss": 2.4011,
+      "step": 1877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1129233837127686,
+      "learning_rate": 7.209154891863557e-05,
+      "loss": 2.3726,
+      "step": 1878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.089972972869873,
+      "learning_rate": 7.206263258068804e-05,
+      "loss": 2.5567,
+      "step": 1879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9959055185317993,
+      "learning_rate": 7.203370707661048e-05,
+      "loss": 2.3389,
+      "step": 1880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9231062531471252,
+      "learning_rate": 7.200477241842029e-05,
+      "loss": 2.4048,
+      "step": 1881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0919133424758911,
+      "learning_rate": 7.197582861813864e-05,
+      "loss": 2.4467,
+      "step": 1882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.000555396080017,
+      "learning_rate": 7.19468756877905e-05,
+      "loss": 2.5386,
+      "step": 1883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1174827814102173,
+      "learning_rate": 7.191791363940466e-05,
+      "loss": 2.4926,
+      "step": 1884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0921218395233154,
+      "learning_rate": 7.188894248501363e-05,
+      "loss": 2.407,
+      "step": 1885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.04496169090271,
+      "learning_rate": 7.185996223665379e-05,
+      "loss": 2.363,
+      "step": 1886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8943964242935181,
+      "learning_rate": 7.183097290636525e-05,
+      "loss": 2.3459,
+      "step": 1887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1562391519546509,
+      "learning_rate": 7.18019745061919e-05,
+      "loss": 2.3579,
+      "step": 1888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.992927610874176,
+      "learning_rate": 7.177296704818142e-05,
+      "loss": 2.5116,
+      "step": 1889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9252989292144775,
+      "learning_rate": 7.17439505443852e-05,
+      "loss": 2.283,
+      "step": 1890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.088307499885559,
+      "learning_rate": 7.171492500685843e-05,
+      "loss": 2.5015,
+      "step": 1891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1222864389419556,
+      "learning_rate": 7.168589044766008e-05,
+      "loss": 2.2199,
+      "step": 1892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9369937777519226,
+      "learning_rate": 7.165684687885279e-05,
+      "loss": 2.2838,
+      "step": 1893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9874539375305176,
+      "learning_rate": 7.162779431250302e-05,
+      "loss": 1.9334,
+      "step": 1894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0229480266571045,
+      "learning_rate": 7.15987327606809e-05,
+      "loss": 2.3218,
+      "step": 1895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0848695039749146,
+      "learning_rate": 7.156966223546035e-05,
+      "loss": 2.4465,
+      "step": 1896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9145216345787048,
+      "learning_rate": 7.154058274891898e-05,
+      "loss": 2.29,
+      "step": 1897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1953297853469849,
+      "learning_rate": 7.151149431313817e-05,
+      "loss": 2.4823,
+      "step": 1898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.883732795715332,
+      "learning_rate": 7.148239694020293e-05,
+      "loss": 2.3356,
+      "step": 1899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8720072507858276,
+      "learning_rate": 7.145329064220211e-05,
+      "loss": 2.151,
+      "step": 1900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.882093608379364,
+      "learning_rate": 7.142417543122815e-05,
+      "loss": 2.3478,
+      "step": 1901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0163112878799438,
+      "learning_rate": 7.139505131937725e-05,
+      "loss": 2.2583,
+      "step": 1902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8707605004310608,
+      "learning_rate": 7.13659183187493e-05,
+      "loss": 2.0594,
+      "step": 1903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9629480242729187,
+      "learning_rate": 7.133677644144789e-05,
+      "loss": 2.2706,
+      "step": 1904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9019947648048401,
+      "learning_rate": 7.130762569958027e-05,
+      "loss": 2.223,
+      "step": 1905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9047419428825378,
+      "learning_rate": 7.127846610525742e-05,
+      "loss": 2.3067,
+      "step": 1906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0522680282592773,
+      "learning_rate": 7.124929767059395e-05,
+      "loss": 2.2904,
+      "step": 1907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1826410293579102,
+      "learning_rate": 7.122012040770816e-05,
+      "loss": 2.4185,
+      "step": 1908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9610967636108398,
+      "learning_rate": 7.119093432872203e-05,
+      "loss": 2.2133,
+      "step": 1909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0889559984207153,
+      "learning_rate": 7.11617394457612e-05,
+      "loss": 2.325,
+      "step": 1910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0843358039855957,
+      "learning_rate": 7.113253577095495e-05,
+      "loss": 2.3105,
+      "step": 1911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.014501690864563,
+      "learning_rate": 7.110332331643623e-05,
+      "loss": 2.2152,
+      "step": 1912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.117771029472351,
+      "learning_rate": 7.107410209434162e-05,
+      "loss": 2.5167,
+      "step": 1913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0908191204071045,
+      "learning_rate": 7.104487211681136e-05,
+      "loss": 2.4465,
+      "step": 1914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9349441528320312,
+      "learning_rate": 7.101563339598934e-05,
+      "loss": 2.0743,
+      "step": 1915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9997676014900208,
+      "learning_rate": 7.098638594402305e-05,
+      "loss": 2.4956,
+      "step": 1916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9454015493392944,
+      "learning_rate": 7.09571297730636e-05,
+      "loss": 2.465,
+      "step": 1917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.876261830329895,
+      "learning_rate": 7.092786489526578e-05,
+      "loss": 2.2634,
+      "step": 1918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0011814832687378,
+      "learning_rate": 7.089859132278793e-05,
+      "loss": 2.3496,
+      "step": 1919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9676500558853149,
+      "learning_rate": 7.086930906779206e-05,
+      "loss": 2.2497,
+      "step": 1920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2238849401474,
+      "learning_rate": 7.084001814244372e-05,
+      "loss": 2.681,
+      "step": 1921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9545730948448181,
+      "learning_rate": 7.081071855891215e-05,
+      "loss": 2.3416,
+      "step": 1922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0307912826538086,
+      "learning_rate": 7.078141032937012e-05,
+      "loss": 2.3351,
+      "step": 1923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0172101259231567,
+      "learning_rate": 7.075209346599401e-05,
+      "loss": 2.4778,
+      "step": 1924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9569324851036072,
+      "learning_rate": 7.07227679809638e-05,
+      "loss": 2.2771,
+      "step": 1925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9619925022125244,
+      "learning_rate": 7.069343388646302e-05,
+      "loss": 2.266,
+      "step": 1926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9336727857589722,
+      "learning_rate": 7.066409119467882e-05,
+      "loss": 2.2743,
+      "step": 1927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.967343807220459,
+      "learning_rate": 7.063473991780189e-05,
+      "loss": 2.2525,
+      "step": 1928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.008945345878601,
+      "learning_rate": 7.06053800680265e-05,
+      "loss": 2.3603,
+      "step": 1929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9258151650428772,
+      "learning_rate": 7.057601165755046e-05,
+      "loss": 2.0378,
+      "step": 1930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.035955548286438,
+      "learning_rate": 7.054663469857519e-05,
+      "loss": 2.4167,
+      "step": 1931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.164644718170166,
+      "learning_rate": 7.051724920330558e-05,
+      "loss": 2.3872,
+      "step": 1932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1040794849395752,
+      "learning_rate": 7.048785518395016e-05,
+      "loss": 2.2441,
+      "step": 1933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2054561376571655,
+      "learning_rate": 7.045845265272092e-05,
+      "loss": 2.3729,
+      "step": 1934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1438981294631958,
+      "learning_rate": 7.042904162183343e-05,
+      "loss": 2.5763,
+      "step": 1935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1155726909637451,
+      "learning_rate": 7.039962210350681e-05,
+      "loss": 2.4375,
+      "step": 1936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9198666214942932,
+      "learning_rate": 7.037019410996364e-05,
+      "loss": 2.4111,
+      "step": 1937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0576508045196533,
+      "learning_rate": 7.034075765343007e-05,
+      "loss": 2.3171,
+      "step": 1938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8595229387283325,
+      "learning_rate": 7.031131274613575e-05,
+      "loss": 2.1381,
+      "step": 1939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9790457487106323,
+      "learning_rate": 7.028185940031385e-05,
+      "loss": 2.2774,
+      "step": 1940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8667134642601013,
+      "learning_rate": 7.025239762820107e-05,
+      "loss": 2.4286,
+      "step": 1941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9495657086372375,
+      "learning_rate": 7.022292744203755e-05,
+      "loss": 2.3188,
+      "step": 1942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9178842306137085,
+      "learning_rate": 7.019344885406696e-05,
+      "loss": 2.1602,
+      "step": 1943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.146416187286377,
+      "learning_rate": 7.016396187653645e-05,
+      "loss": 2.5431,
+      "step": 1944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2337157726287842,
+      "learning_rate": 7.01344665216967e-05,
+      "loss": 2.4217,
+      "step": 1945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9706616997718811,
+      "learning_rate": 7.01049628018018e-05,
+      "loss": 2.4069,
+      "step": 1946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9983075857162476,
+      "learning_rate": 7.007545072910939e-05,
+      "loss": 2.3568,
+      "step": 1947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9811784625053406,
+      "learning_rate": 7.00459303158805e-05,
+      "loss": 2.241,
+      "step": 1948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9459620714187622,
+      "learning_rate": 7.001640157437968e-05,
+      "loss": 2.4022,
+      "step": 1949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9831576943397522,
+      "learning_rate": 6.998686451687493e-05,
+      "loss": 2.3025,
+      "step": 1950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9504924416542053,
+      "learning_rate": 6.995731915563772e-05,
+      "loss": 2.2159,
+      "step": 1951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0905296802520752,
+      "learning_rate": 6.99277655029429e-05,
+      "loss": 2.4068,
+      "step": 1952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.02506685256958,
+      "learning_rate": 6.989820357106886e-05,
+      "loss": 2.4193,
+      "step": 1953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.881464958190918,
+      "learning_rate": 6.986863337229735e-05,
+      "loss": 2.1517,
+      "step": 1954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.05882728099823,
+      "learning_rate": 6.983905491891362e-05,
+      "loss": 2.3527,
+      "step": 1955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.023546576499939,
+      "learning_rate": 6.980946822320631e-05,
+      "loss": 2.4336,
+      "step": 1956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.091468334197998,
+      "learning_rate": 6.977987329746749e-05,
+      "loss": 2.4512,
+      "step": 1957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1171419620513916,
+      "learning_rate": 6.975027015399267e-05,
+      "loss": 2.455,
+      "step": 1958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8383733630180359,
+      "learning_rate": 6.972065880508072e-05,
+      "loss": 2.006,
+      "step": 1959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9669481515884399,
+      "learning_rate": 6.969103926303398e-05,
+      "loss": 2.5064,
+      "step": 1960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9148940443992615,
+      "learning_rate": 6.966141154015818e-05,
+      "loss": 2.3514,
+      "step": 1961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8779053688049316,
+      "learning_rate": 6.963177564876241e-05,
+      "loss": 2.2903,
+      "step": 1962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8760914206504822,
+      "learning_rate": 6.960213160115921e-05,
+      "loss": 2.2645,
+      "step": 1963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9167511463165283,
+      "learning_rate": 6.957247940966448e-05,
+      "loss": 2.2152,
+      "step": 1964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0847175121307373,
+      "learning_rate": 6.954281908659745e-05,
+      "loss": 2.4489,
+      "step": 1965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9281463623046875,
+      "learning_rate": 6.951315064428086e-05,
+      "loss": 2.2219,
+      "step": 1966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.02683687210083,
+      "learning_rate": 6.94834740950407e-05,
+      "loss": 2.3016,
+      "step": 1967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0726292133331299,
+      "learning_rate": 6.945378945120638e-05,
+      "loss": 2.2506,
+      "step": 1968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0985912084579468,
+      "learning_rate": 6.942409672511067e-05,
+      "loss": 2.2787,
+      "step": 1969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1478863954544067,
+      "learning_rate": 6.93943959290897e-05,
+      "loss": 2.3013,
+      "step": 1970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.068804383277893,
+      "learning_rate": 6.936468707548296e-05,
+      "loss": 2.3063,
+      "step": 1971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.069574236869812,
+      "learning_rate": 6.933497017663324e-05,
+      "loss": 2.254,
+      "step": 1972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0655707120895386,
+      "learning_rate": 6.930524524488674e-05,
+      "loss": 2.2795,
+      "step": 1973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9645271897315979,
+      "learning_rate": 6.927551229259295e-05,
+      "loss": 2.2971,
+      "step": 1974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9800378084182739,
+      "learning_rate": 6.92457713321047e-05,
+      "loss": 2.1733,
+      "step": 1975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9926590323448181,
+      "learning_rate": 6.921602237577816e-05,
+      "loss": 2.034,
+      "step": 1976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1322658061981201,
+      "learning_rate": 6.918626543597283e-05,
+      "loss": 2.4054,
+      "step": 1977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0951309204101562,
+      "learning_rate": 6.915650052505149e-05,
+      "loss": 2.3559,
+      "step": 1978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9194061756134033,
+      "learning_rate": 6.912672765538029e-05,
+      "loss": 2.3061,
+      "step": 1979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9342663884162903,
+      "learning_rate": 6.909694683932861e-05,
+      "loss": 2.0073,
+      "step": 1980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0094574689865112,
+      "learning_rate": 6.90671580892692e-05,
+      "loss": 2.2467,
+      "step": 1981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9711092710494995,
+      "learning_rate": 6.903736141757804e-05,
+      "loss": 2.2973,
+      "step": 1982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1370465755462646,
+      "learning_rate": 6.90075568366345e-05,
+      "loss": 2.4145,
+      "step": 1983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1352934837341309,
+      "learning_rate": 6.897774435882114e-05,
+      "loss": 2.418,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.055026650428772,
+      "learning_rate": 6.894792399652383e-05,
+      "loss": 2.3529,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9283546805381775,
+      "learning_rate": 6.891809576213175e-05,
+      "loss": 2.1083,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0603094100952148,
+      "learning_rate": 6.88882596680373e-05,
+      "loss": 2.2533,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9512935280799866,
+      "learning_rate": 6.885841572663618e-05,
+      "loss": 2.3593,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9863986968994141,
+      "learning_rate": 6.882856395032731e-05,
+      "loss": 2.2434,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0063016414642334,
+      "learning_rate": 6.879870435151295e-05,
+      "loss": 2.2904,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1802427768707275,
+      "learning_rate": 6.87688369425985e-05,
+      "loss": 2.1984,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8608043193817139,
+      "learning_rate": 6.873896173599267e-05,
+      "loss": 2.3941,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9798101186752319,
+      "learning_rate": 6.870907874410744e-05,
+      "loss": 2.34,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9803988337516785,
+      "learning_rate": 6.867918797935793e-05,
+      "loss": 2.3386,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8761933445930481,
+      "learning_rate": 6.864928945416256e-05,
+      "loss": 2.22,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9836278557777405,
+      "learning_rate": 6.861938318094298e-05,
+      "loss": 2.2783,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0706288814544678,
+      "learning_rate": 6.858946917212399e-05,
+      "loss": 2.3368,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1138883829116821,
+      "learning_rate": 6.855954744013368e-05,
+      "loss": 2.483,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8299687504768372,
+      "learning_rate": 6.85296179974033e-05,
+      "loss": 2.1987,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9227612614631653,
+      "learning_rate": 6.849968085636735e-05,
+      "loss": 2.317,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8624487519264221,
+      "learning_rate": 6.846973602946348e-05,
+      "loss": 2.2318,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.016144871711731,
+      "learning_rate": 6.843978352913253e-05,
+      "loss": 2.4197,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.02602219581604,
+      "learning_rate": 6.840982336781858e-05,
+      "loss": 2.3632,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9254372715950012,
+      "learning_rate": 6.837985555796885e-05,
+      "loss": 2.2199,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9654033780097961,
+      "learning_rate": 6.834988011203375e-05,
+      "loss": 2.3823,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.03053617477417,
+      "learning_rate": 6.83198970424669e-05,
+      "loss": 2.0467,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2155038118362427,
+      "learning_rate": 6.8289906361725e-05,
+      "loss": 2.5357,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0382225513458252,
+      "learning_rate": 6.825990808226799e-05,
+      "loss": 2.5321,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9617961049079895,
+      "learning_rate": 6.822990221655892e-05,
+      "loss": 2.4034,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8806459903717041,
+      "learning_rate": 6.819988877706405e-05,
+      "loss": 2.2643,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0166072845458984,
+      "learning_rate": 6.816986777625269e-05,
+      "loss": 2.3042,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3165175914764404,
+      "learning_rate": 6.81398392265974e-05,
+      "loss": 2.5489,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8895251750946045,
+      "learning_rate": 6.810980314057382e-05,
+      "loss": 2.1225,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.017021656036377,
+      "learning_rate": 6.807975953066069e-05,
+      "loss": 2.4495,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0848194360733032,
+      "learning_rate": 6.804970840933995e-05,
+      "loss": 2.3013,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0447602272033691,
+      "learning_rate": 6.801964978909661e-05,
+      "loss": 2.3218,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0969734191894531,
+      "learning_rate": 6.79895836824188e-05,
+      "loss": 2.2071,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9825399518013,
+      "learning_rate": 6.79595101017978e-05,
+      "loss": 2.2125,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.042770504951477,
+      "learning_rate": 6.792942905972794e-05,
+      "loss": 2.2947,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9319426417350769,
+      "learning_rate": 6.789934056870668e-05,
+      "loss": 2.2736,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.158310055732727,
+      "learning_rate": 6.786924464123455e-05,
+      "loss": 2.2642,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1341897249221802,
+      "learning_rate": 6.783914128981523e-05,
+      "loss": 2.5299,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1291477680206299,
+      "learning_rate": 6.780903052695541e-05,
+      "loss": 2.4125,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9972942471504211,
+      "learning_rate": 6.777891236516489e-05,
+      "loss": 2.3909,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0319344997406006,
+      "learning_rate": 6.774878681695656e-05,
+      "loss": 2.3553,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1839104890823364,
+      "learning_rate": 6.771865389484637e-05,
+      "loss": 2.2057,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0909045934677124,
+      "learning_rate": 6.768851361135331e-05,
+      "loss": 2.3174,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.026576280593872,
+      "learning_rate": 6.765836597899946e-05,
+      "loss": 2.5135,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9381416440010071,
+      "learning_rate": 6.762821101030994e-05,
+      "loss": 2.1658,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.057152271270752,
+      "learning_rate": 6.759804871781292e-05,
+      "loss": 2.3878,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1768274307250977,
+      "learning_rate": 6.75678791140396e-05,
+      "loss": 2.5656,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1033854484558105,
+      "learning_rate": 6.753770221152423e-05,
+      "loss": 2.4307,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9449601769447327,
+      "learning_rate": 6.75075180228041e-05,
+      "loss": 2.373,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9695514440536499,
+      "learning_rate": 6.74773265604195e-05,
+      "loss": 2.1684,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0132147073745728,
+      "learning_rate": 6.744712783691376e-05,
+      "loss": 2.2036,
+      "step": 2035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0681016445159912,
+      "learning_rate": 6.741692186483326e-05,
+      "loss": 2.3619,
+      "step": 2036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0439531803131104,
+      "learning_rate": 6.738670865672734e-05,
+      "loss": 2.1556,
+      "step": 2037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9893102049827576,
+      "learning_rate": 6.735648822514834e-05,
+      "loss": 2.2133,
+      "step": 2038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9395050406455994,
+      "learning_rate": 6.732626058265165e-05,
+      "loss": 2.299,
+      "step": 2039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0079295635223389,
+      "learning_rate": 6.729602574179563e-05,
+      "loss": 2.3837,
+      "step": 2040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9109761118888855,
+      "learning_rate": 6.726578371514163e-05,
+      "loss": 2.1867,
+      "step": 2041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0518262386322021,
+      "learning_rate": 6.723553451525396e-05,
+      "loss": 2.3097,
+      "step": 2042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.006900429725647,
+      "learning_rate": 6.720527815469999e-05,
+      "loss": 2.1597,
+      "step": 2043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0063409805297852,
+      "learning_rate": 6.717501464604996e-05,
+      "loss": 2.2774,
+      "step": 2044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9496919512748718,
+      "learning_rate": 6.714474400187716e-05,
+      "loss": 2.2178,
+      "step": 2045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9483358263969421,
+      "learning_rate": 6.711446623475781e-05,
+      "loss": 2.3397,
+      "step": 2046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9885390400886536,
+      "learning_rate": 6.708418135727109e-05,
+      "loss": 2.1667,
+      "step": 2047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0352132320404053,
+      "learning_rate": 6.705388938199914e-05,
+      "loss": 2.3577,
+      "step": 2048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0144520998001099,
+      "learning_rate": 6.702359032152708e-05,
+      "loss": 2.1827,
+      "step": 2049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8596828579902649,
+      "learning_rate": 6.699328418844288e-05,
+      "loss": 2.1633,
+      "step": 2050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0513088703155518,
+      "learning_rate": 6.696297099533757e-05,
+      "loss": 2.3119,
+      "step": 2051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.084269642829895,
+      "learning_rate": 6.693265075480503e-05,
+      "loss": 2.1636,
+      "step": 2052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0379847288131714,
+      "learning_rate": 6.69023234794421e-05,
+      "loss": 2.2578,
+      "step": 2053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0158923864364624,
+      "learning_rate": 6.687198918184851e-05,
+      "loss": 2.4868,
+      "step": 2054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9715821743011475,
+      "learning_rate": 6.684164787462695e-05,
+      "loss": 2.2638,
+      "step": 2055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0037987232208252,
+      "learning_rate": 6.6811299570383e-05,
+      "loss": 2.1836,
+      "step": 2056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0476551055908203,
+      "learning_rate": 6.678094428172514e-05,
+      "loss": 2.1808,
+      "step": 2057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0895642042160034,
+      "learning_rate": 6.67505820212648e-05,
+      "loss": 2.405,
+      "step": 2058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9857387542724609,
+      "learning_rate": 6.672021280161624e-05,
+      "loss": 2.3252,
+      "step": 2059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0298118591308594,
+      "learning_rate": 6.668983663539663e-05,
+      "loss": 2.2232,
+      "step": 2060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8861029744148254,
+      "learning_rate": 6.665945353522606e-05,
+      "loss": 2.0568,
+      "step": 2061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.202353835105896,
+      "learning_rate": 6.662906351372747e-05,
+      "loss": 2.1874,
+      "step": 2062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2553598880767822,
+      "learning_rate": 6.659866658352669e-05,
+      "loss": 2.2584,
+      "step": 2063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1088649034500122,
+      "learning_rate": 6.656826275725239e-05,
+      "loss": 2.4089,
+      "step": 2064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1088649034500122,
+      "learning_rate": 6.656826275725239e-05,
+      "loss": 2.2665,
+      "step": 2065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9473740458488464,
+      "learning_rate": 6.653785204753615e-05,
+      "loss": 2.1903,
+      "step": 2066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.946965217590332,
+      "learning_rate": 6.650743446701238e-05,
+      "loss": 2.2611,
+      "step": 2067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9346373677253723,
+      "learning_rate": 6.647701002831834e-05,
+      "loss": 2.1273,
+      "step": 2068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0149379968643188,
+      "learning_rate": 6.644657874409417e-05,
+      "loss": 2.2952,
+      "step": 2069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9972168207168579,
+      "learning_rate": 6.641614062698282e-05,
+      "loss": 2.3427,
+      "step": 2070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.065118670463562,
+      "learning_rate": 6.638569568963013e-05,
+      "loss": 2.4515,
+      "step": 2071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.944165050983429,
+      "learning_rate": 6.635524394468468e-05,
+      "loss": 2.0573,
+      "step": 2072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.144840121269226,
+      "learning_rate": 6.6324785404798e-05,
+      "loss": 2.3152,
+      "step": 2073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0445878505706787,
+      "learning_rate": 6.629432008262433e-05,
+      "loss": 2.3385,
+      "step": 2074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9817285537719727,
+      "learning_rate": 6.626384799082079e-05,
+      "loss": 2.267,
+      "step": 2075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9845954775810242,
+      "learning_rate": 6.623336914204735e-05,
+      "loss": 2.1366,
+      "step": 2076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0155348777770996,
+      "learning_rate": 6.620288354896664e-05,
+      "loss": 2.2416,
+      "step": 2077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9235448837280273,
+      "learning_rate": 6.617239122424426e-05,
+      "loss": 2.2019,
+      "step": 2078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9976829290390015,
+      "learning_rate": 6.614189218054852e-05,
+      "loss": 2.3324,
+      "step": 2079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0142080783843994,
+      "learning_rate": 6.611138643055051e-05,
+      "loss": 2.0407,
+      "step": 2080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0657060146331787,
+      "learning_rate": 6.608087398692418e-05,
+      "loss": 2.1483,
+      "step": 2081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0807374715805054,
+      "learning_rate": 6.605035486234619e-05,
+      "loss": 2.3016,
+      "step": 2082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9735080599784851,
+      "learning_rate": 6.601982906949599e-05,
+      "loss": 2.1046,
+      "step": 2083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9906128644943237,
+      "learning_rate": 6.598929662105582e-05,
+      "loss": 2.3492,
+      "step": 2084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1213064193725586,
+      "learning_rate": 6.595875752971068e-05,
+      "loss": 2.3404,
+      "step": 2085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0322107076644897,
+      "learning_rate": 6.592821180814834e-05,
+      "loss": 2.2961,
+      "step": 2086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.153718113899231,
+      "learning_rate": 6.589765946905927e-05,
+      "loss": 2.5881,
+      "step": 2087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.097139596939087,
+      "learning_rate": 6.586710052513676e-05,
+      "loss": 2.2664,
+      "step": 2088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.010186791419983,
+      "learning_rate": 6.58365349890768e-05,
+      "loss": 2.3186,
+      "step": 2089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0088762044906616,
+      "learning_rate": 6.580596287357814e-05,
+      "loss": 2.278,
+      "step": 2090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9239149689674377,
+      "learning_rate": 6.577538419134223e-05,
+      "loss": 2.0787,
+      "step": 2091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.923526406288147,
+      "learning_rate": 6.574479895507332e-05,
+      "loss": 2.2585,
+      "step": 2092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.051278829574585,
+      "learning_rate": 6.57142071774783e-05,
+      "loss": 2.3197,
+      "step": 2093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8610481023788452,
+      "learning_rate": 6.568360887126683e-05,
+      "loss": 2.1062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1033543348312378,
+      "learning_rate": 6.565300404915125e-05,
+      "loss": 2.2692,
+      "step": 2095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1326587200164795,
+      "learning_rate": 6.562239272384665e-05,
+      "loss": 2.3728,
+      "step": 2096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.927254855632782,
+      "learning_rate": 6.559177490807076e-05,
+      "loss": 2.3092,
+      "step": 2097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9317123889923096,
+      "learning_rate": 6.556115061454409e-05,
+      "loss": 2.3125,
+      "step": 2098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8561984896659851,
+      "learning_rate": 6.553051985598976e-05,
+      "loss": 2.1496,
+      "step": 2099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0486624240875244,
+      "learning_rate": 6.549988264513361e-05,
+      "loss": 2.3402,
+      "step": 2100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0355058908462524,
+      "learning_rate": 6.546923899470416e-05,
+      "loss": 2.443,
+      "step": 2101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9090213179588318,
+      "learning_rate": 6.543858891743262e-05,
+      "loss": 2.1145,
+      "step": 2102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0126415491104126,
+      "learning_rate": 6.540793242605284e-05,
+      "loss": 2.3294,
+      "step": 2103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8979871273040771,
+      "learning_rate": 6.537726953330135e-05,
+      "loss": 2.043,
+      "step": 2104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2439504861831665,
+      "learning_rate": 6.534660025191735e-05,
+      "loss": 2.4003,
+      "step": 2105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0128694772720337,
+      "learning_rate": 6.531592459464269e-05,
+      "loss": 2.3149,
+      "step": 2106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0449320077896118,
+      "learning_rate": 6.528524257422185e-05,
+      "loss": 2.4901,
+      "step": 2107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.064292311668396,
+      "learning_rate": 6.525455420340196e-05,
+      "loss": 2.2646,
+      "step": 2108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0236588716506958,
+      "learning_rate": 6.522385949493282e-05,
+      "loss": 2.1586,
+      "step": 2109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9835452437400818,
+      "learning_rate": 6.519315846156682e-05,
+      "loss": 2.2121,
+      "step": 2110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9713760018348694,
+      "learning_rate": 6.516245111605902e-05,
+      "loss": 2.3667,
+      "step": 2111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.005453109741211,
+      "learning_rate": 6.5131737471167e-05,
+      "loss": 2.1832,
+      "step": 2112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.183020830154419,
+      "learning_rate": 6.510101753965113e-05,
+      "loss": 2.6332,
+      "step": 2113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9999881982803345,
+      "learning_rate": 6.507029133427425e-05,
+      "loss": 2.3876,
+      "step": 2114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9726799130439758,
+      "learning_rate": 6.503955886780185e-05,
+      "loss": 2.2894,
+      "step": 2115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9953005909919739,
+      "learning_rate": 6.500882015300204e-05,
+      "loss": 2.1257,
+      "step": 2116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9523982405662537,
+      "learning_rate": 6.49780752026455e-05,
+      "loss": 2.237,
+      "step": 2117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.023040533065796,
+      "learning_rate": 6.494732402950554e-05,
+      "loss": 2.195,
+      "step": 2118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9973800182342529,
+      "learning_rate": 6.4916566646358e-05,
+      "loss": 2.3301,
+      "step": 2119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9764662384986877,
+      "learning_rate": 6.488580306598131e-05,
+      "loss": 2.3029,
+      "step": 2120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.229904055595398,
+      "learning_rate": 6.485503330115653e-05,
+      "loss": 2.2208,
+      "step": 2121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0190060138702393,
+      "learning_rate": 6.482425736466724e-05,
+      "loss": 2.4219,
+      "step": 2122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9587200880050659,
+      "learning_rate": 6.479347526929957e-05,
+      "loss": 2.3016,
+      "step": 2123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1549015045166016,
+      "learning_rate": 6.476268702784226e-05,
+      "loss": 2.5847,
+      "step": 2124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9411564469337463,
+      "learning_rate": 6.473189265308657e-05,
+      "loss": 2.1017,
+      "step": 2125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9558781385421753,
+      "learning_rate": 6.470109215782631e-05,
+      "loss": 2.1337,
+      "step": 2126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9888544082641602,
+      "learning_rate": 6.467028555485784e-05,
+      "loss": 2.335,
+      "step": 2127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.130926251411438,
+      "learning_rate": 6.463947285698007e-05,
+      "loss": 2.1866,
+      "step": 2128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2253402471542358,
+      "learning_rate": 6.460865407699439e-05,
+      "loss": 2.3034,
+      "step": 2129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.040682077407837,
+      "learning_rate": 6.457782922770481e-05,
+      "loss": 2.2598,
+      "step": 2130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.938464879989624,
+      "learning_rate": 6.454699832191776e-05,
+      "loss": 2.1551,
+      "step": 2131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2095153331756592,
+      "learning_rate": 6.451616137244225e-05,
+      "loss": 2.4554,
+      "step": 2132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1014858484268188,
+      "learning_rate": 6.448531839208978e-05,
+      "loss": 2.2598,
+      "step": 2133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0924184322357178,
+      "learning_rate": 6.445446939367436e-05,
+      "loss": 2.4438,
+      "step": 2134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7775828242301941,
+      "learning_rate": 6.44236143900125e-05,
+      "loss": 2.0983,
+      "step": 2135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0272600650787354,
+      "learning_rate": 6.43927533939232e-05,
+      "loss": 2.116,
+      "step": 2136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1159563064575195,
+      "learning_rate": 6.436188641822795e-05,
+      "loss": 2.3178,
+      "step": 2137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2411421537399292,
+      "learning_rate": 6.433101347575075e-05,
+      "loss": 2.4551,
+      "step": 2138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.978100597858429,
+      "learning_rate": 6.430013457931804e-05,
+      "loss": 2.2784,
+      "step": 2139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.108717679977417,
+      "learning_rate": 6.426924974175874e-05,
+      "loss": 2.2799,
+      "step": 2140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9828939437866211,
+      "learning_rate": 6.423835897590426e-05,
+      "loss": 2.1162,
+      "step": 2141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.111527442932129,
+      "learning_rate": 6.420746229458848e-05,
+      "loss": 2.3411,
+      "step": 2142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0890864133834839,
+      "learning_rate": 6.417655971064769e-05,
+      "loss": 2.1445,
+      "step": 2143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.412643313407898,
+      "learning_rate": 6.414565123692068e-05,
+      "loss": 2.4563,
+      "step": 2144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0636101961135864,
+      "learning_rate": 6.411473688624865e-05,
+      "loss": 2.3016,
+      "step": 2145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1218340396881104,
+      "learning_rate": 6.408381667147527e-05,
+      "loss": 2.1278,
+      "step": 2146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2945888042449951,
+      "learning_rate": 6.405289060544665e-05,
+      "loss": 2.3813,
+      "step": 2147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1130119562149048,
+      "learning_rate": 6.40219587010113e-05,
+      "loss": 2.5307,
+      "step": 2148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0865347385406494,
+      "learning_rate": 6.399102097102019e-05,
+      "loss": 2.3317,
+      "step": 2149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9397939443588257,
+      "learning_rate": 6.396007742832667e-05,
+      "loss": 2.3292,
+      "step": 2150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0613363981246948,
+      "learning_rate": 6.392912808578656e-05,
+      "loss": 2.3713,
+      "step": 2151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.963397204875946,
+      "learning_rate": 6.389817295625802e-05,
+      "loss": 2.1882,
+      "step": 2152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9702935218811035,
+      "learning_rate": 6.386721205260166e-05,
+      "loss": 2.4187,
+      "step": 2153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9399896264076233,
+      "learning_rate": 6.383624538768052e-05,
+      "loss": 2.1147,
+      "step": 2154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.100060224533081,
+      "learning_rate": 6.380527297435996e-05,
+      "loss": 2.3605,
+      "step": 2155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9859282970428467,
+      "learning_rate": 6.377429482550777e-05,
+      "loss": 2.2774,
+      "step": 2156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.968751072883606,
+      "learning_rate": 6.374331095399412e-05,
+      "loss": 2.2791,
+      "step": 2157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9888980388641357,
+      "learning_rate": 6.371232137269154e-05,
+      "loss": 2.2881,
+      "step": 2158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9797753691673279,
+      "learning_rate": 6.368132609447495e-05,
+      "loss": 2.518,
+      "step": 2159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0205283164978027,
+      "learning_rate": 6.365032513222164e-05,
+      "loss": 2.2968,
+      "step": 2160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.107801914215088,
+      "learning_rate": 6.361931849881127e-05,
+      "loss": 2.3751,
+      "step": 2161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.050668478012085,
+      "learning_rate": 6.35883062071258e-05,
+      "loss": 2.1868,
+      "step": 2162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9591382145881653,
+      "learning_rate": 6.355728827004961e-05,
+      "loss": 2.2224,
+      "step": 2163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0764795541763306,
+      "learning_rate": 6.352626470046939e-05,
+      "loss": 2.4608,
+      "step": 2164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9811908602714539,
+      "learning_rate": 6.349523551127417e-05,
+      "loss": 2.4133,
+      "step": 2165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0333083868026733,
+      "learning_rate": 6.346420071535533e-05,
+      "loss": 2.2537,
+      "step": 2166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9955871105194092,
+      "learning_rate": 6.343316032560658e-05,
+      "loss": 2.1028,
+      "step": 2167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.000325083732605,
+      "learning_rate": 6.34021143549239e-05,
+      "loss": 2.1417,
+      "step": 2168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9159572124481201,
+      "learning_rate": 6.337106281620569e-05,
+      "loss": 2.1306,
+      "step": 2169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0105916261672974,
+      "learning_rate": 6.334000572235256e-05,
+      "loss": 2.2855,
+      "step": 2170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9785203337669373,
+      "learning_rate": 6.33089430862675e-05,
+      "loss": 2.5957,
+      "step": 2171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0428968667984009,
+      "learning_rate": 6.327787492085577e-05,
+      "loss": 2.1414,
+      "step": 2172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0057944059371948,
+      "learning_rate": 6.324680123902495e-05,
+      "loss": 2.2187,
+      "step": 2173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9485662579536438,
+      "learning_rate": 6.321572205368488e-05,
+      "loss": 2.1216,
+      "step": 2174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9571912884712219,
+      "learning_rate": 6.318463737774769e-05,
+      "loss": 2.3854,
+      "step": 2175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9733611345291138,
+      "learning_rate": 6.315354722412782e-05,
+      "loss": 2.1331,
+      "step": 2176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8365288972854614,
+      "learning_rate": 6.312245160574198e-05,
+      "loss": 1.9553,
+      "step": 2177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.219465970993042,
+      "learning_rate": 6.30913505355091e-05,
+      "loss": 2.5366,
+      "step": 2178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1044403314590454,
+      "learning_rate": 6.306024402635047e-05,
+      "loss": 2.7092,
+      "step": 2179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0588257312774658,
+      "learning_rate": 6.302913209118953e-05,
+      "loss": 2.2548,
+      "step": 2180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1325033903121948,
+      "learning_rate": 6.299801474295207e-05,
+      "loss": 2.1904,
+      "step": 2181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0532841682434082,
+      "learning_rate": 6.296689199456605e-05,
+      "loss": 2.2866,
+      "step": 2182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1801691055297852,
+      "learning_rate": 6.293576385896175e-05,
+      "loss": 2.3612,
+      "step": 2183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.057409644126892,
+      "learning_rate": 6.290463034907162e-05,
+      "loss": 2.2403,
+      "step": 2184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.038885474205017,
+      "learning_rate": 6.287349147783039e-05,
+      "loss": 2.1227,
+      "step": 2185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9879895448684692,
+      "learning_rate": 6.2842347258175e-05,
+      "loss": 2.206,
+      "step": 2186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9924969673156738,
+      "learning_rate": 6.28111977030446e-05,
+      "loss": 2.0357,
+      "step": 2187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0205665826797485,
+      "learning_rate": 6.278004282538056e-05,
+      "loss": 2.3217,
+      "step": 2188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0260761976242065,
+      "learning_rate": 6.27488826381265e-05,
+      "loss": 2.2149,
+      "step": 2189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0987837314605713,
+      "learning_rate": 6.271771715422819e-05,
+      "loss": 2.6002,
+      "step": 2190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0560429096221924,
+      "learning_rate": 6.268654638663364e-05,
+      "loss": 2.4213,
+      "step": 2191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8696759939193726,
+      "learning_rate": 6.265537034829302e-05,
+      "loss": 2.0358,
+      "step": 2192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0081989765167236,
+      "learning_rate": 6.262418905215875e-05,
+      "loss": 2.2689,
+      "step": 2193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0428459644317627,
+      "learning_rate": 6.259300251118537e-05,
+      "loss": 1.908,
+      "step": 2194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9351733326911926,
+      "learning_rate": 6.256181073832959e-05,
+      "loss": 2.2441,
+      "step": 2195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1892249584197998,
+      "learning_rate": 6.25306137465504e-05,
+      "loss": 2.3864,
+      "step": 2196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1206074953079224,
+      "learning_rate": 6.249941154880882e-05,
+      "loss": 2.3522,
+      "step": 2197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0860079526901245,
+      "learning_rate": 6.246820415806815e-05,
+      "loss": 2.0551,
+      "step": 2198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.082932949066162,
+      "learning_rate": 6.243699158729376e-05,
+      "loss": 2.3076,
+      "step": 2199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.030318021774292,
+      "learning_rate": 6.240577384945324e-05,
+      "loss": 2.1117,
+      "step": 2200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8953393697738647,
+      "learning_rate": 6.237455095751626e-05,
+      "loss": 2.1186,
+      "step": 2201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0202581882476807,
+      "learning_rate": 6.234332292445471e-05,
+      "loss": 2.3063,
+      "step": 2202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1890692710876465,
+      "learning_rate": 6.231208976324252e-05,
+      "loss": 2.4445,
+      "step": 2203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1805052757263184,
+      "learning_rate": 6.228085148685586e-05,
+      "loss": 2.4577,
+      "step": 2204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0207278728485107,
+      "learning_rate": 6.224960810827292e-05,
+      "loss": 2.2037,
+      "step": 2205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1782259941101074,
+      "learning_rate": 6.221835964047408e-05,
+      "loss": 2.405,
+      "step": 2206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0606392621994019,
+      "learning_rate": 6.21871060964418e-05,
+      "loss": 2.1647,
+      "step": 2207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.914368212223053,
+      "learning_rate": 6.21558474891607e-05,
+      "loss": 2.2601,
+      "step": 2208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9489256739616394,
+      "learning_rate": 6.212458383161741e-05,
+      "loss": 2.2066,
+      "step": 2209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9632739424705505,
+      "learning_rate": 6.209331513680075e-05,
+      "loss": 2.2714,
+      "step": 2210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9624541997909546,
+      "learning_rate": 6.206204141770159e-05,
+      "loss": 2.1933,
+      "step": 2211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9768065214157104,
+      "learning_rate": 6.203076268731288e-05,
+      "loss": 2.2703,
+      "step": 2212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1270946264266968,
+      "learning_rate": 6.199947895862969e-05,
+      "loss": 2.2695,
+      "step": 2213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9200823903083801,
+      "learning_rate": 6.196819024464911e-05,
+      "loss": 2.2697,
+      "step": 2214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1265591382980347,
+      "learning_rate": 6.193689655837036e-05,
+      "loss": 2.3851,
+      "step": 2215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9828023314476013,
+      "learning_rate": 6.190559791279467e-05,
+      "loss": 2.2968,
+      "step": 2216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0357253551483154,
+      "learning_rate": 6.18742943209254e-05,
+      "loss": 2.4802,
+      "step": 2217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.057938575744629,
+      "learning_rate": 6.184298579576789e-05,
+      "loss": 2.2558,
+      "step": 2218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9825299382209778,
+      "learning_rate": 6.181167235032958e-05,
+      "loss": 2.1259,
+      "step": 2219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0658671855926514,
+      "learning_rate": 6.178035399761992e-05,
+      "loss": 2.3834,
+      "step": 2220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.099267601966858,
+      "learning_rate": 6.174903075065043e-05,
+      "loss": 2.2984,
+      "step": 2221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9123343229293823,
+      "learning_rate": 6.171770262243466e-05,
+      "loss": 2.2955,
+      "step": 2222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9118330478668213,
+      "learning_rate": 6.168636962598817e-05,
+      "loss": 2.0882,
+      "step": 2223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0769330263137817,
+      "learning_rate": 6.165503177432855e-05,
+      "loss": 2.1853,
+      "step": 2224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0480010509490967,
+      "learning_rate": 6.16236890804754e-05,
+      "loss": 2.3912,
+      "step": 2225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1588715314865112,
+      "learning_rate": 6.159234155745035e-05,
+      "loss": 2.3656,
+      "step": 2226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.136955976486206,
+      "learning_rate": 6.156098921827702e-05,
+      "loss": 2.5085,
+      "step": 2227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0678693056106567,
+      "learning_rate": 6.152963207598103e-05,
+      "loss": 2.2109,
+      "step": 2228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0059723854064941,
+      "learning_rate": 6.149827014359002e-05,
+      "loss": 2.0969,
+      "step": 2229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9654774069786072,
+      "learning_rate": 6.14669034341336e-05,
+      "loss": 2.2956,
+      "step": 2230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1771332025527954,
+      "learning_rate": 6.143553196064334e-05,
+      "loss": 2.3485,
+      "step": 2231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1909273862838745,
+      "learning_rate": 6.140415573615285e-05,
+      "loss": 2.2605,
+      "step": 2232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.138645052909851,
+      "learning_rate": 6.137277477369767e-05,
+      "loss": 2.4658,
+      "step": 2233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1211977005004883,
+      "learning_rate": 6.134138908631531e-05,
+      "loss": 2.1222,
+      "step": 2234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1403050422668457,
+      "learning_rate": 6.130999868704526e-05,
+      "loss": 2.1357,
+      "step": 2235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.957892894744873,
+      "learning_rate": 6.127860358892895e-05,
+      "loss": 2.4389,
+      "step": 2236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2669246196746826,
+      "learning_rate": 6.124720380500979e-05,
+      "loss": 2.4124,
+      "step": 2237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0976136922836304,
+      "learning_rate": 6.12157993483331e-05,
+      "loss": 2.2964,
+      "step": 2238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9335986971855164,
+      "learning_rate": 6.118439023194615e-05,
+      "loss": 2.0483,
+      "step": 2239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9235994219779968,
+      "learning_rate": 6.11529764688982e-05,
+      "loss": 2.1601,
+      "step": 2240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0277663469314575,
+      "learning_rate": 6.112155807224034e-05,
+      "loss": 2.3137,
+      "step": 2241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.009811520576477,
+      "learning_rate": 6.109013505502566e-05,
+      "loss": 2.1311,
+      "step": 2242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9442591667175293,
+      "learning_rate": 6.105870743030918e-05,
+      "loss": 2.0179,
+      "step": 2243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8638668656349182,
+      "learning_rate": 6.102727521114777e-05,
+      "loss": 2.2283,
+      "step": 2244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9701653718948364,
+      "learning_rate": 6.099583841060026e-05,
+      "loss": 2.3022,
+      "step": 2245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0524864196777344,
+      "learning_rate": 6.0964397041727384e-05,
+      "loss": 2.2887,
+      "step": 2246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9672285318374634,
+      "learning_rate": 6.093295111759171e-05,
+      "loss": 2.1625,
+      "step": 2247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9884511828422546,
+      "learning_rate": 6.090150065125778e-05,
+      "loss": 2.2552,
+      "step": 2248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.993460476398468,
+      "learning_rate": 6.087004565579197e-05,
+      "loss": 2.2693,
+      "step": 2249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.037294864654541,
+      "learning_rate": 6.083858614426259e-05,
+      "loss": 2.2251,
+      "step": 2250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9611127376556396,
+      "learning_rate": 6.080712212973975e-05,
+      "loss": 2.3056,
+      "step": 2251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.182281494140625,
+      "learning_rate": 6.077565362529549e-05,
+      "loss": 2.3599,
+      "step": 2252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9786458015441895,
+      "learning_rate": 6.074418064400371e-05,
+      "loss": 2.3954,
+      "step": 2253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0382485389709473,
+      "learning_rate": 6.071270319894015e-05,
+      "loss": 2.1769,
+      "step": 2254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0950396060943604,
+      "learning_rate": 6.0681221303182415e-05,
+      "loss": 2.4015,
+      "step": 2255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0789319276809692,
+      "learning_rate": 6.0649734969809956e-05,
+      "loss": 2.1811,
+      "step": 2256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0877901315689087,
+      "learning_rate": 6.0618244211904064e-05,
+      "loss": 2.3044,
+      "step": 2257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0584704875946045,
+      "learning_rate": 6.058674904254788e-05,
+      "loss": 2.052,
+      "step": 2258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0548897981643677,
+      "learning_rate": 6.055524947482638e-05,
+      "loss": 2.0956,
+      "step": 2259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1302752494812012,
+      "learning_rate": 6.0523745521826324e-05,
+      "loss": 2.3242,
+      "step": 2260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1218180656433105,
+      "learning_rate": 6.049223719663637e-05,
+      "loss": 2.2948,
+      "step": 2261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2134108543395996,
+      "learning_rate": 6.0460724512346914e-05,
+      "loss": 2.3752,
+      "step": 2262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9019377827644348,
+      "learning_rate": 6.042920748205022e-05,
+      "loss": 2.0318,
+      "step": 2263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1271978616714478,
+      "learning_rate": 6.039768611884034e-05,
+      "loss": 2.1899,
+      "step": 2264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0544445514678955,
+      "learning_rate": 6.036616043581311e-05,
+      "loss": 2.5069,
+      "step": 2265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9965691566467285,
+      "learning_rate": 6.03346304460662e-05,
+      "loss": 2.4829,
+      "step": 2266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0951404571533203,
+      "learning_rate": 6.0303096162699015e-05,
+      "loss": 2.4129,
+      "step": 2267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0896196365356445,
+      "learning_rate": 6.027155759881279e-05,
+      "loss": 2.1629,
+      "step": 2268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.966179609298706,
+      "learning_rate": 6.0240014767510497e-05,
+      "loss": 2.3305,
+      "step": 2269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0358151197433472,
+      "learning_rate": 6.020846768189693e-05,
+      "loss": 2.2176,
+      "step": 2270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0353220701217651,
+      "learning_rate": 6.0176916355078604e-05,
+      "loss": 2.2188,
+      "step": 2271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0015580654144287,
+      "learning_rate": 6.014536080016384e-05,
+      "loss": 2.1984,
+      "step": 2272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0143706798553467,
+      "learning_rate": 6.0113801030262674e-05,
+      "loss": 2.3246,
+      "step": 2273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9529539346694946,
+      "learning_rate": 6.008223705848692e-05,
+      "loss": 2.3242,
+      "step": 2274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0235626697540283,
+      "learning_rate": 6.0050668897950134e-05,
+      "loss": 2.3329,
+      "step": 2275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0063844919204712,
+      "learning_rate": 6.0019096561767604e-05,
+      "loss": 2.2139,
+      "step": 2276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0131332874298096,
+      "learning_rate": 5.998752006305635e-05,
+      "loss": 2.1222,
+      "step": 2277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8340829610824585,
+      "learning_rate": 5.9955939414935135e-05,
+      "loss": 1.969,
+      "step": 2278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1093595027923584,
+      "learning_rate": 5.992435463052444e-05,
+      "loss": 2.4252,
+      "step": 2279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0108311176300049,
+      "learning_rate": 5.9892765722946464e-05,
+      "loss": 2.0922,
+      "step": 2280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1923776865005493,
+      "learning_rate": 5.9861172705325106e-05,
+      "loss": 2.1947,
+      "step": 2281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1983928680419922,
+      "learning_rate": 5.982957559078599e-05,
+      "loss": 2.3237,
+      "step": 2282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0826976299285889,
+      "learning_rate": 5.979797439245645e-05,
+      "loss": 2.3686,
+      "step": 2283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0424575805664062,
+      "learning_rate": 5.976636912346548e-05,
+      "loss": 2.1655,
+      "step": 2284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.051666498184204,
+      "learning_rate": 5.973475979694382e-05,
+      "loss": 2.2054,
+      "step": 2285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0823357105255127,
+      "learning_rate": 5.970314642602384e-05,
+      "loss": 2.2324,
+      "step": 2286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1843786239624023,
+      "learning_rate": 5.967152902383961e-05,
+      "loss": 2.3624,
+      "step": 2287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1710737943649292,
+      "learning_rate": 5.96399076035269e-05,
+      "loss": 2.4552,
+      "step": 2288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.058386206626892,
+      "learning_rate": 5.960828217822312e-05,
+      "loss": 2.2723,
+      "step": 2289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8874415755271912,
+      "learning_rate": 5.9576652761067374e-05,
+      "loss": 2.3867,
+      "step": 2290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9444225430488586,
+      "learning_rate": 5.954501936520036e-05,
+      "loss": 2.2279,
+      "step": 2291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0734716653823853,
+      "learning_rate": 5.951338200376451e-05,
+      "loss": 2.0972,
+      "step": 2292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0395317077636719,
+      "learning_rate": 5.948174068990384e-05,
+      "loss": 2.1403,
+      "step": 2293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9498971700668335,
+      "learning_rate": 5.9450095436764043e-05,
+      "loss": 1.9603,
+      "step": 2294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0689033269882202,
+      "learning_rate": 5.941844625749245e-05,
+      "loss": 2.5009,
+      "step": 2295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0162070989608765,
+      "learning_rate": 5.938679316523801e-05,
+      "loss": 2.2613,
+      "step": 2296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0734304189682007,
+      "learning_rate": 5.9355136173151284e-05,
+      "loss": 2.2013,
+      "step": 2297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0628132820129395,
+      "learning_rate": 5.932347529438449e-05,
+      "loss": 2.2205,
+      "step": 2298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9910916686058044,
+      "learning_rate": 5.929181054209143e-05,
+      "loss": 2.3721,
+      "step": 2299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8472923040390015,
+      "learning_rate": 5.9260141929427526e-05,
+      "loss": 2.026,
+      "step": 2300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1242260932922363,
+      "learning_rate": 5.9228469469549806e-05,
+      "loss": 2.3974,
+      "step": 2301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1843485832214355,
+      "learning_rate": 5.91967931756169e-05,
+      "loss": 2.2006,
+      "step": 2302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0849052667617798,
+      "learning_rate": 5.916511306078902e-05,
+      "loss": 2.1252,
+      "step": 2303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0051435232162476,
+      "learning_rate": 5.9133429138227955e-05,
+      "loss": 2.2693,
+      "step": 2304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1396090984344482,
+      "learning_rate": 5.910174142109711e-05,
+      "loss": 2.3839,
+      "step": 2305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9908087253570557,
+      "learning_rate": 5.907004992256144e-05,
+      "loss": 2.0956,
+      "step": 2306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1127233505249023,
+      "learning_rate": 5.9038354655787485e-05,
+      "loss": 2.468,
+      "step": 2307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0172871351242065,
+      "learning_rate": 5.900665563394334e-05,
+      "loss": 2.1868,
+      "step": 2308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0174964666366577,
+      "learning_rate": 5.897495287019866e-05,
+      "loss": 2.2764,
+      "step": 2309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0883628129959106,
+      "learning_rate": 5.894324637772466e-05,
+      "loss": 2.1508,
+      "step": 2310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.084130048751831,
+      "learning_rate": 5.891153616969411e-05,
+      "loss": 2.3899,
+      "step": 2311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9476576447486877,
+      "learning_rate": 5.8879822259281324e-05,
+      "loss": 2.212,
+      "step": 2312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0509998798370361,
+      "learning_rate": 5.884810465966213e-05,
+      "loss": 2.5159,
+      "step": 2313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9988600611686707,
+      "learning_rate": 5.8816383384013915e-05,
+      "loss": 2.1514,
+      "step": 2314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0895625352859497,
+      "learning_rate": 5.878465844551557e-05,
+      "loss": 2.4518,
+      "step": 2315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9921971559524536,
+      "learning_rate": 5.8752929857347536e-05,
+      "loss": 2.3393,
+      "step": 2316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1141127347946167,
+      "learning_rate": 5.872119763269176e-05,
+      "loss": 2.2197,
+      "step": 2317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.071019172668457,
+      "learning_rate": 5.868946178473168e-05,
+      "loss": 2.2156,
+      "step": 2318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1200246810913086,
+      "learning_rate": 5.8657722326652265e-05,
+      "loss": 2.5677,
+      "step": 2319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.980599045753479,
+      "learning_rate": 5.8625979271639976e-05,
+      "loss": 2.1073,
+      "step": 2320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0579314231872559,
+      "learning_rate": 5.859423263288276e-05,
+      "loss": 2.1392,
+      "step": 2321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1410402059555054,
+      "learning_rate": 5.856248242357008e-05,
+      "loss": 2.3255,
+      "step": 2322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1488001346588135,
+      "learning_rate": 5.8530728656892816e-05,
+      "loss": 2.2384,
+      "step": 2323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.212266206741333,
+      "learning_rate": 5.849897134604341e-05,
+      "loss": 2.3695,
+      "step": 2324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0283424854278564,
+      "learning_rate": 5.8467210504215744e-05,
+      "loss": 2.4603,
+      "step": 2325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.954558253288269,
+      "learning_rate": 5.8435446144605125e-05,
+      "loss": 2.1858,
+      "step": 2326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9829445481300354,
+      "learning_rate": 5.840367828040837e-05,
+      "loss": 2.0767,
+      "step": 2327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0847017765045166,
+      "learning_rate": 5.837190692482376e-05,
+      "loss": 2.1788,
+      "step": 2328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.045385718345642,
+      "learning_rate": 5.834013209105098e-05,
+      "loss": 2.1675,
+      "step": 2329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0157382488250732,
+      "learning_rate": 5.830835379229122e-05,
+      "loss": 2.253,
+      "step": 2330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0601003170013428,
+      "learning_rate": 5.8276572041747036e-05,
+      "loss": 2.3814,
+      "step": 2331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.965325117111206,
+      "learning_rate": 5.824478685262248e-05,
+      "loss": 2.4011,
+      "step": 2332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1596111059188843,
+      "learning_rate": 5.8212998238123006e-05,
+      "loss": 2.4676,
+      "step": 2333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1549735069274902,
+      "learning_rate": 5.818120621145551e-05,
+      "loss": 2.2352,
+      "step": 2334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0137028694152832,
+      "learning_rate": 5.8149410785828274e-05,
+      "loss": 2.3138,
+      "step": 2335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9978339076042175,
+      "learning_rate": 5.8117611974451e-05,
+      "loss": 2.3959,
+      "step": 2336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0896414518356323,
+      "learning_rate": 5.8085809790534816e-05,
+      "loss": 2.2739,
+      "step": 2337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2692291736602783,
+      "learning_rate": 5.805400424729226e-05,
+      "loss": 2.2984,
+      "step": 2338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.085280179977417,
+      "learning_rate": 5.802219535793721e-05,
+      "loss": 2.3384,
+      "step": 2339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9687846302986145,
+      "learning_rate": 5.799038313568501e-05,
+      "loss": 2.2798,
+      "step": 2340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9319024085998535,
+      "learning_rate": 5.795856759375232e-05,
+      "loss": 2.2836,
+      "step": 2341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0726532936096191,
+      "learning_rate": 5.792674874535722e-05,
+      "loss": 2.026,
+      "step": 2342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.102683424949646,
+      "learning_rate": 5.7894926603719147e-05,
+      "loss": 2.1784,
+      "step": 2343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2278518676757812,
+      "learning_rate": 5.786310118205893e-05,
+      "loss": 2.3606,
+      "step": 2344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.071347713470459,
+      "learning_rate": 5.783127249359872e-05,
+      "loss": 2.1268,
+      "step": 2345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9608170390129089,
+      "learning_rate": 5.779944055156207e-05,
+      "loss": 2.3143,
+      "step": 2346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1765340566635132,
+      "learning_rate": 5.776760536917383e-05,
+      "loss": 2.456,
+      "step": 2347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.061861515045166,
+      "learning_rate": 5.7735766959660266e-05,
+      "loss": 2.383,
+      "step": 2348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0791264772415161,
+      "learning_rate": 5.770392533624891e-05,
+      "loss": 2.2218,
+      "step": 2349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0799822807312012,
+      "learning_rate": 5.767208051216869e-05,
+      "loss": 2.2321,
+      "step": 2350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.113734483718872,
+      "learning_rate": 5.7640232500649826e-05,
+      "loss": 2.3321,
+      "step": 2351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3395144939422607,
+      "learning_rate": 5.76083813149239e-05,
+      "loss": 2.4012,
+      "step": 2352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.186967372894287,
+      "learning_rate": 5.757652696822376e-05,
+      "loss": 2.3551,
+      "step": 2353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0989201068878174,
+      "learning_rate": 5.754466947378363e-05,
+      "loss": 2.2941,
+      "step": 2354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.06804358959198,
+      "learning_rate": 5.751280884483898e-05,
+      "loss": 2.2519,
+      "step": 2355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0938410758972168,
+      "learning_rate": 5.748094509462663e-05,
+      "loss": 2.3842,
+      "step": 2356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0626455545425415,
+      "learning_rate": 5.744907823638465e-05,
+      "loss": 2.214,
+      "step": 2357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.05101478099823,
+      "learning_rate": 5.741720828335246e-05,
+      "loss": 2.1299,
+      "step": 2358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9704009890556335,
+      "learning_rate": 5.738533524877072e-05,
+      "loss": 2.2309,
+      "step": 2359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0254722833633423,
+      "learning_rate": 5.735345914588138e-05,
+      "loss": 2.2032,
+      "step": 2360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1339313983917236,
+      "learning_rate": 5.7321579987927675e-05,
+      "loss": 2.3066,
+      "step": 2361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0867952108383179,
+      "learning_rate": 5.728969778815411e-05,
+      "loss": 2.1945,
+      "step": 2362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.147448182106018,
+      "learning_rate": 5.7257812559806425e-05,
+      "loss": 2.203,
+      "step": 2363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1121718883514404,
+      "learning_rate": 5.7225924316131665e-05,
+      "loss": 2.332,
+      "step": 2364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8399686217308044,
+      "learning_rate": 5.71940330703781e-05,
+      "loss": 2.1474,
+      "step": 2365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1496496200561523,
+      "learning_rate": 5.716213883579524e-05,
+      "loss": 2.3638,
+      "step": 2366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.031838059425354,
+      "learning_rate": 5.7130241625633866e-05,
+      "loss": 1.9089,
+      "step": 2367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9755390882492065,
+      "learning_rate": 5.7098341453145944e-05,
+      "loss": 2.2843,
+      "step": 2368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1183704137802124,
+      "learning_rate": 5.7066438331584725e-05,
+      "loss": 2.5521,
+      "step": 2369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9158876538276672,
+      "learning_rate": 5.7034532274204675e-05,
+      "loss": 2.01,
+      "step": 2370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0092593431472778,
+      "learning_rate": 5.7002623294261446e-05,
+      "loss": 2.1665,
+      "step": 2371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.059857964515686,
+      "learning_rate": 5.697071140501193e-05,
+      "loss": 2.3088,
+      "step": 2372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.133827567100525,
+      "learning_rate": 5.693879661971423e-05,
+      "loss": 2.3299,
+      "step": 2373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1038013696670532,
+      "learning_rate": 5.690687895162764e-05,
+      "loss": 2.3106,
+      "step": 2374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9750762581825256,
+      "learning_rate": 5.687495841401267e-05,
+      "loss": 2.2287,
+      "step": 2375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1319153308868408,
+      "learning_rate": 5.6843035020131e-05,
+      "loss": 2.2303,
+      "step": 2376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.06272292137146,
+      "learning_rate": 5.6811108783245516e-05,
+      "loss": 2.2731,
+      "step": 2377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.021862506866455,
+      "learning_rate": 5.677917971662027e-05,
+      "loss": 2.292,
+      "step": 2378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0783511400222778,
+      "learning_rate": 5.674724783352049e-05,
+      "loss": 2.094,
+      "step": 2379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0759806632995605,
+      "learning_rate": 5.671531314721259e-05,
+      "loss": 2.11,
+      "step": 2380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9890070557594299,
+      "learning_rate": 5.668337567096414e-05,
+      "loss": 2.0527,
+      "step": 2381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1057968139648438,
+      "learning_rate": 5.665143541804384e-05,
+      "loss": 2.2269,
+      "step": 2382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.947693407535553,
+      "learning_rate": 5.6619492401721605e-05,
+      "loss": 2.2065,
+      "step": 2383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.07021164894104,
+      "learning_rate": 5.658754663526845e-05,
+      "loss": 2.1877,
+      "step": 2384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1083215475082397,
+      "learning_rate": 5.655559813195656e-05,
+      "loss": 2.3212,
+      "step": 2385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1732569932937622,
+      "learning_rate": 5.6523646905059225e-05,
+      "loss": 2.2247,
+      "step": 2386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.088223934173584,
+      "learning_rate": 5.649169296785089e-05,
+      "loss": 2.2489,
+      "step": 2387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0008341073989868,
+      "learning_rate": 5.645973633360713e-05,
+      "loss": 2.1983,
+      "step": 2388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0112793445587158,
+      "learning_rate": 5.642777701560462e-05,
+      "loss": 2.3326,
+      "step": 2389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0483134984970093,
+      "learning_rate": 5.639581502712117e-05,
+      "loss": 2.2066,
+      "step": 2390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.039129376411438,
+      "learning_rate": 5.636385038143568e-05,
+      "loss": 2.2838,
+      "step": 2391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0869569778442383,
+      "learning_rate": 5.633188309182817e-05,
+      "loss": 2.4684,
+      "step": 2392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0876009464263916,
+      "learning_rate": 5.6299913171579786e-05,
+      "loss": 2.1016,
+      "step": 2393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0779485702514648,
+      "learning_rate": 5.626794063397269e-05,
+      "loss": 2.3154,
+      "step": 2394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9571425914764404,
+      "learning_rate": 5.62359654922902e-05,
+      "loss": 2.0209,
+      "step": 2395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1168889999389648,
+      "learning_rate": 5.6203987759816665e-05,
+      "loss": 2.2624,
+      "step": 2396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0143100023269653,
+      "learning_rate": 5.617200744983757e-05,
+      "loss": 2.1169,
+      "step": 2397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0126361846923828,
+      "learning_rate": 5.614002457563943e-05,
+      "loss": 2.3257,
+      "step": 2398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0071935653686523,
+      "learning_rate": 5.610803915050982e-05,
+      "loss": 2.0275,
+      "step": 2399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0726008415222168,
+      "learning_rate": 5.60760511877374e-05,
+      "loss": 2.25,
+      "step": 2400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1387009620666504,
+      "learning_rate": 5.604406070061187e-05,
+      "loss": 2.259,
+      "step": 2401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.189996600151062,
+      "learning_rate": 5.601206770242399e-05,
+      "loss": 2.2586,
+      "step": 2402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0166211128234863,
+      "learning_rate": 5.598007220646554e-05,
+      "loss": 2.1121,
+      "step": 2403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2152832746505737,
+      "learning_rate": 5.5948074226029356e-05,
+      "loss": 2.3188,
+      "step": 2404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1469526290893555,
+      "learning_rate": 5.59160737744093e-05,
+      "loss": 2.5087,
+      "step": 2405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9470880627632141,
+      "learning_rate": 5.5884070864900274e-05,
+      "loss": 2.1673,
+      "step": 2406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9345309138298035,
+      "learning_rate": 5.585206551079818e-05,
+      "loss": 2.126,
+      "step": 2407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9392812252044678,
+      "learning_rate": 5.5820057725399965e-05,
+      "loss": 2.1231,
+      "step": 2408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0305763483047485,
+      "learning_rate": 5.578804752200354e-05,
+      "loss": 2.2998,
+      "step": 2409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0777441263198853,
+      "learning_rate": 5.5756034913907865e-05,
+      "loss": 2.3491,
+      "step": 2410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8323517441749573,
+      "learning_rate": 5.572401991441286e-05,
+      "loss": 2.1724,
+      "step": 2411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0944236516952515,
+      "learning_rate": 5.569200253681949e-05,
+      "loss": 2.1548,
+      "step": 2412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9519125819206238,
+      "learning_rate": 5.565998279442966e-05,
+      "loss": 2.1934,
+      "step": 2413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1495665311813354,
+      "learning_rate": 5.562796070054629e-05,
+      "loss": 2.261,
+      "step": 2414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1793828010559082,
+      "learning_rate": 5.559593626847326e-05,
+      "loss": 2.324,
+      "step": 2415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9944158792495728,
+      "learning_rate": 5.55639095115154e-05,
+      "loss": 2.1598,
+      "step": 2416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8898161053657532,
+      "learning_rate": 5.5531880442978554e-05,
+      "loss": 2.2561,
+      "step": 2417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1506081819534302,
+      "learning_rate": 5.54998490761695e-05,
+      "loss": 2.3729,
+      "step": 2418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2081480026245117,
+      "learning_rate": 5.5467815424395964e-05,
+      "loss": 2.1348,
+      "step": 2419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0896615982055664,
+      "learning_rate": 5.543577950096664e-05,
+      "loss": 1.9966,
+      "step": 2420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0872379541397095,
+      "learning_rate": 5.540374131919115e-05,
+      "loss": 2.4759,
+      "step": 2421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0616167783737183,
+      "learning_rate": 5.5371700892380066e-05,
+      "loss": 2.1137,
+      "step": 2422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9578679800033569,
+      "learning_rate": 5.533965823384487e-05,
+      "loss": 2.2324,
+      "step": 2423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0651428699493408,
+      "learning_rate": 5.530761335689801e-05,
+      "loss": 2.0321,
+      "step": 2424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2522410154342651,
+      "learning_rate": 5.5275566274852844e-05,
+      "loss": 2.0666,
+      "step": 2425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2052727937698364,
+      "learning_rate": 5.52435170010236e-05,
+      "loss": 2.5533,
+      "step": 2426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0634469985961914,
+      "learning_rate": 5.521146554872548e-05,
+      "loss": 2.1132,
+      "step": 2427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9900693893432617,
+      "learning_rate": 5.517941193127456e-05,
+      "loss": 2.2171,
+      "step": 2428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1087092161178589,
+      "learning_rate": 5.514735616198782e-05,
+      "loss": 2.4248,
+      "step": 2429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0739015340805054,
+      "learning_rate": 5.5115298254183114e-05,
+      "loss": 2.1463,
+      "step": 2430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.215565800666809,
+      "learning_rate": 5.508323822117922e-05,
+      "loss": 2.3747,
+      "step": 2431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1229560375213623,
+      "learning_rate": 5.505117607629579e-05,
+      "loss": 2.241,
+      "step": 2432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.053659439086914,
+      "learning_rate": 5.501911183285333e-05,
+      "loss": 2.0721,
+      "step": 2433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0946427583694458,
+      "learning_rate": 5.498704550417324e-05,
+      "loss": 2.1861,
+      "step": 2434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1437902450561523,
+      "learning_rate": 5.4954977103577786e-05,
+      "loss": 2.361,
+      "step": 2435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1825858354568481,
+      "learning_rate": 5.492290664439009e-05,
+      "loss": 2.3296,
+      "step": 2436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1713136434555054,
+      "learning_rate": 5.489083413993413e-05,
+      "loss": 2.2767,
+      "step": 2437
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0578572750091553,
+      "learning_rate": 5.485875960353472e-05,
+      "loss": 2.165,
+      "step": 2438
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.205674648284912,
+      "learning_rate": 5.4826683048517534e-05,
+      "loss": 2.3075,
+      "step": 2439
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1073272228240967,
+      "learning_rate": 5.479460448820909e-05,
+      "loss": 2.2557,
+      "step": 2440
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9776976704597473,
+      "learning_rate": 5.476252393593672e-05,
+      "loss": 2.1767,
+      "step": 2441
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2096604108810425,
+      "learning_rate": 5.473044140502859e-05,
+      "loss": 2.2069,
+      "step": 2442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0737159252166748,
+      "learning_rate": 5.4698356908813705e-05,
+      "loss": 2.3703,
+      "step": 2443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9930804967880249,
+      "learning_rate": 5.466627046062187e-05,
+      "loss": 2.2059,
+      "step": 2444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1545132398605347,
+      "learning_rate": 5.46341820737837e-05,
+      "loss": 2.3605,
+      "step": 2445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2247344255447388,
+      "learning_rate": 5.4602091761630626e-05,
+      "loss": 2.3471,
+      "step": 2446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0682978630065918,
+      "learning_rate": 5.4569999537494865e-05,
+      "loss": 2.1733,
+      "step": 2447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0409482717514038,
+      "learning_rate": 5.453790541470946e-05,
+      "loss": 2.272,
+      "step": 2448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1454033851623535,
+      "learning_rate": 5.450580940660818e-05,
+      "loss": 2.1302,
+      "step": 2449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0584505796432495,
+      "learning_rate": 5.447371152652564e-05,
+      "loss": 2.2615,
+      "step": 2450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1126294136047363,
+      "learning_rate": 5.444161178779721e-05,
+      "loss": 2.4287,
+      "step": 2451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.235378623008728,
+      "learning_rate": 5.4409510203759016e-05,
+      "loss": 2.257,
+      "step": 2452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.179336428642273,
+      "learning_rate": 5.437740678774796e-05,
+      "loss": 2.3458,
+      "step": 2453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.324066400527954,
+      "learning_rate": 5.4345301553101746e-05,
+      "loss": 2.49,
+      "step": 2454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.969918966293335,
+      "learning_rate": 5.4313194513158774e-05,
+      "loss": 2.2449,
+      "step": 2455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0969445705413818,
+      "learning_rate": 5.4281085681258215e-05,
+      "loss": 2.3004,
+      "step": 2456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.927386999130249,
+      "learning_rate": 5.424897507073999e-05,
+      "loss": 2.2089,
+      "step": 2457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0980873107910156,
+      "learning_rate": 5.421686269494477e-05,
+      "loss": 2.3772,
+      "step": 2458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9382542967796326,
+      "learning_rate": 5.418474856721394e-05,
+      "loss": 2.0598,
+      "step": 2459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.013615369796753,
+      "learning_rate": 5.415263270088962e-05,
+      "loss": 2.3279,
+      "step": 2460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8925917148590088,
+      "learning_rate": 5.412051510931464e-05,
+      "loss": 2.3848,
+      "step": 2461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.033079981803894,
+      "learning_rate": 5.408839580583256e-05,
+      "loss": 2.1921,
+      "step": 2462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0885909795761108,
+      "learning_rate": 5.405627480378767e-05,
+      "loss": 2.4066,
+      "step": 2463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1899850368499756,
+      "learning_rate": 5.402415211652492e-05,
+      "loss": 2.3485,
+      "step": 2464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0921618938446045,
+      "learning_rate": 5.399202775738998e-05,
+      "loss": 2.2417,
+      "step": 2465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.116931438446045,
+      "learning_rate": 5.3959901739729256e-05,
+      "loss": 2.3552,
+      "step": 2466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0450005531311035,
+      "learning_rate": 5.392777407688978e-05,
+      "loss": 2.2754,
+      "step": 2467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.059498906135559,
+      "learning_rate": 5.3895644782219314e-05,
+      "loss": 2.3662,
+      "step": 2468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1072646379470825,
+      "learning_rate": 5.386351386906625e-05,
+      "loss": 2.36,
+      "step": 2469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1296688318252563,
+      "learning_rate": 5.383138135077971e-05,
+      "loss": 2.2185,
+      "step": 2470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1418070793151855,
+      "learning_rate": 5.379924724070943e-05,
+      "loss": 2.2629,
+      "step": 2471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.100908637046814,
+      "learning_rate": 5.376711155220584e-05,
+      "loss": 2.3142,
+      "step": 2472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9671291708946228,
+      "learning_rate": 5.373497429862001e-05,
+      "loss": 2.0687,
+      "step": 2473
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9962815046310425,
+      "learning_rate": 5.370283549330368e-05,
+      "loss": 2.1659,
+      "step": 2474
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1401292085647583,
+      "learning_rate": 5.367069514960922e-05,
+      "loss": 2.3037,
+      "step": 2475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9728932976722717,
+      "learning_rate": 5.363855328088963e-05,
+      "loss": 2.1245,
+      "step": 2476
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1346549987792969,
+      "learning_rate": 5.3606409900498555e-05,
+      "loss": 2.0995,
+      "step": 2477
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0783462524414062,
+      "learning_rate": 5.357426502179027e-05,
+      "loss": 2.2686,
+      "step": 2478
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2617290019989014,
+      "learning_rate": 5.3542118658119676e-05,
+      "loss": 2.3513,
+      "step": 2479
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.017901062965393,
+      "learning_rate": 5.350997082284228e-05,
+      "loss": 1.9973,
+      "step": 2480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0675183534622192,
+      "learning_rate": 5.347782152931421e-05,
+      "loss": 2.0782,
+      "step": 2481
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.144954800605774,
+      "learning_rate": 5.34456707908922e-05,
+      "loss": 2.3724,
+      "step": 2482
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0356035232543945,
+      "learning_rate": 5.3413518620933564e-05,
+      "loss": 2.2949,
+      "step": 2483
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0484577417373657,
+      "learning_rate": 5.338136503279623e-05,
+      "loss": 2.2499,
+      "step": 2484
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0190857648849487,
+      "learning_rate": 5.3349210039838705e-05,
+      "loss": 2.2478,
+      "step": 2485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.062890648841858,
+      "learning_rate": 5.3317053655420105e-05,
+      "loss": 2.2267,
+      "step": 2486
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1329113245010376,
+      "learning_rate": 5.328489589290009e-05,
+      "loss": 2.2082,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0427322387695312,
+      "learning_rate": 5.325273676563891e-05,
+      "loss": 2.1744,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.124898076057434,
+      "learning_rate": 5.322057628699739e-05,
+      "loss": 2.3034,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1518089771270752,
+      "learning_rate": 5.318841447033689e-05,
+      "loss": 2.2121,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1152760982513428,
+      "learning_rate": 5.3156251329019354e-05,
+      "loss": 2.1049,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0527405738830566,
+      "learning_rate": 5.312408687640727e-05,
+      "loss": 2.3949,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1132731437683105,
+      "learning_rate": 5.3091921125863644e-05,
+      "loss": 2.0656,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1509168148040771,
+      "learning_rate": 5.305975409075204e-05,
+      "loss": 1.9041,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2525471448898315,
+      "learning_rate": 5.3027585784436595e-05,
+      "loss": 2.2778,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.062935471534729,
+      "learning_rate": 5.2995416220281904e-05,
+      "loss": 1.9403,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.200419545173645,
+      "learning_rate": 5.296324541165312e-05,
+      "loss": 2.1903,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1227291822433472,
+      "learning_rate": 5.2931073371915916e-05,
+      "loss": 2.0264,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1248441934585571,
+      "learning_rate": 5.289890011443648e-05,
+      "loss": 2.2074,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0719256401062012,
+      "learning_rate": 5.2866725652581486e-05,
+      "loss": 2.2594,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1185513734817505,
+      "learning_rate": 5.283454999971811e-05,
+      "loss": 2.3009,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1492599248886108,
+      "learning_rate": 5.2802373169214056e-05,
+      "loss": 2.212,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.924199104309082,
+      "learning_rate": 5.277019517443749e-05,
+      "loss": 1.9999,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2401320934295654,
+      "learning_rate": 5.2738016028757046e-05,
+      "loss": 2.2762,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0675532817840576,
+      "learning_rate": 5.2705835745541875e-05,
+      "loss": 2.1993,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.102575421333313,
+      "learning_rate": 5.267365433816157e-05,
+      "loss": 2.3064,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.278714895248413,
+      "learning_rate": 5.26414718199862e-05,
+      "loss": 2.325,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.023869514465332,
+      "learning_rate": 5.260928820438631e-05,
+      "loss": 2.1955,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1627167463302612,
+      "learning_rate": 5.257710350473287e-05,
+      "loss": 2.3604,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9298085570335388,
+      "learning_rate": 5.254491773439734e-05,
+      "loss": 2.1662,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9941608309745789,
+      "learning_rate": 5.251273090675158e-05,
+      "loss": 2.1003,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.047812581062317,
+      "learning_rate": 5.248054303516794e-05,
+      "loss": 2.3178,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1716901063919067,
+      "learning_rate": 5.2448354133019164e-05,
+      "loss": 2.4804,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.05693781375885,
+      "learning_rate": 5.241616421367843e-05,
+      "loss": 2.281,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9803346991539001,
+      "learning_rate": 5.238397329051936e-05,
+      "loss": 2.0251,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2060540914535522,
+      "learning_rate": 5.235178137691596e-05,
+      "loss": 2.4718,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1247016191482544,
+      "learning_rate": 5.2319588486242686e-05,
+      "loss": 2.319,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0047340393066406,
+      "learning_rate": 5.2287394631874355e-05,
+      "loss": 2.0385,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2944616079330444,
+      "learning_rate": 5.225519982718623e-05,
+      "loss": 2.2517,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0792415142059326,
+      "learning_rate": 5.222300408555394e-05,
+      "loss": 2.1705,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3378055095672607,
+      "learning_rate": 5.2190807420353504e-05,
+      "loss": 2.3328,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0445810556411743,
+      "learning_rate": 5.2158609844961346e-05,
+      "loss": 2.2425,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0028311014175415,
+      "learning_rate": 5.2126411372754256e-05,
+      "loss": 2.14,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2918621301651,
+      "learning_rate": 5.209421201710938e-05,
+      "loss": 2.1844,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1223700046539307,
+      "learning_rate": 5.2062011791404256e-05,
+      "loss": 2.1989,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.03181791305542,
+      "learning_rate": 5.202981070901677e-05,
+      "loss": 1.9245,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.223561406135559,
+      "learning_rate": 5.199760878332518e-05,
+      "loss": 2.5056,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1539108753204346,
+      "learning_rate": 5.196540602770805e-05,
+      "loss": 2.207,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.003443956375122,
+      "learning_rate": 5.193320245554435e-05,
+      "loss": 2.3866,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1555861234664917,
+      "learning_rate": 5.190099808021335e-05,
+      "loss": 2.3411,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.075884222984314,
+      "learning_rate": 5.186879291509465e-05,
+      "loss": 2.0459,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.035159945487976,
+      "learning_rate": 5.1836586973568224e-05,
+      "loss": 2.2295,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9219068288803101,
+      "learning_rate": 5.1804380269014296e-05,
+      "loss": 2.1858,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0942752361297607,
+      "learning_rate": 5.177217281481348e-05,
+      "loss": 1.9814,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9139752984046936,
+      "learning_rate": 5.173996462434665e-05,
+      "loss": 2.2228,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1825475692749023,
+      "learning_rate": 5.170775571099502e-05,
+      "loss": 2.3134,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.250906229019165,
+      "learning_rate": 5.167554608814007e-05,
+      "loss": 2.0547,
+      "step": 2537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2897554636001587,
+      "learning_rate": 5.1643335769163606e-05,
+      "loss": 2.4089,
+      "step": 2538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1308863162994385,
+      "learning_rate": 5.1611124767447693e-05,
+      "loss": 2.0957,
+      "step": 2539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.028605580329895,
+      "learning_rate": 5.157891309637473e-05,
+      "loss": 2.198,
+      "step": 2540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.141920804977417,
+      "learning_rate": 5.154670076932733e-05,
+      "loss": 2.2578,
+      "step": 2541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0863796472549438,
+      "learning_rate": 5.1514487799688414e-05,
+      "loss": 1.8969,
+      "step": 2542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1235265731811523,
+      "learning_rate": 5.1482274200841184e-05,
+      "loss": 2.2934,
+      "step": 2543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0120298862457275,
+      "learning_rate": 5.145005998616908e-05,
+      "loss": 2.1713,
+      "step": 2544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.190248966217041,
+      "learning_rate": 5.141784516905578e-05,
+      "loss": 2.1445,
+      "step": 2545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9952515959739685,
+      "learning_rate": 5.138562976288527e-05,
+      "loss": 2.2495,
+      "step": 2546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.042124629020691,
+      "learning_rate": 5.135341378104171e-05,
+      "loss": 2.2606,
+      "step": 2547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.221734881401062,
+      "learning_rate": 5.132119723690957e-05,
+      "loss": 2.1044,
+      "step": 2548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0093833208084106,
+      "learning_rate": 5.12889801438735e-05,
+      "loss": 2.4603,
+      "step": 2549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.140610694885254,
+      "learning_rate": 5.1256762515318416e-05,
+      "loss": 2.2676,
+      "step": 2550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0337181091308594,
+      "learning_rate": 5.12245443646294e-05,
+      "loss": 2.2202,
+      "step": 2551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2382378578186035,
+      "learning_rate": 5.119232570519182e-05,
+      "loss": 2.2168,
+      "step": 2552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.086612343788147,
+      "learning_rate": 5.116010655039121e-05,
+      "loss": 2.2495,
+      "step": 2553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1526801586151123,
+      "learning_rate": 5.112788691361333e-05,
+      "loss": 2.1882,
+      "step": 2554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9488018155097961,
+      "learning_rate": 5.1095666808244123e-05,
+      "loss": 2.0735,
+      "step": 2555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0479086637496948,
+      "learning_rate": 5.106344624766975e-05,
+      "loss": 2.2217,
+      "step": 2556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2198848724365234,
+      "learning_rate": 5.103122524527656e-05,
+      "loss": 2.1791,
+      "step": 2557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0002027750015259,
+      "learning_rate": 5.099900381445105e-05,
+      "loss": 2.1593,
+      "step": 2558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1423838138580322,
+      "learning_rate": 5.096678196857992e-05,
+      "loss": 2.2054,
+      "step": 2559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.043610692024231,
+      "learning_rate": 5.093455972105006e-05,
+      "loss": 2.2389,
+      "step": 2560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.139237403869629,
+      "learning_rate": 5.090233708524849e-05,
+      "loss": 2.0799,
+      "step": 2561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.137156367301941,
+      "learning_rate": 5.0870114074562424e-05,
+      "loss": 2.2405,
+      "step": 2562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.155000925064087,
+      "learning_rate": 5.08378907023792e-05,
+      "loss": 2.2265,
+      "step": 2563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0503755807876587,
+      "learning_rate": 5.080566698208634e-05,
+      "loss": 2.2985,
+      "step": 2564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1522451639175415,
+      "learning_rate": 5.077344292707148e-05,
+      "loss": 2.1844,
+      "step": 2565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1361162662506104,
+      "learning_rate": 5.074121855072242e-05,
+      "loss": 2.2525,
+      "step": 2566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3352183103561401,
+      "learning_rate": 5.070899386642709e-05,
+      "loss": 2.209,
+      "step": 2567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2009083032608032,
+      "learning_rate": 5.067676888757353e-05,
+      "loss": 2.1619,
+      "step": 2568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2592493295669556,
+      "learning_rate": 5.064454362754991e-05,
+      "loss": 2.0422,
+      "step": 2569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1160496473312378,
+      "learning_rate": 5.061231809974454e-05,
+      "loss": 2.3637,
+      "step": 2570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.226593255996704,
+      "learning_rate": 5.058009231754579e-05,
+      "loss": 2.2637,
+      "step": 2571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1354622840881348,
+      "learning_rate": 5.054786629434219e-05,
+      "loss": 2.2376,
+      "step": 2572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1305807828903198,
+      "learning_rate": 5.051564004352235e-05,
+      "loss": 2.2173,
+      "step": 2573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.054132342338562,
+      "learning_rate": 5.0483413578474945e-05,
+      "loss": 2.259,
+      "step": 2574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9862446784973145,
+      "learning_rate": 5.045118691258879e-05,
+      "loss": 2.1942,
+      "step": 2575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0588676929473877,
+      "learning_rate": 5.041896005925275e-05,
+      "loss": 2.1414,
+      "step": 2576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0771385431289673,
+      "learning_rate": 5.038673303185577e-05,
+      "loss": 2.1566,
+      "step": 2577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0995625257492065,
+      "learning_rate": 5.0354505843786894e-05,
+      "loss": 2.3372,
+      "step": 2578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0467884540557861,
+      "learning_rate": 5.032227850843522e-05,
+      "loss": 2.2383,
+      "step": 2579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9930281043052673,
+      "learning_rate": 5.029005103918986e-05,
+      "loss": 2.0106,
+      "step": 2580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1796607971191406,
+      "learning_rate": 5.0257823449440034e-05,
+      "loss": 2.2506,
+      "step": 2581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1392539739608765,
+      "learning_rate": 5.022559575257503e-05,
+      "loss": 2.1651,
+      "step": 2582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.114362359046936,
+      "learning_rate": 5.0193367961984116e-05,
+      "loss": 2.4223,
+      "step": 2583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1151282787322998,
+      "learning_rate": 5.016114009105667e-05,
+      "loss": 2.1634,
+      "step": 2584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.079990029335022,
+      "learning_rate": 5.012891215318202e-05,
+      "loss": 1.977,
+      "step": 2585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9454202651977539,
+      "learning_rate": 5.0096684161749595e-05,
+      "loss": 2.0546,
+      "step": 2586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9897853136062622,
+      "learning_rate": 5.006445613014883e-05,
+      "loss": 1.9721,
+      "step": 2587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1861988306045532,
+      "learning_rate": 5.003222807176915e-05,
+      "loss": 2.3194,
+      "step": 2588
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2646193504333496,
+      "learning_rate": 5e-05,
+      "loss": 2.1529,
+      "step": 2589
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1220669746398926,
+      "learning_rate": 4.996777192823087e-05,
+      "loss": 2.0956,
+      "step": 2590
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2959846258163452,
+      "learning_rate": 4.993554386985118e-05,
+      "loss": 2.4138,
+      "step": 2591
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1543713808059692,
+      "learning_rate": 4.990331583825041e-05,
+      "loss": 2.3132,
+      "step": 2592
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2334351539611816,
+      "learning_rate": 4.9871087846817986e-05,
+      "loss": 2.2382,
+      "step": 2593
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0511740446090698,
+      "learning_rate": 4.983885990894335e-05,
+      "loss": 2.2274,
+      "step": 2594
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2143988609313965,
+      "learning_rate": 4.9806632038015895e-05,
+      "loss": 2.1022,
+      "step": 2595
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9817091226577759,
+      "learning_rate": 4.977440424742499e-05,
+      "loss": 2.103,
+      "step": 2596
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.970143735408783,
+      "learning_rate": 4.974217655055997e-05,
+      "loss": 2.0302,
+      "step": 2597
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.075971245765686,
+      "learning_rate": 4.970994896081016e-05,
+      "loss": 2.2699,
+      "step": 2598
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.152275562286377,
+      "learning_rate": 4.96777214915648e-05,
+      "loss": 2.1895,
+      "step": 2599
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.054585576057434,
+      "learning_rate": 4.9645494156213104e-05,
+      "loss": 2.1979,
+      "step": 2600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0910075902938843,
+      "learning_rate": 4.961326696814422e-05,
+      "loss": 2.2837,
+      "step": 2601
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9086835384368896,
+      "learning_rate": 4.9581039940747255e-05,
+      "loss": 2.0896,
+      "step": 2602
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3382318019866943,
+      "learning_rate": 4.954881308741121e-05,
+      "loss": 2.6671,
+      "step": 2603
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1451923847198486,
+      "learning_rate": 4.951658642152505e-05,
+      "loss": 2.1742,
+      "step": 2604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9834786653518677,
+      "learning_rate": 4.9484359956477666e-05,
+      "loss": 2.1047,
+      "step": 2605
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9927077889442444,
+      "learning_rate": 4.945213370565781e-05,
+      "loss": 2.0151,
+      "step": 2606
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.119503140449524,
+      "learning_rate": 4.941990768245421e-05,
+      "loss": 2.2849,
+      "step": 2607
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.023508071899414,
+      "learning_rate": 4.9387681900255466e-05,
+      "loss": 2.1368,
+      "step": 2608
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0839546918869019,
+      "learning_rate": 4.935545637245008e-05,
+      "loss": 2.161,
+      "step": 2609
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1457449197769165,
+      "learning_rate": 4.932323111242647e-05,
+      "loss": 2.3428,
+      "step": 2610
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2061474323272705,
+      "learning_rate": 4.9291006133572924e-05,
+      "loss": 2.3735,
+      "step": 2611
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.158387303352356,
+      "learning_rate": 4.925878144927759e-05,
+      "loss": 2.1201,
+      "step": 2612
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9952234625816345,
+      "learning_rate": 4.922655707292854e-05,
+      "loss": 1.9672,
+      "step": 2613
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1044412851333618,
+      "learning_rate": 4.9194333017913684e-05,
+      "loss": 2.147,
+      "step": 2614
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.268934965133667,
+      "learning_rate": 4.916210929762082e-05,
+      "loss": 2.2707,
+      "step": 2615
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1374739408493042,
+      "learning_rate": 4.91298859254376e-05,
+      "loss": 2.2827,
+      "step": 2616
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3687907457351685,
+      "learning_rate": 4.909766291475153e-05,
+      "loss": 2.2662,
+      "step": 2617
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2223460674285889,
+      "learning_rate": 4.906544027894996e-05,
+      "loss": 2.2967,
+      "step": 2618
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1467280387878418,
+      "learning_rate": 4.9033218031420104e-05,
+      "loss": 2.2608,
+      "step": 2619
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.064272403717041,
+      "learning_rate": 4.9000996185548974e-05,
+      "loss": 2.1038,
+      "step": 2620
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1694166660308838,
+      "learning_rate": 4.896877475472346e-05,
+      "loss": 2.1404,
+      "step": 2621
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9216558337211609,
+      "learning_rate": 4.893655375233025e-05,
+      "loss": 2.1316,
+      "step": 2622
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0884166955947876,
+      "learning_rate": 4.890433319175589e-05,
+      "loss": 2.1793,
+      "step": 2623
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0512974262237549,
+      "learning_rate": 4.8872113086386686e-05,
+      "loss": 2.6694,
+      "step": 2624
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9388218522071838,
+      "learning_rate": 4.88398934496088e-05,
+      "loss": 1.9963,
+      "step": 2625
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9282180666923523,
+      "learning_rate": 4.880767429480819e-05,
+      "loss": 2.1092,
+      "step": 2626
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1426091194152832,
+      "learning_rate": 4.877545563537061e-05,
+      "loss": 2.3429,
+      "step": 2627
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9738643765449524,
+      "learning_rate": 4.8743237484681596e-05,
+      "loss": 1.9339,
+      "step": 2628
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0926381349563599,
+      "learning_rate": 4.871101985612651e-05,
+      "loss": 1.9509,
+      "step": 2629
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1418894529342651,
+      "learning_rate": 4.8678802763090435e-05,
+      "loss": 2.1996,
+      "step": 2630
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1092438697814941,
+      "learning_rate": 4.864658621895829e-05,
+      "loss": 2.1621,
+      "step": 2631
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1143807172775269,
+      "learning_rate": 4.861437023711475e-05,
+      "loss": 2.2908,
+      "step": 2632
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1852971315383911,
+      "learning_rate": 4.8582154830944225e-05,
+      "loss": 2.3196,
+      "step": 2633
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1799108982086182,
+      "learning_rate": 4.8549940013830934e-05,
+      "loss": 2.2027,
+      "step": 2634
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1496182680130005,
+      "learning_rate": 4.851772579915883e-05,
+      "loss": 2.3287,
+      "step": 2635
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2083065509796143,
+      "learning_rate": 4.848551220031159e-05,
+      "loss": 2.26,
+      "step": 2636
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.230102777481079,
+      "learning_rate": 4.845329923067268e-05,
+      "loss": 2.4077,
+      "step": 2637
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1535024642944336,
+      "learning_rate": 4.8421086903625276e-05,
+      "loss": 2.3621,
+      "step": 2638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.047376275062561,
+      "learning_rate": 4.838887523255232e-05,
+      "loss": 2.0731,
+      "step": 2639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1810579299926758,
+      "learning_rate": 4.835666423083641e-05,
+      "loss": 2.1323,
+      "step": 2640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9855197668075562,
+      "learning_rate": 4.832445391185994e-05,
+      "loss": 2.1892,
+      "step": 2641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0868531465530396,
+      "learning_rate": 4.829224428900499e-05,
+      "loss": 2.0752,
+      "step": 2642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3391833305358887,
+      "learning_rate": 4.826003537565335e-05,
+      "loss": 2.406,
+      "step": 2643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0650008916854858,
+      "learning_rate": 4.822782718518652e-05,
+      "loss": 2.0204,
+      "step": 2644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0211480855941772,
+      "learning_rate": 4.8195619730985695e-05,
+      "loss": 2.117,
+      "step": 2645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0072457790374756,
+      "learning_rate": 4.816341302643178e-05,
+      "loss": 2.1574,
+      "step": 2646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3205819129943848,
+      "learning_rate": 4.8131207084905346e-05,
+      "loss": 2.332,
+      "step": 2647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0824506282806396,
+      "learning_rate": 4.809900191978665e-05,
+      "loss": 2.2095,
+      "step": 2648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0513261556625366,
+      "learning_rate": 4.806679754445565e-05,
+      "loss": 2.2478,
+      "step": 2649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9969863295555115,
+      "learning_rate": 4.803459397229196e-05,
+      "loss": 2.1249,
+      "step": 2650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1259515285491943,
+      "learning_rate": 4.800239121667484e-05,
+      "loss": 2.1693,
+      "step": 2651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9839857220649719,
+      "learning_rate": 4.797018929098323e-05,
+      "loss": 2.2027,
+      "step": 2652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0986284017562866,
+      "learning_rate": 4.793798820859575e-05,
+      "loss": 2.1549,
+      "step": 2653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0597199201583862,
+      "learning_rate": 4.7905787982890644e-05,
+      "loss": 2.1518,
+      "step": 2654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2141317129135132,
+      "learning_rate": 4.787358862724577e-05,
+      "loss": 2.1785,
+      "step": 2655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.172964096069336,
+      "learning_rate": 4.784139015503867e-05,
+      "loss": 2.2568,
+      "step": 2656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1262918710708618,
+      "learning_rate": 4.7809192579646514e-05,
+      "loss": 2.069,
+      "step": 2657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0286924839019775,
+      "learning_rate": 4.777699591444609e-05,
+      "loss": 2.0881,
+      "step": 2658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.118626594543457,
+      "learning_rate": 4.77448001728138e-05,
+      "loss": 2.2548,
+      "step": 2659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.054107427597046,
+      "learning_rate": 4.771260536812567e-05,
+      "loss": 2.2589,
+      "step": 2660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9808487892150879,
+      "learning_rate": 4.768041151375734e-05,
+      "loss": 2.0144,
+      "step": 2661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.113303542137146,
+      "learning_rate": 4.764821862308405e-05,
+      "loss": 1.9994,
+      "step": 2662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1462305784225464,
+      "learning_rate": 4.7616026709480655e-05,
+      "loss": 2.454,
+      "step": 2663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.09292471408844,
+      "learning_rate": 4.758383578632158e-05,
+      "loss": 2.2825,
+      "step": 2664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3024530410766602,
+      "learning_rate": 4.755164586698085e-05,
+      "loss": 2.3675,
+      "step": 2665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.326647162437439,
+      "learning_rate": 4.7519456964832067e-05,
+      "loss": 2.4049,
+      "step": 2666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3551032543182373,
+      "learning_rate": 4.7487269093248424e-05,
+      "loss": 2.4945,
+      "step": 2667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0647035837173462,
+      "learning_rate": 4.745508226560267e-05,
+      "loss": 2.1964,
+      "step": 2668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0151616334915161,
+      "learning_rate": 4.7422896495267136e-05,
+      "loss": 1.8517,
+      "step": 2669
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.009636640548706,
+      "learning_rate": 4.73907117956137e-05,
+      "loss": 2.2377,
+      "step": 2670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1767066717147827,
+      "learning_rate": 4.735852818001381e-05,
+      "loss": 2.1789,
+      "step": 2671
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0628044605255127,
+      "learning_rate": 4.732634566183844e-05,
+      "loss": 2.1634,
+      "step": 2672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3006330728530884,
+      "learning_rate": 4.729416425445814e-05,
+      "loss": 2.5624,
+      "step": 2673
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1123629808425903,
+      "learning_rate": 4.7261983971242966e-05,
+      "loss": 2.2053,
+      "step": 2674
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2658381462097168,
+      "learning_rate": 4.722980482556253e-05,
+      "loss": 2.3402,
+      "step": 2675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1371991634368896,
+      "learning_rate": 4.719762683078595e-05,
+      "loss": 2.2699,
+      "step": 2676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0259774923324585,
+      "learning_rate": 4.71654500002819e-05,
+      "loss": 2.0061,
+      "step": 2677
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2392221689224243,
+      "learning_rate": 4.713327434741853e-05,
+      "loss": 2.1646,
+      "step": 2678
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0668394565582275,
+      "learning_rate": 4.710109988556354e-05,
+      "loss": 2.2257,
+      "step": 2679
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0527206659317017,
+      "learning_rate": 4.7068926628084096e-05,
+      "loss": 2.248,
+      "step": 2680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2277740240097046,
+      "learning_rate": 4.7036754588346894e-05,
+      "loss": 2.3633,
+      "step": 2681
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2017061710357666,
+      "learning_rate": 4.700458377971811e-05,
+      "loss": 2.3388,
+      "step": 2682
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.256166696548462,
+      "learning_rate": 4.6972414215563416e-05,
+      "loss": 2.3736,
+      "step": 2683
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2040871381759644,
+      "learning_rate": 4.694024590924796e-05,
+      "loss": 2.1851,
+      "step": 2684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9613001942634583,
+      "learning_rate": 4.6908078874136374e-05,
+      "loss": 2.2685,
+      "step": 2685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1157244443893433,
+      "learning_rate": 4.687591312359274e-05,
+      "loss": 2.1027,
+      "step": 2686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2101187705993652,
+      "learning_rate": 4.684374867098065e-05,
+      "loss": 2.164,
+      "step": 2687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.127551555633545,
+      "learning_rate": 4.681158552966311e-05,
+      "loss": 2.0598,
+      "step": 2688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0937916040420532,
+      "learning_rate": 4.677942371300261e-05,
+      "loss": 2.1638,
+      "step": 2689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.085304856300354,
+      "learning_rate": 4.6747263234361085e-05,
+      "loss": 2.1857,
+      "step": 2690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.040910243988037,
+      "learning_rate": 4.67151041070999e-05,
+      "loss": 2.153,
+      "step": 2691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.171785593032837,
+      "learning_rate": 4.6682946344579893e-05,
+      "loss": 2.1079,
+      "step": 2692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.12923264503479,
+      "learning_rate": 4.665078996016129e-05,
+      "loss": 2.3779,
+      "step": 2693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2049013376235962,
+      "learning_rate": 4.661863496720378e-05,
+      "loss": 2.3916,
+      "step": 2694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1487796306610107,
+      "learning_rate": 4.658648137906645e-05,
+      "loss": 2.0559,
+      "step": 2695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1767491102218628,
+      "learning_rate": 4.655432920910782e-05,
+      "loss": 2.2197,
+      "step": 2696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2331112623214722,
+      "learning_rate": 4.652217847068579e-05,
+      "loss": 2.285,
+      "step": 2697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9188472628593445,
+      "learning_rate": 4.649002917715772e-05,
+      "loss": 2.1941,
+      "step": 2698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1066975593566895,
+      "learning_rate": 4.645788134188034e-05,
+      "loss": 2.2158,
+      "step": 2699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.034701943397522,
+      "learning_rate": 4.642573497820975e-05,
+      "loss": 2.035,
+      "step": 2700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1800949573516846,
+      "learning_rate": 4.639359009950147e-05,
+      "loss": 2.2097,
+      "step": 2701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9022451639175415,
+      "learning_rate": 4.6361446719110405e-05,
+      "loss": 2.0525,
+      "step": 2702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1205979585647583,
+      "learning_rate": 4.632930485039081e-05,
+      "loss": 2.2126,
+      "step": 2703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0287922620773315,
+      "learning_rate": 4.629716450669634e-05,
+      "loss": 2.0348,
+      "step": 2704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.23370361328125,
+      "learning_rate": 4.626502570138001e-05,
+      "loss": 2.3524,
+      "step": 2705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1167563199996948,
+      "learning_rate": 4.623288844779418e-05,
+      "loss": 2.2781,
+      "step": 2706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1214942932128906,
+      "learning_rate": 4.620075275929059e-05,
+      "loss": 1.9327,
+      "step": 2707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2049275636672974,
+      "learning_rate": 4.61686186492203e-05,
+      "loss": 2.1178,
+      "step": 2708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1712466478347778,
+      "learning_rate": 4.6136486130933766e-05,
+      "loss": 2.1955,
+      "step": 2709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0110836029052734,
+      "learning_rate": 4.6104355217780704e-05,
+      "loss": 2.2886,
+      "step": 2710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1214884519577026,
+      "learning_rate": 4.6072225923110225e-05,
+      "loss": 2.1433,
+      "step": 2711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0610171556472778,
+      "learning_rate": 4.604009826027075e-05,
+      "loss": 1.8672,
+      "step": 2712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0059003829956055,
+      "learning_rate": 4.6007972242610023e-05,
+      "loss": 2.3849,
+      "step": 2713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1734795570373535,
+      "learning_rate": 4.59758478834751e-05,
+      "loss": 2.2872,
+      "step": 2714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.044341802597046,
+      "learning_rate": 4.594372519621235e-05,
+      "loss": 2.0976,
+      "step": 2715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2016968727111816,
+      "learning_rate": 4.591160419416744e-05,
+      "loss": 2.1079,
+      "step": 2716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1239283084869385,
+      "learning_rate": 4.5879484890685376e-05,
+      "loss": 2.2734,
+      "step": 2717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0556504726409912,
+      "learning_rate": 4.5847367299110394e-05,
+      "loss": 1.9758,
+      "step": 2718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0920689105987549,
+      "learning_rate": 4.581525143278608e-05,
+      "loss": 2.2189,
+      "step": 2719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1241929531097412,
+      "learning_rate": 4.578313730505524e-05,
+      "loss": 2.3436,
+      "step": 2720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1320151090621948,
+      "learning_rate": 4.5751024929260014e-05,
+      "loss": 2.3238,
+      "step": 2721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2731056213378906,
+      "learning_rate": 4.5718914318741803e-05,
+      "loss": 2.1817,
+      "step": 2722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1884522438049316,
+      "learning_rate": 4.5686805486841244e-05,
+      "loss": 2.2553,
+      "step": 2723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0355480909347534,
+      "learning_rate": 4.5654698446898266e-05,
+      "loss": 2.3216,
+      "step": 2724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2326463460922241,
+      "learning_rate": 4.5622593212252044e-05,
+      "loss": 2.0269,
+      "step": 2725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2242099046707153,
+      "learning_rate": 4.5590489796240996e-05,
+      "loss": 2.2235,
+      "step": 2726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1396737098693848,
+      "learning_rate": 4.55583882122028e-05,
+      "loss": 2.2257,
+      "step": 2727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0328422784805298,
+      "learning_rate": 4.552628847347437e-05,
+      "loss": 2.1307,
+      "step": 2728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1723183393478394,
+      "learning_rate": 4.5494190593391826e-05,
+      "loss": 2.2261,
+      "step": 2729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1576533317565918,
+      "learning_rate": 4.5462094585290554e-05,
+      "loss": 2.17,
+      "step": 2730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2776761054992676,
+      "learning_rate": 4.5430000462505126e-05,
+      "loss": 2.302,
+      "step": 2731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.084293007850647,
+      "learning_rate": 4.5397908238369366e-05,
+      "loss": 2.0248,
+      "step": 2732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.097307562828064,
+      "learning_rate": 4.5365817926216296e-05,
+      "loss": 2.0829,
+      "step": 2733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.036377191543579,
+      "learning_rate": 4.533372953937813e-05,
+      "loss": 2.2001,
+      "step": 2734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.049536108970642,
+      "learning_rate": 4.53016430911863e-05,
+      "loss": 1.9382,
+      "step": 2735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.221959114074707,
+      "learning_rate": 4.526955859497141e-05,
+      "loss": 2.2969,
+      "step": 2736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.106056571006775,
+      "learning_rate": 4.523747606406329e-05,
+      "loss": 2.0876,
+      "step": 2737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0941306352615356,
+      "learning_rate": 4.520539551179091e-05,
+      "loss": 2.3479,
+      "step": 2738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9729526042938232,
+      "learning_rate": 4.5173316951482485e-05,
+      "loss": 1.9955,
+      "step": 2739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1575934886932373,
+      "learning_rate": 4.51412403964653e-05,
+      "loss": 2.3067,
+      "step": 2740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1984082460403442,
+      "learning_rate": 4.510916586006589e-05,
+      "loss": 2.3909,
+      "step": 2741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2523044347763062,
+      "learning_rate": 4.507709335560992e-05,
+      "loss": 2.1401,
+      "step": 2742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0183086395263672,
+      "learning_rate": 4.504502289642222e-05,
+      "loss": 1.8875,
+      "step": 2743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1312308311462402,
+      "learning_rate": 4.5012954495826773e-05,
+      "loss": 2.1506,
+      "step": 2744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1302073001861572,
+      "learning_rate": 4.498088816714669e-05,
+      "loss": 2.3319,
+      "step": 2745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.130051612854004,
+      "learning_rate": 4.494882392370423e-05,
+      "loss": 1.9676,
+      "step": 2746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1508311033248901,
+      "learning_rate": 4.49167617788208e-05,
+      "loss": 2.2367,
+      "step": 2747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.19626784324646,
+      "learning_rate": 4.4884701745816905e-05,
+      "loss": 2.21,
+      "step": 2748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2065848112106323,
+      "learning_rate": 4.4852643838012205e-05,
+      "loss": 2.343,
+      "step": 2749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9878059029579163,
+      "learning_rate": 4.482058806872545e-05,
+      "loss": 2.1632,
+      "step": 2750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9911506175994873,
+      "learning_rate": 4.478853445127452e-05,
+      "loss": 2.066,
+      "step": 2751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1859633922576904,
+      "learning_rate": 4.47564829989764e-05,
+      "loss": 2.6003,
+      "step": 2752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9985003471374512,
+      "learning_rate": 4.472443372514716e-05,
+      "loss": 1.8455,
+      "step": 2753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2183936834335327,
+      "learning_rate": 4.4692386643101994e-05,
+      "loss": 2.2731,
+      "step": 2754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1699811220169067,
+      "learning_rate": 4.466034176615514e-05,
+      "loss": 2.1167,
+      "step": 2755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9188485741615295,
+      "learning_rate": 4.462829910761995e-05,
+      "loss": 1.983,
+      "step": 2756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2464858293533325,
+      "learning_rate": 4.459625868080887e-05,
+      "loss": 1.8144,
+      "step": 2757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.013093113899231,
+      "learning_rate": 4.4564220499033375e-05,
+      "loss": 2.2621,
+      "step": 2758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.12288236618042,
+      "learning_rate": 4.453218457560405e-05,
+      "loss": 2.1451,
+      "step": 2759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.078503966331482,
+      "learning_rate": 4.450015092383051e-05,
+      "loss": 2.1762,
+      "step": 2760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0371326208114624,
+      "learning_rate": 4.446811955702146e-05,
+      "loss": 2.2349,
+      "step": 2761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0648269653320312,
+      "learning_rate": 4.443609048848461e-05,
+      "loss": 2.1024,
+      "step": 2762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.141663908958435,
+      "learning_rate": 4.440406373152676e-05,
+      "loss": 2.0313,
+      "step": 2763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1983346939086914,
+      "learning_rate": 4.437203929945373e-05,
+      "loss": 2.2046,
+      "step": 2764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2857317924499512,
+      "learning_rate": 4.434001720557035e-05,
+      "loss": 2.3218,
+      "step": 2765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.073123812675476,
+      "learning_rate": 4.430799746318052e-05,
+      "loss": 2.2519,
+      "step": 2766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2683120965957642,
+      "learning_rate": 4.427598008558715e-05,
+      "loss": 2.2768,
+      "step": 2767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2051763534545898,
+      "learning_rate": 4.4243965086092154e-05,
+      "loss": 2.3418,
+      "step": 2768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0827112197875977,
+      "learning_rate": 4.421195247799647e-05,
+      "loss": 1.9788,
+      "step": 2769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0083296298980713,
+      "learning_rate": 4.4179942274600054e-05,
+      "loss": 2.0702,
+      "step": 2770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.088995099067688,
+      "learning_rate": 4.414793448920182e-05,
+      "loss": 2.0335,
+      "step": 2771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0133614540100098,
+      "learning_rate": 4.4115929135099724e-05,
+      "loss": 2.0235,
+      "step": 2772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.996837854385376,
+      "learning_rate": 4.40839262255907e-05,
+      "loss": 2.1438,
+      "step": 2773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2043745517730713,
+      "learning_rate": 4.4051925773970656e-05,
+      "loss": 2.1025,
+      "step": 2774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0532684326171875,
+      "learning_rate": 4.401992779353447e-05,
+      "loss": 2.4215,
+      "step": 2775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1713353395462036,
+      "learning_rate": 4.398793229757602e-05,
+      "loss": 2.2465,
+      "step": 2776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.187928318977356,
+      "learning_rate": 4.395593929938813e-05,
+      "loss": 2.1042,
+      "step": 2777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.18411386013031,
+      "learning_rate": 4.3923948812262596e-05,
+      "loss": 2.2124,
+      "step": 2778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1316043138504028,
+      "learning_rate": 4.389196084949018e-05,
+      "loss": 1.8961,
+      "step": 2779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.244913101196289,
+      "learning_rate": 4.3859975424360574e-05,
+      "loss": 2.3289,
+      "step": 2780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0544663667678833,
+      "learning_rate": 4.382799255016243e-05,
+      "loss": 2.3337,
+      "step": 2781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2765873670578003,
+      "learning_rate": 4.379601224018335e-05,
+      "loss": 2.0444,
+      "step": 2782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.201769232749939,
+      "learning_rate": 4.376403450770983e-05,
+      "loss": 2.1028,
+      "step": 2783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2618476152420044,
+      "learning_rate": 4.373205936602733e-05,
+      "loss": 2.3014,
+      "step": 2784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1092052459716797,
+      "learning_rate": 4.370008682842023e-05,
+      "loss": 2.2256,
+      "step": 2785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.251831293106079,
+      "learning_rate": 4.366811690817183e-05,
+      "loss": 2.2931,
+      "step": 2786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9401856660842896,
+      "learning_rate": 4.363614961856433e-05,
+      "loss": 2.1106,
+      "step": 2787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9786458015441895,
+      "learning_rate": 4.360418497287885e-05,
+      "loss": 2.138,
+      "step": 2788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9347699880599976,
+      "learning_rate": 4.35722229843954e-05,
+      "loss": 2.0951,
+      "step": 2789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3065866231918335,
+      "learning_rate": 4.354026366639289e-05,
+      "loss": 2.3797,
+      "step": 2790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0089865922927856,
+      "learning_rate": 4.350830703214913e-05,
+      "loss": 2.3348,
+      "step": 2791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2500706911087036,
+      "learning_rate": 4.347635309494079e-05,
+      "loss": 2.1808,
+      "step": 2792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1672199964523315,
+      "learning_rate": 4.3444401868043454e-05,
+      "loss": 2.1167,
+      "step": 2793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.983588457107544,
+      "learning_rate": 4.3412453364731557e-05,
+      "loss": 2.2841,
+      "step": 2794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0634911060333252,
+      "learning_rate": 4.33805075982784e-05,
+      "loss": 2.0993,
+      "step": 2795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1356382369995117,
+      "learning_rate": 4.3348564581956163e-05,
+      "loss": 2.2276,
+      "step": 2796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1552454233169556,
+      "learning_rate": 4.3316624329035874e-05,
+      "loss": 2.2012,
+      "step": 2797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0222834348678589,
+      "learning_rate": 4.328468685278742e-05,
+      "loss": 2.0893,
+      "step": 2798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0604536533355713,
+      "learning_rate": 4.3252752166479524e-05,
+      "loss": 1.9965,
+      "step": 2799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2812671661376953,
+      "learning_rate": 4.322082028337975e-05,
+      "loss": 2.3636,
+      "step": 2800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.257215976715088,
+      "learning_rate": 4.3188891216754496e-05,
+      "loss": 2.1628,
+      "step": 2801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1110063791275024,
+      "learning_rate": 4.3156964979869006e-05,
+      "loss": 2.1827,
+      "step": 2802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2058603763580322,
+      "learning_rate": 4.3125041585987344e-05,
+      "loss": 2.005,
+      "step": 2803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0316931009292603,
+      "learning_rate": 4.3093121048372364e-05,
+      "loss": 2.1021,
+      "step": 2804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.141912817955017,
+      "learning_rate": 4.3061203380285784e-05,
+      "loss": 2.3102,
+      "step": 2805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0971494913101196,
+      "learning_rate": 4.302928859498808e-05,
+      "loss": 2.0064,
+      "step": 2806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2006888389587402,
+      "learning_rate": 4.2997376705738565e-05,
+      "loss": 2.1795,
+      "step": 2807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.031239628791809,
+      "learning_rate": 4.2965467725795336e-05,
+      "loss": 2.1285,
+      "step": 2808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.065553903579712,
+      "learning_rate": 4.2933561668415286e-05,
+      "loss": 2.1152,
+      "step": 2809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9840570092201233,
+      "learning_rate": 4.290165854685407e-05,
+      "loss": 1.9307,
+      "step": 2810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2986780405044556,
+      "learning_rate": 4.286975837436615e-05,
+      "loss": 2.3036,
+      "step": 2811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1440057754516602,
+      "learning_rate": 4.283786116420476e-05,
+      "loss": 2.1069,
+      "step": 2812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1023457050323486,
+      "learning_rate": 4.28059669296219e-05,
+      "loss": 2.1606,
+      "step": 2813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2141016721725464,
+      "learning_rate": 4.277407568386833e-05,
+      "loss": 1.9847,
+      "step": 2814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1982697248458862,
+      "learning_rate": 4.274218744019357e-05,
+      "loss": 2.1816,
+      "step": 2815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9759746193885803,
+      "learning_rate": 4.27103022118459e-05,
+      "loss": 2.0019,
+      "step": 2816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0526723861694336,
+      "learning_rate": 4.267842001207233e-05,
+      "loss": 2.1267,
+      "step": 2817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0894958972930908,
+      "learning_rate": 4.2646540854118624e-05,
+      "loss": 2.3243,
+      "step": 2818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9469379186630249,
+      "learning_rate": 4.261466475122929e-05,
+      "loss": 1.8731,
+      "step": 2819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9933145046234131,
+      "learning_rate": 4.258279171664755e-05,
+      "loss": 2.1067,
+      "step": 2820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1602191925048828,
+      "learning_rate": 4.255092176361535e-05,
+      "loss": 2.2499,
+      "step": 2821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2459923028945923,
+      "learning_rate": 4.251905490537338e-05,
+      "loss": 2.2134,
+      "step": 2822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9828397631645203,
+      "learning_rate": 4.2487191155161024e-05,
+      "loss": 2.0333,
+      "step": 2823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9360442161560059,
+      "learning_rate": 4.245533052621637e-05,
+      "loss": 1.8555,
+      "step": 2824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.183509349822998,
+      "learning_rate": 4.242347303177625e-05,
+      "loss": 2.2081,
+      "step": 2825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.354470133781433,
+      "learning_rate": 4.2391618685076126e-05,
+      "loss": 2.1942,
+      "step": 2826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.176853060722351,
+      "learning_rate": 4.235976749935019e-05,
+      "loss": 2.1778,
+      "step": 2827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1098893880844116,
+      "learning_rate": 4.232791948783134e-05,
+      "loss": 2.1431,
+      "step": 2828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1329100131988525,
+      "learning_rate": 4.229607466375111e-05,
+      "loss": 2.0313,
+      "step": 2829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1379848718643188,
+      "learning_rate": 4.226423304033976e-05,
+      "loss": 2.3384,
+      "step": 2830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2040475606918335,
+      "learning_rate": 4.2232394630826184e-05,
+      "loss": 1.9573,
+      "step": 2831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.261744499206543,
+      "learning_rate": 4.220055944843795e-05,
+      "loss": 2.2487,
+      "step": 2832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1729544401168823,
+      "learning_rate": 4.216872750640129e-05,
+      "loss": 2.2375,
+      "step": 2833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2258720397949219,
+      "learning_rate": 4.2136898817941087e-05,
+      "loss": 2.0603,
+      "step": 2834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.117484211921692,
+      "learning_rate": 4.210507339628086e-05,
+      "loss": 2.326,
+      "step": 2835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0966858863830566,
+      "learning_rate": 4.207325125464279e-05,
+      "loss": 2.0488,
+      "step": 2836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.228640079498291,
+      "learning_rate": 4.2041432406247696e-05,
+      "loss": 2.2467,
+      "step": 2837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1861058473587036,
+      "learning_rate": 4.2009616864315e-05,
+      "loss": 2.0664,
+      "step": 2838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.25771164894104,
+      "learning_rate": 4.197780464206279e-05,
+      "loss": 2.3382,
+      "step": 2839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1249022483825684,
+      "learning_rate": 4.194599575270776e-05,
+      "loss": 1.9555,
+      "step": 2840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1024919748306274,
+      "learning_rate": 4.191419020946519e-05,
+      "loss": 2.1884,
+      "step": 2841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1212260723114014,
+      "learning_rate": 4.188238802554901e-05,
+      "loss": 2.2317,
+      "step": 2842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0594757795333862,
+      "learning_rate": 4.1850589214171744e-05,
+      "loss": 2.0435,
+      "step": 2843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1304869651794434,
+      "learning_rate": 4.181879378854451e-05,
+      "loss": 2.2836,
+      "step": 2844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0967012643814087,
+      "learning_rate": 4.1787001761877e-05,
+      "loss": 2.0622,
+      "step": 2845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3830442428588867,
+      "learning_rate": 4.175521314737753e-05,
+      "loss": 2.2912,
+      "step": 2846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1173629760742188,
+      "learning_rate": 4.1723427958252976e-05,
+      "loss": 1.9762,
+      "step": 2847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1283148527145386,
+      "learning_rate": 4.1691646207708794e-05,
+      "loss": 2.1414,
+      "step": 2848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9861019849777222,
+      "learning_rate": 4.165986790894902e-05,
+      "loss": 1.9106,
+      "step": 2849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0260334014892578,
+      "learning_rate": 4.162809307517625e-05,
+      "loss": 2.2868,
+      "step": 2850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0887340307235718,
+      "learning_rate": 4.1596321719591634e-05,
+      "loss": 2.094,
+      "step": 2851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.928357720375061,
+      "learning_rate": 4.1564553855394894e-05,
+      "loss": 1.7729,
+      "step": 2852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9530646800994873,
+      "learning_rate": 4.1532789495784274e-05,
+      "loss": 2.1613,
+      "step": 2853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0110509395599365,
+      "learning_rate": 4.15010286539566e-05,
+      "loss": 1.9499,
+      "step": 2854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.070439338684082,
+      "learning_rate": 4.1469271343107196e-05,
+      "loss": 2.0659,
+      "step": 2855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.103806734085083,
+      "learning_rate": 4.143751757642994e-05,
+      "loss": 2.0041,
+      "step": 2856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.141443133354187,
+      "learning_rate": 4.1405767367117245e-05,
+      "loss": 2.2194,
+      "step": 2857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.049987554550171,
+      "learning_rate": 4.137402072836002e-05,
+      "loss": 2.2481,
+      "step": 2858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1695691347122192,
+      "learning_rate": 4.134227767334773e-05,
+      "loss": 2.0937,
+      "step": 2859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2293835878372192,
+      "learning_rate": 4.1310538215268315e-05,
+      "loss": 2.182,
+      "step": 2860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0207219123840332,
+      "learning_rate": 4.127880236730824e-05,
+      "loss": 2.0752,
+      "step": 2861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.13922119140625,
+      "learning_rate": 4.124707014265246e-05,
+      "loss": 2.4074,
+      "step": 2862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2285854816436768,
+      "learning_rate": 4.121534155448443e-05,
+      "loss": 2.1372,
+      "step": 2863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1116665601730347,
+      "learning_rate": 4.11836166159861e-05,
+      "loss": 2.0542,
+      "step": 2864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1064773797988892,
+      "learning_rate": 4.115189534033787e-05,
+      "loss": 2.2587,
+      "step": 2865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0167003870010376,
+      "learning_rate": 4.112017774071868e-05,
+      "loss": 2.3136,
+      "step": 2866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1449366807937622,
+      "learning_rate": 4.108846383030589e-05,
+      "loss": 2.3843,
+      "step": 2867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.112859845161438,
+      "learning_rate": 4.105675362227536e-05,
+      "loss": 2.1569,
+      "step": 2868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1947486400604248,
+      "learning_rate": 4.1025047129801365e-05,
+      "loss": 2.2917,
+      "step": 2869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9947771430015564,
+      "learning_rate": 4.0993344366056686e-05,
+      "loss": 2.4436,
+      "step": 2870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9930851459503174,
+      "learning_rate": 4.0961645344212534e-05,
+      "loss": 2.0528,
+      "step": 2871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0020256042480469,
+      "learning_rate": 4.092995007743858e-05,
+      "loss": 2.076,
+      "step": 2872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0670757293701172,
+      "learning_rate": 4.089825857890291e-05,
+      "loss": 2.2648,
+      "step": 2873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1888692378997803,
+      "learning_rate": 4.0866570861772056e-05,
+      "loss": 2.0523,
+      "step": 2874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1078568696975708,
+      "learning_rate": 4.0834886939211e-05,
+      "loss": 2.1451,
+      "step": 2875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0585168600082397,
+      "learning_rate": 4.080320682438311e-05,
+      "loss": 2.2092,
+      "step": 2876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1469738483428955,
+      "learning_rate": 4.07715305304502e-05,
+      "loss": 2.1446,
+      "step": 2877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9719735383987427,
+      "learning_rate": 4.073985807057249e-05,
+      "loss": 1.9412,
+      "step": 2878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.248894214630127,
+      "learning_rate": 4.070818945790859e-05,
+      "loss": 2.2417,
+      "step": 2879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.150815486907959,
+      "learning_rate": 4.0676524705615526e-05,
+      "loss": 2.0147,
+      "step": 2880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1760647296905518,
+      "learning_rate": 4.064486382684873e-05,
+      "loss": 1.9932,
+      "step": 2881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1746759414672852,
+      "learning_rate": 4.061320683476201e-05,
+      "loss": 2.0536,
+      "step": 2882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2906413078308105,
+      "learning_rate": 4.058155374250756e-05,
+      "loss": 2.1732,
+      "step": 2883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0394985675811768,
+      "learning_rate": 4.054990456323596e-05,
+      "loss": 2.1378,
+      "step": 2884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.206418514251709,
+      "learning_rate": 4.0518259310096176e-05,
+      "loss": 2.3353,
+      "step": 2885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1012951135635376,
+      "learning_rate": 4.048661799623551e-05,
+      "loss": 2.0032,
+      "step": 2886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2287952899932861,
+      "learning_rate": 4.045498063479965e-05,
+      "loss": 2.1119,
+      "step": 2887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1646031141281128,
+      "learning_rate": 4.042334723893264e-05,
+      "loss": 2.2885,
+      "step": 2888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1748591661453247,
+      "learning_rate": 4.0391717821776886e-05,
+      "loss": 2.2473,
+      "step": 2889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.18471360206604,
+      "learning_rate": 4.036009239647311e-05,
+      "loss": 2.2851,
+      "step": 2890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2976069450378418,
+      "learning_rate": 4.03284709761604e-05,
+      "loss": 2.2516,
+      "step": 2891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.310422420501709,
+      "learning_rate": 4.029685357397618e-05,
+      "loss": 2.1702,
+      "step": 2892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1384907960891724,
+      "learning_rate": 4.02652402030562e-05,
+      "loss": 2.2528,
+      "step": 2893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1138052940368652,
+      "learning_rate": 4.0233630876534525e-05,
+      "loss": 2.0762,
+      "step": 2894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1359188556671143,
+      "learning_rate": 4.0202025607543564e-05,
+      "loss": 2.2394,
+      "step": 2895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2143911123275757,
+      "learning_rate": 4.0170424409214014e-05,
+      "loss": 2.2001,
+      "step": 2896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1571825742721558,
+      "learning_rate": 4.0138827294674905e-05,
+      "loss": 2.2555,
+      "step": 2897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0436348915100098,
+      "learning_rate": 4.010723427705354e-05,
+      "loss": 1.9591,
+      "step": 2898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9170854091644287,
+      "learning_rate": 4.0075645369475564e-05,
+      "loss": 1.9067,
+      "step": 2899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1465580463409424,
+      "learning_rate": 4.004406058506487e-05,
+      "loss": 2.172,
+      "step": 2900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2224793434143066,
+      "learning_rate": 4.001247993694365e-05,
+      "loss": 2.0118,
+      "step": 2901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.072432518005371,
+      "learning_rate": 3.99809034382324e-05,
+      "loss": 2.003,
+      "step": 2902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1481987237930298,
+      "learning_rate": 3.9949331102049864e-05,
+      "loss": 2.318,
+      "step": 2903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3016895055770874,
+      "learning_rate": 3.991776294151307e-05,
+      "loss": 2.1561,
+      "step": 2904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.207659125328064,
+      "learning_rate": 3.9886198969737324e-05,
+      "loss": 2.3229,
+      "step": 2905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1565788984298706,
+      "learning_rate": 3.9854639199836156e-05,
+      "loss": 1.9051,
+      "step": 2906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9898173213005066,
+      "learning_rate": 3.982308364492139e-05,
+      "loss": 1.8925,
+      "step": 2907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1060086488723755,
+      "learning_rate": 3.9791532318103075e-05,
+      "loss": 2.1405,
+      "step": 2908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2063525915145874,
+      "learning_rate": 3.9759985232489515e-05,
+      "loss": 2.1916,
+      "step": 2909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9966272115707397,
+      "learning_rate": 3.972844240118724e-05,
+      "loss": 2.0065,
+      "step": 2910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0934768915176392,
+      "learning_rate": 3.9696903837301004e-05,
+      "loss": 2.167,
+      "step": 2911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2021089792251587,
+      "learning_rate": 3.9665369553933815e-05,
+      "loss": 2.2411,
+      "step": 2912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.179885745048523,
+      "learning_rate": 3.96338395641869e-05,
+      "loss": 2.1955,
+      "step": 2913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0141338109970093,
+      "learning_rate": 3.960231388115968e-05,
+      "loss": 2.1438,
+      "step": 2914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0514180660247803,
+      "learning_rate": 3.9570792517949794e-05,
+      "loss": 1.9725,
+      "step": 2915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9945041537284851,
+      "learning_rate": 3.9539275487653105e-05,
+      "loss": 1.9674,
+      "step": 2916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1254409551620483,
+      "learning_rate": 3.950776280336366e-05,
+      "loss": 2.0912,
+      "step": 2917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0682120323181152,
+      "learning_rate": 3.947625447817369e-05,
+      "loss": 2.0856,
+      "step": 2918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.174100399017334,
+      "learning_rate": 3.944475052517364e-05,
+      "loss": 2.0566,
+      "step": 2919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0892690420150757,
+      "learning_rate": 3.941325095745213e-05,
+      "loss": 2.0625,
+      "step": 2920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1360410451889038,
+      "learning_rate": 3.938175578809594e-05,
+      "loss": 2.2134,
+      "step": 2921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.006296157836914,
+      "learning_rate": 3.9350265030190056e-05,
+      "loss": 1.9684,
+      "step": 2922
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2756166458129883,
+      "learning_rate": 3.93187786968176e-05,
+      "loss": 2.2594,
+      "step": 2923
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3170961141586304,
+      "learning_rate": 3.9287296801059856e-05,
+      "loss": 2.2497,
+      "step": 2924
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2610149383544922,
+      "learning_rate": 3.9255819355996296e-05,
+      "loss": 2.1293,
+      "step": 2925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1776492595672607,
+      "learning_rate": 3.9224346374704515e-05,
+      "loss": 2.2928,
+      "step": 2926
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2300310134887695,
+      "learning_rate": 3.9192877870260264e-05,
+      "loss": 2.3696,
+      "step": 2927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1859444379806519,
+      "learning_rate": 3.916141385573743e-05,
+      "loss": 2.0711,
+      "step": 2928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0355595350265503,
+      "learning_rate": 3.9129954344208034e-05,
+      "loss": 2.0769,
+      "step": 2929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1259721517562866,
+      "learning_rate": 3.909849934874224e-05,
+      "loss": 1.9003,
+      "step": 2930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1183993816375732,
+      "learning_rate": 3.90670488824083e-05,
+      "loss": 2.1884,
+      "step": 2931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.212632656097412,
+      "learning_rate": 3.9035602958272634e-05,
+      "loss": 2.1921,
+      "step": 2932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0603406429290771,
+      "learning_rate": 3.9004161589399747e-05,
+      "loss": 2.0468,
+      "step": 2933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1597102880477905,
+      "learning_rate": 3.897272478885224e-05,
+      "loss": 2.1359,
+      "step": 2934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.111085295677185,
+      "learning_rate": 3.894129256969083e-05,
+      "loss": 2.2451,
+      "step": 2935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0846775770187378,
+      "learning_rate": 3.890986494497434e-05,
+      "loss": 2.1275,
+      "step": 2936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.070550560951233,
+      "learning_rate": 3.887844192775968e-05,
+      "loss": 1.9552,
+      "step": 2937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1369173526763916,
+      "learning_rate": 3.884702353110182e-05,
+      "loss": 2.302,
+      "step": 2938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0911873579025269,
+      "learning_rate": 3.881560976805385e-05,
+      "loss": 2.1418,
+      "step": 2939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3088383674621582,
+      "learning_rate": 3.878420065166691e-05,
+      "loss": 2.1417,
+      "step": 2940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1180495023727417,
+      "learning_rate": 3.875279619499022e-05,
+      "loss": 2.3123,
+      "step": 2941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.120854139328003,
+      "learning_rate": 3.8721396411071046e-05,
+      "loss": 2.2777,
+      "step": 2942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0766147375106812,
+      "learning_rate": 3.869000131295474e-05,
+      "loss": 2.1622,
+      "step": 2943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2317938804626465,
+      "learning_rate": 3.865861091368469e-05,
+      "loss": 2.1339,
+      "step": 2944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1316431760787964,
+      "learning_rate": 3.862722522630234e-05,
+      "loss": 2.2125,
+      "step": 2945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0452659130096436,
+      "learning_rate": 3.8595844263847146e-05,
+      "loss": 2.2208,
+      "step": 2946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1254700422286987,
+      "learning_rate": 3.856446803935666e-05,
+      "loss": 1.9667,
+      "step": 2947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.113021969795227,
+      "learning_rate": 3.853309656586641e-05,
+      "loss": 2.277,
+      "step": 2948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0045725107192993,
+      "learning_rate": 3.850172985640998e-05,
+      "loss": 2.0869,
+      "step": 2949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0286333560943604,
+      "learning_rate": 3.8470367924018965e-05,
+      "loss": 1.8567,
+      "step": 2950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0150840282440186,
+      "learning_rate": 3.843901078172298e-05,
+      "loss": 1.9239,
+      "step": 2951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2076079845428467,
+      "learning_rate": 3.8407658442549656e-05,
+      "loss": 2.3077,
+      "step": 2952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2184566259384155,
+      "learning_rate": 3.8376310919524623e-05,
+      "loss": 1.9846,
+      "step": 2953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.220082402229309,
+      "learning_rate": 3.834496822567148e-05,
+      "loss": 2.1953,
+      "step": 2954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0584285259246826,
+      "learning_rate": 3.8313630374011844e-05,
+      "loss": 2.2272,
+      "step": 2955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.173874020576477,
+      "learning_rate": 3.828229737756535e-05,
+      "loss": 1.9842,
+      "step": 2956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2322007417678833,
+      "learning_rate": 3.825096924934958e-05,
+      "loss": 2.0307,
+      "step": 2957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.173226237297058,
+      "learning_rate": 3.82196460023801e-05,
+      "loss": 2.0493,
+      "step": 2958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2535656690597534,
+      "learning_rate": 3.818832764967045e-05,
+      "loss": 1.9972,
+      "step": 2959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1459277868270874,
+      "learning_rate": 3.815701420423213e-05,
+      "loss": 1.9108,
+      "step": 2960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.212562084197998,
+      "learning_rate": 3.812570567907462e-05,
+      "loss": 1.9954,
+      "step": 2961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2123018503189087,
+      "learning_rate": 3.809440208720534e-05,
+      "loss": 2.2966,
+      "step": 2962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0660669803619385,
+      "learning_rate": 3.806310344162966e-05,
+      "loss": 2.1189,
+      "step": 2963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2345613241195679,
+      "learning_rate": 3.80318097553509e-05,
+      "loss": 2.3694,
+      "step": 2964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1217395067214966,
+      "learning_rate": 3.800052104137033e-05,
+      "loss": 2.0442,
+      "step": 2965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1095170974731445,
+      "learning_rate": 3.796923731268712e-05,
+      "loss": 2.2264,
+      "step": 2966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0918705463409424,
+      "learning_rate": 3.793795858229842e-05,
+      "loss": 2.171,
+      "step": 2967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0936399698257446,
+      "learning_rate": 3.7906684863199266e-05,
+      "loss": 2.109,
+      "step": 2968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0871905088424683,
+      "learning_rate": 3.7875416168382605e-05,
+      "loss": 1.9891,
+      "step": 2969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2915701866149902,
+      "learning_rate": 3.784415251083932e-05,
+      "loss": 2.1709,
+      "step": 2970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1743053197860718,
+      "learning_rate": 3.7812893903558206e-05,
+      "loss": 2.2043,
+      "step": 2971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2228480577468872,
+      "learning_rate": 3.778164035952594e-05,
+      "loss": 2.3561,
+      "step": 2972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3193913698196411,
+      "learning_rate": 3.77503918917271e-05,
+      "loss": 2.2684,
+      "step": 2973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9961354732513428,
+      "learning_rate": 3.771914851314417e-05,
+      "loss": 2.0346,
+      "step": 2974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0459060668945312,
+      "learning_rate": 3.768791023675748e-05,
+      "loss": 2.0285,
+      "step": 2975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1308026313781738,
+      "learning_rate": 3.765667707554531e-05,
+      "loss": 2.0192,
+      "step": 2976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0578926801681519,
+      "learning_rate": 3.762544904248374e-05,
+      "loss": 2.2264,
+      "step": 2977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.090437412261963,
+      "learning_rate": 3.759422615054678e-05,
+      "loss": 2.1112,
+      "step": 2978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0519940853118896,
+      "learning_rate": 3.7563008412706254e-05,
+      "loss": 2.3481,
+      "step": 2979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3366960287094116,
+      "learning_rate": 3.753179584193187e-05,
+      "loss": 2.3525,
+      "step": 2980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.133583664894104,
+      "learning_rate": 3.750058845119119e-05,
+      "loss": 2.0268,
+      "step": 2981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2235387563705444,
+      "learning_rate": 3.7469386253449624e-05,
+      "loss": 2.4423,
+      "step": 2982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9832154512405396,
+      "learning_rate": 3.743818926167041e-05,
+      "loss": 1.9819,
+      "step": 2983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.334920883178711,
+      "learning_rate": 3.740699748881465e-05,
+      "loss": 2.4792,
+      "step": 2984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1447644233703613,
+      "learning_rate": 3.737581094784126e-05,
+      "loss": 2.0879,
+      "step": 2985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1102259159088135,
+      "learning_rate": 3.7344629651706974e-05,
+      "loss": 2.1366,
+      "step": 2986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.264297604560852,
+      "learning_rate": 3.731345361336636e-05,
+      "loss": 2.2569,
+      "step": 2987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1890358924865723,
+      "learning_rate": 3.728228284577181e-05,
+      "loss": 2.2219,
+      "step": 2988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.011131763458252,
+      "learning_rate": 3.72511173618735e-05,
+      "loss": 1.857,
+      "step": 2989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1817939281463623,
+      "learning_rate": 3.7219957174619435e-05,
+      "loss": 2.026,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1728813648223877,
+      "learning_rate": 3.7188802296955403e-05,
+      "loss": 2.0071,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1379050016403198,
+      "learning_rate": 3.7157652741825e-05,
+      "loss": 2.1484,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.204978585243225,
+      "learning_rate": 3.712650852216961e-05,
+      "loss": 2.1745,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2144535779953003,
+      "learning_rate": 3.709536965092837e-05,
+      "loss": 2.2668,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1340166330337524,
+      "learning_rate": 3.706423614103825e-05,
+      "loss": 2.1805,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2222732305526733,
+      "learning_rate": 3.703310800543397e-05,
+      "loss": 2.1153,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0987876653671265,
+      "learning_rate": 3.700198525704796e-05,
+      "loss": 2.2365,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2468786239624023,
+      "learning_rate": 3.6970867908810494e-05,
+      "loss": 2.4814,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2057362794876099,
+      "learning_rate": 3.693975597364956e-05,
+      "loss": 2.1787,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0984792709350586,
+      "learning_rate": 3.6908649464490915e-05,
+      "loss": 2.2556,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0627754926681519,
+      "learning_rate": 3.687754839425804e-05,
+      "loss": 2.2373,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1246107816696167,
+      "learning_rate": 3.6846452775872194e-05,
+      "loss": 2.3314,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0255030393600464,
+      "learning_rate": 3.6815362622252334e-05,
+      "loss": 1.8578,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1040009260177612,
+      "learning_rate": 3.6784277946315146e-05,
+      "loss": 2.3319,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1248080730438232,
+      "learning_rate": 3.675319876097506e-05,
+      "loss": 1.9496,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.02542245388031,
+      "learning_rate": 3.672212507914423e-05,
+      "loss": 1.9353,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0602060556411743,
+      "learning_rate": 3.669105691373251e-05,
+      "loss": 2.1458,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1893417835235596,
+      "learning_rate": 3.665999427764745e-05,
+      "loss": 2.297,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1057038307189941,
+      "learning_rate": 3.6628937183794324e-05,
+      "loss": 2.1198,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.161980390548706,
+      "learning_rate": 3.6597885645076104e-05,
+      "loss": 2.1473,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3231292963027954,
+      "learning_rate": 3.656683967439344e-05,
+      "loss": 2.3151,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2175874710083008,
+      "learning_rate": 3.653579928464467e-05,
+      "loss": 2.1172,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1974048614501953,
+      "learning_rate": 3.650476448872584e-05,
+      "loss": 2.2253,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1278289556503296,
+      "learning_rate": 3.647373529953063e-05,
+      "loss": 2.0833,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.214345932006836,
+      "learning_rate": 3.64427117299504e-05,
+      "loss": 2.2522,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2579439878463745,
+      "learning_rate": 3.641169379287421e-05,
+      "loss": 2.2475,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1279019117355347,
+      "learning_rate": 3.6380681501188755e-05,
+      "loss": 2.2014,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1462270021438599,
+      "learning_rate": 3.634967486777836e-05,
+      "loss": 2.1593,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0903986692428589,
+      "learning_rate": 3.631867390552506e-05,
+      "loss": 1.8783,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1151155233383179,
+      "learning_rate": 3.6287678627308475e-05,
+      "loss": 2.1396,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2958154678344727,
+      "learning_rate": 3.62566890460059e-05,
+      "loss": 2.3247,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2420563697814941,
+      "learning_rate": 3.6225705174492245e-05,
+      "loss": 2.285,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1221368312835693,
+      "learning_rate": 3.6194727025640055e-05,
+      "loss": 2.4191,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.099870204925537,
+      "learning_rate": 3.616375461231949e-05,
+      "loss": 2.071,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0777273178100586,
+      "learning_rate": 3.613278794739834e-05,
+      "loss": 2.1042,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1725934743881226,
+      "learning_rate": 3.6101827043741995e-05,
+      "loss": 2.2251,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0281099081039429,
+      "learning_rate": 3.6070871914213456e-05,
+      "loss": 2.0371,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1553140878677368,
+      "learning_rate": 3.603992257167334e-05,
+      "loss": 2.2429,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.125572681427002,
+      "learning_rate": 3.600897902897982e-05,
+      "loss": 2.0336,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3807004690170288,
+      "learning_rate": 3.59780412989887e-05,
+      "loss": 2.1991,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.187697410583496,
+      "learning_rate": 3.5947109394553346e-05,
+      "loss": 1.9952,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1063741445541382,
+      "learning_rate": 3.5916183328524724e-05,
+      "loss": 2.1312,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1260043382644653,
+      "learning_rate": 3.588526311375136e-05,
+      "loss": 2.0348,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1589512825012207,
+      "learning_rate": 3.5854348763079336e-05,
+      "loss": 2.1124,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2998945713043213,
+      "learning_rate": 3.582344028935232e-05,
+      "loss": 2.227,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.198371410369873,
+      "learning_rate": 3.5792537705411527e-05,
+      "loss": 2.1121,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2556592226028442,
+      "learning_rate": 3.5761641024095735e-05,
+      "loss": 2.3375,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1458193063735962,
+      "learning_rate": 3.573075025824128e-05,
+      "loss": 2.0622,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2213597297668457,
+      "learning_rate": 3.569986542068199e-05,
+      "loss": 2.2226,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1277366876602173,
+      "learning_rate": 3.566898652424927e-05,
+      "loss": 1.8852,
+      "step": 3040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1191434860229492,
+      "learning_rate": 3.563811358177206e-05,
+      "loss": 1.9036,
+      "step": 3041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1300182342529297,
+      "learning_rate": 3.560724660607681e-05,
+      "loss": 2.1477,
+      "step": 3042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2154111862182617,
+      "learning_rate": 3.557638560998752e-05,
+      "loss": 2.2287,
+      "step": 3043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1706537008285522,
+      "learning_rate": 3.5545530606325654e-05,
+      "loss": 2.0907,
+      "step": 3044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9757716059684753,
+      "learning_rate": 3.5514681607910236e-05,
+      "loss": 2.3721,
+      "step": 3045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0770155191421509,
+      "learning_rate": 3.5483838627557764e-05,
+      "loss": 2.1455,
+      "step": 3046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.144429326057434,
+      "learning_rate": 3.5453001678082255e-05,
+      "loss": 2.2312,
+      "step": 3047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2857710123062134,
+      "learning_rate": 3.542217077229521e-05,
+      "loss": 2.1552,
+      "step": 3048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.239700436592102,
+      "learning_rate": 3.5391345923005613e-05,
+      "loss": 2.2055,
+      "step": 3049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.03708815574646,
+      "learning_rate": 3.536052714301995e-05,
+      "loss": 2.0606,
+      "step": 3050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2818922996520996,
+      "learning_rate": 3.532971444514217e-05,
+      "loss": 2.0816,
+      "step": 3051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2894648313522339,
+      "learning_rate": 3.52989078421737e-05,
+      "loss": 2.3303,
+      "step": 3052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1407487392425537,
+      "learning_rate": 3.526810734691345e-05,
+      "loss": 2.4039,
+      "step": 3053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0559332370758057,
+      "learning_rate": 3.5237312972157754e-05,
+      "loss": 2.265,
+      "step": 3054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9834976196289062,
+      "learning_rate": 3.520652473070044e-05,
+      "loss": 2.0831,
+      "step": 3055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9270179867744446,
+      "learning_rate": 3.5175742635332774e-05,
+      "loss": 2.1743,
+      "step": 3056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1945832967758179,
+      "learning_rate": 3.514496669884347e-05,
+      "loss": 2.2087,
+      "step": 3057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2346687316894531,
+      "learning_rate": 3.5114196934018703e-05,
+      "loss": 2.2328,
+      "step": 3058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2124534845352173,
+      "learning_rate": 3.508343335364202e-05,
+      "loss": 2.1322,
+      "step": 3059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.241300106048584,
+      "learning_rate": 3.505267597049447e-05,
+      "loss": 2.1248,
+      "step": 3060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2074027061462402,
+      "learning_rate": 3.50219247973545e-05,
+      "loss": 2.2396,
+      "step": 3061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2159794569015503,
+      "learning_rate": 3.499117984699797e-05,
+      "loss": 2.2076,
+      "step": 3062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0514881610870361,
+      "learning_rate": 3.4960441132198164e-05,
+      "loss": 2.1289,
+      "step": 3063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1267411708831787,
+      "learning_rate": 3.492970866572577e-05,
+      "loss": 2.2687,
+      "step": 3064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1312416791915894,
+      "learning_rate": 3.4898982460348885e-05,
+      "loss": 2.3251,
+      "step": 3065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0103620290756226,
+      "learning_rate": 3.4868262528833e-05,
+      "loss": 2.1767,
+      "step": 3066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1353061199188232,
+      "learning_rate": 3.483754888394101e-05,
+      "loss": 2.2418,
+      "step": 3067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1467231512069702,
+      "learning_rate": 3.480684153843319e-05,
+      "loss": 2.1912,
+      "step": 3068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2458595037460327,
+      "learning_rate": 3.4776140505067176e-05,
+      "loss": 2.1771,
+      "step": 3069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3903307914733887,
+      "learning_rate": 3.474544579659803e-05,
+      "loss": 2.3196,
+      "step": 3070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0940121412277222,
+      "learning_rate": 3.471475742577815e-05,
+      "loss": 1.903,
+      "step": 3071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1216938495635986,
+      "learning_rate": 3.468407540535731e-05,
+      "loss": 2.0349,
+      "step": 3072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1551539897918701,
+      "learning_rate": 3.465339974808264e-05,
+      "loss": 2.1318,
+      "step": 3073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1730247735977173,
+      "learning_rate": 3.462273046669865e-05,
+      "loss": 2.217,
+      "step": 3074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.292612075805664,
+      "learning_rate": 3.459206757394716e-05,
+      "loss": 2.4625,
+      "step": 3075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.066978096961975,
+      "learning_rate": 3.4561411082567383e-05,
+      "loss": 2.1599,
+      "step": 3076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1204233169555664,
+      "learning_rate": 3.4530761005295845e-05,
+      "loss": 1.8755,
+      "step": 3077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.133314847946167,
+      "learning_rate": 3.450011735486641e-05,
+      "loss": 2.2638,
+      "step": 3078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1562806367874146,
+      "learning_rate": 3.4469480144010256e-05,
+      "loss": 2.1076,
+      "step": 3079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1632139682769775,
+      "learning_rate": 3.4438849385455916e-05,
+      "loss": 2.266,
+      "step": 3080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.201830267906189,
+      "learning_rate": 3.440822509192925e-05,
+      "loss": 2.3369,
+      "step": 3081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.171052098274231,
+      "learning_rate": 3.437760727615337e-05,
+      "loss": 1.9825,
+      "step": 3082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2836536169052124,
+      "learning_rate": 3.434699595084877e-05,
+      "loss": 2.0465,
+      "step": 3083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.134320855140686,
+      "learning_rate": 3.4316391128733197e-05,
+      "loss": 2.0553,
+      "step": 3084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2252637147903442,
+      "learning_rate": 3.4285792822521726e-05,
+      "loss": 2.167,
+      "step": 3085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1480376720428467,
+      "learning_rate": 3.42552010449267e-05,
+      "loss": 2.0102,
+      "step": 3086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3658033609390259,
+      "learning_rate": 3.422461580865778e-05,
+      "loss": 2.1959,
+      "step": 3087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0973422527313232,
+      "learning_rate": 3.419403712642188e-05,
+      "loss": 2.2394,
+      "step": 3088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0328969955444336,
+      "learning_rate": 3.416346501092322e-05,
+      "loss": 1.9258,
+      "step": 3089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0990405082702637,
+      "learning_rate": 3.4132899474863256e-05,
+      "loss": 1.8763,
+      "step": 3090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0280392169952393,
+      "learning_rate": 3.410234053094074e-05,
+      "loss": 2.0412,
+      "step": 3091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.202352523803711,
+      "learning_rate": 3.407178819185167e-05,
+      "loss": 2.2758,
+      "step": 3092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2695547342300415,
+      "learning_rate": 3.4041242470289325e-05,
+      "loss": 2.1457,
+      "step": 3093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1508245468139648,
+      "learning_rate": 3.401070337894419e-05,
+      "loss": 2.1218,
+      "step": 3094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.116688847541809,
+      "learning_rate": 3.398017093050402e-05,
+      "loss": 2.152,
+      "step": 3095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2512667179107666,
+      "learning_rate": 3.394964513765383e-05,
+      "loss": 2.2865,
+      "step": 3096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1920404434204102,
+      "learning_rate": 3.3919126013075834e-05,
+      "loss": 2.2713,
+      "step": 3097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1282846927642822,
+      "learning_rate": 3.388861356944949e-05,
+      "loss": 1.9476,
+      "step": 3098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0807522535324097,
+      "learning_rate": 3.38581078194515e-05,
+      "loss": 2.2917,
+      "step": 3099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0330047607421875,
+      "learning_rate": 3.382760877575575e-05,
+      "loss": 1.9205,
+      "step": 3100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1820330619812012,
+      "learning_rate": 3.3797116451033375e-05,
+      "loss": 2.364,
+      "step": 3101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.172626256942749,
+      "learning_rate": 3.376663085795268e-05,
+      "loss": 2.1738,
+      "step": 3102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0901702642440796,
+      "learning_rate": 3.3736152009179214e-05,
+      "loss": 2.0253,
+      "step": 3103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2348275184631348,
+      "learning_rate": 3.370567991737567e-05,
+      "loss": 2.1187,
+      "step": 3104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0797226428985596,
+      "learning_rate": 3.367521459520201e-05,
+      "loss": 1.9211,
+      "step": 3105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.240177869796753,
+      "learning_rate": 3.364475605531532e-05,
+      "loss": 2.107,
+      "step": 3106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2258336544036865,
+      "learning_rate": 3.3614304310369886e-05,
+      "loss": 2.4295,
+      "step": 3107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2435377836227417,
+      "learning_rate": 3.358385937301718e-05,
+      "loss": 1.8824,
+      "step": 3108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2559667825698853,
+      "learning_rate": 3.3553421255905845e-05,
+      "loss": 2.0935,
+      "step": 3109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0209184885025024,
+      "learning_rate": 3.352298997168167e-05,
+      "loss": 2.0356,
+      "step": 3110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1813994646072388,
+      "learning_rate": 3.349256553298764e-05,
+      "loss": 2.084,
+      "step": 3111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1589832305908203,
+      "learning_rate": 3.3462147952463865e-05,
+      "loss": 2.2051,
+      "step": 3112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.214340329170227,
+      "learning_rate": 3.343173724274763e-05,
+      "loss": 2.2292,
+      "step": 3113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1358150243759155,
+      "learning_rate": 3.340133341647333e-05,
+      "loss": 2.0576,
+      "step": 3114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1226097345352173,
+      "learning_rate": 3.337093648627254e-05,
+      "loss": 2.2006,
+      "step": 3115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0518732070922852,
+      "learning_rate": 3.334054646477395e-05,
+      "loss": 2.1147,
+      "step": 3116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.233135461807251,
+      "learning_rate": 3.331016336460338e-05,
+      "loss": 2.6531,
+      "step": 3117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2094101905822754,
+      "learning_rate": 3.3279787198383773e-05,
+      "loss": 2.2832,
+      "step": 3118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1978839635849,
+      "learning_rate": 3.32494179787352e-05,
+      "loss": 2.0658,
+      "step": 3119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1823810338974,
+      "learning_rate": 3.321905571827485e-05,
+      "loss": 2.2472,
+      "step": 3120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1168357133865356,
+      "learning_rate": 3.3188700429617004e-05,
+      "loss": 2.2488,
+      "step": 3121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1452237367630005,
+      "learning_rate": 3.315835212537305e-05,
+      "loss": 2.3243,
+      "step": 3122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0321886539459229,
+      "learning_rate": 3.312801081815151e-05,
+      "loss": 2.0485,
+      "step": 3123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0571107864379883,
+      "learning_rate": 3.309767652055793e-05,
+      "loss": 2.1043,
+      "step": 3124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3033229112625122,
+      "learning_rate": 3.3067349245194986e-05,
+      "loss": 2.5071,
+      "step": 3125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2223877906799316,
+      "learning_rate": 3.303702900466244e-05,
+      "loss": 2.1988,
+      "step": 3126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1825019121170044,
+      "learning_rate": 3.300671581155713e-05,
+      "loss": 2.3727,
+      "step": 3127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.22109854221344,
+      "learning_rate": 3.297640967847295e-05,
+      "loss": 2.3273,
+      "step": 3128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2024574279785156,
+      "learning_rate": 3.294611061800087e-05,
+      "loss": 2.315,
+      "step": 3129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1196943521499634,
+      "learning_rate": 3.291581864272893e-05,
+      "loss": 2.227,
+      "step": 3130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1583483219146729,
+      "learning_rate": 3.288553376524222e-05,
+      "loss": 2.139,
+      "step": 3131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0342150926589966,
+      "learning_rate": 3.2855255998122856e-05,
+      "loss": 1.93,
+      "step": 3132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0325300693511963,
+      "learning_rate": 3.2824985353950055e-05,
+      "loss": 1.8828,
+      "step": 3133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0582820177078247,
+      "learning_rate": 3.279472184530004e-05,
+      "loss": 2.1219,
+      "step": 3134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0427874326705933,
+      "learning_rate": 3.2764465484746045e-05,
+      "loss": 2.0653,
+      "step": 3135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1720659732818604,
+      "learning_rate": 3.2734216284858386e-05,
+      "loss": 2.1703,
+      "step": 3136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.266379952430725,
+      "learning_rate": 3.270397425820438e-05,
+      "loss": 2.0213,
+      "step": 3137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.265608549118042,
+      "learning_rate": 3.267373941734836e-05,
+      "loss": 2.41,
+      "step": 3138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1742353439331055,
+      "learning_rate": 3.264351177485167e-05,
+      "loss": 1.9882,
+      "step": 3139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.099360704421997,
+      "learning_rate": 3.2613291343272676e-05,
+      "loss": 2.0372,
+      "step": 3140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1127774715423584,
+      "learning_rate": 3.258307813516674e-05,
+      "loss": 2.1838,
+      "step": 3141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3018437623977661,
+      "learning_rate": 3.2552872163086234e-05,
+      "loss": 2.4319,
+      "step": 3142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1400641202926636,
+      "learning_rate": 3.2522673439580513e-05,
+      "loss": 2.2081,
+      "step": 3143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1655497550964355,
+      "learning_rate": 3.249248197719592e-05,
+      "loss": 2.1902,
+      "step": 3144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0686534643173218,
+      "learning_rate": 3.246229778847578e-05,
+      "loss": 2.172,
+      "step": 3145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0453447103500366,
+      "learning_rate": 3.243212088596042e-05,
+      "loss": 2.3014,
+      "step": 3146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.117150068283081,
+      "learning_rate": 3.2401951282187094e-05,
+      "loss": 1.9937,
+      "step": 3147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0770418643951416,
+      "learning_rate": 3.2371788989690075e-05,
+      "loss": 2.3083,
+      "step": 3148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3093047142028809,
+      "learning_rate": 3.234163402100055e-05,
+      "loss": 2.1971,
+      "step": 3149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1859861612319946,
+      "learning_rate": 3.23114863886467e-05,
+      "loss": 1.9829,
+      "step": 3150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1883084774017334,
+      "learning_rate": 3.228134610515364e-05,
+      "loss": 2.0839,
+      "step": 3151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1883084774017334,
+      "learning_rate": 3.228134610515364e-05,
+      "loss": 2.2295,
+      "step": 3152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1301331520080566,
+      "learning_rate": 3.225121318304345e-05,
+      "loss": 1.8257,
+      "step": 3153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.211375117301941,
+      "learning_rate": 3.222108763483512e-05,
+      "loss": 2.0377,
+      "step": 3154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.300515055656433,
+      "learning_rate": 3.219096947304461e-05,
+      "loss": 1.9759,
+      "step": 3155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1109033823013306,
+      "learning_rate": 3.2160858710184784e-05,
+      "loss": 2.1161,
+      "step": 3156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1109033823013306,
+      "learning_rate": 3.2160858710184784e-05,
+      "loss": 2.0588,
+      "step": 3157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2990118265151978,
+      "learning_rate": 3.2130755358765444e-05,
+      "loss": 2.4035,
+      "step": 3158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2160468101501465,
+      "learning_rate": 3.2100659431293324e-05,
+      "loss": 1.918,
+      "step": 3159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2777607440948486,
+      "learning_rate": 3.2070570940272066e-05,
+      "loss": 1.9149,
+      "step": 3160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.210695505142212,
+      "learning_rate": 3.2040489898202197e-05,
+      "loss": 2.0615,
+      "step": 3161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2806577682495117,
+      "learning_rate": 3.201041631758119e-05,
+      "loss": 2.1799,
+      "step": 3162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.988571286201477,
+      "learning_rate": 3.198035021090339e-05,
+      "loss": 1.9988,
+      "step": 3163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0952222347259521,
+      "learning_rate": 3.1950291590660055e-05,
+      "loss": 2.1506,
+      "step": 3164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3533520698547363,
+      "learning_rate": 3.192024046933931e-05,
+      "loss": 1.9274,
+      "step": 3165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1861109733581543,
+      "learning_rate": 3.189019685942619e-05,
+      "loss": 2.1127,
+      "step": 3166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2303154468536377,
+      "learning_rate": 3.186016077340259e-05,
+      "loss": 2.2696,
+      "step": 3167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2069785594940186,
+      "learning_rate": 3.1830132223747297e-05,
+      "loss": 2.1262,
+      "step": 3168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.175464391708374,
+      "learning_rate": 3.180011122293598e-05,
+      "loss": 1.9767,
+      "step": 3169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0859519243240356,
+      "learning_rate": 3.177009778344109e-05,
+      "loss": 2.1838,
+      "step": 3170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.122432827949524,
+      "learning_rate": 3.174009191773203e-05,
+      "loss": 1.8703,
+      "step": 3171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.265146017074585,
+      "learning_rate": 3.171009363827502e-05,
+      "loss": 2.1106,
+      "step": 3172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2760982513427734,
+      "learning_rate": 3.168010295753312e-05,
+      "loss": 2.1594,
+      "step": 3173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1076937913894653,
+      "learning_rate": 3.165011988796626e-05,
+      "loss": 2.2118,
+      "step": 3174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1700499057769775,
+      "learning_rate": 3.162014444203117e-05,
+      "loss": 2.0997,
+      "step": 3175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3898290395736694,
+      "learning_rate": 3.159017663218144e-05,
+      "loss": 2.5207,
+      "step": 3176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1518590450286865,
+      "learning_rate": 3.156021647086749e-05,
+      "loss": 2.2853,
+      "step": 3177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1180380582809448,
+      "learning_rate": 3.1530263970536555e-05,
+      "loss": 1.9604,
+      "step": 3178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.11062753200531,
+      "learning_rate": 3.150031914363267e-05,
+      "loss": 2.3345,
+      "step": 3179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3269715309143066,
+      "learning_rate": 3.1470382002596705e-05,
+      "loss": 2.0409,
+      "step": 3180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1370004415512085,
+      "learning_rate": 3.144045255986633e-05,
+      "loss": 2.0668,
+      "step": 3181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.190418004989624,
+      "learning_rate": 3.141053082787602e-05,
+      "loss": 2.194,
+      "step": 3182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3042305707931519,
+      "learning_rate": 3.1380616819057034e-05,
+      "loss": 2.1391,
+      "step": 3183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1703323125839233,
+      "learning_rate": 3.135071054583745e-05,
+      "loss": 2.057,
+      "step": 3184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2317429780960083,
+      "learning_rate": 3.1320812020642084e-05,
+      "loss": 2.2836,
+      "step": 3185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4658677577972412,
+      "learning_rate": 3.129092125589258e-05,
+      "loss": 2.3151,
+      "step": 3186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2464203834533691,
+      "learning_rate": 3.126103826400733e-05,
+      "loss": 2.2345,
+      "step": 3187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2525060176849365,
+      "learning_rate": 3.123116305740151e-05,
+      "loss": 2.3823,
+      "step": 3188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.176871418952942,
+      "learning_rate": 3.120129564848707e-05,
+      "loss": 2.1454,
+      "step": 3189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0958044528961182,
+      "learning_rate": 3.11714360496727e-05,
+      "loss": 1.9969,
+      "step": 3190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1029804944992065,
+      "learning_rate": 3.114158427336385e-05,
+      "loss": 2.1906,
+      "step": 3191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.045389175415039,
+      "learning_rate": 3.1111740331962714e-05,
+      "loss": 2.0016,
+      "step": 3192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0808098316192627,
+      "learning_rate": 3.108190423786827e-05,
+      "loss": 2.1157,
+      "step": 3193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1432369947433472,
+      "learning_rate": 3.105207600347617e-05,
+      "loss": 2.0555,
+      "step": 3194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1836111545562744,
+      "learning_rate": 3.1022255641178875e-05,
+      "loss": 2.0127,
+      "step": 3195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9941319823265076,
+      "learning_rate": 3.099244316336551e-05,
+      "loss": 1.9212,
+      "step": 3196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1276236772537231,
+      "learning_rate": 3.096263858242196e-05,
+      "loss": 2.0513,
+      "step": 3197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1445965766906738,
+      "learning_rate": 3.093284191073082e-05,
+      "loss": 1.9671,
+      "step": 3198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2364290952682495,
+      "learning_rate": 3.09030531606714e-05,
+      "loss": 2.0163,
+      "step": 3199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.119657278060913,
+      "learning_rate": 3.0873272344619717e-05,
+      "loss": 1.9884,
+      "step": 3200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3854775428771973,
+      "learning_rate": 3.08434994749485e-05,
+      "loss": 2.2064,
+      "step": 3201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1750646829605103,
+      "learning_rate": 3.081373456402717e-05,
+      "loss": 1.8914,
+      "step": 3202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.297029972076416,
+      "learning_rate": 3.078397762422184e-05,
+      "loss": 2.0069,
+      "step": 3203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2778465747833252,
+      "learning_rate": 3.07542286678953e-05,
+      "loss": 2.4795,
+      "step": 3204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1434937715530396,
+      "learning_rate": 3.0724487707407065e-05,
+      "loss": 2.1872,
+      "step": 3205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2102333307266235,
+      "learning_rate": 3.069475475511326e-05,
+      "loss": 2.2633,
+      "step": 3206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2206382751464844,
+      "learning_rate": 3.0665029823366757e-05,
+      "loss": 2.3755,
+      "step": 3207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.163025140762329,
+      "learning_rate": 3.063531292451705e-05,
+      "loss": 2.2095,
+      "step": 3208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2557752132415771,
+      "learning_rate": 3.0605604070910286e-05,
+      "loss": 2.3245,
+      "step": 3209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.161940336227417,
+      "learning_rate": 3.0575903274889325e-05,
+      "loss": 2.2691,
+      "step": 3210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2551060914993286,
+      "learning_rate": 3.054621054879364e-05,
+      "loss": 2.4936,
+      "step": 3211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9778130054473877,
+      "learning_rate": 3.0516525904959325e-05,
+      "loss": 2.1533,
+      "step": 3212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1208423376083374,
+      "learning_rate": 3.0486849355719162e-05,
+      "loss": 2.0465,
+      "step": 3213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9677853584289551,
+      "learning_rate": 3.0457180913402562e-05,
+      "loss": 2.2176,
+      "step": 3214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1083478927612305,
+      "learning_rate": 3.0427520590335558e-05,
+      "loss": 2.1367,
+      "step": 3215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.156612515449524,
+      "learning_rate": 3.03978683988408e-05,
+      "loss": 2.2426,
+      "step": 3216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0157347917556763,
+      "learning_rate": 3.0368224351237594e-05,
+      "loss": 2.2176,
+      "step": 3217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2536413669586182,
+      "learning_rate": 3.0338588459841832e-05,
+      "loss": 2.2382,
+      "step": 3218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3270387649536133,
+      "learning_rate": 3.030896073696603e-05,
+      "loss": 2.4326,
+      "step": 3219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0081274509429932,
+      "learning_rate": 3.02793411949193e-05,
+      "loss": 2.219,
+      "step": 3220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1447783708572388,
+      "learning_rate": 3.0249729846007352e-05,
+      "loss": 2.1165,
+      "step": 3221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0474432706832886,
+      "learning_rate": 3.0220126702532515e-05,
+      "loss": 1.9471,
+      "step": 3222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1293374300003052,
+      "learning_rate": 3.0190531776793695e-05,
+      "loss": 1.9399,
+      "step": 3223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2466857433319092,
+      "learning_rate": 3.016094508108639e-05,
+      "loss": 2.1301,
+      "step": 3224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1445633172988892,
+      "learning_rate": 3.013136662770266e-05,
+      "loss": 1.9958,
+      "step": 3225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2026582956314087,
+      "learning_rate": 3.010179642893116e-05,
+      "loss": 2.1397,
+      "step": 3226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1790103912353516,
+      "learning_rate": 3.0072234497057117e-05,
+      "loss": 2.1509,
+      "step": 3227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2894244194030762,
+      "learning_rate": 3.0042680844362305e-05,
+      "loss": 2.0596,
+      "step": 3228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2558839321136475,
+      "learning_rate": 3.0013135483125077e-05,
+      "loss": 2.1931,
+      "step": 3229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3021998405456543,
+      "learning_rate": 2.998359842562033e-05,
+      "loss": 2.193,
+      "step": 3230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2305383682250977,
+      "learning_rate": 2.9954069684119513e-05,
+      "loss": 2.1253,
+      "step": 3231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2355146408081055,
+      "learning_rate": 2.9924549270890622e-05,
+      "loss": 2.3047,
+      "step": 3232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0367801189422607,
+      "learning_rate": 2.98950371981982e-05,
+      "loss": 2.0841,
+      "step": 3233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0727219581604004,
+      "learning_rate": 2.9865533478303308e-05,
+      "loss": 2.1911,
+      "step": 3234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0394127368927002,
+      "learning_rate": 2.9836038123463557e-05,
+      "loss": 1.8414,
+      "step": 3235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1524858474731445,
+      "learning_rate": 2.980655114593306e-05,
+      "loss": 2.193,
+      "step": 3236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0478854179382324,
+      "learning_rate": 2.977707255796247e-05,
+      "loss": 2.049,
+      "step": 3237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.26347017288208,
+      "learning_rate": 2.9747602371798944e-05,
+      "loss": 2.3708,
+      "step": 3238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0783021450042725,
+      "learning_rate": 2.9718140599686156e-05,
+      "loss": 1.9788,
+      "step": 3239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.209764003753662,
+      "learning_rate": 2.9688687253864265e-05,
+      "loss": 2.3015,
+      "step": 3240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0339103937149048,
+      "learning_rate": 2.9659242346569948e-05,
+      "loss": 2.0799,
+      "step": 3241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0658990144729614,
+      "learning_rate": 2.962980589003638e-05,
+      "loss": 2.102,
+      "step": 3242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2140477895736694,
+      "learning_rate": 2.9600377896493203e-05,
+      "loss": 2.3246,
+      "step": 3243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2150501012802124,
+      "learning_rate": 2.9570958378166567e-05,
+      "loss": 2.1697,
+      "step": 3244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9891424775123596,
+      "learning_rate": 2.9541547347279087e-05,
+      "loss": 2.2469,
+      "step": 3245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2906655073165894,
+      "learning_rate": 2.9512144816049846e-05,
+      "loss": 2.1362,
+      "step": 3246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2736527919769287,
+      "learning_rate": 2.9482750796694414e-05,
+      "loss": 2.5341,
+      "step": 3247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3342067003250122,
+      "learning_rate": 2.945336530142482e-05,
+      "loss": 2.4649,
+      "step": 3248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3574109077453613,
+      "learning_rate": 2.9423988342449537e-05,
+      "loss": 2.0472,
+      "step": 3249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.072100281715393,
+      "learning_rate": 2.939461993197351e-05,
+      "loss": 2.1969,
+      "step": 3250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2049839496612549,
+      "learning_rate": 2.936526008219812e-05,
+      "loss": 2.3542,
+      "step": 3251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1547746658325195,
+      "learning_rate": 2.9335908805321182e-05,
+      "loss": 2.1699,
+      "step": 3252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.254075288772583,
+      "learning_rate": 2.9306566113536983e-05,
+      "loss": 2.0672,
+      "step": 3253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.189421534538269,
+      "learning_rate": 2.9277232019036228e-05,
+      "loss": 2.2341,
+      "step": 3254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0738204717636108,
+      "learning_rate": 2.9247906534006007e-05,
+      "loss": 1.8212,
+      "step": 3255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0931024551391602,
+      "learning_rate": 2.9218589670629898e-05,
+      "loss": 2.2205,
+      "step": 3256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1673890352249146,
+      "learning_rate": 2.9189281441087857e-05,
+      "loss": 1.9966,
+      "step": 3257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.091571569442749,
+      "learning_rate": 2.9159981857556285e-05,
+      "loss": 2.1894,
+      "step": 3258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1453078985214233,
+      "learning_rate": 2.9130690932207983e-05,
+      "loss": 2.0338,
+      "step": 3259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1274535655975342,
+      "learning_rate": 2.91014086772121e-05,
+      "loss": 1.942,
+      "step": 3260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1736867427825928,
+      "learning_rate": 2.9072135104734257e-05,
+      "loss": 2.0529,
+      "step": 3261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0778629779815674,
+      "learning_rate": 2.9042870226936424e-05,
+      "loss": 2.1094,
+      "step": 3262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.056606411933899,
+      "learning_rate": 2.901361405597698e-05,
+      "loss": 2.0587,
+      "step": 3263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2633581161499023,
+      "learning_rate": 2.898436660401067e-05,
+      "loss": 2.2168,
+      "step": 3264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.140537977218628,
+      "learning_rate": 2.8955127883188644e-05,
+      "loss": 2.241,
+      "step": 3265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2646980285644531,
+      "learning_rate": 2.892589790565839e-05,
+      "loss": 1.9555,
+      "step": 3266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1808679103851318,
+      "learning_rate": 2.8896676683563783e-05,
+      "loss": 1.9183,
+      "step": 3267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.28373122215271,
+      "learning_rate": 2.886746422904506e-05,
+      "loss": 2.1591,
+      "step": 3268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2036153078079224,
+      "learning_rate": 2.8838260554238806e-05,
+      "loss": 2.1437,
+      "step": 3269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1729240417480469,
+      "learning_rate": 2.8809065671277967e-05,
+      "loss": 1.9726,
+      "step": 3270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.323307991027832,
+      "learning_rate": 2.877987959229185e-05,
+      "loss": 2.2199,
+      "step": 3271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2161145210266113,
+      "learning_rate": 2.8750702329406066e-05,
+      "loss": 2.3176,
+      "step": 3272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.204622507095337,
+      "learning_rate": 2.8721533894742593e-05,
+      "loss": 2.2693,
+      "step": 3273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3548904657363892,
+      "learning_rate": 2.8692374300419732e-05,
+      "loss": 2.1374,
+      "step": 3274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2658134698867798,
+      "learning_rate": 2.8663223558552123e-05,
+      "loss": 2.0252,
+      "step": 3275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.147983193397522,
+      "learning_rate": 2.8634081681250702e-05,
+      "loss": 2.0828,
+      "step": 3276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3287144899368286,
+      "learning_rate": 2.8604948680622756e-05,
+      "loss": 2.0735,
+      "step": 3277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3575403690338135,
+      "learning_rate": 2.8575824568771857e-05,
+      "loss": 2.1823,
+      "step": 3278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.182281732559204,
+      "learning_rate": 2.854670935779792e-05,
+      "loss": 1.914,
+      "step": 3279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2428412437438965,
+      "learning_rate": 2.851760305979708e-05,
+      "loss": 1.9483,
+      "step": 3280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.124841332435608,
+      "learning_rate": 2.848850568686186e-05,
+      "loss": 2.0192,
+      "step": 3281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.302913784980774,
+      "learning_rate": 2.8459417251081034e-05,
+      "loss": 2.225,
+      "step": 3282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2872365713119507,
+      "learning_rate": 2.8430337764539672e-05,
+      "loss": 2.1112,
+      "step": 3283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3106504678726196,
+      "learning_rate": 2.840126723931912e-05,
+      "loss": 2.266,
+      "step": 3284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1952452659606934,
+      "learning_rate": 2.8372205687496998e-05,
+      "loss": 2.0818,
+      "step": 3285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2315983772277832,
+      "learning_rate": 2.8343153121147214e-05,
+      "loss": 2.174,
+      "step": 3286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0574779510498047,
+      "learning_rate": 2.8314109552339927e-05,
+      "loss": 1.9083,
+      "step": 3287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2876168489456177,
+      "learning_rate": 2.8285074993141564e-05,
+      "loss": 2.1448,
+      "step": 3288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3282440900802612,
+      "learning_rate": 2.8256049455614807e-05,
+      "loss": 2.426,
+      "step": 3289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.247812271118164,
+      "learning_rate": 2.8227032951818587e-05,
+      "loss": 1.9267,
+      "step": 3290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2607022523880005,
+      "learning_rate": 2.81980254938081e-05,
+      "loss": 1.8905,
+      "step": 3291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1012412309646606,
+      "learning_rate": 2.8169027093634747e-05,
+      "loss": 1.8109,
+      "step": 3292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1986079216003418,
+      "learning_rate": 2.814003776334621e-05,
+      "loss": 2.2316,
+      "step": 3293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3495522737503052,
+      "learning_rate": 2.811105751498637e-05,
+      "loss": 2.1904,
+      "step": 3294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.331513524055481,
+      "learning_rate": 2.808208636059535e-05,
+      "loss": 2.1841,
+      "step": 3295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3004357814788818,
+      "learning_rate": 2.805312431220949e-05,
+      "loss": 2.1623,
+      "step": 3296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1404954195022583,
+      "learning_rate": 2.8024171381861364e-05,
+      "loss": 2.0083,
+      "step": 3297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.978043794631958,
+      "learning_rate": 2.799522758157971e-05,
+      "loss": 1.8331,
+      "step": 3298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1134780645370483,
+      "learning_rate": 2.7966292923389513e-05,
+      "loss": 2.0457,
+      "step": 3299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.194579005241394,
+      "learning_rate": 2.7937367419311966e-05,
+      "loss": 1.9911,
+      "step": 3300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1336421966552734,
+      "learning_rate": 2.7908451081364428e-05,
+      "loss": 2.1291,
+      "step": 3301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.03590989112854,
+      "learning_rate": 2.787954392156048e-05,
+      "loss": 2.0757,
+      "step": 3302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3903228044509888,
+      "learning_rate": 2.785064595190987e-05,
+      "loss": 2.115,
+      "step": 3303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1758348941802979,
+      "learning_rate": 2.7821757184418552e-05,
+      "loss": 1.9384,
+      "step": 3304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2432326078414917,
+      "learning_rate": 2.7792877631088592e-05,
+      "loss": 2.0337,
+      "step": 3305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9622214436531067,
+      "learning_rate": 2.7764007303918304e-05,
+      "loss": 2.0095,
+      "step": 3306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0028340816497803,
+      "learning_rate": 2.7735146214902146e-05,
+      "loss": 1.9703,
+      "step": 3307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2362873554229736,
+      "learning_rate": 2.770629437603072e-05,
+      "loss": 2.3267,
+      "step": 3308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2478605508804321,
+      "learning_rate": 2.7677451799290792e-05,
+      "loss": 1.9836,
+      "step": 3309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3041118383407593,
+      "learning_rate": 2.76486184966653e-05,
+      "loss": 2.129,
+      "step": 3310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0522397756576538,
+      "learning_rate": 2.7619794480133297e-05,
+      "loss": 1.8429,
+      "step": 3311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1974029541015625,
+      "learning_rate": 2.7590979761670003e-05,
+      "loss": 1.9325,
+      "step": 3312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3031870126724243,
+      "learning_rate": 2.7562174353246767e-05,
+      "loss": 2.2369,
+      "step": 3313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3466713428497314,
+      "learning_rate": 2.753337826683107e-05,
+      "loss": 2.2946,
+      "step": 3314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1482619047164917,
+      "learning_rate": 2.7504591514386507e-05,
+      "loss": 2.3661,
+      "step": 3315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4295260906219482,
+      "learning_rate": 2.747581410787282e-05,
+      "loss": 2.3529,
+      "step": 3316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3010493516921997,
+      "learning_rate": 2.744704605924585e-05,
+      "loss": 2.2599,
+      "step": 3317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.313651204109192,
+      "learning_rate": 2.7418287380457563e-05,
+      "loss": 2.2699,
+      "step": 3318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1357613801956177,
+      "learning_rate": 2.738953808345601e-05,
+      "loss": 1.8909,
+      "step": 3319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0895934104919434,
+      "learning_rate": 2.736079818018538e-05,
+      "loss": 1.9548,
+      "step": 3320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.167822241783142,
+      "learning_rate": 2.7332067682585927e-05,
+      "loss": 2.2412,
+      "step": 3321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1073156595230103,
+      "learning_rate": 2.7303346602594e-05,
+      "loss": 2.0822,
+      "step": 3322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2051808834075928,
+      "learning_rate": 2.727463495214206e-05,
+      "loss": 2.0887,
+      "step": 3323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1608835458755493,
+      "learning_rate": 2.724593274315864e-05,
+      "loss": 2.3123,
+      "step": 3324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1947855949401855,
+      "learning_rate": 2.721723998756832e-05,
+      "loss": 1.9865,
+      "step": 3325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.139821171760559,
+      "learning_rate": 2.718855669729179e-05,
+      "loss": 2.1547,
+      "step": 3326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0599219799041748,
+      "learning_rate": 2.715988288424579e-05,
+      "loss": 1.9754,
+      "step": 3327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.028922200202942,
+      "learning_rate": 2.713121856034314e-05,
+      "loss": 2.2565,
+      "step": 3328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.141190767288208,
+      "learning_rate": 2.710256373749269e-05,
+      "loss": 2.4646,
+      "step": 3329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1700505018234253,
+      "learning_rate": 2.7073918427599377e-05,
+      "loss": 2.0688,
+      "step": 3330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0706321001052856,
+      "learning_rate": 2.704528264256415e-05,
+      "loss": 2.0308,
+      "step": 3331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1114405393600464,
+      "learning_rate": 2.7016656394284023e-05,
+      "loss": 2.1081,
+      "step": 3332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2177112102508545,
+      "learning_rate": 2.6988039694652045e-05,
+      "loss": 2.4283,
+      "step": 3333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2327845096588135,
+      "learning_rate": 2.695943255555729e-05,
+      "loss": 2.2712,
+      "step": 3334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0380595922470093,
+      "learning_rate": 2.6930834988884867e-05,
+      "loss": 1.9967,
+      "step": 3335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2641569375991821,
+      "learning_rate": 2.6902247006515913e-05,
+      "loss": 2.1936,
+      "step": 3336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.214482307434082,
+      "learning_rate": 2.6873668620327573e-05,
+      "loss": 1.8527,
+      "step": 3337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0687963962554932,
+      "learning_rate": 2.6845099842193007e-05,
+      "loss": 2.1049,
+      "step": 3338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.277461290359497,
+      "learning_rate": 2.6816540683981384e-05,
+      "loss": 1.9995,
+      "step": 3339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.256076455116272,
+      "learning_rate": 2.6787991157557897e-05,
+      "loss": 2.1196,
+      "step": 3340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1049776077270508,
+      "learning_rate": 2.6759451274783674e-05,
+      "loss": 2.0905,
+      "step": 3341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2376770973205566,
+      "learning_rate": 2.673092104751591e-05,
+      "loss": 2.0939,
+      "step": 3342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1897879838943481,
+      "learning_rate": 2.670240048760776e-05,
+      "loss": 1.95,
+      "step": 3343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1944849491119385,
+      "learning_rate": 2.667388960690835e-05,
+      "loss": 1.9916,
+      "step": 3344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1553385257720947,
+      "learning_rate": 2.6645388417262797e-05,
+      "loss": 2.09,
+      "step": 3345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2443220615386963,
+      "learning_rate": 2.6616896930512203e-05,
+      "loss": 2.0167,
+      "step": 3346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1409354209899902,
+      "learning_rate": 2.6588415158493605e-05,
+      "loss": 2.0987,
+      "step": 3347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1810201406478882,
+      "learning_rate": 2.6559943113040054e-05,
+      "loss": 2.1393,
+      "step": 3348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.17218017578125,
+      "learning_rate": 2.6531480805980523e-05,
+      "loss": 2.0955,
+      "step": 3349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2502883672714233,
+      "learning_rate": 2.650302824913993e-05,
+      "loss": 2.0688,
+      "step": 3350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1386139392852783,
+      "learning_rate": 2.6474585454339164e-05,
+      "loss": 2.0175,
+      "step": 3351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0521570444107056,
+      "learning_rate": 2.6446152433395056e-05,
+      "loss": 2.3007,
+      "step": 3352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1121586561203003,
+      "learning_rate": 2.641772919812038e-05,
+      "loss": 2.0012,
+      "step": 3353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0306673049926758,
+      "learning_rate": 2.638931576032384e-05,
+      "loss": 1.842,
+      "step": 3354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0795143842697144,
+      "learning_rate": 2.6360912131810066e-05,
+      "loss": 2.2761,
+      "step": 3355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.036980152130127,
+      "learning_rate": 2.6332518324379607e-05,
+      "loss": 2.0784,
+      "step": 3356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.400142788887024,
+      "learning_rate": 2.630413434982896e-05,
+      "loss": 2.2607,
+      "step": 3357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2432442903518677,
+      "learning_rate": 2.627576021995049e-05,
+      "loss": 2.3911,
+      "step": 3358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1127339601516724,
+      "learning_rate": 2.6247395946532517e-05,
+      "loss": 1.9977,
+      "step": 3359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1018630266189575,
+      "learning_rate": 2.6219041541359248e-05,
+      "loss": 1.8527,
+      "step": 3360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0926493406295776,
+      "learning_rate": 2.6190697016210776e-05,
+      "loss": 2.15,
+      "step": 3361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1011204719543457,
+      "learning_rate": 2.616236238286311e-05,
+      "loss": 2.1976,
+      "step": 3362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1705886125564575,
+      "learning_rate": 2.6134037653088144e-05,
+      "loss": 2.1584,
+      "step": 3363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1695728302001953,
+      "learning_rate": 2.6105722838653657e-05,
+      "loss": 2.1394,
+      "step": 3364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2505104541778564,
+      "learning_rate": 2.60774179513233e-05,
+      "loss": 2.1444,
+      "step": 3365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.174416184425354,
+      "learning_rate": 2.604912300285662e-05,
+      "loss": 2.1551,
+      "step": 3366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3489195108413696,
+      "learning_rate": 2.6020838005009008e-05,
+      "loss": 2.0429,
+      "step": 3367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.098331093788147,
+      "learning_rate": 2.599256296953174e-05,
+      "loss": 2.0172,
+      "step": 3368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.179227352142334,
+      "learning_rate": 2.5964297908171968e-05,
+      "loss": 2.0358,
+      "step": 3369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2146906852722168,
+      "learning_rate": 2.5936042832672647e-05,
+      "loss": 2.1713,
+      "step": 3370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2098674774169922,
+      "learning_rate": 2.5907797754772622e-05,
+      "loss": 1.9669,
+      "step": 3371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0979505777359009,
+      "learning_rate": 2.5879562686206586e-05,
+      "loss": 2.133,
+      "step": 3372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2448256015777588,
+      "learning_rate": 2.5851337638705064e-05,
+      "loss": 1.9035,
+      "step": 3373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.158939242362976,
+      "learning_rate": 2.5823122623994427e-05,
+      "loss": 2.0793,
+      "step": 3374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1762070655822754,
+      "learning_rate": 2.5794917653796867e-05,
+      "loss": 2.347,
+      "step": 3375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2338398694992065,
+      "learning_rate": 2.5766722739830396e-05,
+      "loss": 2.2603,
+      "step": 3376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3077367544174194,
+      "learning_rate": 2.5738537893808877e-05,
+      "loss": 1.9788,
+      "step": 3377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0816043615341187,
+      "learning_rate": 2.5710363127441954e-05,
+      "loss": 2.1956,
+      "step": 3378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.107358455657959,
+      "learning_rate": 2.5682198452435114e-05,
+      "loss": 1.9345,
+      "step": 3379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1692085266113281,
+      "learning_rate": 2.565404388048963e-05,
+      "loss": 2.216,
+      "step": 3380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1764472723007202,
+      "learning_rate": 2.562589942330259e-05,
+      "loss": 1.9915,
+      "step": 3381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.327494502067566,
+      "learning_rate": 2.5597765092566893e-05,
+      "loss": 2.2034,
+      "step": 3382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2146849632263184,
+      "learning_rate": 2.556964089997118e-05,
+      "loss": 2.0304,
+      "step": 3383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2631030082702637,
+      "learning_rate": 2.5541526857199928e-05,
+      "loss": 2.1489,
+      "step": 3384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.075334072113037,
+      "learning_rate": 2.5513422975933376e-05,
+      "loss": 1.8599,
+      "step": 3385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1492429971694946,
+      "learning_rate": 2.548532926784757e-05,
+      "loss": 2.209,
+      "step": 3386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1115727424621582,
+      "learning_rate": 2.5457245744614284e-05,
+      "loss": 2.0628,
+      "step": 3387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2788714170455933,
+      "learning_rate": 2.5429172417901093e-05,
+      "loss": 2.3627,
+      "step": 3388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1019207239151,
+      "learning_rate": 2.5401109299371328e-05,
+      "loss": 2.1066,
+      "step": 3389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1694954633712769,
+      "learning_rate": 2.537305640068408e-05,
+      "loss": 2.0739,
+      "step": 3390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2126094102859497,
+      "learning_rate": 2.5345013733494196e-05,
+      "loss": 2.0673,
+      "step": 3391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2479236125946045,
+      "learning_rate": 2.5316981309452252e-05,
+      "loss": 2.156,
+      "step": 3392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1895098686218262,
+      "learning_rate": 2.52889591402046e-05,
+      "loss": 2.12,
+      "step": 3393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1115283966064453,
+      "learning_rate": 2.526094723739333e-05,
+      "loss": 2.1929,
+      "step": 3394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.231583833694458,
+      "learning_rate": 2.5232945612656207e-05,
+      "loss": 2.0917,
+      "step": 3395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.181118130683899,
+      "learning_rate": 2.5204954277626803e-05,
+      "loss": 1.8313,
+      "step": 3396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3142009973526,
+      "learning_rate": 2.5176973243934375e-05,
+      "loss": 2.2146,
+      "step": 3397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2661306858062744,
+      "learning_rate": 2.514900252320391e-05,
+      "loss": 2.0276,
+      "step": 3398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.156557321548462,
+      "learning_rate": 2.5121042127056115e-05,
+      "loss": 1.7123,
+      "step": 3399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2904441356658936,
+      "learning_rate": 2.5093092067107405e-05,
+      "loss": 1.902,
+      "step": 3400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.328402042388916,
+      "learning_rate": 2.506515235496988e-05,
+      "loss": 2.3528,
+      "step": 3401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.016461968421936,
+      "learning_rate": 2.5037223002251374e-05,
+      "loss": 2.1305,
+      "step": 3402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1606156826019287,
+      "learning_rate": 2.50093040205554e-05,
+      "loss": 2.0724,
+      "step": 3403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.204656720161438,
+      "learning_rate": 2.4981395421481156e-05,
+      "loss": 2.0481,
+      "step": 3404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1604260206222534,
+      "learning_rate": 2.495349721662354e-05,
+      "loss": 2.1515,
+      "step": 3405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2191580533981323,
+      "learning_rate": 2.4925609417573127e-05,
+      "loss": 2.2237,
+      "step": 3406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.259335994720459,
+      "learning_rate": 2.4897732035916167e-05,
+      "loss": 2.004,
+      "step": 3407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2065887451171875,
+      "learning_rate": 2.486986508323457e-05,
+      "loss": 2.0155,
+      "step": 3408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3458106517791748,
+      "learning_rate": 2.4842008571105953e-05,
+      "loss": 2.2064,
+      "step": 3409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2277822494506836,
+      "learning_rate": 2.4814162511103546e-05,
+      "loss": 2.3869,
+      "step": 3410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2867904901504517,
+      "learning_rate": 2.478632691479626e-05,
+      "loss": 2.138,
+      "step": 3411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1251308917999268,
+      "learning_rate": 2.475850179374867e-05,
+      "loss": 1.9588,
+      "step": 3412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2571005821228027,
+      "learning_rate": 2.473068715952097e-05,
+      "loss": 2.1981,
+      "step": 3413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.097315788269043,
+      "learning_rate": 2.470288302366905e-05,
+      "loss": 1.9576,
+      "step": 3414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9353156685829163,
+      "learning_rate": 2.4675089397744355e-05,
+      "loss": 1.9574,
+      "step": 3415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1069005727767944,
+      "learning_rate": 2.464730629329403e-05,
+      "loss": 2.0946,
+      "step": 3416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2039896249771118,
+      "learning_rate": 2.4619533721860828e-05,
+      "loss": 2.3056,
+      "step": 3417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9974215030670166,
+      "learning_rate": 2.4591771694983128e-05,
+      "loss": 2.0548,
+      "step": 3418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0523213148117065,
+      "learning_rate": 2.4564020224194932e-05,
+      "loss": 2.2125,
+      "step": 3419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2504395246505737,
+      "learning_rate": 2.453627932102585e-05,
+      "loss": 2.1961,
+      "step": 3420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0200324058532715,
+      "learning_rate": 2.4508548997001102e-05,
+      "loss": 2.0496,
+      "step": 3421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1559205055236816,
+      "learning_rate": 2.448082926364151e-05,
+      "loss": 2.1994,
+      "step": 3422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2485624551773071,
+      "learning_rate": 2.4453120132463504e-05,
+      "loss": 2.0702,
+      "step": 3423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1439924240112305,
+      "learning_rate": 2.442542161497911e-05,
+      "loss": 2.2911,
+      "step": 3424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2523905038833618,
+      "learning_rate": 2.4397733722695946e-05,
+      "loss": 2.1811,
+      "step": 3425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1082782745361328,
+      "learning_rate": 2.437005646711719e-05,
+      "loss": 2.0216,
+      "step": 3426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3258651494979858,
+      "learning_rate": 2.4342389859741625e-05,
+      "loss": 2.1143,
+      "step": 3427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1543132066726685,
+      "learning_rate": 2.431473391206361e-05,
+      "loss": 2.1503,
+      "step": 3428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0844066143035889,
+      "learning_rate": 2.428708863557307e-05,
+      "loss": 2.1715,
+      "step": 3429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1483145952224731,
+      "learning_rate": 2.4259454041755503e-05,
+      "loss": 2.2339,
+      "step": 3430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2792038917541504,
+      "learning_rate": 2.4231830142091948e-05,
+      "loss": 2.0597,
+      "step": 3431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2527892589569092,
+      "learning_rate": 2.420421694805904e-05,
+      "loss": 2.2562,
+      "step": 3432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2273582220077515,
+      "learning_rate": 2.4176614471128917e-05,
+      "loss": 1.9843,
+      "step": 3433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.324568748474121,
+      "learning_rate": 2.414902272276931e-05,
+      "loss": 2.181,
+      "step": 3434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3886842727661133,
+      "learning_rate": 2.412144171444346e-05,
+      "loss": 2.2948,
+      "step": 3435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2613803148269653,
+      "learning_rate": 2.4093871457610167e-05,
+      "loss": 2.1136,
+      "step": 3436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3148837089538574,
+      "learning_rate": 2.4066311963723747e-05,
+      "loss": 2.1117,
+      "step": 3437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3787257671356201,
+      "learning_rate": 2.4038763244234063e-05,
+      "loss": 2.2657,
+      "step": 3438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2453417778015137,
+      "learning_rate": 2.40112253105865e-05,
+      "loss": 2.066,
+      "step": 3439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1916533708572388,
+      "learning_rate": 2.3983698174221924e-05,
+      "loss": 1.9593,
+      "step": 3440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2974342107772827,
+      "learning_rate": 2.3956181846576763e-05,
+      "loss": 2.1609,
+      "step": 3441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.187440037727356,
+      "learning_rate": 2.392867633908294e-05,
+      "loss": 2.0549,
+      "step": 3442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2634189128875732,
+      "learning_rate": 2.3901181663167867e-05,
+      "loss": 2.3176,
+      "step": 3443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1905274391174316,
+      "learning_rate": 2.3873697830254478e-05,
+      "loss": 2.1809,
+      "step": 3444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.136976718902588,
+      "learning_rate": 2.3846224851761185e-05,
+      "loss": 2.092,
+      "step": 3445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2870666980743408,
+      "learning_rate": 2.3818762739101908e-05,
+      "loss": 2.258,
+      "step": 3446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2269023656845093,
+      "learning_rate": 2.3791311503686032e-05,
+      "loss": 2.001,
+      "step": 3447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0422815084457397,
+      "learning_rate": 2.376387115691844e-05,
+      "loss": 2.1386,
+      "step": 3448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.115127444267273,
+      "learning_rate": 2.3736441710199475e-05,
+      "loss": 2.0571,
+      "step": 3449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.149420142173767,
+      "learning_rate": 2.3709023174924978e-05,
+      "loss": 2.034,
+      "step": 3450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.192733883857727,
+      "learning_rate": 2.368161556248622e-05,
+      "loss": 2.2271,
+      "step": 3451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2757207155227661,
+      "learning_rate": 2.3654218884269963e-05,
+      "loss": 2.1705,
+      "step": 3452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.259885311126709,
+      "learning_rate": 2.3626833151658423e-05,
+      "loss": 2.6257,
+      "step": 3453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1721830368041992,
+      "learning_rate": 2.359945837602925e-05,
+      "loss": 2.1762,
+      "step": 3454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.995424747467041,
+      "learning_rate": 2.357209456875556e-05,
+      "loss": 2.0071,
+      "step": 3455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.223223090171814,
+      "learning_rate": 2.3544741741205912e-05,
+      "loss": 2.3844,
+      "step": 3456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2155178785324097,
+      "learning_rate": 2.3517399904744287e-05,
+      "loss": 2.0828,
+      "step": 3457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0229766368865967,
+      "learning_rate": 2.349006907073012e-05,
+      "loss": 2.0394,
+      "step": 3458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.167786717414856,
+      "learning_rate": 2.3462749250518274e-05,
+      "loss": 1.9761,
+      "step": 3459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3153791427612305,
+      "learning_rate": 2.3435440455459003e-05,
+      "loss": 2.3946,
+      "step": 3460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1345432996749878,
+      "learning_rate": 2.3408142696898013e-05,
+      "loss": 2.1126,
+      "step": 3461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.105927586555481,
+      "learning_rate": 2.3380855986176414e-05,
+      "loss": 2.158,
+      "step": 3462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1281957626342773,
+      "learning_rate": 2.3353580334630742e-05,
+      "loss": 2.002,
+      "step": 3463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2178541421890259,
+      "learning_rate": 2.3326315753592916e-05,
+      "loss": 2.3034,
+      "step": 3464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2159948348999023,
+      "learning_rate": 2.3299062254390264e-05,
+      "loss": 2.1387,
+      "step": 3465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.157670497894287,
+      "learning_rate": 2.3271819848345518e-05,
+      "loss": 1.9185,
+      "step": 3466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1643574237823486,
+      "learning_rate": 2.324458854677679e-05,
+      "loss": 1.85,
+      "step": 3467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1052184104919434,
+      "learning_rate": 2.32173683609976e-05,
+      "loss": 1.9566,
+      "step": 3468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1593255996704102,
+      "learning_rate": 2.3190159302316794e-05,
+      "loss": 2.0693,
+      "step": 3469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0666685104370117,
+      "learning_rate": 2.3162961382038658e-05,
+      "loss": 2.1524,
+      "step": 3470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1494935750961304,
+      "learning_rate": 2.3135774611462817e-05,
+      "loss": 2.0642,
+      "step": 3471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.311722755432129,
+      "learning_rate": 2.3108599001884284e-05,
+      "loss": 2.0139,
+      "step": 3472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2731173038482666,
+      "learning_rate": 2.3081434564593418e-05,
+      "loss": 2.2497,
+      "step": 3473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2750704288482666,
+      "learning_rate": 2.305428131087594e-05,
+      "loss": 1.9958,
+      "step": 3474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2233006954193115,
+      "learning_rate": 2.302713925201293e-05,
+      "loss": 2.183,
+      "step": 3475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.155893325805664,
+      "learning_rate": 2.300000839928081e-05,
+      "loss": 1.8292,
+      "step": 3476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2156445980072021,
+      "learning_rate": 2.2972888763951357e-05,
+      "loss": 2.0433,
+      "step": 3477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1050035953521729,
+      "learning_rate": 2.2945780357291675e-05,
+      "loss": 1.862,
+      "step": 3478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.124853491783142,
+      "learning_rate": 2.291868319056421e-05,
+      "loss": 2.2447,
+      "step": 3479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2468905448913574,
+      "learning_rate": 2.2891597275026737e-05,
+      "loss": 2.1239,
+      "step": 3480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.136094570159912,
+      "learning_rate": 2.286452262193236e-05,
+      "loss": 2.0415,
+      "step": 3481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2157008647918701,
+      "learning_rate": 2.2837459242529498e-05,
+      "loss": 2.1956,
+      "step": 3482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.150514841079712,
+      "learning_rate": 2.28104071480619e-05,
+      "loss": 2.1788,
+      "step": 3483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1859058141708374,
+      "learning_rate": 2.278336634976859e-05,
+      "loss": 1.9063,
+      "step": 3484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3259600400924683,
+      "learning_rate": 2.275633685888394e-05,
+      "loss": 1.9958,
+      "step": 3485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0826069116592407,
+      "learning_rate": 2.2729318686637608e-05,
+      "loss": 2.1752,
+      "step": 3486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.128329873085022,
+      "learning_rate": 2.270231184425454e-05,
+      "loss": 2.0518,
+      "step": 3487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0676186084747314,
+      "learning_rate": 2.2675316342955e-05,
+      "loss": 1.8877,
+      "step": 3488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1359302997589111,
+      "learning_rate": 2.2648332193954515e-05,
+      "loss": 1.9452,
+      "step": 3489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2571144104003906,
+      "learning_rate": 2.262135940846391e-05,
+      "loss": 2.34,
+      "step": 3490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0278241634368896,
+      "learning_rate": 2.259439799768928e-05,
+      "loss": 1.9997,
+      "step": 3491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0454822778701782,
+      "learning_rate": 2.2567447972832013e-05,
+      "loss": 1.8258,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.418929100036621,
+      "learning_rate": 2.2540509345088734e-05,
+      "loss": 2.0979,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1863402128219604,
+      "learning_rate": 2.2513582125651356e-05,
+      "loss": 2.0985,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1562347412109375,
+      "learning_rate": 2.2486666325707058e-05,
+      "loss": 1.9612,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2933623790740967,
+      "learning_rate": 2.245976195643825e-05,
+      "loss": 2.2341,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1080174446105957,
+      "learning_rate": 2.2432869029022624e-05,
+      "loss": 1.6458,
+      "step": 3497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3441821336746216,
+      "learning_rate": 2.240598755463309e-05,
+      "loss": 2.0617,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1854298114776611,
+      "learning_rate": 2.2379117544437817e-05,
+      "loss": 2.0928,
+      "step": 3499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1928367614746094,
+      "learning_rate": 2.2352259009600198e-05,
+      "loss": 2.1527,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5025,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 3.621074911834931e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}